Amsterdam, 18 October 2019

EMA/595311/2019

Withdrawal Assessment report

Nuzyra (omadacycline tosylate)

Procedure No. EMEA/H/C/4715

Note
Assessment report as adopted by the CHMP with all information of a commercially confidential nature
deleted

Official address Domenico Scarlattilaan 6 ● 1083 HS Amsterdam ● The Netherlands

Address for visits and deliveries Refer to www.ema.europa.eu/how-to-find-us
Send us a question Go to www.ema.europa.eu/contact Telephone +31 (0)88 781 6000 An agency of the European Union

© European Medicines Agency, 2020. Reproduction is authorised provided the source is acknowledged.
Table of contents
1. CHMP Recommendation ........................................................................... 8
2. Executive summary ................................................................................. 9
2.1. Problem statement ............................................................................................... 9
2.1.1. Disease or condition........................................................................................... 9
2.1.2. Epidemiology .................................................................................................... 9
2.1.3. Aetiology and pathogenesis ................................................................................ 9
2.1.4. Clinical presentation, diagnosis............................................................................ 9
2.1.5. Management ................................................................................................... 10
2.2. About the product .............................................................................................. 10
2.3. The development programme/compliance with CHMP guidance/scientific advice ......... 11
2.4. General comments on compliance with GMP, GLP, GCP ........................................... 11
2.5. Type of application and other comments on the submitted dossier............................ 11

3. Scientific overview and discussion ........................................................ 12

3.1. Quality aspects .................................................................................................. 12
3.1.1. Introduction .................................................................................................... 12
3.1.2. Active Substance ............................................................................................. 12
3.1.3. Finished Medicinal Product - Powder for Concentrate for Solution for Infusion.......... 13
3.1.4. Finished Medicinal Product – Film-coated tablets ................................................. 14
3.1.5. Discussion and conclusions on chemical, pharmaceutical and biological aspects ....... 15
3.2. Non-clinical aspects ............................................................................................ 15
3.2.1. Pharmacology ................................................................................................. 15
3.2.2. Pharmacokinetics............................................................................................. 16
3.2.3. Toxicology ...................................................................................................... 18
3.2.4. Discussion on non-clinical aspects...................................................................... 20
3.2.5. Conclusion on non-clinical aspects ..................................................................... 22
3.3. Clinical aspects .................................................................................................. 22
3.3.1. Pharmacokinetics............................................................................................. 22
3.3.2. Pharmacodynamics .......................................................................................... 32
3.3.3. Discussion on clinical pharmacology ................................................................... 46
3.3.4. Conclusions on clinical pharmacology ................................................................. 47
3.3.5. Clinical efficacy ............................................................................................... 49
3.3.6. Analysis performed across trials (pooled analyses) .............................................. 86
3.3.7. Clinical studies in special populations ................................................................. 87
3.3.8. Supportive studies ........................................................................................... 88
3.3.9. Discussion on clinical efficacy ............................................................................ 89
3.3.10. Clinical safety ................................................................................................ 91
3.3.11. Discussion on clinical safety .......................................................................... 111
3.3.12. Conclusions on clinical safety ........................................................................ 113
3.4. Risk management plan...................................................................................... 113
3.4.1. Safety Specification ....................................................................................... 113
3.4.2. Discussion on safety specification .................................................................... 113
3.4.3. Conclusions on the safety specification ............................................................. 113
3.4.4. Pharmacovigilance plan .................................................................................. 114

EMA/595311/2019 Page 2/123

3.4.5. Risk minimisation measures ............................................................................ 114
3.4.6. Conclusion on the RMP ................................................................................... 115
3.5. Pharmacovigilance system ................................................................................. 115

4. Benefit risk assessment....................................................................... 115

4.1. Therapeutic Context ......................................................................................... 115
4.1.1. Disease or condition ....................................................................................... 115
4.1.2. Available therapies and unmet medical need ..................................................... 115
4.1.3. Main clinical studies ....................................................................................... 116
4.2. Favourable effects ............................................................................................ 116
4.3. Uncertainties and limitations about favourable effects ........................................... 117
4.4. Unfavourable effects ......................................................................................... 117
4.5. Uncertainties and limitations about unfavourable effects ....................................... 118
4.6. Effects Table .................................................................................................... 118
4.7. Benefit-risk assessment and discussion ............................................................... 118
4.7.1. Importance of favourable and unfavourable effects ............................................ 118
4.7.2. Balance of benefits and risks ........................................................................... 119
4.8. Conclusions ..................................................................................................... 119

6. QRD checklist for the review of user testing results ............................ 120
1. Technical assessment ........................................................................ 121
1.1 Recruitment.................................................................................................... 121
1.2 Questionnaire ................................................................................................. 121
1.3 Time aspects .................................................................................................. 121
1.4 Procedural aspects .......................................................................................... 121
1.5 Interview aspects ............................................................................................ 121

2. Evaluation of responses ..................................................................... 121

2.1 Evaluation system ........................................................................................... 122
2.2 Question rating system .................................................................................... 122

3. Data processing ................................................................................. 122

4. Quality aspects .................................................................................. 122
4.1 Evaluation of diagnostic questions ..................................................................... 122
4.2 Evaluation of layout and design ......................................................................... 122

5. Diagnostic quality/evaluation............................................................ 123

6. Conclusions ....................................................................................... 123

EMA/595311/2019 Page 3/123

Abbreviations
ABSSSI Acute bacterial skin and skin structure infections
AC Alveolar cell
AE Adverse event
ALP Alkaline phosphatase
ALT Alanine aminotransferase
AP Action potential
AST Aspartate aminotransferase
ATC Anatomic Therapeutic Chemical
ATCC American Type Culture Collection
ATS American Thoracic Society
AUC Area Under Curve
B/R Benefit/Risk
BAL Bronchiolar lavage
BBB Blood brain barrier
BCS Biopharmaceutics Classification System
BMI Body mass index
BSA Body surface area
BUN Blood urea nitrogen
CABP Community-acquired bacterial pneumonia
CAP Community Acquired Pneumonia
CE Clinically evaluable
CEP Certificate of suitability of Monographs of the European Pharmacopoeia
CF Cystic fibrosis
CHMP Committee for Medicinal Products for Human Use
CI Confidence interval
CL Clearance
CLr Renal clearance
CLSI Clinical & Laboratory Standards Institute
Cmax Maximum concentration
COPD Chronic obstructive pulmonary disease
CrCl Creatinine clearance
CSR Clinical study report
cSSSI Complicated skin and skin structure infections
CT Computed tomography
CV Coefficient of variation
CYP Cytochrome P450
DBL Database lock
DBP Diastolic blood pressure
ECG Electrocardiogram
ECR Early Clinical Response
EDTA Ethylenediaminetetraacetic acid
ELF Epithelial lining fluid
EMA European Medicines Agency
EOT End of treatment
ESBL Extended-spectrum beta-lactamase
ESRD End-stage renal disease
EU European Union

EMA/595311/2019 Page 4/123

EUCAST European Committee on Antimicrobial Susceptibility Testing
FCC Food Chemicals Codex
FDA Food and Drug Administration
FIC Fractional inhibitory concentrations
FK Karl Fisher
FTIR Fourier-Transform Infrared Spectroscopy
G- Gram negative
G+ Gram positive
GCP Good Clinical Practice
GCP Gas Chromatography
GGT Gamma-glutamyl transferase
GI Gastrointestinal
GLP Good Laboratory Practice
GMP Good Manufacturing Practice
HAP Hospital-acquired pneumonia
HCAP Health-care associated pneumonia
HDPE High-density polyethylene
HPLC High-performance liquid chromatography
HPLC-UV High-performance liquid chromatography - Ultraviolet
HR Heart rate
IACR Investigator’s assessment of clinical response
IC50 Half maximal inhibitory concentration
International Conference on Harmonisation of Technical Requirements for Registration of
ICH
Pharmaceuticals for Human Use
ICU Intensive care unit
INN International non-proprietary name
ITT Intention to treat
IV Intravenous
LC-MS/MS Liquid chromatography with tandem mass spectrometric detection
LDPE Low-density polyethylene
LLOQ Lower Limit of Quantification
LoQ List of Questions
MAA Marketing authorisation application
MDR Multi-drug resistance
MDRSP Multidrug-resistant S. pneumoniae
ME Microbiologically-Evaluable
MIC Minimum Inhibitory Concentration
MIC50 Minimum inhibitory concentration against 50% of the isolates
MIC90 Minimum inhibitory concentration against 90% of the isolates
Micro-
Microbiological modified intent-to-treat
mITT
MIEC Minimal inhibitory extracellular concentration
mITT Modified intent to treat
MO Major Objection
MOA Monoamine oxidase
MRSA Methicillin resistant Staphylococcus aureus
MSSA Methicillin susceptible Staphylococcus aureus
MTD Maximum tolerated dose
N/A Not Applicable

EMA/595311/2019 Page 5/123

ND Not determined
NDA New Drug Application
NF National Formulary
NMR Nuclear magnetic resonance
NRU Neutral Red Uptake
OC Other Concern
OMC Omadacycline
PD Pharmacodynamic
PEC Predicted environmental concentration
Ph. Eur. European Pharmacopoeia
PI Product information
PIP Paediatric Investigation Plan
PK Pharmacokinetic
PL Product leaflet
PO Per oral
popPK Population pharmacokinetics
PORT Pneumonia Outcomes Research Team
PRAC Pharmacovigilance Risk Assessment Committee
PRSP Penicillin resistant Streptococcus pneumoniae
PSSP Penicillin susceptible Streptococcus pneumoniae
PSUR Periodic Safety Update Report
PT Preferred Term
PTA Probability of Target Attainment
PTE Post-therapy evaluation
PXR Pregnane X receptor
q.s. quantity sufficient
q12h Every 12 hours
q24h Every 24 hours
QD Four times daily
RBC Red blood cell
RHD Recommended human dose
RMP Risk Management Plan
ROI Residue on Ignition
SAE Serious adverse event
SAP Statistical analysis plan
SB Single blind
SBP Systolic blood pressure
SEM Structural equation modeling
SIR Systemic inflammatory response
SIRS Systemic inflammatory response syndrome
SmPC Summary of Product Characteristics
SOC System Organ Class
SPN Streptococcus pneumoniae
t1/2 Terminal half-life
TB Tuberculosis
TEAE Treatment emergent adverse event
TK Toxicokinetic
TK Time-kill

EMA/595311/2019 Page 6/123

tmax Time of maximum plasma concentration
TOC Test of cure
ULOQ Upper limit of quantitation
US United States
USP United States Pharmacopeia
uUTI Uncomplicated urinary tract infection
V Volume of distribution
VAP Ventilator-acquired pneumonia
WBC White blood cell
VPC Visual predictive check
VRE Vancomycin-resistant enterococci
VSE Vancomycin-susceptible enterococci
Vss Steady-state volume of distribution
XRPD X-ray powder diffraction

EMA/595311/2019 Page 7/123

1. CHMP Recommendation
Based on the review of the data on quality, safety, efficacy, the application for Nuzyra (previously
proposed named Nualto) in the treatment of adults with the following infections (see sections 4.4 and
5.1):

- Community-acquired pneumonia (CAP)

- Acute bacterial skin and skin structure infections (ABSSSI)

is not approvable since a "major objection" has been identified, which preclude a recommendation for
marketing authorisation at the present time. The details of this major objection are provided in the List
of Questions.

The major objection precluding a recommendation of marketing authorisation pertains to the following
principal deficiencies:

Benefit/Risk
A positive B/R balance for omadacycline in CAP cannot be inferred based on available data. The
evidence for this indication comes solely from a single pivotal study. Moreover, an unexplained
difference in mortality rate was noted between treatment arms, enhancing the uncertainties
due to the single pivotal trial. Additional efficacy and safety data from the planned further
study in CAP will be required to provide a comprehensive basis for evaluating the B/R balance
for omadacycline in CAP, for the purposes of authorisation for this indication.

Inspection issues

GMP inspection

Inspections of the drug substance manufacturing site and/or the drug product manufacturing site
and/or the batch release site are not considered necessary for finalization of assessment of Module3.

GCP inspection

A routine request for GCP inspection was adopted by CHMP for the following clinical studies: PTK0796-
ABSI-1108 and PTK0796-CABP-1200 (ref INS/GCP/2018/031). The routine inspection included one site
for each of the studies (located in Poland and South Africa) plus the contract research organisation. In
summary, the conclusion was that the clinical trial data are reliable enough to support the application
submitted for Nuzyra.

New active substance status

Based on the review of data on the quality of the active substance, the Rapporteur considers that
omadacycline is to be qualified as a new active substance in itself.

EMA/595311/2019 Page 8/123

2. Executive summary

2.1. Problem statement

2.1.1. Disease or condition

The indications applied for are treatment of adults with:

• Acute bacterial skin and skin structure infections (ABSSSI)

• Community-acquired bacterial pneumonia (CABP)

2.1.2. Epidemiology

Acute bacterial skin and skin structure infections (ABSSSI) can be serious, limb- or life-threatening
conditions, often requiring systemic antibiotic therapy and possibly surgical management and
hospitalisation. In 2012, there were an estimated 850,000 skin infections across the 5 major EU
markets (France, Germany, Italy, Spain, and United Kingdom). The economic impact of CAP is
substantial, with an annual cost in excess of $10 billion in Europe.

Community-acquired pneumonia is the most common infectious disease leading to hospitalisation and
mortality among all age groups and is a leading cause of morbidity and mortality and throughout the
world. The annual incidence of CAP across Europe ranges from 1.54 to 1.7 per 1,000 adults with over 3
million ambulatory cases and approximately 1 million hospital admissions per year. Mortality rates are
5% to 15% in hospitalised patients. Between 1.2% and 10% of adults admitted to hospital with CABP
are managed in an intensive care unit, and for these patients mortality is more than 40%. Even after
successful treatment, residual symptoms such as cough, chest pain and fatigue may persist for up to 6
months.

2.1.3. Aetiology and pathogenesis

Gram-positive bacteria, especially S. aureus and Streptococcus spp., are the most frequent cause of
ABSSSI. Patients with necrotising skin infections, infections associated with human or animal bites,
traumatic or surgical wound infections, and skin infections associated with iv drug use, as well as
immunosuppressed patients with ABSSSI, may require treatment with antibiotics that provide more
than narrow spectrum anti-staphylococcal and anti-streptococcal activity.

Typical bacterial pathogens implicated in the majority of cases of CABP are S. pneumoniae and
Haemophilus influenzae. Atypical pathogens, including Mycoplasma pneumoniae, Legionella
pneumophila and Chlamydia pneumoniae, affect a considerable proportion of CABP patients.

2.1.4. Clinical presentation, diagnosis

ABSSSI comprises several types, including cellulitis/erysipelas, major abscess and wound infection, as
well as rarer conditions such as infected lower extremity ulcers. In some instances, ABSSSI may arise
related to a foreign body, animal or human bite, trauma or surgical wound, or iv drug abuse, while in
others no obvious cause may be found. Patients with pre-existing immunosuppression or diabetes
mellitus may be more prone to ABSSSI.

Pneumonia may present in different ways in different patients, ranging from mild symptoms of lower
respiratory tract infection, with or without fever, to severe sepsis and systemic collapse. CABP is
classified separately from Hospital-acquired pneumonia (HAP) (developing ≥48 hours after hospital
admission), Ventilator-acquired pneumonia (VAP) or Health-care associated pneumonia (HCAP), which

EMA/595311/2019 Page 9/123

develop in hospitalised and ventilated patients or residents of long term care facilities, tend to be
caused by a different range of pathogens, in particular a higher frequency of multi-drug resistant
organisms, and are associated with increasing morbidity and mortality.

2.1.5. Management

The management of ABSSSI includes topical and/or systemic antibiotics generally aimed at Gram
positive pathogens (including beta-lactams, lipopeptides, oxazolidinones, glycopeptides and
tetracyclines), combined with other topical treatments and dressings, and often with adjunctive
surgical management to remove any causative foreign body, debride diseased tissue and drain any
associated collection. Addition of a further antibiotic with a different profile may be required where
Gram negative or anaerobic pathogens are suspected.

The management of CABP includes systemic antibiotics (including beta-lactams, macrolides,

fluoroquinolones, oxazolidinones and tetracyclines), combined with supportive care e.g. fluid
management, supplementary oxygen, ventilatory support and, occasionally, adjunctive procedures e.g.
drainage of an associated pleural effusion. Empiric therapy, expected to cover the most common
organisms and most common resistance patterns for the local area, may be indicated where a
causative pathogen cannot be identified from microbiological samples. Only fluoroquinolones,
macrolides and tetracyclines provide reliable cover where atypical pathogens are suspected or
confirmed.

The emergence of resistance, including multi-drug resistance (MDR), against first line antibiotics has,
as in other areas, made treatment of both ABSSSI and CABP more challenging and there is an unmet
need for additional antibiotic treatment options.

2.2. About the product

Omadacycline exhibits a bacteriostatic effect by binding to the 30S subunit of the bacterial ribosome,
thereby blocking bacterial protein synthesis.

Pharmacotherapeutic group: Antibacterials for systemic use, ATC code not yet assigned.

The proposed indication is:

Omadacycline is indicated for the treatment of adults with the following infections:

• Community acquired bacterial pneumonia (CABP)

• Acute bacterial skin and skin structure infections (ABSSSI)

The proposed posology is:

CAP and ABSSSI: Treatment duration is 7 to 14 days.

Infection Loading Doses Maintenance Dose

CAP Day 1: 200 mg 100 mg by intravenous infusion

by intravenous infusion over 60 over 30 minutes once daily
minutes
OR
OR
300 mg orally once daily
100 mg by intravenous infusion
over 30 minutes twice

EMA/595311/2019 Page 10/123

 ABSSSI Day 1: 200 mg 100 mg by intravenous infusion
by intravenous infusion over 60 over 30 minutes once daily
minutes
OR
OR
300 mg orally once daily
100 mg by intravenous infusion
over 30 minutes twice

OR

Day 1 and Day 2:

450 mg orally once daily
Hepatic or Renal Impairment

No dose adjustment is necessary in patients with mild, moderate, or severe hepatic impairment (Child
Pugh classes A, B, or C) (see section 5.2).

No dose adjustment is necessary in patients with mild, moderate, or severe renal impairment
(including End Stage Renal Disease) (see section 5.2).

2.3. The development programme/compliance with CHMP

guidance/scientific advice

The applicant sought scientific advice on the overall development programme for acute bacterial skin
and skin structure infections (ABSSSI) and complicated skin and skin structure infections (cSSSI),
community-acquired bacterial pneumonia (CABP), from national agencies (FR, UK, DE) and CHMP in
2008 and 2009.

A US pivotal Phase 3 study in cSSSI was initiated but then discontinued by the applicant following a
change in FDA Guidance on the development of treatments for ABSSSI. A planned EU pivotal study in
cSSSI was subsequently delayed. The applicant later sought advice from CHMP on the design of new
Phase 3 studies in ABSSSI and CABP in 2015 in two parallel procedures.

Scientific advice was generally consistent between CHMP and FDA, with the exception of primary
efficacy endpoint timing in Phase 3 clinical studies (Post-Treatment Evaluation (PTE) for CHMP, Early
Clinical Response (ECR) for FDA), which was addressed by separate Statistical Analysis Plans.

2.4. General comments on compliance with GMP, GLP, GCP

No specific concerns had been identified by the assessment at the time of adoption of the inspection
request. Critical observations raised at a site in study 16301, identified by the sponsor and reported in
the FDA NDA, led to exclusion of data from this site for the purposes FDA analysis, as highlighted at
D120. According to the FDA NDA assessment, a sensitivity analysis excluding the patients from this
site was consistent with the results from primary analysis.

2.5. Type of application and other comments on the submitted dossier

Legal basis

The legal basis for this application refers to:

Article 8.3 of Directive 2001/83/EC, as amended - complete and independent application.

EMA/595311/2019 Page 11/123

New active substance status
Based on the review of data on the quality of the active substance, the Rapporteur considers that
omadacycline is to be qualified as a new active substance in itself.

Orphan designation - Not Applicable

Similarity with orphan medicinal products

The application did not contain a critical report pursuant to Article 8 of Regulation (EC) No. 141/2000
and Article 3 of Commission Regulation (EC) No 847/2000, addressing the possible similarity with
authorised orphan medicinal products.

Information on paediatric requirements

Pursuant to Article 7 of Regulation (EC) No 1901/2006, the application included an EMA Decision(s)
P/0269/2018 on the agreement of a paediatric investigation plan (PIP). At the time of submission of
the application, the PIP P/0269/2018 was not yet completed as some measures were deferred.

3. Scientific overview and discussion

3.1. Quality aspects

3.1.1. Introduction

Nuzyra (omadacycline) has been developed in two dosage forms, powder for concentrate for solution
for infusion (IV product) and film-coated tablets (oral tablets).

3.1.2. Active Substance

General Information

The International non-proprietary name (INN) is omadacycline and the salt form omadacycline tosylate
is the drug substance. The drug substance is a crystalline compound with three forms, Form 3 being
the desired polymorphic form. Aqueous solubility of the drug substance is highly soluble across the pH
range 1-7, with a solubility of ≥ 2 g/mL for all tested solutions.

N N
H H
OH
H
N NH2
OH
OH O OH O O SO3H

Molecular formula: Salt form on anhydrous basis: C29H40N4O7 ∙ C7H8O3S

Relative molecular mass:Salt form: 728.9

Manufacture, process controls and characterisation

The synthesis of omadacycline tosylate was developed and is sufficiently robust to provide assurance
that the process produces the drug substance of consistent quality, complying with the designed
specification.

Starting materials

EMA/595311/2019 Page 12/123

The starting material was agreed through the pre-submission MAA meetings advice with the Co-
Rapporteur on 03-May-2018 (reference is being made to the Co-Rapporteur meeting minutes) and the
Rapporteur on 08-Jun-2018 (reference is being made to the Rapporteur meeting minutes) and
supported by CHMP.
Intermediate
There are intermediates defined in the omadacycline tosylate drug substance process.
The route of synthesis is sufficiently described, and the major phases in the synthesis of omadacycline
tosylate are controlled during the reaction.

Characterisation
The structure of Omadacycline p-Toluenesulfonate (omadacycline tosylate) drug substance has been
confirmed through Fourier-Transform Infrared Spectroscopy (FTIR), UV, Mass Spectrum, NMR
(including multiplicity, coupling constants and integral values), and X-ray Crystallography. In addition,
Structure and XRPD were used to confirm Form 3 of the tosylate salt.

Impurities
Potential organic related substances and degradants are detected using stability indicating HPLC
methods.

Specification, analytical procedures, reference standards, batch analysis,

and container closure
The specifications for the drug substance are based on batch analyses of several batches of drug
substance prepared by the commercial process, and batches used clinically and toxicologically, as well
as stability data. The methodology has been validated to meet the general requirements of Ph Eur
(where relevant) and the ICH guideline Q2B, Validation of Analytical Procedures; Methodology.

Omadacycline p-toluenesulfonate (omadacycline tosylate) drug substance is packaged in an

appropriate container closure system. The specification is considered acceptable.

Stability

Stability data are available for 3 primary registration/stability batches of omadacycline tosylate drug
substance through 18 months at long term conditions and 6 months at accelerated conditions. The
results through 18 months at long term conditions and 6 months at accelerated conditions are
consistent, remain unchanged and meet the acceptance criteria for appearance, polymorphic form,
water, assay and related substances. Individual data are presented indicate that the drug substance is
stable at the intended storage condition, refrigeration (2°C to 8°C). The stability data provided support
the retest period of 24 months when stored below 2°C to 8°C provided that the issues identified have
been clarified.

3.1.3. Finished Medicinal Product - Powder for Concentrate for Solution for
Infusion

Description of the product and Pharmaceutical Development

Omadacycline 100 mg Powder for Concentrate for Solution for Infusion is a sterilized, lyophilized
powder reconstituted to a solution for infusion. It is available as a 100 mg unit dose in a clear glass
vial with grey elastomer stopper, and aluminium crimped-on seal with tear-off cap.

The compatibility of the chosen excipients with omadacycline has been demonstrated by in use stability
of reconstituted lyophilizate and solid-state drug product stability studies. The final formula consists of

EMA/595311/2019 Page 13/123

omadacycline tosylate, sucrose, hydrochloric acid concentrated, sodium hydroxide and water for
injection. The container closure system (clear glass vial with a grey elastomer stopper, secured with
aluminium crimped-on seal with a tear-off cap) is also in common use for packaging of commercial
sterile pharmaceutical products and meets USP/Ph. Eur. requirements. The pharmaceutical
development has adequately been described and it is considered to be acceptable. The choice of sterile
processing has been justified.

Manufacture of the product and process controls

The process for the manufacturing of the finished product follows conventional pharmaceutical
practices.
It is considered that the manufacture is sufficiently robust to provide assurance that the process
produces the finished product 100 mg powder for solution for infusion of consistent quality, complying
with the designed specification.

Product specification, analytical procedures, batch analysis

The release and stability (shelf-life) specifications presented cover relevant parameters for the dosage
form: appearance , identification, assay, related substances, visible particles of the reconstituted
solution, sub-visible particles of the reconstituted solution, uniformity of dosage units, osmolality of the
reconstituted solution, pH of the reconstituted solution, water content, bacterial endotoxins, sterility.
The methodology is validated to meet the general requirements of Ph Eur (where relevant) and the
ICH guideline Q2B, Validation of Analytical Procedures; Methodology.

Stability of the product

Stability data for omadacycline IV was provided and support the proposed shelf life. The results
provided through 24 months at long term conditions and 6 months at accelerated conditions support
the proposed shelf-life provided and the issues identified have been clarified by the applicant.

3.1.4. Finished Medicinal Product – Film-coated tablets

Description of the product and Pharmaceutical Development

The drug product is described as an immediate release, yellow-coloured, diamond-shaped, film-coated

tablet containing the drub substance omadacycline tosylate equivalent to 150 mg of omadacycline.

All excipients utilized to manufacture omadacycline tablets are commonly used in pharmaceutical
applications. The dissolution method development has been provided. The dissolution conditions are
justified.

The final tablet to be proposed for the commercial market is a yellow, diamond shaped, film-coated
tablet with “OMC” on one side and “150” on the other side.

The manufacturing process development has been described in detail. The process of manufacture is a
standard process. The conditions setting, and in-process controls have been discussed.

Microbial testing on one batch per year on release as proposed by the applicant is considered justified.
No changes in microbial growth have been observed up to date.

Manufacture of the product and process controls

The process is considered a standard manufacturing process and robustness of the process have been
demonstrated during development.

Product specification, analytical procedures, batch analysis

EMA/595311/2019 Page 14/123

The release and stability (shelf-life) specifications presented cover relevant parameters for the dosage
form: appearance, identification, assay, related substances, dissolution, water content, uniformity of
dosage units, microbial enumeration, sodium bisulphite identification.

A risk assessment with respect to the potential presence of elemental impurities in the drug product
based on the general principles outlined in Section 5.1 of ICH-guideline Q3D has been performed.

The analytical methods have been described and summary of validation is presented in alignment with
the general requirements of Ph Eur (where relevant) and the ICH guideline Q2B, Validation of
Analytical Procedures; Methodology. Microbiological examination test is performed according to the
methods described in European Pharmacopeia 2.6.12 and 2.6.13.

Batch analysis results comply with the proposed specification and confirm that batches are of
consistent quality.

The proposed commercial packaging is an aluminium blister with an integrated desiccant and
aluminium lidding blister. The packaging materials have been adequately described.

Stability of the product

Updated long term stability data is presented through 18 months at long term conditions. The data
remains consistent and meets the acceptance criteria for appearance, assay, related substances,
dissolution and water content.

Based on the stability data, the shelf life and storage condition proposed by the applicant is judged
acceptable.

Adventitious agents – N/A

GMO – N/A

3.1.5. Discussion and conclusions on chemical, pharmaceutical and

biological aspects

The information provided on drug substance and drug product is considered sufficient.

The choice of aseptic process has been justified.

The application is recommended for approval from a chemical and pharmaceutical point of view.

3.2. Non-clinical aspects

3.2.1. Pharmacology

Primary pharmacodynamics – see section 3.3.2. Pharmacodynamics.

Secondary pharmacodynamics

Potential off-target effects of omadacycline was evaluated in batteries of receptors and ion channels.
Omadacycline was found to inhibit the muscarinic M2 receptor, with an IC50 value of 4.25 µM. In
addition, IC50 values between 10 and 30 μM were found in nine binding assays (adenosine 3 receptor,
alpha 2 adrenergic 2B receptor, dopamine D3 receptor, histamine H3 receptor, muscarinic M1 receptor,
opiate delta receptor, serotonin transporter, serotonin 5HT3 ion channel, pregnane X receptor (PXR)).

Safety pharmacology

EMA/595311/2019 Page 15/123

In vitro studies

In general, omadacycline caused depolarisation of the resting membrane potential, reduced AP

amplitude, slight AP prolongation, reduced upstroke and rate of depolarisation at concentrations >100
µM, indicating low clinical relevance.

Omadacycline was considered a low-potency blocker of hERG K+ current (IC25 166 μg/mL) and INa
(IC50 1.5 mM). Omadacycline inhibited the Na+/K+ ATPase pump current at approximately 10-fold
lower concentrations (IC50 of 0.29-0.52 mM) than those that affected hERG K+ and INa. Thus, the
observed effects on hERG K+ and INa could be secondary to depolarization of the resting potential.
Due to exposure margins of ≥ 97, these findings are considered of low clinical relevance.

In the rabbit SA-node, increased cycle length and reduced spontaneous beating rate induced by the
pan-muscarinic M2 agonist carbamylcholine was reversed by omadacycline concentrations of 8-390
μM. This is consistent with omadacycline being a muscarinic M2 antagonist, with an IC50 value of 4.25
µM, and may indicate a potential to inhibit vagal regulation of the SA node.

In vivo studies

In anaesthetized monkeys, omadacycline led to decreased dP/dtmax+, dP/dtmax-, and left ventricular
dP/dt.P-1 in the first 15-60 minutes of omadacycline infusion at the highest dose of 90 mg/kg (14
times the clinical relevant AUC).

In conscious monkeys, omadacycline resulted in a slight increase in blood pressure (systolic, diastolic
and mean) and a moderate increase in heart rate at all dose levels (increased with 37, 58, and 27 BPM
at 5, 20 and 40 mg/kg, respectively). While blood pressure normalized within 30 minutes, the
increase in heart rate persisted for ≥ 4.5 hours post-dose initiation. The findings were consistent with
M2 antagonist activity seen in binding assays, and with reversal of M2 agonism effects (i.e., reduced
spontaneous beat rate) in the rabbit sinoatrial node study in vitro. No effects were seen on ECG
parameters, body temperature, or pulmonary/respiratory parameters.

No omadacycline-related effects were seen on renal parameters in rats. In a seizure threshold study in
rats, omadacycline had a small anticonvulsive effect at 25 and 50 mg/kg. No effects were seen on
locomotor activity in an Irwin test in rats. A significant increase in latency to pain response was seen
in high-dose animals (50 mg/kg) at 0.5 and 24 hour time points. The range of individual responses
was however wide and overlapped with data in the control and other treated groups, and the
importance of this finding is unclear.

Pharmacodynamic drug interactions

No pharmacodynamic drug interactions were evaluated.

3.2.2. Pharmacokinetics

Methods of analysis

The liquid chromatography with tandem mass spectrometric detection (LC-MS/MS) methods used for
bioanalysis are considered adequately validated, with acceptable selectivity, calibration ranges,
accuracy and precision. Precision and accuracy were within the acceptable limits for the dilution,
indicating that samples above the ULOQ (upper limit of quantitation) can be accurately quantified by
diluting with control plasma to a concentration within the range of the calibration curve.

Absorption

EMA/595311/2019 Page 16/123

In vitro data indicate that omadacycline can be classified as a poor passive permeability compound,
and a substrate for P-gp transport across Caco-2 monolayers, which is likely to limit the overall
omadacycline absorption and, in particular, distribution to tissues with a high P-gp expression.

After single-dose IV administration, plasma exposure was dose-related and plasma elimination linear,
with short terminal elimination disposition half-lives in mice and rats (2.0 to 5.5 hours in mice and 3.5
to 5.7 hours in rats), and longer in monkeys (14 to 21 hours). Correspondingly, higher clearance was
observed in mice and rats (0.9-1.2 L/h/kg) than in monkeys (0.3 L/h/kg). Steady-state volume of
distribution (Vss) was similar across species (3-7 L/kg). After oral administration, Cmax was reached in
0.5-3 hours and 2-4 hours in rats and monkeys respectively.

Oral absorption was low in rats (2.9 to 4.2% in plasma and blood respectively), and an absolute oral
bioavailability was as low as 0.2 % in fasted animals, slightly higher in non-fasted animals. Based on
total recovery of radioactivity, total absorption has been re-estimated to about 0.5%. This low
bioavailability may be due to an interaction with bile salts in the rat intestine (Lin et al, Antimicrob
Agents Chemother 61:e01784-16 https://doi.org/10.1128/AAC.01784-16).

Low plasma exposure levels were also achieved in monkeys following PO administration of
omadacycline as HCl salt and as tosylate salt. There were no differences in PK parameters between the
male and female monkeys, or between salt forms. In humans, bioavailability is around 35 %.

Based on TK data from repeat-dose IV toxicity studies, Cmax and AUC levels increase in a dose-related
manner in rats and monkeys, without sex-related differences in exposure. There were no significant
accumulation with repeat daily dosing up to 13 weeks (accumulation ratios <1.8). After oral
administration, systemic exposures were lower but displayed similar characteristics.

Distribution

Omadacycline (10-10 000 ng/mL) exhibited low binding to plasma proteins in plasma from mouse, rat,
monkey, and human (15-21%), with no major species differences.

Distribution data indicate that omadacycline is rapidly and widely distributed following both IV and PO
administration in pigmented and albino male rats, with levels above blood levels in most tissues,
including lung and skin. At 24h post IV dose, highest levels indicating tissue retention were seen in
bone mineral, Harderian gland and thyroid gland. After PO administration, highest levels were
observed in bone mineral, liver, GI-tract, spleen, Harderian gland and salivary gland. Low levels in the
CNS indicate limited distribution across the BBB. At 24h post-dose, uveal tract levels were not
measurable, indicating lack of retention to melanin. While very high levels were seen in bile 5 minutes
after IV administration, the levels at 24h were not measurable. Distribution to female reproductive
organs (including trans-placental) has not been addressed but is expected.

Metabolism

Omadacycline was metabolically stable following in vitro incubation with rat, dog, monkey and human
liver microsomes and hepatocytes.

In intact rat (study 1000137A), unchanged omadacycline and the C4-epimer were major components
in plasma, faeces and urine. In plasma, omadacycline/ C4-epimer ratio was about 0.6 following IV
dosing. In bile-duct cannulated rats (study R1000163), unchanged omadacycline/C-4 epimer were the
major components in bile (27.3% of dose), urine (38.4% of dose) and faeces (30.2% of dose).

Study PH-34172 is the only study addressing metabolism and excretion in monkeys, using [3H]
omadacycline. Total recovery in excreta is low, partly because radioactivity is related to tritiated water.
This is supported by a long t1/2 for total radioactivity in plasma, and the fact that 36.5% of total
radioactivity was lost as water following freeze-drying of excreta. It is further concluded that the mass

EMA/595311/2019 Page 17/123

balance data have to be interpreted with caution as residual radioactivity in the carcass was not
investigated. Similar metabolic profiles were seen in monkey and rat, with few, minor metabolites in all
three compartments.

Excretion

In intact rats, excretion of radioactivity following IV administration of [14C]omadacycline was complete

within 48 hours, with a total recovery of 109%. Excretion was mainly via faeces (about 80%), and to a
lesser extent via urine (about 30%). In bile-duct cannulated rats, excretion was via bile (24.1%), urine
(29.6%) and faeces (24.5%).

In monkey, total recovery was only 73% within 336 hours following IV administration of
[3H]omadacycline. Due to a substantial loss of radioactivity, the validity of the study is questioned,
and the excretion data should be interpreted with caution. However, an excretion pattern similar to rat
is indicated, with 36.0% of the recovered radioactivity in urine and 30.0% in faeces.

Pharmacokinetic drug interactions

Omadacycline showed very little or no inhibition for CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9,
CYP2C19, CYP2D6, CYP2E1 and CYP3A4/5 in vitro. Results from in vitro preincubation experiments
indicated that omadacycline showed no apparent time-dependent inhibition of CYP1A2, CYP2C9,
CYP2D6 or CYP3A4/5. With respect to the hepatic exposure, it can be concluded that, omadacycline is
unlikely to inhibit the metabolic clearance of potential concomitant medications metabolized by
CYP1A2, CYP2A6, CYP2C9, CYP2D6, CYP2E1 and CYP3A4/5 in humans. There are concerns about
clinical relevance of no inhibition potential for CYP2B6, 2C8 and 2C19 observed in vitro (OC, for more
details, see Clinical AR). With regards to the intestinal exposure, the Applicant is invited to justify that
there is no clinical relevance on intestinal inhibition of CYP3A4 and CYP2C9 by omadacycline (OC).

Omadacycline was found not to be an in vitro inducer of CYP2C9, CYP2C19, CYP3A, or UGT1A1
activities or mRNA. There was a slight mRNA induction response for CYP1A1/2, CYP2B6, and CYP2C8;
however, this did not translate into appreciable induction of CYP activity. In addition, omadacycline
was not an in vitro inducer of CYP1B1 and CYP2J2 and mRNAs. It is not likely that omadacycline would
act as an inducer of major drug metabolizing enzymes in humans.

Omadacycline was not substrate for OATP1B1, OATP1B3, OAT1, OAT3 and OCT in vitro. Omadacycline
is not expected to alter the distribution or elimination of medicinal products that are substrates of
human OAT1, OAT3 and OCT transporter proteins. However, for OATP1B1 and OATP1B3 new in vitro
studies at relevant concentrations are foreseen (OC, for more details, see Clinical AR).

Moreover, the clinical study confirmed that omadacyline is P-gp substrate (For more details, see
Clinical AR).

Omadacycline had no inhibition effect on the P-gp, OATP1B1, OATP1B3, OCT2, OAT3 and BCRP activity
in vitro. Omadacycline weakly inhibited the activity of OAT1 (30% at 25 μM) in vitro. For OAT1B1 and
1B3 and OCT2, no in vivo inhibition is expected. However, it cannot be excluded that omadacycline
may inhibit OAT1, OAT3 BCRP and P-gp in humans, new in vitro studies at relevant concentrations are
foreseen (OC, for more details, see Clinical AR).

Omadacycline is not an in vitro inducer of P-gp and MRP2.

3.2.3. Toxicology

The nonclinical safety program for omadacycline was addressed in a comprehensive battery of
nonclinical toxicology studies evaluating the IV and PO routes of administration. The rat and the
monkey were selected as the main toxicology species and are considered relevant species. All pivotal
toxicity studies were conducted in compliance with Good Laboratory Practice (GLP) regulations and
according to current guidelines.

EMA/595311/2019 Page 18/123

Single and repeat-dose toxicity

Single IV dose toxicity studies with IV administrations were performed in mice and rats and repeat
dose toxicity studies with daily IV and PO administration were performed in rats and monkeys.

Target organs for toxicity were the hematopoietic and GI systems, adrenal glands, thymus, and liver,
with low to non-existing margins of safety. In the rat, effects on the testes and spermatogenesis were
also noted, while in the monkey, there was evidence of effects on the heart. Pigment discoloration
(with or without iron) was noted in many tissues.

Reversible haematology findings in rats and monkeys included decreased red cell mass, haemoglobin,
haematocrit, increased reticulocytes and extramedullary haematopoiesis in the spleen. The findings
indicate increased erythrocyte turnover, without formation of anti-RBC antibodies. The haematology
findings were noted in studies from 14 days of exposure in rats, and in the 13-week studies in
monkeys.

Reversible GI related findings without safety margins were noted in all monkey studies independent of
route of administration, and included emesis, diarrhoea, and faecal changes (soft/mucoid/watery).
Microscopic changes including inflammatory cell infiltration in the mucosa and lamina propria, villus
stunting and fusion in the small intestine, mucosal oedema, lymphoid depletion, were seen in a 14-day
IV study at a high dose with exposure levels 6-7 times the clinically relevant dose. In rats, minor GI-
related findings without histopathological correlates were only observed following oral administration.
Similar findings are well known for other tetracyclines.

Increased adrenal gland weights were observed in both rats and monkeys that generally correlated
with hypertrophy of the zona fasciculate, albeit at higher exposure levels in monkeys. Effects on
thymus and spleen were observed in both rats and monkeys included reduced relative thymus weight,
reduced spleen weight and lymphoid depletion in thymus and spleen. In rats, involution/atrophy of the
thymus was also noted. In monkeys, microscopic findings, lymphoid depletion and hypocellularity of
the bone marrow was observed together with reduced spleen weight at an exposure 7.3 times the
clinically relevant exposure.

Immunotoxicity was not evaluated in an independent study, but additional immunotoxicology testing
was performed in the pivotal 4-week IV monkey study. In this study lymph node/lymphoid depletion
splenic lymphoid depletion and (mild) focal necrosis together with thymus gland lymphoid depletion,
was observed. In the high-dose group (45 mg/kg/day, above MTD), increased total lymphocyte
numbers and T cell subsets (mature T cells, helper T cells, and cytotoxic T cells) in males and a
decrease in natural killer cell activity in females were seen.

Liver-related reversible findings were observed in both rats and monkeys (including increased AST, ALT
and bilirubin), in general without histopathological findings. In a 37-day IV study in rats, however,
clinical chemistry changes were accompanied by histopathological changes (increased liver weight,
Kupffer cell activation, cellular infiltration), indicating inflammation. At exposure levels 7 times human
exposure level at RHD, minimal to slight hepatocellular necrosis was observed.

Histopathological changes in cardiac tissue were observed in IV studies of 4-13 weeks duration in
monkeys. The reversible findings included cardiac myofiber degeneration or myocardial vacuolar
degeneration and occurred at high exposures.

Partly reversible pigmentation was identified in both rats and monkeys, with or without iron. The iron-
containing pigment based on positive Perls’ stain, suggesting the presence of hemosiderin, was
probably a result of increased erythrocyte breakdown. The pigment deposition and discoloration of
some tissues is considered a class effect seen with other tetracyclines.

EMA/595311/2019 Page 19/123

Genotoxicity

Omadacycline was weakly mutagenic in L5178Y cells at concentrations at concentrations extending into
the toxic range and demonstrated a weakly positive clastogenic signal in two in vitro cytogenetics
assays. In the micronucleus tests in mice and rats at IV doses up to 150 mg/kg/day and 80
mg/kg/day, respectively, omadacycline is not considered genotoxic.

Carcinogenicity

No carcinogenicity studies were performed with omadacycline.

Reproductive and developmental toxicity

While no effects were observed on reproductive organs in repeat-dose toxicity studies in monkeys,
testicular atrophy, oligospermia/aspermia and reduced sperm motility were observed in rats. In an IV
fertility study in rats, sperm-related effects as increased weight of seminal vesicles were seen together
with reduced motile sperm, sperm-count and density in the high dose group (20 mg/kg/day). At the
same dose, a reduced number of corpora lutea and implantation sites, increased post-implantation loss
and reduced foetal viability were observed in female rats.

In embryo-foetal toxicity studies in rats and rabbits, omadacycline-related effects on foetal

development included embryo-lethality and/or increased post-implantation loss, whole body oedema,
reduced foetal body weights, delayed ossification, cardiovascular and skeletal defects (rabbits only).
Reduced foetal weights and mean viable foetuses were also observed in rabbits together with
increased post-implantation loss.

Potential effects of omadacycline on pre- and postnatal development have been studied in rats. A slight
reduction in pup weights seen at 30 mg/kg/day were not considered an adverse effect. There were no
omadacycline-related effects on survival, postnatal development, behaviour or reproductive capability
of F1 offspring, and the NOAEL was 30 mg/kg/day (3X the RHD AUC).

Local tolerance

Local tolerance in rabbits was good, indicating low potential for irritation.

Phototoxicity

Omadacycline is classed as “probably phototoxic” based on an in vitro 3T3 NRU phototoxicity test.

3.2.4. Discussion on non-clinical aspects

Pharmacology

Omadacycline inhibited the muscarinic M2 receptor with an IC50 value of 4.25 µM. Based on human
Cmax levels (free fraction) achieved after 100 mg IV omadacycline at steady state, there is only a very
limited margin of safety to inhibition of the muscarinic M2 receptor. Safety pharmacology studies have
demonstrated increased heart rate in vivo and reversal of M2 agonistic effects on the SA node in vitro,
indicating a potential to inhibit vagal regulation of the SA node.

IC50 values between 10 and 30 μM were found for adenosine 3 receptor, alpha 2 adrenergic 2B
receptor, dopamine D3 receptor, histamine H3 receptor, muscarinic M1 receptor, opiate delta receptor,
serotonin transporter, serotonin 5HT3 ion channel, pregnane X receptor (PXR).

The clinical relevance potential targets have not been addressed by the applicant. This is, however, not
considered to be a safety concern due to lack of relevant findings in non-clinical studies, and low
distribution across the blood-brain barrier (BBB).

EMA/595311/2019 Page 20/123

At high concentrations, omadacycline resulted in statistically significant depolarisation of the
membrane potential in Purkinje fibres from rabbits (at 300 µM) and dogs (at 250 µM), and in SA node
from rabbit (1.5 mM). In the Langendorff preparation from guinea pigs, a slight but non-significant
reduction in heart rate was observed, possibly related to a reduced slope phase 4 depolarization of the
sinoatrial (SA) node. The effects on the Langendorff preparation form guinea pig was, however,
minimal, not significantly different from vehicle control, and occurred at concentrations 33 times the
free fraction in human plasma at intended dosing. Although other mechanisms cannot be entirely
excluded, the observed depolarisation in Purkinje fibres and SA nodes occurs at test concentrations
where an inhibitory effect on the Na/K-ATPase is expected (IC50 values 300-500 µM). However, due
to substantial exposure margins relative to expected Cmax in patients, minor effects on heart rate and
no effects on QTc in monkeys, the in vitro findings are considered to be of low clinical relevance.

Pharmacokinetics

A 10 to 20-fold difference between absorption and bioavailability was observed in rats following oral
administration. It is acknowledged that different distribution properties of the active substance and
metabolites could lead to erroneous estimates of total absorption relative to bioavailability. Based on
total recovery of radioactivity, however, total absorption has been re-estimated to about 0.5%, slightly
higher than bioavailability.

Distribution studies were conducted in males only, thus potential distribution to female reproductive
organs has not been evaluated. Omadacycline is widely distributed, and distribution to female
reproductive organs is expected. In addition, reproductive toxicity studies have shown effects on
fertility and embryo-foetal toxicity. Taken together, the lack of distribution data in female animals is
considered acceptable.

In study PH-34172 using [3H]omadacycline, total recovery in excreta is low, partly because 36.5% of
total radioactivity is related to tritiated water. It is possible that tritiated water that was formed in vivo
may have been lost during respiration (expired air was not collected) and via ambient evaporation
from urine and faeces prior to having been collected from the metabolism cages. Additional tritiated
water likely remained in the animals, distributed in body water, beyond the final collection interval
(336 hours), which represents slightly less than 3 half-lives of tritiated water (93.2 hours). However,
there is little concern of sustained retention of drug-related material in the body, based on the full
recovery observed in rats (~ 80% and ~ 30% recovery in faeces and urine, respectively after IV
administration of [14C]-omadacycline; Study R1000137A) and, more importantly, the 95.5% recovery
of 14C-radioactivity in humans (Study CPTK796A2101).

Repeat-dose toxicology

In some organs the pigment deposition correlated with cellular damage e.g., 14-week study in
monkeys with canicular bile stasis. Pigment deposition was dose-related and persistent after
discontinuation of treatment in most of the organs. The Applicant clarified that this statement was
made on the basis of staining for the presence of bile and not on any histological evidence of bile
stasis. Moreover, it was demonstrated that no histologic changes were noted to the bile duct or the bile
ductules within the liver that would indicate bile stasis. The Fouchet's stain on the liver samples
confirmed the accumulation of intracanalicular bile. Canalicular bile stasis is seen as a primary form of
drug toxicity but can also be seen as a secondary event in a variety of pathologic processes, including
gram negative septicemia (Arias et. al., 1994). The latter was considered as particularly prominent
possibility in the present study, given the histologic alterations in the intestine.

Severe cardiac myofiber degeneration or myocardial vacuolar degeneration has been observed in 4-
week and 13-week studies in monkeys with the lowest 3.8 multiple of systemic exposures (AUC) as in
therapeutic use. As no adequate risk mitigation measures can be proposed, this should be reflected in

EMA/595311/2019 Page 21/123

the SmPC to complement toxicology profile of the drug. The SmPC has been updated with additional
data in preclinical section 5.3 on cardiac myofiber degeneration and myocardial vacuolar degeneration
observed in studies with monkeys. However, text should be further revised to include exact safety
margins, method of administration and duration of the other study as current information is too vague.

In general, dose-related exposure levels have been observed in monkeys at IV doses up to 40

mg/kg/day. In study report no. 1116-007, however, a 17-fold increase in AUC, and a 46-fold increase
in Cmax was observed between 5 and 45 mg/kg at study day 1. Considering the atypical TK
parameters in this study, leading to moribundity and deaths at 45 mg/kg/day, dosing errors cannot be
excluded.

Genotoxicity and carcinogenicity

The lack of carcinogenicity studies is considered acceptable, due to lack of genotoxic potential, and the
intended short-term treatment (up to 14 days).

Phototoxicity

Phototoxicity is known as a class effect of tetracyclines, and omadacycline is classified as “probably

phototoxic”. However, the skin was in the group of tissues showing the lowest exposure 2 hours after
omadacycline administration, and very low residual radioactivity within 24 hours after administration.
Omadacycline was not considered to have a significant phototoxicity risk in humans and no
photosensitivity was reported in omadacycline treated subjects. Therefore, no specific warning for
phototoxicity is warranted in the SmPC.

3.2.5. Conclusion on non-clinical aspects

Nuzyra may be granted a marketing authorisation from a non-clinical point of view.

3.3. Clinical aspects

3.3.1. Pharmacokinetics

Across 22 clinical biopharmaceutic and clinical pharmacology studies, 695 subjects were exposed to
omadacycline. The maximal doses of omadacycline investigated were: 600-mg as a single iv and oral
dose; 200-mg once given daily for 7 consecutive days as a multiple iv dose regimen; and 600-mg
every 24 hours (q24h) for 5 consecutive days as multiple oral dosing regimen.

A population PK model for omadacycline was developed using data from 14 Phase 1 studies, two Phase
3 studies conducted in patients with ABSSSI (Studies PTK0796-CSSI-0804 and PTK0796-ABSI-1108)
and one Phase 3 study conducted in patients with CABP (Study PTK0796-CABP-1200).

Bioanalysis:

During the development, omadacycline concentrations were assayed in several biological media,
plasma, urine, dialysate, bronchiolar lavage fluid and cellular pellets. In addition, tigecycline, urea and
verapamil were analysed. The performance of the bioanalytical methods was demonstrated by
validation performed in compliance with principles of GLP and in accordance with requirements of the
FDA Bioanalytical Method Validation Guidance for Industry and Guideline on Bioanalytical Method
Validation (EMEA/CHMP/EWP/192217/2009 Rev1).

For all the clinical studies, inter-day accuracy and precision data demonstrated the reliability of the
assays. Samples were stored at -70°C, and the analytes were shown to be stable under these

EMA/595311/2019 Page 22/123

conditions in the validation studies. Incurred sample reanalysis (ISR) was not performed for
omadacycline in several of the early studies. It was also not performed in some instances where only a
small number of samples were analysed (in accordance with the sample analysis plan). For the
remaining studies, at least 2.5%-10% of the samples were re-assayed as ISR and at least 67% of the
results were within 20% of the original assay result. Overall, all bioanalytical methods used were
robust and reliable.

Absorption:

Following an oral dose of 300 mg the absolute bioavailability is around 34%; a value that probably is
representative of the fraction absorbed from the gastrointestinal tract (based on a low non-renal
clearance). Thus, the oral omadacycline drug product can be tentatively classified as a BCS 3 product,
with overall absorption likely limited by low permeability, not dissolution or solubility.

Food has a large negative effect on the systemic exposure of oral omadacycline, especially when food
is consumed close in time compared to the oral dosing. A high fat meal has a larger negative impact on
the systemic exposure than a low fat meal, and the presence of divalent cations in the food (e.g., dairy
products) seems to reduce the exposure further.

Table 1. Summary of food effect results.

In the proposed SmPC it is stated: “The film-coated tablets are to be taken after fasting for 4 hours
and can be taken with water. After taking the tablet, no food or drink (except water) is to be consumed
for 2 hours and no dairy products, antacids, or multivitamins for 4 hours…”

The possible negative impact of divalent cations on the exposure of omadacycline when taken post-
dose has not been studied.

EMA/595311/2019 Page 23/123

It has not been studied in a dedicated PK study whether the different loading dose options (oral or iv)
lead to similar AUCs. The AUCs achieved by different maintenance dose options, 100 mg IV infusion
QD or an oral 300 mg dose QD, were compared in a cross-over study in healthy subjects.

Table 2. Geometric mean ratio and 90% confidence intervals for PK parameters

Distribution:

Across several studies employing iv omadacycline (50-, 100- or 200-mg), the volume of distribution
estimates were approximately 200-300 L, with preferential compartmentalisation to plasma.
Omadacycline is weakly bound (~20 %) to human plasma proteins in the concentration range of 10-
10000 ng/mL.

Study PTK0796-BAL-15104 was an open-label, parallel group, multiple IV dose study to assess intra-
pulmonary steady-state concentrations in epithelial lining fluid (ELF) and alveolar cells (AC), and
pulmonary distribution time-course, of omadacycline and tigecycline in healthy adult subjects with the
aid of bronchoscopy.

Table 3. Summary of the AUCAC, AUCELF, and AUCPlasma, Ratio of AUCAC/AUCplasma,

and Ratio of AUCELF/AUCplasma by Treatment

Elimination:

The systemic clearance of omadacycline in healthy subjects is approximately 11 L/h and half-life of
omadacycline of 17- 24 hours was observed in clinical studies.

Study CPTK796A2101 was an open-label study to assess the absorption, distribution, metabolism, and
elimination of [14C]-labelled PTK796, and the safety and tolerability of PTK796 and metabolites, in 6
healthy male subjects following a single oral dose of 300 mg PTK796.

EMA/595311/2019 Page 24/123

 Table 4. Excretion of total radioactivity (% of dose) in humans following an oral dose
of 300 mg [14C]PTK796

No metabolites were detected. Several impurities and degradation products were characterized in both
the dosing solution and biological samples (plasma, urine and faeces). In all bio fluids, the C-4 epimer
of PTK796 (known to form upon standing) was observed. PTK796 in plasma reached Cmax between 1
and 4 h.

Renal excretion of omadacycline was evaluated in four clinical pharmacology studies:

Table 5. Summary of Mean Percent of Dose Excreted in the Urine and Renal Clearance
of Omadacycline Across Studies

No discernible omadacycline metabolism was observed in studies of human liver microsomes,

hepatocytes, liver sub-cellular fractions, or recombinant preparations of human drug metabolizing
enzymes.

Dose proportionality for the oral formulation was addressed via the following studies:

EMA/595311/2019 Page 25/123

 Table 6. Summary of Mean Oral Pharmacokinetic Parameters of the Tosylate Salt
Polymorphic Form 3 of Omadacycline Across Studies in Healthy Subjects

A summary of PK parameters following single iv doses of omadacycline ranging from 25- to 600-mg is
summarized below:

Table 7. Summary of Mean (SD) Pharmacokinetic Parameters (PTK0796-SDES-0501)

Omadacycline exhibits proportional PK over the entire dose range (25 to 600 mg) as assessed by area
under the concentration time curve (AUC).

Potential time-dependency in PK and estimates of intra and inter individual variability was not
presented.

EMA/595311/2019 Page 26/123

The multiple dose IV and oral administration with once a day dose setting resulted in similar
accumulation levels of omadacycline with accumulation ratio of 1.5 pointing on week accumulation of
the substance.

Special populations:

Impaired renal impairment was studied in study PTK0796-RENL-15102 which was an open-label,
single-dose, two period, parallel group study. ESRD subjects on stable hemodialysis (n=8) received
single dose omadacycline 100-mg via a 30-min iv infusion followed by an additional iv infused dose of
omadacycline after a washout period of 10 to 20 days, and matched healthy subjects (n=8) received
single dose omadacycline 100-mg iv via a 30-min infusion.

Table 8. Statistical Comparison of Pharmacokinetic Parameters Between ESRD

Subjects on Stable Hemodialysis and Matched Healthy Control Subjects (PTK0796-
RENL-15102)

Ratio of
PK Geometric 90% Confidence
parameter Cohort Geometric Mean Mean (%) Interval
Cohort 1 Period 1
AUClast 9210 103 85.8 - 124.3
(Test)
(ng∙h/mL)
Cohort 2 (Reference) 8910
Cohort 1 Period 1
AUC∞ 10100 105 87.7 - 125.8
(Test)
(ng∙h/mL)
Cohort 2 (Reference) 9610
Cohort 1 Period 1
Cmax 1780 94.3 72.4 - 122.7
(Test)
(ng/mL)
Cohort 2 (Reference) 1880
Cohort 1 Period 1 (Test): ESRD subjects on stable hemodialysis (dosing after dialysis).
Cohort 2 (Reference): Healthy subjects.
Source: PTK0796-RENL-15102 CSR Table 11-6.

Table 9. Statistical Comparison of Pharmacokinetic Parameters Between ESRD

Subjects on Stable Hemodialysis for Period 2 (Test) and Period 1 (Reference)
(PTK0796-RENL-15102)

Ratio of 90%
Geometric Geometric Confidence
PK parameter Cohort Mean Mean (%) Interval
Period 2 (Test) 94.8 -
9090 98.8
AUClast (ng∙h/mL) 102.9
Period 1 (Reference) 9210
Period 2 (Test) 10000 99.5 96.1 – 103
AUC∞ (ng∙h/mL)
Period 1 (Reference) 10100
Period 2 (Test) 98.3 -
2180 123
Cmax (ng/mL) 153.6
Period 1 (Reference) 1780
Cohort 1 Period 1: Omadacycline 100 mg ESRD subjects on stable hemodialysis (dosing after dialysis); Cohort 1 Period 2: Omadacycline 100 mg ESRD
subjects on stable hemodialysis (dosing before dialysis).
Source: PTK0796-RENL-15102 CSR Table 11-7.

18 hepatically impaired subjects and 12 healthy subjects were studied in study CPTK796A2201, an
open-label, fixed-sequence study. During period 1, subjects with mild hepatic impairment (Group 1)
received a single 100-mg iv dose of omadacycline, and subjects with moderate and severe hepatic
impairment (Groups 2 and 3, respectively) received a single 50-mg iv dose of omadacycline. During
period 2, after 7 days’ washout, Groups 1 and 2 received a single oral dose of 300-mg or 150-mg of
omadacycline, respectively, after a 10-hour fast; fasting continued until 4 hours after dose
administration.

EMA/595311/2019 Page 27/123

 Table 10. Dose-Normalized AUClast and Cmax of PTK796 vs. Child-Turcotte-Pugh score,
IV infusion. Healthy Subjects Were Assigned a Score of Zero

Table 11. Geometric Mean Ratio Comparison of and 90% Confidence Intervals for Dose
normalized Primary PK Parameters (Hepatic Impaired Subjects/Healthy Subjects)
(CPTK796A2201)

Group 11 Group 22 Group 33

100-mg iv 300-mg oral 50-mg iv 150-mg oral 50-mg iv
Parameter Omadacycline Omadacycline Omadacycline Omadacycline Omadacycline
AUClast 0.90 0.79 0.85 1.02 1.08
(ng·h/mL) (0.73, 1.11) (0.5, 1.24) (0.75, 0.97) (0.75, 1.4) (0.91, 1.27)
AUC∞ 0.86 0.79 0.88 1.02 1.08
(ng·h/mL) (0.69, 1.07) (0.5, 1.24) (0.78, 0.99) (0.75, 1.4) (0.91, 1.27)
Cmax 1.42 0.96 1.02 1.24 1.08
(ng/mL) (1.1, 1.84) (0.64, 1.42) (0.84, 1.25) (0.94, 1.65) (0.89, 1.31)
1 Group 1: mild hepatic impairment vs. Matched healthy subjects.
2 Group 2: moderate hepatic impairment vs. Matched healthy subjects.
3 Group 3: Severe hepatic impairment vs. healthy subjects matched to group 2, receiving 50-mg iv PTK796.
Source: CPTK796A2201 CSR Table 11.

Race was not found to be a significant covariate in the popPk analysis. With only 15 Asians in the
datasets, even with an improved popPK model, it is difficult to draw conclusions on the similarity of PK
in Asians compared to other groups.

Body Weight was not included in the final popPk model, despite being expected to affect omadacycline
PK. Sex was included as a covariate on clearances and volumes, but may be highly correlated to
weight and it would be appropriate to update the popPk model to include body weight instead (see
popPK section and list of questions).

Age was not found to be a significant covariate in the popPK analysis. Since exposure is not changed in
renal impairment, CrCl or age were not expected to be significant covariates.

Interactions:

Study PTK0796-DDI-17106 was a three-period, open-label, single-sequence study to evaluate the

effect of verapamil ER on the PK of a single oral dose of 300 mg omadacycline in healthy adult

EMA/595311/2019 Page 28/123

subjects. The verapamil vs control geometric mean ratio of AUC∞ and Cmax were 124.58% and
113.55%, respectively (90% CI [108.457, 143.118], and [101.249, 127.361]).

popPK model

The applicant has developed a popPK model for Omadacycline in several steps described in reports
ICPD 00397-1 and ICPD 00397-2.

Data

Data from ten Phase 1 studies (Studies PTK 0796-OBAV-0502, TK 0796-BEQU-0801, PTK 0796-BEQV-
0806 , PTK 0796-BAVA- 0810 , PTK 0796-MDOR-0901 , CPTK796A2103 , CPTK796A2201,
CPTK796A2104 , CPTK796A2101, and PTK0796-FDEF-15101 ) were utilized to develop the original
omadacycline population PK model. Data from these studies were pooled with data from three
additional Phase 1 studies (Studies PTK0796-RENL- 15102 , PTK0796-BAL-15104 and PTK0796-
MDPO-16105), a Phase 1b uncomplicated urinary tract infection (uUTI) patients (Study PTK0796-UUTI-
15103 ), two Phase 3 studies conducted in patients with skin infections (Studies PTK0 796-CSSI-0804
and PTK0796-ABSI-1108 and one Phase 3 study conducted in patients with CABP (Study PTK0796-
CABP-1200). The final model was further evaluated by using data from Study PTK0796-ABSI-16301, a
dataset which was not utilized during model development. PK samples with omadacycline
concentrations below the LLOQ were excluded from the population PK analysis.

Table 12. Summary of the number of subjects and plasma omadacycline concentrations
available and included in the PK analysis

EMA/595311/2019 Page 29/123

 Table 13. Summary statistics of subject demographics, clinical laboratory measures,
and disease-related indices for the overall PK analysis population

Final model

The final population PK model for omadacycline was a linear, three-compartment model with zero
order IV input and first-order absorption using transit compartments to account for a delay in oral
absorption following administration of the tablet or capsule formulations. ELF concentrations were
modelled as a subcompartment of Vp1. Sex was estimated as a covariate for CL, CLd1, VP1 and Vp2.

Sex and body weight are likely to be correlated. Many covariates that are correlated were tested. It
would have been useful if the applicant had plotted the covariates against each other and provided the
R2 value prior to the covariate analysis to limit covariates to be tested (i.e. chose one of body weight
or BSA instead of testing sex, body weight, BSA, BMI and height).

Overall, the final popPK analysis shows model misspecification. The pc-VPC for plasma concentration
show the simulated 50th percentile is lower than the observed data. The simulated lower percentile is
also lower than observed while the upper percentile looks adequate (figure 19). The pc-VPC for ELF
show model misspecification and the lower and upper percentiles of the observed data are not included
(figure 20). The external validation VPCs also indicates model misspecification (figure 21 and 22).
These issues need to be solved for the probability of target attainment simulations to be trustworthy.

EMA/595311/2019 Page 30/123

 Table 14. Population PK parameter estimates for the final population PK model

Figure 20. pc-VPC for ELF data

EMA/595311/2019 Page 31/123

3.3.2. Pharmacodynamics

3.3.2.1. Mechanism of action

The available genetic, biochemical and structural data support that omadacycline is similar in structure
to minocycline and binds to the primary tetracycline binding site of the 30S ribosomal subunit to inhibit
bacterial protein synthesis.

3.3.2.2. Bactericidal vs bacteriostatic activity of omadacycline

Minimal bactericidal concentration and time-kill data suggest that omadacycline is bacteriostatic
against Enterococcus faecalis, Enterococcus faecium, S. aureus and Escherichia coli, but displays slow
bactericidal activity against M. catarrhalis and rapid bactericidal activity against Haemophilus
influenzae and S. pneumoniae. This spectrum of activity is in line with the tetracycline class, which is
generally considered bacteriostatic rather than bactericidal.

3.3.2.3. Post-antibiotic effect

There are sparse, variable in vitro and in vivo data suggesting a possible post-antibiotic effect of 2-3
hours in S. aureus and S. pneumoniae.

EMA/595311/2019 Page 32/123

3.3.2.4. Intracellular activity against Legionella pneumophila

An in vitro study demonstrated activity of omadacycline against intracellular L. pneumophila and S.

aureus at minimal inhibitory extracellular concentration (MIEC) to MIC ratio of ≤¼.

3.3.2.5. Activity in the presence of body fluids or additives

Omadacycline MIC for S. aureus, S. pneumoniae, E. coli and H. influenza is not affected in vitro by the
presence of human serum or bovine surfactant but appears to be negatively affected for selected
organisms (E. coli, K. pneumoniae, S. saprophyticus, S. aureus) in the presence of human urine (8-fold
higher) or high magnesium ion concentrations (16-fold higher). The mechanism for this effect is not
clear but may be an important factor in the food effect (notably milk and cation-containing antacids
and multivitamins) on bioavailability (see section 3.3.1. Pharmacokinetics).

3.3.2.6. Resistance

In vitro testing of global Gram positive and negative SENTRY surveillance isolates (the majority from
2016) expressing one or more tet genes indicates that omadacycline may be able to withstand many of
the classical tetracycline-specific resistance mechanisms of both efflux
(tet(A),(B),(C),(D),(G),(J),(K),(L)) and ribosomal protection (tet(M),(O),(W)). This is supported by a
number of smaller in vitro studies of tetracycline-resistant organisms. However, like other
tetracyclines, omadacycline is inactivated via hydrolysation by tetracycline monooxygenase tet(X).

One of the mechanisms of tigecycline resistance is alteration in the conserved KYKD motif of RpsJ, a
ribosomal structure protein. In an in vitro study (CICbioGUNE_2018 milestone), the mode of
omadacycline binding to the KYKD motif was shown to be different than that of tigecycline.
Additionally, during SENTRY surveillance studies 2009-216, the incidence of these mutations was rare.
It is considered unlikely that these mutations will develop and spread rapidly during clinical use.

Omadacycline activity may be reduced in the presence and/or up-regulation of non-specific efflux
pumps, such as the RamA-controlled RND-type efflux pump AcrAB, commonly found in Gram negative
species, and MexXY, expressed by P. aeruginosa. However, the tetracycline class antibiotics have not
been reported to be a prominent substrate of the AbcA efflux system.

Gly180Val mutation leads to an increase in mexC expression which probably has an impact on the
expression level of MexCD efflux pump. These mutations are considered to be specific for P. aeruginosa
but could lead to omadacycline resistance development in other organisms as a result of nfxB-homolog
alteration. Nonetheless, OMC is not intended for infection caused by P. aeruginosa and other organisms
possessing this mutation are not known or their frequency is rare so far. Thus, this is considered as a
hypothetical risk.

3.3.2.7. Cross resistance

The in vitro activity of omadacycline was not affected in in vitro studies against S. aureus, H.
influenzae, and S. pneumoniae strains with macrolide-, ciprofloxacin- or erythromycin-resistance, and
E. coli expressing beta-lactamases.

Clinical isolates displaying resistance to other antibiotics were selected from the 2016 SENTRY
surveillance collection. The correlation (presented as R2 values) between MIC for omadacycline and
other antibiotics was low for most antibiotics (highest R2 of 0.5865 vs tigecycline, for E. coli and K.
pneumoniae isolates).

EMA/595311/2019 Page 33/123

3.3.2.8. Laboratory selection of resistance

Single-step (4x, 8x and 16x MIC of omadacycline) and serial passage (10 days’ sub-MIC) selection of
resistance was investigated in four strains of S. aureus with (3) and without (1) pre-existing tet(K), (L)
or (M) resistance genes. The resistance frequency was determined to be <1.5x10-8.

In another study, ≤2.9 x 10-9 CFU/mL inocula of 5 H. influenzae strains failed to produce a drug-
resistant (5x MIC) sub-population when exposed to 3x or 5x MIC over 48 hours in the one-
compartment model.

3.3.2.9. Pharmacodynamic interactions with other medicinal products

Overall, no consistent trends of antagonism or synergy with other antibiotics classes were observed
according to fractional inhibitory concentrations (FIC) (checkerboard analysis) using CLSI broth micro-
dilution methodology for E. coli, S. aureus, Enterococcus spp., and S. pneumoniae strains.

3.3.2.10. In vitro spectrum of activity

The in vitro activity of omadacycline was tested against over 100,000 bacterial isolates, collected by
global SENTRY surveillance programs (of which approximately one third came from Europe and one
half from North America) between 2009 and 2016.

Table 15. Multi-year Global Surveillance Omadacycline Distribution, Gram(+)Isolates

MIC50 MIC90
Speciesa Total
(μg/mL) (μg/mL)
Staphylococcus aureus (all) 29280 0.12 0.25
Staphylococcus aureus (methicillin-susceptible) 17339 0.12 0.25
Staphylococcus aureus (methicillin-resistant) 11941 0.12 0.5
2474 0.12 0.5
Staphylococcus aureus (tetracycline-resistant)

Coagulase-negative staphylococci (all)b 4928 0.25 1

Coagulase-negative staphylococci (methicillin-susceptible) 1415 0.12 0.5
Coagulase-negative staphylococci (methicillin-resistant) 3513 0.25 1
Enterococcus faecalis (all) 5711 0.12 0.5
Enterococcus faecalis (vancomycin susceptible) 5548 0.12 0.5
Enterococcus faecalis (vancomycin resistant) 163 0.12 0.5
Enterococcus faecium (all) 3195 0.06 0.25
Enterococcus faecium (vancomycin susceptible) 1593 0.06 0.25
Enterococcus faecium (vancomycin resistant) 1602 0.06 0.25
Streptococcus pneumoniae (all) 9980 0.06 0.12
Streptococcus pneumoniae (penicillin susceptible) 6089 0.06 0.12
Streptococcus pneumoniae (penicillin resistant) 1030 0.06 0.12
Streptococcus pneumoniae multidrug resistant (MDRSP) 3078 0.06 0.12
2873 0.06 0.12
Streptococcus pneumoniae (tetracycline-resistant)

Streptococcus agalactiae (all) 2554 0.06 0.12

Streptococcus agalactiae (macrolide resistant) 1009 0.12 0.12
Streptococcus pyogenes (all) 2561 0.06 0.06
Streptococcus pyogenes (macrolide resistant) 332 0.06 0.12
Streptococcus spp. viridans group (all)c 1387 0.06 0.12
Streptococcus spp. viridans group (penicillin susceptible) 1597 0.06 0.12
Streptococcus spp. viridans group (penicillin resistant) 104 0.06 0.12
Streptococcus anginosus group (all)d 465 0 06 0 12

EMA/595311/2019 Page 34/123

 MIC50 MIC90
Speciesa Total
(μg/mL) (μg/mL)

EUCAST = European Committee on Antimicrobial Susceptibility Testing, MIC = minimum

inhibitory concentration, MIC50 = minimum inhibitory concentration against 50% of the isolates,
MIC90 = minimum inhibitory concentration for at least 90% of the isolates, No. = number.
aEUCAST breakpoints applied for species groupings.

Source: 2.7.2.4. adapted from applicant’s Summary of Clinical Pharmacology Studies – Special Studies- Micro

Table 16. Multi-year Global Surveillance Omadacycline Distribution, Gram(-) Isolates

MIC50 MIC90
Species Total
(μg/mL) (μg/mL)
Enterobacteriaceae (all) 32563 2 >4
Enterobacteriaceae (ESBL-phenotype) 5601 2 >4
Escherichia coli (all) 14091 1 2
Escherichia coli (ESBL-phenotype) 2953 1 4
Klebsiella pneumoniae (all) 6792 2 >4
Klebsiella pneumoniae (ESBL-phenotype) 2214 2 >4
Enterobacter cloacae 2703 2 >4
Haemophilus influenzae 4683 1 2
Moraxella catarrhalisa 408 0.25 0.25
Acinetobacter baumannii 2754 2 8
Stenotrophomonas maltophilia 1023 2 8
Pseudomonas 1986 32 >32
ESBL = extended-spectrum beta-lactamase, MIC = minimum inhibitory concentration, MIC50 = minimum inhibitory
concentration against 50% of the isolates, MIC90 = minimum inhibitory concentration for at least 90% of the isolates.
Source: 2.7.2.4. adapted from applicant’s Summary of Clinical Pharmacology Studies – Special Studies – Micro

Omadacycline has variable activity against Acinetobacter spp., and Stenotrophomonas spp. (≤8
µg/mL), and is not active against Proteus spp., Providencia spp., Morganella spp., and P. aeruginosa.

3.3.2.11. Non-clinical PK/PD

Neutropenic mouse thigh model

In an initial study (Craig 2006), neutropenic mice (2-3 test animals per dose) injected with test strains
were administered fractionated SC omadacycline doses ranging from 0.156 to 40 mg/kg using dosing
intervals of 3, 6, 12 or 24 hours. Similar dose-response relationships were observed with 3-, 6-, and
12-hourly intervals. The 24-hr regimen was least effective. AUC0-24:MIC had the highest R2 and least
variability, particularly at high and low extremes, than Cmax/MIC or Time>MIC, identifying AUC:MIC
as the index best correlated to efficacy (this is true for other tetracyclines):

Figure 1. Relationship for different PK/PD indices in S. pneumoniae (A), S. aureus (B)
and E. coli (C) in the neutropenic mouse thigh model (Craig 2006)

A. Streptococcus pneumoniae B. Staphylococcus aureus

EMA/595311/2019 Page 35/123

 B. Escherichia coli (2)

Source: submitted report Craig 2006 Figures 8-10.

Following this, 12-hourly dosing regimens were tested against strains with varying omadacycline MIC
(3 MSSA, 2 CA-MRSA, 12 S. pneumoniae, 3 E. coli and 1 K. pneumoniae). Across strains, the dose
required for bacterial stasis varied over 200-fold (from 0.368 to 74.4 mg/kg/12h), and the AUC:MIC
for stasis varied 5.5-fold (14.9 to 81.8), with the lowest values obtained for S. pneumoniae and E. coli
strains and the highest for some S. aureus strains and for K. pneumoniae.

In a second study, Andes et al. reported R2 of 0.92, 0.87 and 0.88 for AUC:MIC for S. aureus, S.
pneumoniae and E. coli, respectively, however no comparison to other potential PK/PD indices were
presented for comparison.

EMA/595311/2019 Page 36/123

 Table 17. PD targets in the neutropenic murine thigh model, (Craig and Andes studies)

Craig 2006 Andes Thigh 2017

Stasis Stasis 1 log kill
MIC Static Dose Plasma 24- MIC Static Dose Plasma 24- 1 log kill dose Plasma 24-
Organism
(µg/mL) (mg/kg/12h) Hr (µg/mL) (mg/kg/12h)a Hr (mg/kg/12h)a Hr
AUC/MIC AUC/MIC AUC/MIC
S. aureus ATCC
0.25 11.9 81.8 0.25 5.835 29.64 12.1 58.83
29213
S. aureus ATCC
0.5 11.4 39.3 0.25 10.44 51.13 64 302.51
Smith
S. aureus ATCC
0.5 13.6 46.3 0.5 6.42 16.19 23.81 56.61
33591
S. aureus CA-MRSA
0.25 8.88 62.5 0.5 9.39 23.12 22.035 52.49
MW2
S. aureus CA-MRSA
0.25 5.02 37.8 0.5 8.77 21.68 26.16 62.06
R2527 tetR
S. aureus 6538P 0.25 4.215 22.05 9.97 48.95
S. aureus ATCC
0.25 4.36 22.71 12.7 61.63
25923
S. aureus WIS-1 0.5 5.4 13.80 17.725 42.48
S. aureus LSI 1848 0.5 8.22 20.41 26.5 62.86
S. aureus 307109 0.5 6.56 16.52 13.285 32.17
S. aureus Mean 10.2 ± 3.3 53.5 ± 18.6 6.96 23.73 ± 10.61 22.83 78.06 ± 79.47

S. pneumoniae ATCC 0.900 25.4

0.06 0.06 2.25 53.36 3.765 83.01
10813 1.02 28.8
S. pneumoniae ATCC
0.06 0.799 22.6 0.06 0.495 17.53 1.115 30.40
49619
S. pneumoniae CDC
0.06 1.25 35.3 0.12 2.71 31.20 4.875 52.35
1020 tetR
S. pneumoniae CDC
0.03 0.368 20.8 0.06 1.145 31.00 2.885 65.78
1199 tetR
S. pneumoniae CDC
0.06 1.45 38.7 0.12 2.905 33.13 7.4 77.03
1293 tetR
S. pneumoniae CDC
0.03 0.561 31.7
1329
S. pneumoniae CDC
0.06 0.820 23.2
1396 tetR
S. pneumoniae CDC
0.06 0.798 22.6
146
S. pneumoniae CDC
0.03 0.408 23.1
673
S. pneumoniae CDC
0.06 0.636 19.1
1325 tetR
S. pneumoniae ATCC
0.03 0.540 30.5
6301
S. pneumoniae ATCC
0.06 0.526 14.9
6303
S. pneumoniae Mean 0.759 ± 0.330 23.8 ± 9.6 1.9 33.25 ± 12.85 4.005 61.71 ± 21.05

E. coli ATCC 25922 0.5 7.03 25.3 0.5 41.82 98.84 NA NA

E. coli 1-741 tetR 0.5 5.86 21.6 0.5 32.38 76.53 NA NA

E. coli 1-894 tetR 0.5 6.00 22.0 1 23.88 28.39 37.81 44.68
E. coli 681 1 44.705 52.83 NA NA

K pneumoniae ATCC 71.1 59.4

2.0
43816 77.8 65.0
Enterobacteriaceae
23.3 ± 34.1 32.8 ± 19.7 35.695 64.14 ± 30.35
Mean
a
Calculated from Static Dose (mg/kg/24h) presented in original report.

EMA/595311/2019 Page 37/123

ATCC = American Type Culture Collection, AUC = area under the plasma concentration time curve, CFU = colony-
forming units, MIC = minimum inhibitory concentration.
Source: adapted from applicant’s Summary of Clinical Pharmacology Studies – Special Studies – Microbiology.

EMA/595311/2019 Page 38/123

Mouse neutropenic lung model

In an initial study (report Nicolau C370), neutropenic mice were inoculated nasally with test strain (1
penR mefA+ S. pneumoniae #100 and 1 penS SPN #22; 6 test animals per strain), before being
administered fractionated SC omadacycline doses of between 0.25 and 25 mg/kg (dosing intervals 6,
12 or 24 hours), although the tested regimens varied (inexplicably) for each strain, resulting in little to
no fractionation practice amongst the assays for SPN #22.

The report states that a high correlation between the plasma free drug AUC/MIC and efficacy was
noted for both strains, however the reported R2=0.904 and 0.699, respectively, suggests this
relationship was somewhat less robust for SPN #100:

Figure 2. Relationship between plasma free drug AUC/MIC and antimicrobial activity of
omadacycline against S. pneumoniae #100 in the neutropenic mouse lung model
(Nicolau C370)

Source: submitted report Nicolau C370, Figure 3.

Lepak et al. (2017) also reported dose-response data for 4-fold increasing SC doses ranging from 0.1
to 25.6 mg/kg/12h omadacycline against 4 S. pneumoniae test strains in the neutropenic mouse lung
model. The relationships for plasma AUC/MIC and ELF AUC/MIC were moderately robust (R2=0.74 and
0.75, respectively). The unusually high values obtained for strain 1293 are not explained.

EMA/595311/2019 Page 39/123

 Table 18. 24-h static and 1-log and 2-log kill doses and associated plasma free drug
AUC/MIC values for S. pneumoniae (4 strains) in the neutropenic mouse lung model
(Lepak 2017)

Source: Lepak et al. (2017) Table 2.

One-compartment in vitro model

The total-drug ELF AUC0-24:MIC ratio associated with efficacy against a panel of 5 clinical H.
influenzae isolates (omadacycline MIC 1 to 2 µg/mL) was determined in a dose-ranging study in a one-
compartment in vitro infection model delivering 24-hour omadacycline concentration-time profiles
representing those observed in human epithelial lining fluid (ELF) after IV administration of q12h doses
from 12.5 to 400 mg. The target selected to be used in PTA analysis was the median ELF AUC0-24:MIC
ratio corresponding to 1-log kill AUC/MIC = 8.91.

Figure 3. Relationship between change in log10 CFU/ml from baseline at 24 hours and
omadacycline total-drug ELF AUC:MIC ratio for H. influenzae (5 strains) in the one-
compartment in vitro model (ICPD-00396)

Source: applicant’s submitted report ICPD-00396.

3.3.2.12. Probability of target attainment analyses

The applicant performed Monte Carlo Simulation to assess the probability of target attainment with
four different dosing regimens. The popPK model is discussed in section 3.3.1. Pharmacokinetics.

Univariable analysis

Clinical PK/PD relationships for efficacy were explored for both CABP and ABSSSI using data from
Phase 3 patients in the Microbiologically-Evaluable (ME) population who had PK data available, for each
indication (182 subjects and 10 subjects for ABSSSI and CABP, respectively).

The only significant univariable relationship identified was between AUC/MIC and ECR (p=0.011) in the
subset of ABSSSI subjects with S. aureus at baseline. Model-predicted percent probability of clinical

EMA/595311/2019 Page 40/123

success at ECR in ABSSSI ranged from 87.2 to 95.6% for S. aureus up to MIC 0.5 µg/mL, when clinical
response at ECR was modelled as a continuous, two-group (free AUC:MIC threshold 13.2) or three-
group (free AUC:MIC thresholds 14.5 and 70.9) variable.

PTA analysis

For ABSSSI, non-clinical PK-PD targets were initially selected based on the median plasma free drug
AUC/MIC ratio targets associated with stasis for S. aureus, S. pneumoniae (use to inform targets for
Streptococcus spp.), and E. coli, in the neutropenic mouse thigh model (Andes Thigh 2017). For CABP,
PK-PD targets were initially selected based on the median free-drug ELF AUC/MIC ratio targets
associated with a 1- log10 CFU reduction from baseline for S. pneumoniae in the neutropenic mouse
lung model (Lepak 2017), excluding isolate 1293, and for H. influenzae in a 24-hour one-compartment
in vitro infection model.

Table 19. Omadacycline free-drug plasma and ELF AUC/MIC targets selected by the
Applicant for PTA analysis

Median Static Median 1 log kill

Organism
Plasma AUC/MIC ELF AUC/MIC
ABSSSI
S. aureus (10) 21.9
S. pneumoniae (5) 31.2
E. coli (4) 64.7
CABP
S. pneumoniae (3) 13.3
H. influenzae (5) 8.91
ELF = epithelial lining fluid. Source data: submitted report ICPD-00398-1.

The applicant performed target attainment analyses for both ABSSSI and CABP using AUC/MIC ratio
targets randomly assigned based on an estimated truncated log normal distribution of AUC/MIC ratio
targets for each pathogen. The observed mean and standard deviation among the non-clinical PK-PD
targets were used as the parameters for the log normal distributions. Such an approach, while
appropriate for some types of analysis (e.g. popPK), is less desirable from the regulatory standpoint
for PTA analyses, where the primary concern is adequate coverage of the least susceptible target
organisms when they are eventually encountered in clinical practice. At the request of CHMP, PTA
analysis was also run using the median, highest and second highest (excluding high outlier for S.
pneumoniae) targets observed in non-clinical models for the key organisms for the CAP indication (S.
pneumoniae and H. influenzae).

The breakpoints used to interpret the PTA analysis results were MIC90 values from European SENTRY
surveillance data. The MIC90 values from the presented European surveillance data for S. aureus and
S. pneumoniae reflect the breakpoints agreed so far with EUCAST (further discussion re: H. influenzae
is to follow).

ABSSSI

S. aureus

Both iv-to-po and po dosing regimens with a loading dose (either 100 mg q12h or 200 mg q24h on
Day 1 for iv loading) provided >90% PTA on Days 1 to 2 and switch day, using randomly-assigned
free-drug plasma AUC:MIC targets, only up to MIC 0.12 µg/mL. At the European surveillance 2016
MIC90 value/ EUCAST breakpoint for S. aureus of 0.25 µg/mL, PTA was as low as 60.8% for the po
dosing regimen. When assessed against the European surveillance 2016 distribution for S. aureus
there was a >90% PTA for iv-to-po (100 mg iv q12h) and 86.9% probability of PTA for po dosing
regimen.

EMA/595311/2019 Page 41/123

Streptococcus species

Similarly, the probability of achieving stasis for S. pneumoniae was >90% at the European surveillance
2016 MIC90/ EUCAST breakpoint value for S. pneumoniae (proxy) 0.12 µg/mL, but only for the for the
“IV-to-PO” dosing regimens with loading dose.

The Applicant has not conducted re-analyses for S. aureus or Streptococcus species using median and
highest target values. Given the sufficient evidence for good clinical outcomes in ABSSSI, further
discussion of the PTA support for this indication is not pursued.

CAP

S. pneumoniae

Both iv-to-po and po dosing regimens with a loading dose (either 100 mg q12h or 200 mg q24h on
Day 1 for iv loading) provided >90% PTA on Days 1 to 2 and switch day, at the European surveillance
2016 MIC90 value/ EUCAST breakpoint for S. pneumoniae of 0.12 µg/mL, using both randomly-
assigned total-drug ELF AUC:MIC targets and randomly-assigned free-drug plasma AUC:MIC targets.
When assessed against the European surveillance 2016 distribution for S. pneumoniae there was a
>99% PTA for all dosing regimens for both total-drug ELF and free-drug plasma AUC:MIC ratios.

As expected, PTA using a single conservative target (highest observed) was lower than PTA using
randomly assigned or median targets for S. pneumoniae and does not cover the European surveillance
2016 MIC90/ EUCAST breakpoint in either ELF or plasma. As noted by the Applicant, these predictions
do not agree with the clinical efficacy data from study CABP-1200, in which higher MIC values were
successfully treated. Furthermore, PTA using the second highest observed (excluding S. pneumoniae
1293 as an outlier) achieved >90% coverage up to the same MIC value as median target or a
randomly assigned target, that is to say the three approaches produced the same highest MIC value
with >90% coverage.

H. influenzae

Iv-to-po regimens with a loading dose provided >90% PTA on Days 1 to 2 and switch day, at the
European surveillance 2016 MIC90 value for H. influenzae of 1 µg/mL, using randomly-assigned total-
drug ELF AUC:MIC targets. Meanwhile, PTA for the po regimen at the same MIC was 82.7%. PTA was
notably lower for randomly-assigned free-drug plasma AUC:MIC targets. At the European surveillance
2016 MIC90 value of 1 µg/mL, PTA for plasma targets was as low as 35.0% for the po dosing regimen.
When assessed against the European surveillance 2016 distribution for H. influenzae there was a
>89.9% PTA for iv-to-po (100 mg iv q12h) and 86.1% probability of PTA for po dosing regimen, but
only using total-drug ELF.

PTA using a single conservative target (highest observed) for H. influenzae achieved >90% coverage
up to the same MIC value as median target or a randomly assigned target. Adequate coverage is
achieved at a higher MIC (covering the European surveillance 2016 MIC90) for ELF targets than plasma
targets (not reaching the European surveillance 2016 MIC90). Notably, there is better agreement
between clinical efficacy data for H. influenzae and the PTA results for ELF targets than plasma targets,
which is interesting.

There appears to be a disconnect (in favour of clinical outcomes) between the non-clinical predictions
and clinical outcomes observed in study CABP-1200. The Applicant asserts that this indicates that the
highest plasma PK-PD target is not a good predictor of clinical outcome for in the CABP-1200 study.
However, this lack of agreement weakens rather than strengthens the available evidence for efficacy in
CAP, which comes from a single pivotal study. Given the conclusion that a positive B/R balance cannot
be concluded for omadacycline in CAP without additional data from a further study, additional

EMA/595311/2019 Page 42/123

discussion regarding the PTA support for this indication does not need to be further pursued at this
time.

Table 20. Percent probabilities of PK-PD target attainment by MIC based on randomly
assigned plasma or ELF AUC/MIC ratio targets associated with stasis for S. aureus
and Streptococcus species, among simulated patients after administration of IV to
PO or PO omadacycline dosing regimens

S. aureus (ABSSSI) – net stasis

Streptococcus spp. (ABSSSI) – net stasis

a Target value randomly assigned based on an estimated log normal distribution of PK-PD targets associated with
the same endpoint in the same species.
b. Shaded cells indicate PK-PD target attainment values ≥90%.

EMA/595311/2019 Page 43/123

 Table 21. Clinical Response (CABP-1200 study) and Percent Probabilities of PK-PD Target Attainment by MIC for Total-drug ELF or Free-
drug Plasma AUC:MIC Ratio Targets Associated with a 1-log10 CFU Reduction for S. pneumoniae, Simulated Patients, Omadacycline
100 mg iv q12h on Day 1 Followed by 100 mg iv q24h on Day 2 with a po Switch to 300 mg po q24h on Day 3

Percent probability of PK-PD target attainment by MIC on Days 1 to 2 among simulated patientsb,c
Assessment of free-drug plasma exposures and AUC:MIC ratio
Assessment of total-drug ELF exposures and AUC:MIC ratio targets targets
Randomly assigned Median of all Highest Second highest Randomly assigned Median of all Highest PK- Second highest
Percentage of based on all PK-PD PK-PD targets, PK-PD target, PK-PD target, based on all PK-PD PK-PD targets, PD target, PK-PD target,
MIC clinical success at targets, excluding excluding the including the including the targets, excluding excluding the including the including the
(mg/L) PTE (n/N), N=28a the outlierd outliere outlierf outlierg the outlierd outliere outlierf outlierg
0.015 100 (2/2) 100 100 100 100 100 100 100 100
0.03 85.7 (12/14) 100 100 100 100 100 100 99.5 100
0.06 100 (10/10) 100 100 98.5 100 100 100 26.0 100
0.12* 50 (1/2) 100 100 15.0 100 100 100 0 100
0.25 NA 100 100 0 100 97.5 100 0 99.9
0.5 NA 99.3 100 0 100 75.4 85.2 0 42.2
1 NA 82.9 96.3 0 67.0 34.3 0.96 0 0.02

a. Based on data from patients with S. pneumoniae at baseline in the microITT population of Study PTK0796-CABP-1200.
b. Assessed using total-drug ELF or a free-drug plasma AUC:MIC ratio target associated with a 1-log10 CFU reduction from baseline for S. pneumoniae based on data from a
neutropenic murine lung-infection model.
c. Based on the assessment of average total-drug ELF or free-drug plasma AUC0-24 on Days 1 and 2.
d. Based on data for all S. pneumoniae isolates excluding the outlier (S. pneumoniae 1293), the total-drug ELF and free-drug plasma AUC:MIC ratio targets associated with a 1-
log10 CFU reduction from baseline were randomly assigned based on an estimated log normal distributions of AUC:MIC ratio targets associated with the same endpoint.
e. Based on data for all S. pneumoniae isolates excluding the outlier (S. pneumoniae 1293), the median total-drug ELF and free-drug plasma AUC:MIC ratio targets associated
with a 1-log10 CFU reduction from baseline were 13.3 and 15.2, respectively.
f. Based on data for all four S. pneumoniae isolates studied, the highest total-drug ELF and free-drug plasma AUC:MIC ratio targets associated with a 1-log10 CFU reduction
from baseline were 200.6 and 180.0, respectively.
g. Based on data for all four S. pneumoniae isolates studied, the second highest total-drug ELF and free-drug plasma AUC:MIC ratio targets associated with a 1-log10 CFU
reduction from baseline were 17.6 and 19.7, respectively.
*European surveillance 2016 MIC90.

EMA/595311/2019 Page 44/123

 Table 22. Clinical Response (CABP-1200 study) and Percent Probabilities of PK-PD Target Attainment by MIC for Total-drug ELF or Free-
drug Plasma AUC:MIC Ratio Targets Associated with a 1-log10 CFU Reduction for H. influenzae, Simulated Patients, Omadacycline 100
mg iv q12h on Day 1 Followed by 100 mg iv q24h on Day 2 with a po Switch to 300 mg po q24h on Day 3

Percent probability of PK-PD target attainment among simulated patientsb,c

Assessment of total-drug ELF exposures Assessment of free-drug plasma exposures
by AUC:MIC ratio targets by AUC:MIC ratio targets
Median Median
Randomly assigned of all Highest Randomly assigned of all Highest
MIC Percentage of Clinical based on PK-PD PK-PD Second highest based on PK-PD PK-PD Second highest
(mg/L) Success at PTE (n/N), N=32a all PK-PD targetsd targetse targetf PK-PD targetg all PK-PD targetsd targetse targetf PK-PD targetg
0.25 NA 100 100 100 100 100 100 100 100

0.5 100 (1/1) 100 100 100 100 99.5 100 99.0 99.9

1* 88.9 (16/18) 99.5 100 99.2 99.9 67.0 61.2 16.2 44.2

2** 66.7 (8/12) 68.8 65.1 19.0 48.6 7.36 0.14 0 0.02

4 100 (1/1) 8.42 0.22 0 0.04 0 0 0 0

a.Based on data from patients with H. influenzae at baseline in the microITT population of Study PTK0796-CABP-1200.
b. Assessed using total-drug ELF or a free-drug plasma AUC:MIC ratio target associated with a 1-log10 CFU reduction from baseline for H. influenzae based on an one-
compartment in vitro infection model were randomly assigned based on an estimated log normal distribution of AUC:MIC ratio targets associated with the same endpoint.
c.Based on the assessment of average total-drug ELF or free-drug plasma AUC0-24 on Days 1 and 2.
d. Based on data for all five H. influenzae isolates, the total-drug ELF or free-drug plasma AUC:MIC ratio targets associated with a 1-log10 CFU reduction from baseline
were randomly assigned based on an estimated log normal distributions of AUC:MIC ratio targets associated with the same endpoint.
e.Based on data for all five H. influenzae isolates, the median total-drug ELF or free-drug plasma AUC:MIC ratio target associated with a 1-log10 CFU reduction from baseline
was 8.91.
f. Based on data for all five H. influenzae isolates studied, the highest total-drug ELF or free-drug plasma AUC:MIC ratio targets associated with a 1-log10 CFU reduction from
baseline was 11.6.
g. Based on data for all five H. influenzae isolates studied, the second highest total-drug ELF or free-drug plasma AUC:MIC ratio targets associated with a 1-log10 CFU
reduction from baseline was 9.73.
*European surveillance MIC90 using 2016 data, **European surveillance MIC90 using 2009-2016 data, see Q193.

EMA/595311/2019 Page 45/123

Susceptibility breakpoints

EUCAST have suggested susceptible breakpoints based on ECOFF and PTA data for S. aureus (≤0.25
mg/L), S. lugdunensis (≤0.25 mg/L), S. pneumoniae (≤0.125 mg/L), S. agalactiae (≤0.25 mg/L), S.
pyogenes (≤0.25 mg/L), S. anginosus group (≤0.125 mg/L) and H. influenzae (≤2 mg/L).

The committee did not agree there was sufficient evidence to set breakpoints for additional species
(Enterococcus spp., Enterobacterales and anaerobes).

3.3.3. Discussion on clinical pharmacology

ADME

Biopharmaceutical data indicate a rather pronounced food effect, which has been adequately
characterised.

Omadacycline has been shown to be metabolically stable both in vivo and in vitro. 25-45% of
omadacycline’s total clearance is due to renal excretion mechanisms. Urinary clearance (CLr) following
iv administration was estimated in several studies at 3 L/h. Renal filtration is estimated to be
approximately 6 L/h. Consequently, renal reabsorption via active transport seems a possibility. Biliary
excretion of unchanged drug constitutes the remainder, ca 55-75% of total clearance. Across iv
studies, total clearance was around 10 L/h, and terminal half-life 15-25 h. Daily dosing results in
accumulation of approximately 50% over the initial dose exposure.

Special populations

One study was performed to evaluate the effect of renal function on omadacycline exposure. The
results did not indicate any impact on omadacycline AUC in the ESRD group.

Regarding the effect of decreased hepatic function, a dedicated hepatic impairment study including
both oral and iv omadacycline did not show a difference in AUC across the different cohorts; healthy
subjects, Child Pugh classes A, B and C.

Interactions

Effect of other medicines on omadacycline:

The risk for clinically relevant PK interactions is deemed low based on the available data. Some
uncertainty still remains though for a possible scenario where omadacycline turns out to be a substrate
for a transport protein involved in its biliary secretion. Some transport protein substrate assays
(OTP1B1 and 1B3) are requested to be repeated as several concentrations of the possible substrate is
recommended to be tested. In the performed studies, only one concentration of omadacycline was
used. PAM

Effect omadacycline on other medicines:

Omadacycline is deemed to have low potential as perpetrator in PK interactions. No competitive CYP

inhibition is expected in vivo based on the in vitro assay. The same conclusion may be drawn for CYP
induction. TDI data for all CYPs has not been presented and is requested. LoQ For the transport
protein perpetrator assays, some uncertainty remain regarding if the results should be seen as valid as
the studied concentration range may have been too low in comparison with regulatory cut-offs and
new studies are requested. PAM

MoA

EMA/595311/2019 Page 46/123

The mechanism of action and pharmacological properties of omadacycline have been generally well
characterised through a series of in vitro studies. The data presented describing laboratory selection of
resistance is sparse. Resistance was not detected in any isolate in the Phase 3 studies, in any of the
Phase 3 trials, although the Applicant acknowledges that collection of post-baseline isolates, and
therefore the opportunity to detect on-treatment resistance emergence, was limited.

PK/PD index

It appears that AUC/MIC is the most relevant index, as for other tetracyclines.

PD target

The data underpinning selection of the target are sparse and highly variable. The lack of positive
control means there is no clear bridge or comparison possible between results of the different models.
At the request of CHMP, the Applicant has presented PTA analyses using randomly assigned, median,
highest and second highest (excluding high outlier) targets for key pathogens for CABP. The same re-
analysis have not been produced for ABSSSI.

PTA analysis

The Applicant’s emphasis on the clinical PDT (ECR), given that this clinical PK/PD relationship is based
upon a very small number of clinical failures in the clinical trials, the role of other factors (such as
adjunctive surgery) is unclear, and the relationship does not relate to the primary efficacy endpoint
required by EMA.

The applicant also emphasises the important of ELF over plasma free drug AUC/MIC targets in CABP,
however, the values for ELF penetration in both non-clinical studies and healthy volunteer human PK
studies show high inter-individual variability that may, in part, be an artefact of the sampling
approach.

A significant proportion of CABP patients had infections caused by atypical infections, coverage of
which cannot be addressed by PTA analysis due to the lack of established non-clinical models for
characterising PK-PD relationships in these species. Evidence for efficacy in atypical pathogens is
addressed in section 3.3.5. Clinical efficacy.

Notwithstanding the above, the presented PTA results are not favourable for the highest MIC values of
interest across all key pathogens and cannot, therefore, support efficacy of the dosing regimen used in
the Phase 3 pivotal studies, nor inclusion of all the species proposed by the Applicant for the list of
organisms in SmPC 5.1 expected to be susceptible on the basis of in vitro data. There appears to be a
disconnect (in favour of clinical outcomes) between the non-clinical predictions and clinical outcomes
observed in study CABP-1200, in which organisms with higher MIC values than those predicted by PTA
analysis were successfully treated. The Applicant considers that nonclinical PK-PD PTA data may not be
as predictive for certain classes of antibiotics (tetracyclines for instance).

3.3.4. Conclusions on clinical pharmacology

Regarding PK, the applicant has agreed to perform three studies post approval, which is deemed to be
an acceptable approach. The studies are:

1) in vitro transport substrate studies for the transporters OATP1B1 and OATP1B3

2) in vitro CYP3A4 and CYP2C9 inhibition assessment at the theoretical maximal gut concentration

3) in vitro inhibition evaluation of the transporters OAT1, OAT3, P-gp and BCRP

EMA/595311/2019 Page 47/123

There is also one remaining question in the LoQ where The applicant is requested to fulfil the guideline
requirements and provide in vitro data on TDI for CYP2B6, 2C8 and 2C19.

Otherwise, all PK related issues are resolved.

Overall, given the sufficient evidence from two adequately-sized clinical studies for good clinical
outcomes in ABSSSI at suitable MIC values, further discussion of the lack of comprehensive PTA
support for this indication is not pursued. Furthermore, given that a positive BR balance cannot be
concluded for omadacycline in CAP without additional data from a further study, additional discussion
regarding the PTA support for this indication does not need to be further pursued at this time.

EMA/595311/2019 Page 48/123

3.3.5. Clinical efficacy

Table 23. Overview of Pivotal and Supportive Efficacy Studies

M/F
N study Study Primary Study N Diagnosis Primary
Study ID Design N per am arm Duration Mean
centres Posology Objective total Incl. criteria Endpoint
Age

Studies in cSSSI
Omadacycline
100mg iv q24h Safety and cSSSI (major abscess,
Up to 7d IACR at
PTK0796- Phase 2, R,  200mg po tolerability vs Omadacycline=111 infected lower extremity
IV, up to 56%M TOC visit in
CSSI- 11 (US) SB, active q24h; linezolid in 219 ulcer, cellulitis, infection
14d total 45y mITT and
0702 comparator Linezolid 600mg adults with Linezolid=108 associated with trauma or
(iv+po) CE-TOC
iv q12h  cSSSI removable foreign body)
600mg po q12h
Omadacycline cSSSI (major abscess,
100mg iv q24h cellulitis, infection
Non-inferiority Up to 7d IACR at
PTK0796- Phase 3, R,  300mg po Omadacycline=68 associated with trauma or
to linezolid in IV, up to 66%M EOT and
CSSI- 6 (US) SB, active q24h; 143 removable foreign body) +
adults with 14d total 39y TOC visit in
0804a comparator Linezolid 600mg Linezolid=72 evidence of SIR,
cSSSI (iv+po) ITT and CE
iv q12h  immunosuppression or
600mg po q12h specified comorbidity
Studies in ABSSSI
55 Omadacycline
Overall
(N. 100mg iv q12h
Non-inferiority Up to 7d ABSSSI ≥75cm2 (wound clinical
PTK0796- America, E. Phase 3, R, x2 iv q24h  Omadacycline=329
to linezolid in IV, up to 65%M infection, response at
ABSI- Europe, W. DB, active 300mg po q24h; 645
adults with 14d total 47y cellulitis/erysipelas, major PTE visit in
1108 Europe, comparator Linezolid 600mg Linezolid=326
ABSSSI (iv+po) abscess) + evidence of SIR mITT and
Latin iv q12h 
CE
America ) 600mg po q12h
Omadacycline Overall
450mg po q24h Non-inferiority ABSSSI ≥75cm2 (wound clinical
PTK0796- Phase 3, R, Omadacycline=368
x2 then 300mg to linezolid in 63%M infection, response at
ABSI- 33 (US) DB, active 735 7d 14d po
po q24h; adults with 44y cellulitis/erysipelas, major PTE visit in
16301 comparator Linezolid=367
Linezolid 600mg ABSSSI abscess) + evidence of SIR mITT and
po q12h CE
Studies in CABP
86 Omadacycline
Non-inferiority Overall
(N. 100mg iv q12h Up to 7d
PTK0796- Phase 3, R, to moxifloxacin Omadacycline=329 Radiographically confirmed clinical
America, x2 then q24h  IV, up to 56%M
CABP- DB, active in adults with 660 CABP requiring iv Abx, response at
W. Europe, 300mg po q24h; 14d total 62y
1200 comparator CABP Port Risk Moxifloxacin=331 PORT Risk Class II-IV PTE visit in
E. Europe, Moxifloxacin (iv+po)
Class III-IV ITT and CE
Rest of 400mg iv q24h

EMA/595311/2019 Page 49/123

 world)  300mg po
q24h
ABSSSI=acute bacterial skin and skin structure infection, CABP=community-acquired bacterial pneumonia, cSSSI=complicated skin and skin structure infections,
IACR=investigator’s assessment of clinical response, R=randomised, SB=single-blind, TOC=test of cure, PTE=post therapy evaluation, SIR=systemic inflammatory response,
mITT=modified intent to treat, CE=clinically evaluable,
a
Terminated early by sponsor. Table compiled by assessor. Source data: applicant’s submitted Summaries of Clinical Efficacy Studies in ABSSSI and CABP.

EMA/595311/2019 Page 50/123

3.3.5.1. Dose response studies

No dose response studies were conducted for the proposed indications. Dose selection for the Phase 3
programme was based on in vitro and in vivo non-clinical PK/PD studies and Monte Carlo simulations of
pharmacodynamic target attainment (see section 3.3.2. Pharmacodynamics).

The proposed iv loading dose of 200 mg q24h was not tested in PTA analysis or in clinical studies of
ABSSSI or CABP, but only in patients with uncomplicated UTI in study PTK0796-UUTI-15103, thus
there is no substantial safety database for it in the target populations.

In Study PTK0796-UUTI-15103, a median increase in heart rate of 20 bpm was seen after
administration of an iv loading dose of 200-mg over 30-45 minutes. A median increase of 20 bpm is
considered potentially clinically significant for the target patient populations, who are generally more
clinically unwell than patients with uUTI and may not tolerate this haemodynamic change well. No
additional details are provided to enable further characterisation of any potential safety concern e.g.
timing and speed of onset of the effect, time of resolution/ duration of the effect, related changes (if
any) in blood pressure, TEAEs reported in association.

The Applicant asserts, on the basis of modelling, that infusing the 200 mg dose over a 60-90 minutes
will reduce median heart rate elevations to below 20 beat per minute. However, there are no clinical
data to support this hypothesis. Support for the safety of the alternative iv loading dose of 200 mg in
the target populations is insufficient to preclude its recommendation in section 4.2. In addition, there is
no clearly established benefit demonstrated in terms of clinical efficacy. The alternative iv loading dose
is therefore not considered acceptable for inclusion in the product information.

3.3.5.2. Main studies

3.3.5.2.1. Study PTK0796-ABSI-1108 - A Phase 3 Randomised, Double-Blind, Multi-Centre

Study to Compare the Safety and Efficacy of Omadacycline IV/PO to Linezolid (Zyvox®)
IV/PO for Treating Adult Subjects with Acute Bacterial Skin and Skin Structure Infection
(ABSSSI)

Methods (Study PTK0796-ABSI-1108)

• Study participants

Subjects 18 to 80 years of age from 55 global centres, with qualifying ABSSSI ≥75 cm2 continuous
surface area (SA), expected to require at least 3 days’ IV treatment:

o Wound infection: an infection characterized by purulent drainage from a wound with

surrounding erythema, edema, and/or induration extending at least 5 cm in the shortest
distance from the peripheral margin of the wound.

o Cellulitis/erysipelas: a diffuse skin infection characterized by spreading areas of erythema,

edema, and/or induration.

o Major abscess: an infection characterized by a collection of pus within the dermis or deeper
with surrounding erythema, edema, and/or induration extending at least 5 cm in the shortest
distance from the peripheral margin of the abscess.

And one of the following, within 24h of randomisation:

• Elevated white blood cell (WBC) count (≥ 10,000 cells/mm3) or leukopenia (≤ 4,000
cells/mm3);

EMA/595311/2019 Page 51/123

 • Elevated immature neutrophils (≥ 15% band forms) regardless of total peripheral WBC count;

• Lymphatic involvement: lymphangitis or lymphadenopathy that was proximal to and in a

location that suggested drainage from the qualifying infection; or

• Fever or hypothermia documented by the investigator (temperature > 38.0°C [100.4°F] or less
than 36.0°C [95.5°F]).

Subjects were stratified at entry by ABSSSI type and geographic area, and major abscess was limited
to 30% of subjects.

Principal exclusion criteria:

(a) having received 1 or more dose of potentially effective systemic antibiotic in the 72 h prior to
the first dose of test article (unless infecting organism shown to be intermediate or resistant to
initial treatment),

(b) having applied a topical antibacterial agent(s) continuously for ≥72 h immediately prior to
randomisation,

(c) infections were there outcome was strongly influenced by other factors, where treatment >14
days was anticipated, when chronic skin lesions impaired determination of response, where the
pathogen was suspected or known to be resistant to test article (e.g. chronic >3 months,
underlying skin condition, burns, life-threatening infections (necrotising fasciitis, gangrene),
underlying vascular insufficiency requiring immediate revascularisation or curative amputation,
bone, joint or other organ involvement, underlying immune deficiency, human or animal bites).

(d) Pre-existing hepatic, renal, immunological or cardiac disease as defined in the protocol

(e) Haemodynamically unstable despite fluid resuscitation, or requiring pressor agents

(f) Receipt of MOA inhibitor within 14 days prior to screening

• Treatments

Subjects were randomised (1:1) to one of the following two treatments, for up to 7 days IV and up to
14 days total IV+PO:

o Omadacycline, 100 mg iv q12h for 2 doses, then 100 mg iv q24h with the option to switch to
300 mg po q24h after a minimum of 3 days.

o Linezolid, 600 mg iv q12h with the option to switch to 600 mg po q12h after a minimum of 3
days.

Subjects could receive adjunctive Gram-negative cover at the discretion of the investigator if a Gram-
negative pathogen, anaerobe or linezolid-resistant pathogen was detected at baseline.

• Objectives

Primary objective: to demonstrate that omadacycline administered iv and po was non-inferior to

linezolid administered iv and po in the treatment of adults with ABSSSI known or suspected to be due
to Gram-positive pathogens.

Secondary objectives:

- to evaluate the safety of omadacycline in adults subjects with ABSSSI.

- to evaluate the clinical response according to the causative pathogen.

- to characterise the PK of omadacycline in adults subjects with ABSSSI.

EMA/595311/2019 Page 52/123

• Outcomes/endpoints

Structured evaluations were conducted at Early Clinical Response (48-72 h after first dose), End of
Treatment (EOT, within 2 d of last dose) and Post Therapy Evaluation (PTE, 7-14 d after last dose).

Co-primary efficacy endpoints: Overall clinical response at PTE (derived from IACR at EOT and PTE) in
the mITT and CE-PTE populations, according to the following criteria:

Table 24. Investigator-assigned clinical success/ failure at the EOT and PTE visits

Clinical Success at the EOT and PTE assessments was defined as meeting the following:
• The subject was alive.
• The infection was sufficiently resolved such that further antibacterial therapy was not needed. These
subjects may have had some residual changes related to infection requiring ancillary (ie, non-antibiotic)
treatment, eg, bandages on a healing wound, debridement of uninfected tissue (ie, necrotic).
Clinical Failure was defined as meeting any of the criteria below:
• The investigator discontinued test article and indicated that the infection had responded inadequately
such that alternative (rescue) antibacterial therapy was needed.
• The subject received antibacterial therapy that may have been effective for the infection under study
for a different infection from the one under study.
• The subject developed an AE that required discontinuation of test article prior to completion of the
planned test article regimen.
• An unplanned major surgical intervention (ie, procedures that would not normally be performed at the
bedside) for the infection under study.
• The subject died before evaluation.
Indeterminate was defined as the clinical response to test article could not be adequately inferred because:
• The subject was not seen for the assessment because the subject withdrew consent, was lost to follow-
up, or other specified reason.
• Other specified reason.
Source: applicant’s Summary of Clinical Efficacy – ABSSSI.

Secondary efficacy endpoints:

 Microbiological response at EOT and overall microbiological response at PTE (derived from
microbiological response at EOT and PTE) per-subject in the micro-mITT and ME-PTE
populations.

 Overall microbiological response at PTE (derived from microbiological response at EOT and
PTE) according to pathogen in the micro-mITT and ME-PTE populations.

Infection site or blood isolates always considered a pathogen were: MSSA, MRSA, Group A, B,C and G
β-haemolytic streptococci, Streptococcus anginosus group, E. faecalis and E. faecium, Staphylococcus
lugdunensis. Isolates always considered contaminants were: fungi, Staphylococcus saprophyticus,
Corynebacterium spp., Bacillus spp., diptheroids, Micrococcus spp., and Propionibacterium spp.. Other
isolates and combinations were considered in a blinded case-by-case- manner.

• Sample size

Assuming an 85% outcome rate in both treatment groups, non-inferiority margin of 10%, and a 1-
sided alpha of 0.0125, with a total of 632 subjects, there was 89% power to show non-inferiority for
investigator’s assessment of clinical response at PTE in the mITT population. With an evaluability rate
of 80%, there were 506 subjects provided 91% power to show non-inferiority in the CE population.

• Blinding (masking)

The investigator and sponsor were blinded to treatment arm assignments throughout the study. Both
the iv and po phases of the study were double-blind and double-dummy.

EMA/595311/2019 Page 53/123

• Statistical methods

Study populations

• The intent-to-treat (ITT) population consisted of all randomized subjects regardless of whether
or not the subject received test article. A subject was considered randomized when the IxRS
provided the test article assignment (ie, completes a randomization transaction).

• The mITT population consisted of all randomized subjects without a baseline sole Gram-
negative ABSSSI pathogen.

• The micro-mITT population consisted of all subjects in the mITT population who had at least 1
Gram-positive causative bacterial pathogen identified at baseline.

• Two CE analysis sets were defined; the CE-EOT and the CE-PTE comprised mITT patients who
received a minimum dose of study drug, had a qualifying ABSSSI, completed an outcome
assessment, and met all other evaluability criteria detailed in the SAP. By definition, subjects
with an indeterminate clinical outcome at the EOT or PTE visits were excluded from the
relevant CE population.

Multiplicity

No adjustments were made. Confidence intervals for secondary outcomes were descriptive only.

Missing data

Subjects were defined as an indeterminate if the investigator could not determine whether the subject
was a clinical success or failure at the EOT or PTE visits or the subject had a missing response. By
definition, subjects with an indeterminate response were included in the denominator for analyses in
the mITT and micro-mITT populations, and thus, were considered clinical failures. Subjects with an
indeterminate response were excluded from the CE-EOT, CE-PTE, ME-EOT, and ME-PTE populations

Results (Study PTK0796-ABSI-1108)

Figure 4. Participant flow, Study PTK0796-ABSI-1108

Randomised
ITT 655

Not treated Not treated

6 4
Omadacycline Linezolid
Safety 323 Safety 322
ITT 329 ITT 326

13 Sole Gram-negative 15
pathogen at Baseline

mITT 316 mITT 311

88 No Gram-positive 84
pathogen identified
at Baseline
Excluded from 47 51 Excluded from
Clinically Evaluable micro-mITT 228 micro-mITT 227 Clinically Evaluable
population population

Clinically
CE-PTE 269 CE-PTE 260
evaluable

Microbiologically
ME-PTE 188 evaluable
ME -PTE 192

CE = clinically evaluable, ITT = intent-to-treat, ME = microbiologically evaluable, mITT = modified intent-to-treat,

micro-mITT = microbiological modified intent-to-treat, PTE = post therapy evaluation.
Source: applicant’s Summary of Clinical Efficacy – ABSSSI.

EMA/595311/2019 Page 54/123

 Table 25. Subject Disposition, ITT population (Study PTK0796-ABSI-1108)

Source: Study PTK0796-ABSI-1108 CSR.

• Conduct of the study

Database lock: 09 June 2016

Data base unblind: 09 June 2016

The SAP was finalised on 24 May 2016, prior to unblinding, and the Applicant states that no knowledge
about results was available when changing the alpha-level in the final SAP.

EMA/595311/2019 Page 55/123

• Baseline data

Table 26. Demographics of Safety population (Study PTK0796-ABSI-1108)

Source: Study PTK0796-ABSI-1108 CSR Table 10.

EMA/595311/2019 Page 56/123

 Table 27. Medical history occurring in >10% of Safety population (Study PTK0796-
ABSI-1108)

Source: Study PTK0796-ABSI-1108 CSR Table 11.

• Causative organisms

The frequency of specific species amongst pooled ABSSSI Phase 3 study data is presented in Section
3.3.6. Analysis performed across trials (pooled analyses). Only 11 subjects in the omadacycline arm
and 9 subjects in the linezolid arm had positive blood cultures (<5% of mITT population).

• Numbers analysed

Table 28. Analysis populations Overall and by Geographic Region (Study PTK0796-
ABSI-1108)
Region Population Omadacycline Linezolid All Subjects
n (%) n (%) n (%)
Overall ITT 329 326 655
Safety 323 ( 98.2) 322 ( 98.8) 645 ( 98.5)
mITT 316 ( 96.0) 311 ( 95.4) 627 ( 95.7)
micro-mITT 228 ( 69.3) 227 ( 69.6) 455 ( 69.5)
CE-EOT 278 ( 84.5) 273 ( 83.7) 551 ( 84.1)
CE-PTE 269 ( 81.8) 260 ( 79.8) 529 ( 80.8)
ME-EOT 195 ( 59.3) 198 ( 60.7) 393 ( 60.0)
ME-PTE 188 ( 57.1) 192 ( 58.9) 380 ( 58.0)
North America (USA) ITT 208 207 415
Safety 202 ( 97.1) 203 ( 98.1) 405 ( 97.6)
mITT 207 ( 99.5) 202 ( 97.6) 409 ( 98.6)
micro-mITT 173 ( 83.2) 166 ( 80.2) 339 ( 81.7)
CE-EOT 171 ( 82.2) 168 ( 81.2) 339 ( 81.7)
CE-PTE 166 ( 79.8) 161 ( 77.8) 327 ( 78.8)
ME-EOT 141 ( 67.8) 139 ( 67.1) 280 ( 67.5)
ME-PTE 136 ( 65.4) 136 ( 65.7) 272 ( 65.5)
Eastern Europe ITT 92 92 184
Safety 92 (1 000) 92 (1 000) 184 (1 000)
mITT 81 ( 88.0) 84 ( 91.3) 165 ( 89.7)
micro-mITT 48 ( 52.2) 56 ( 60.9) 104 ( 56.5)
CE-EOT 81 ( 88.0) 82 ( 89.1) 163 ( 88.6)
CE-PTE 78 ( 84.8) 77 ( 83.7) 155 ( 84.2)
ME-EOT 48 ( 52.2) 54 ( 58.7) 102 ( 55.4)
ME-PTE 46 ( 50.0) 52 ( 56.5) 98 ( 53.3)
Western Europe ITT 24 23 47
Safety 24 (100.0) 23 (100.0) 47 (100.0)
mITT 23 ( 95.8) 21 ( 91.3) 44 ( 93.6)
micro-mITT 7 ( 29.2) 4 ( 17.4) 11 ( 23.4)

EMA/595311/2019 Page 57/123

 CE-EOT 21 ( 87.5) 19 ( 82.6) 40 ( 85.1)
CE-PTE 20 ( 83.3) 18 ( 78.3) 38 ( 80.9)
ME-EOT 6 ( 25.0) 4 ( 17.4) 10 ( 21.3)
ME-PTE 6 ( 25.0) 3 ( 13.0) 9 ( 19.1)
Latin America ITT 5 4 9
Safety 5 (100.0) 4 (100.0) 9 (100.0)
mITT 5 (100.0) 4 (100.0) 9 (100.0)
micro-mITT 0 1 ( 25.0) 1 ( 11.1)
CE-EOT 5 (100.0) 4 (100.0) 9 (100.0)
CE-PTE 5 (100.0) 4 (100.0) 9 (100.0)
ME-EOT 0 1 ( 25.0) 1 ( 11.1)
ME-PTE 0 1 ( 25.0) 1 ( 11.1)
Table compiled by assessor. Source data: Study PTK0796-ABSI-1108 CSR Table 14.1.1.2.

• Outcomes and estimation

Primary efficacy analysis

Table 29. Efficacy: Overall Clinical Response at PTE in mITT, and CE-PTE Populations
(Study PTK0796-ABSI-1108)

Omadacycline Linezolid Difference (95%

Efficacy Outcome n (%) n (%) CI)
mITT N = 316 N = 311
Clinical success 272 (86.1) 260 (83.6) 2.5 (-3.2, 8.1)
Clinical failure or indeterminate 44 (13.9) 51 (16.4)
Clinical failure 20 (6.3) 27 (8.7)
Discontinued test article with a need for rescue
10 (3.2) 10 (3.2)
antibacterial therapy
Potentially effective antibacterial received for
6 (1.9) 5 (1.6)
different indication
Clinical failure at EOT 15 (4.7) 18 (5.8)
Discontinued test article with a need for
12 (3.8) 8 (2.6)
rescue antibacterial therapy
AE requiring discontinuation of test article 12 (3.8 8 (2.6)
Unplanned major surgical intervention 0 1 (0.3)
Indeterminate 24 (7.6) 24 (7.7)
CE-PTE N = 269 N = 260
Clinical success 259 (96.3) 243 (93.5) 2.8 (-0.9, 7.1)
Clinical failure 10 (3.7) 17 (6.5)
CI = confidence interval, CE = clinically evaluable, CSR = clinical study report, EOT = end of treatment, EMA =
European Medicines Agency, mITT = modified intent-to-treat, PTE = post therapy evaluation.
Source adapted from: Study PTK0796-ABSI-1108 EMA CSR Table 2 and Table 14.2.2.2.
Given that the lower limit of the 95% CI for the treatment difference between arms in the mITT and CE
populations was above the pre-specified margin of -10%, omadacycline was considered non-inferior to
linezolid. Sensitivity analysis of the primary outcome conducted by constructing an unstratified 95% CI
yielded results almost identical to the primary analysis in both mITT and CE populations.

Sensitivity analysis of the co-primary efficacy endpoints in the all-treated population (all randomised
subjects receiving at least 1 dose of test article) was provided as recommended in CHMP Scientific
Advice. In the Safety population, clinical success rates were high (87.3% for omadacycline, 85.4% for
linezolid) and comparable between both treatment groups (difference [95% CI]: 1.9 [-3.5, 7.2]. The
95% CI for the sensitivity analysis without stratification was [-3.4, 7.3].

When patients with indeterminate response were imputed with clinical success, sensitivity analysis
produced similar results to the primary analysis.

EMA/595311/2019 Page 58/123

Secondary efficacy analyses

Microbiological response (per-subject)

Nearly all cases of microbiological eradication were “presumed” (i.e. there were very few post-baseline
cultures performed) in the micro-mITT.

Table 30. Microbiological Response at EOT and overall microbiological response at PTE
in micro-mITT, ME-EOT, and ME-PTE Populations (Study PTK0796-ABSI-1108)

Omadacycline Linezolid
Visit/Outcome n (%) n (%) Difference (95% CI)
micro-mITT N = 228 N = 227
Microbiological response at
EOT visit
Favourable 202 (88.6) 199 (87.7) 0.9 (-5.1, 7.0)
Eradication 0 4 (1.8) -
Presumed eradication 202 (88.6) 195 (85.9) -
Unfavourable 7 (3.1) 13 (5.7) -
Persistence 1 (0.4) 2 (0.9) -
Presumed persistence 6 (2.6) 11 (4.8) -
Indeterminate 19 (8.3) 15 (6.6) -
Overall microbiological response
at PTE visit
Favourable 194 (85.1) 189 (83.3) 1.8 (-5.0, 8.6)
Unfavourable 11 (4.8) 19 (8.4) -
Indeterminate 23 (10.1) 19 (8.4) -
ME-EOT N = 195 N = 198
Microbiological response at
EOT visit
Favourable 192 (98.5) 191 (96.5) 2.0 (-1.3, 5.9)
Eradication 0 2 (1.0) -
Presumed eradication 192 (98.5) 189 (95.5) -
Unfavourable 3 (1.5) 7 (3.5) -
Persistence 1 (0.5) 1 (0.5) -
Presumed persistence 2 (1.0) 6 (3.0) -
ME-PTE N = 188 N = 192
Microbiological response at PTE
Visit
Favourable 184 (97.9) 180 (93.8) 4.1 (0.1, 8.7)
Eradication 1 (0.5) 2 (1.0) -
Presumed eradication 183 (97.3) 178 (92.7) -
Unfavourable 4 (2.1) 12 (6.3) -
Presumed persistence 4 (2.1) 12 (6.3) -
Overall microbiological response
at PTE visit
Favourable 184 (97.9) 180 (93.8) 4.1 (0.1, 8.7)
Unfavourable 4 (2.1) 12 (6.3) -
CI = confidence interval, CSR = clinical study report, EMA = European Medicines Agency, EOT = end of treatment;
ME = microbiologically evaluable; micro-mITT = microbiological modified intent-to-treat; PTE = post therapy
evaluation.
Source: Study PTK0796-ABSI-1108 CSR, Table 6.
Microbiological response (per-pathogen)

Similar rates of overall favourable microbiological response were observed for both treatment arms for
S. aureus and pooled non-S. aureus species:

EMA/595311/2019 Page 59/123

 Table 31. Overall microbiological response at PTE, according to pathogen, micro-mITT
and population (Study PTK0796-ABSI-1108)

Favourable overall microbiological response at PTE

Pathogen n/N (%)
Omadacycline (n=228) Linezolid (n=227)
Staphylococcus aureus 130/156 (83) 126/151 (83)
MSSA 74/88 (84) 84/102 (82)
MRSA 57/69 (83) 43/50 (86)
Non-S. aureus organisms 148/177 (84) 133/169 (79)
Table compiled by assessor. Source data: Study PTK0796-ABSI-1108 CSR v2 Table 14.2.13.1.1.

Additional efficacy analyses

Mono- and poly-microbial infections

Table 32. Overall clinical success and overall favourable microbiological response at
PTE, micro-mITT population, according to infection type (Study PTK0796-ABSI-1108)

Overall favourable microbiological

Overall clinical success at PTE
Infection type response at PTE
n/N (%)
n/N (%)
Omadacycline Linezolid Difference
Omadacycline (228) Linezolid (227)
(228) (227) (95% CI)
Mono-
137/156 (88) 145/171 (85) 3.0 (-4.6, 10.6) 137/156 (88) 145/171 (85)
microbial G+
Poly-microbial -7.3 (-28.5,
23/31 (74) 22/27 (82) 24/31 (77) 22/27 (82)
G+ 15.1)
Poly-microbial
33/41 (81) 22/29 (76) 4.6 (-14.7, 25.5) 33/41 (81) 22/29 (76)
mixed
Table complied by assessor. Source data: Study PTK0796-ABSI-1108 CSR Tables 14.2.2.4.1 and 14.2.12.4.1.

Clinical outcome according to baseline pathogen and/or MIC

Pooled clinical response results according to pathogen and baseline MIC are discussed below in Section
3.3.6. Analysis performed across trials (pooled analyses).

New and super-infections

Table 33. New and super-infections (Study PTK0796-ABSI-1108)

Subject Treatment New or super-infection

262-0094
6d iv linezolid Clostridium sordellii (Day 3), super-infection
38M
262-0011
5d iv 6d po omadacycline Streptococcus intermedius (Day 22), new infection
42F
201-0001
10d iv linezolid P. aeruginosa and K. pneumoniae (Day 10), new infection
59F
254-0003
4d iv 7d po linezolid MSSA (Day 17), new infection
61M
Table compiled by assessor. Source data: Study PTK0796-ABSI-1108 EMA CSR.

Concordance of clinical outcome across visits

At both of the later evaluation visits, the omadacycline arm demonstrated numerically better continuity
of cure than linezolid, with 4/268 ECR successes deemed failures at EOT and 7/268 deemed failures at
PTE in the omadacycline arm, vs 9/266 ECR successes deemed failures at EOT and 16/266 deemed
failures at PTE in the linezolid arm. The number of ECR successes later deemed indeterminate
(including subjects not clinically evaluable) was the same in both treatment arms.

EMA/595311/2019 Page 60/123

Sub-group analyses

Most subjects in the mITT population had comparable clinical success by type of infection (wound
infection, cellulitis/erysipelas, and major abscess, ~80%, 85-91% and ~85%, respectively). Perhaps
unsurprisingly, the highest clinical success rates (~88%) were observed in subjects with smallest
lesion size (≤300 cm2) and the lowest (72-82%) in subjects with the greatest lesion size (≥1000 cm2).
There were no obvious differences revealed by sub-group analysis according to geographic region, EU
vs non-EU, history of IV drug use, bacteraemia or SIRS. Some numerical differences are noted, but
small absolute numbers in some of these groups precludes any robust conclusions. Sub-group analysis
was also provided for subjects who switched to PO treatment vs subjects who did not switch. For
subjects in mITT population who did not switch to po treatment, omadacycline was inferior to linezolid
as lower limit of 97.5% confidence interval was lower than -10% (-13.4%). However, this was a small
sub-group (14 omadacycline, 15 linezolid) and only and exploratory analysis.

3.3.5.2.2. Study PTK0796-ABSI-16301 - A Phase 3 Randomised, Double-Blind, Multi-

Centre Study to Compare the Safety and Efficacy of Oral Omadacycline to Oral Linezolid
(Zyvox®) for Treating Adult Subjects with Acute Bacterial Skin and Skin Structure Infection
(ABSSSI)

Methods (Study PTK0796-ABSI-16301)

• Study participants

Subjects 18 to 80 years of age from 53 US centres. Inclusion criteria were as for Study PTK0896-ABSI-
1108. Subjects who received only a single dose of short-acting (standard dosing regimen >once a
day), non-oxazolidinone antibiotic were eligible for enrolment (capped at 25%).

• Treatments

Subjects were randomised (1:1) to one of the following two treatments, for a total of 7 to 14 days:

o Omadacycline, 450 mg po q24h for 2 doses, followed by 300 mg po q24h.

o Linezolid, 600 mg po q12h.

Subjects could receive adjunctive Gram-negative cover at the discretion of the investigator if a Gram-
negative pathogen, anaerobe or linezolid-resistant pathogen was detected at baseline.

• Objectives

Primary objective: to demonstrate that omadacycline administered po for 7 to 14 days was non-inferior
to linezolid administered po for 7 to 14 days in the treatment of adult subjects with ABSSSI known or
suspected to be due to Gram-positive pathogens.

Secondary objectives: as for study PTK0796-ABSI-1108.

• Outcomes/endpoints - As for study PTK0796-ABSI-1108.

• Sample size

Assuming an 85% outcome rate in both treatment groups, a non-inferiority margin of 10%, and a 1-
sided alpha of 0.025, with a total of 704 subjects, there was more than 90% power to show non-
inferiority for IACR at PTE in the mITT population. With an evaluability rate of 80%, there were 564
subjects in the CE population. Assuming a 90% outcome rate in both treatment groups, a non-
inferiority margin of 10%, and a 1-sided alpha of 0.025, 564 subjects provided more than 90% power
to show non-inferiority in the CE population.

EMA/595311/2019 Page 61/123

• Blinding (masking)

Investigator and sponsor were blinded to treatment allocation. Double-dummy was used throughout.

• Statistical methods - As for study PTK0796-ABSI-1108.

Results (Study PTK0796-ABSI-16301)

Figure 5. Participant flow, Study PTK0796-ABSI-16301

CE = clinically evaluable, ITT = intent-to-treat, ME = microbiologically evaluable, mITT = modified intent-to-treat,

micro-mITT = microbiological modified intent-to-treat, PTE = post therapy evaluation.
Source: applicant’s Summary of Clinical Efficacy – ABSSSI.

Table 34. Subject disposition, ITT population (Study PTK0796-ABSI-16301)

Source: Study PTK0796-ABSI-16301 CSR Table 4.

EMA/595311/2019 Page 62/123

• Conduct of the study

Database lock: 30 June 2017

Database unblind: 03 July 2017

• Baseline data

Table 35. Demographics of Safety population (Study PTK0796-ABSI-1108)

Source: Study PTK0796-ABSI-16301 CSR Table 7.

EMA/595311/2019 Page 63/123

 Table 36. Medical history occurring in >10% of Safety population (Study PTK0796-
ABSI-16301)

Source: Study PTK0796-ABSI-16301 CSR Table 8.

• Causative organisms

The frequency of specific species across pooled Phase 3 study data is presented in Section 3.3.6.
Analysis performed across trials (pooled analyses). Only 2 subjects in the omadacycline arm and 8
subjects in the linezolid arm had positive blood cultures (<2% of mITT population).

• Numbers analysed

Table 37. Analysis populations (Study PTK0796-ABSI-16301)

Region Population Omadacycline Linezolid All Subjects

n (%) n (%) n (%)
Overall ITT 368 367 735
Safety 368 ( 100) 367 ( 100) 735 ( 100)
mITT 360 ( 97.8) 360 ( 98.1) 720 ( 98.0)
micro-mITT 276 ( 75.0) 287 ( 78.2) 563 (76.6)
CE-EOT 304 (82.6) 296 (80.7) 600 (81.6)
CE-PTE 284 (77.2) 292 (79.6) 576 (78.4)
ME-EOT 233 (63.3) 229 (62.4) 462 (62.9)
ME-PTE 220 (59.8) 225 (61.3) 445 (60.5)
Source: adapted from Study PTK0796-ABSI-16301 CSR Table 6.

EMA/595311/2019 Page 64/123

• Outcomes and estimation

Primary efficacy analysis

Table 38. Efficacy: Overall Clinical Response at PTE in mITT, and CE-PTE Populations
(Study PTK0796-ABSI-16301)

Omadacycline Linezolid Difference (95%

Efficacy Outcome n (%) n (%) CI)
mITT N = 360 N = 360
Clinical success 303 (84.2) 291 (80.8) 3.3 (-2.2, 9.0)
Clinical failure or indeterminate 57 (15.8) 69 (19.2)
Clinical failure 12 (3.3) 21 (5.8)
Infection required alternative antibacterial 1 (0.3) 1 (0.3)
Other (non-compliance) 0 1 (0.3)
Clinical failure at EOT 11 (3.1) 19 (5.3)
Discontinued test article with a need for 5 (1.4) 11 (3.1)
rescue antibacterial therapy
Potentially effective antibacterial received for 1 (0.3) 0
different indication
AE requiring discontinuation of test article 3 (0.8) 4 (1.1)
Other 3 (0.8) 6 (1.7)
Indeterminate 45 (12.5) 48 (13.3)
CE-PTE N = 284 N = 292
Clinical success 278 (97.9) 279 (95.5) 2.3 (-0.5, 5.8)
Clinical failure 6 (2.1) 13 (4.5)
CI = confidence interval, CE = clinically evaluable, CSR = clinical study report, EOT = end of treatment, EMA =
European Medicines Agency, mITT = modified intent-to-treat, PTE = post therapy evaluation.
Source: adapted from Study PTK0796-ABSI-16301 EMA CSR Table 3 and Table 14.2.2.2.

Given that the lower limit of the 95% CI for the treatment difference between arms in the mITT and CE
populations was above the pre-specified margin of -10%, omadacycline was considered non-inferior to
linezolid. Sensitivity analysis of the primary outcome conducted by constructing an unstratified 95% CI
yielded results almost identical to the primary analysis in both mITT and CE populations.

Sensitivity analysis of the co-primary efficacy endpoints in the all-treated population (all randomised
subjects receiving at least 1 dose of test article) was provided as recommended in CHMP Scientific
Advice. In the Safety population, clinical success rates were high (84.2% for omadacycline, 80.7% for
linezolid) and comparable between both treatment groups (difference [95% CI]: 3.6 [-1.9, 9.1] for the
analysis with and without stratification).

When patients with indeterminate response were imputed with clinical success, sensitivity analysis
produced similar results to the primary analysis.

EMA/595311/2019 Page 65/123

Secondary efficacy analyses

Microbiological response (per-subject)

Table 39. Microbiological Response at EOT and PTE in micro-mITT, ME-EOT, and ME-PTE
Populations (Study PTK0796-ABSI-16301)

CI = confidence interval, CSR = clinical study report, EOT = end of treatment, ME = microbiologically evaluable,
micro-mITT = microbiological modified intent-to-treat, PTE = post therapy evaluation.
Source: Study PTK0796-ABSI-16301 EMA CSR Table 6.

Nearly all cases of microbiological eradication were “presumed” (i.e. there were very few post-baseline
cultures performed) in the micro-mITT.

Microbiological response (per-pathogen)

Better rates of overall favourable microbiological response were observed for omadacycline for S.
aureus and pooled non-S. aureus species:

Table 40. Overall microbiological response at PTE, according to pathogen, micro-mITT

and population (Study PTK0796-ABSI-16301)

Favourable overall microbiological response at PTE

Pathogen n/N (%)
Omadacycline (n=276) Linezolid (n=287)
Staphylococcus aureus 182/220 (83) 187/233 (80)
MSSA 97/120 (81) 104/130 (80)
MRSA 89/104 (86) 85/107 (79)
Non-S. aureus organisms 199/238 (84) 145/197 (74)
Table compiled by assessor. Source data: Study PTK0796-ABSI-16301 CSR Table 14.2.13.1.1.

EMA/595311/2019 Page 66/123

Additional efficacy analyses

Mono- and poly-microbial infections

Table 41. Overall clinical success and overall favourable microbiological response at PTE
in the micro-mITT population according to type of infection (Study PTK0796-ABSI-
16301)

Favourable microbiological
Overall clinical success at PTE
Infection type response
n/N (%)
n/N (%)
Omadacycline Linezolid Difference (95% Omadacycline Linezolid
(276) (287) CI) (276) (276)
Mono-microbial
155/184 (84) 167/212 (79) 5.5 (-2.3, 13.1) 155/184 (84) 167/212 (79)
G+
Poly-microbial
49/60 (82) 26/37 (70) 11.4 (-5.6, 29.7) 49/60 (82) 26/37 (70)
G+
Poly-microbial
25/32 (78) 31/38 (82) -3.5 (-23.3, 15.6) 25/32 (78) 31/38 (82)
mixed
Table complied by assessor. Source data: Study PTK0796-ABSI-16301 CSR Tables 14.2.2.4.1 and 14.2.12.4.1.

Clinical outcome according to baseline pathogen and/or MIC

Pooled clinical response results according to pathogen and baseline MIC are discussed below in Section
3.3.6. Analysis performed across trials (pooled analyses).

New and super-infections

Table 42. New and super-infections (Study PTK0796-ABSI-16301)

Subject Treatment New or super-infection

608-3031
3d omadacycline MSSA (Day 2), super-infection
34M
- E. coli (Day 7), super-infection
610-3046 - Bacteroides thetaiotaomicron, Citrobacter freundii and Morganella
10d omadacycline
52M morganii (Day 11), new infection
- Citrobacter braakii and Anaerococcus tetradius (Day 18), new infections
608-3138
3d linezolid Proteus mirabilis (Day 3), new infection
51M
610-3002
3d linezolid Staphylococcus aureus (Day 6), new infection
48M
636-3013 Klebsiella pneumoniae and Stenotrophomonas maltophilia (Day 18), new
17d linezolid
48M infections
Table compiled by assessor. Source data: Study PTK0796-ABSI-16301 EMA CSR.

Concordance of clinical outcome across visits

At both of the later evaluation visits, the omadacycline arm demonstrated numerically better continuity
of cure than linezolid, with 3/315 ECR successes deemed failures at EOT and 4/315 deemed failures at
PTE in the omadacycline arm, vs 8/297 ECR successes deemed failures at EOT and 10/297 deemed
failures at PTE in the linezolid arm.

Sub-group analyses

Most subjects in the mITT population had comparable clinical success by type of infection (wound
infection, cellulitis/erysipelas, and major abscess, 77-82%, 88-93% and 79-84%, respectively).
Perhaps unsurprisingly, the highest clinical success rates (~88%) were observed in subjects with
smallest lesion size (≤300 cm2) and the lowest (~75%) in subjects with the greatest lesion size
(≥1000 cm2). There were no obvious differences revealed by sub-group analysis according to

EMA/595311/2019 Page 67/123

geographic region, EU vs non-EU, history of IV drug use, bacteraemia or SIRS. Some numerical
differences are noted, but small absolute numbers in these groups precludes robust conclusions.

3.3.5.2.3. Study PTK0796-CABP-1200 - A Phase 3 Randomised, Double-Blind, Multi-

Centre, Non-inferiority Study to Compare the Safety and Efficacy of Omadacycline IV/PO to
Moxifloxacin IV/PO for Treating Adult Subjects with Community-Acquired Bacterial
Pneumonia (CABP)

Methods (Study PTK0796-CABP-1200)

• Study participants

Male and female subjects aged 18 years or older, enrolled across 86 global sites with CABP, fulfilling all
of the below criteria and expected to require at least 3 days’ iv treatment:

1. Had at least 3 of the following symptoms:

o Cough
o Production of purulent sputum
o Dyspnoea
o Pleuritic chest pain
2. Had at least 2 of the following abnormal vital signs:
o Fever or hypothermia documented by the investigator (temperature > 38.0°C [100.4°F] or <
36.0°C [95.5°F])
o Hypotension with systolic blood pressure (SBP) less than 90 mm Hg
o Heart rate (HR) greater than 90 beats per min (bpm)
o Respiratory rate (RR) greater than 20 breaths/min
3. Had at least 1 clinical sign or laboratory finding associated with CABP:
o Hypoxemia (partial pressure of arterial oxygen [PaO2] < 60 mm Hg by arterial blood gas
[ABG] or oxygen saturation < 90% by pulse oximetry)
o Physical examination findings of pulmonary consolidation (eg, dullness on percussion, bronchial
breath sounds, or egophony)
o An elevated total white blood cell (WBC) count (> 12,000 cells/mm3) or leukopenia (WBC <
4,000 cells/mm3) or elevated immature neutrophils (> 15% band forms regardless of total
peripheral WBC count)
4. Radiographically-confirmed pneumonia, ie, new or progressive pulmonary infiltrate(s) on chest X-
ray or chest computed tomography (CT) scan consistent with acute bacterial pneumonia within 24
h prior to the first dose of test article.
5. Had disease categorised as being Pneumonia Outcomes Research Team (PORT) Risk Class II, III,
or IV at Screening.

Principal exclusion criteria:

(a) having received 1 or more dose of potentially effective systemic antibiotic in the 72 h prior to
the first dose of test article (unless infecting organism shown to be intermediate or resistant to
initial treatment, or only a single dose of short-acting antibiotic (dosed more than once daily)
was received (capped at 25%)),

(b) suspected or known to have CABP caused by a pathogen resistant to either test article (eg, K.
pneumoniae, Pseudomonas aeruginosa, Pneumocystis jiroveci, obligate anaerobes,
mycobacteria, fungal pathogens),

(c) suspected or confirmed empyema, lung abscess, hospital-acquired pneumoniae (HAP) (defined
as onset of clinical signs or symptoms ≥48 h after hospitalisation) or healthcare associated

EMA/595311/2019 Page 68/123

 pneumonia (HCAP) (defined as pneumonia acquired in long-term or intermediate healthcare
facility or presenting following a recent hospitalisation, as specified in the protocol),

(d) requiring acute intervention to stabilise blood pressure/ tissue perfusion, or having evidence of
septic shock (defined by all of fever >38 degrees Celsius, hypothermia <36 degrees Celsius,
tachycardia >90 bpm, tachypnoea > 20 bpm, deranged WBC count, systolic hypotension <90
mmHg despite fluid challenge, perfusion abnormalities including lactic acidosis, oliguria, change
in mental status),

(e) haemodynamically unstable despite fluid resuscitation, or requiring pressor agents

(f) known or suspected neoplastic lung disease, aspiration, cystic fibrosis, active tuberculosis,
bronchiectasis, bronchial obstruction, chronic neurological disorder affecting airway clearance,
chronic obstructive pulmonary disease (COPD).

(g) pre-existing hepatic, renal, immunological or cardiac disease as defined in the protocol

Subject randomization was stratified by PORT Risk Class (II or III/IV), receipt of an allowed prior
antibacterial therapy in the 72 hours prior to study treatment, and geographic region. The number of
subjects in PORT Risk Class II was limited to 15% of randomized subjects and subjects with allowed
prior antibacterial therapy limited to 25%.

• Treatments

Subjects were randomised (1:1) to 1 of the following 2 treatment arms, to receive at least 3 days iv
treatment, and between 7 and at most 14 days total treatment:

o Omadacycline, 100 mg iv q12h (first 2 doses), followed by 100 mg iv q24h, followed by

optional switch to 300 mg po q24h after at least 3 days of iv treatment.

o Moxifloxacin, 400 mg iv q24h, followed by optional switch to 400 mg po q24h after at least 3
days of iv treatment.

• Objectives

Primary objective: to demonstrate that omadacycline 100 mg iv q12h for 2 doses, followed by 100 mg
iv/300 mg po q24h, was non-inferior to moxifloxacin 400 mg iv/po q24h in the treatment of adults
with CABP.

Secondary objectives:

- to evaluate the safety of omadacycline in adults subjects with CABP.

- to evaluate the clinical response according to the causative pathogen.

- to characterise the PK of omadacycline in adults subjects with CABP.

• Outcomes/endpoints

Structured evaluations were conducted at Early Clinical Response (48-72 h after first dose), End of
Treatment (EOT, within 2 d of last dose), and Post Therapy Evaluation (PTE, 5-10 d after last dose).

Co-primary efficacy endpoint: overall assessment of clinical response at PTE (derived from IACR at EOT
and PTE) in the ITT and CE-PTE populations for PORT Risk Class III/IV subjects, according to the
following criteria:

EMA/595311/2019 Page 69/123

 Table 43. Investigator-assigned clinical success/ failure at the TOC and PTE visits

Clinical Success at the EOT assessment was defined as meeting the following:
• The subject was alive.
• The infection was sufficiently resolved such that further antibacterial therapy was not needed. These
subjects may have had some residual changes related to infection (e.g. cough) requiring ancillary (ie, non-
antibiotic) treatment, (eg, expectorant).
Clinical Success at the PTE assessment was defined as meeting the following:
• Survival, without receiving any systemic antibacterial other than test article.
• Resolution of signs and symptoms of the infection present at Screening with no new symptoms or
complications attributable to CABP and no need for further antibacterial therapy.
Clinical Failure was defined as meeting any of the criteria below:
• The subject required alternative antibacterial treatment for CABP prior to EOT related to either (a)
progression or development of new symptoms to CABP or (b) development of infectious complications of
CABP (eg, empyema, lung abscess) or (c) subject developed an adverse event (AE) that required
discontinuation of study therapy prior to the EOT.
• The subject received antibacterial therapy that may have been effective for the infection under study
for a different infection from the one under study.
• The subject died before evaluation.
Indeterminate was defined as the clinical response to test article could not be adequately inferred because:
• The subject was not seen for the assessment because the subject withdrew consent, was lost to follow-
up, or other specified reason.
• Other specified reason.
Source: applicant’s Summary of Clinical Efficacy – CABP.

Secondary efficacy endpoints:

 Microbiological response at EOT and overall microbiological response at PTE (derived from
microbiological response at EOT and PTE) per-subject in the micro-mITT and ME-PTE
populations.

 Overall microbiological response at PTE (derived from microbiological response at EOT and
PTE) according to pathogen in the micro-mITT and ME-PTE populations.

In addition to blood and sputum microscopy and culture, urinary antigen screening (L. pneumophila, P.
pneumoniae) and serology (L. pneumophila, M. pneumoniae, C. pneumoniae) were performed. Blood
and sputum isolates always considered a pathogen were: S. pneumoniae, H. influenzae, S. aureus,
M. catarrhalis, L. pneumophila, M. pneumoniae, and C. pneumoniae. Isolates always considered
contaminants were: fungi, Enterococcus spp., viridans streptococci, coagulase-negative staphylococci,
Micrococcus spp., Neisseria spp. other than Neisseria meningitidis, Corynebacterium spp. and other
coryneforms, Lactobacillus spp., Vibrio spp., Capnocytophaga spp., Cardiobacterium spp.,
Flavobacterium spp. Other isolates were considered in a blinded case-by-case- manner.

• Sample size

For the IACR at PTE endpoint in the ITT population (PORT Risk Class III and IV) assuming an outcome
rate of 79% in both treatment groups, non-inferiority margin of 10%, 80% power, and a 1-sided alpha
level of 0.0125 (since 1 CABP study was being conducted), using the sample size determination
method of Farrington and Manning (1990), a total of 638 subjects (PORT Risk Class III and IV) were
required. Assuming an 80% evaluability rate, 510 subjects were available in the CE population.
Assuming an 85% response rate in both treatment groups, a 10% non-inferiority margin, 1-sided
alpha level of 0.0125, there was 81% power to show non-inferiority for the IACR at PTE in the CE
population. If 15% of subjects were in PORT Risk Class II and therefore excluded from the primary
efficacy evaluation, a total of 750 subjects were required.

EMA/595311/2019 Page 70/123

• Blinding (masking)

Investigator and sponsor were blinded to treatment allocation. The iv and po phases of the study were
double-blind and double-dummy.

• Statistical methods

Study populations

• The intent-to-treat (ITT) population consisted of all randomized subjects with PORT Risk Class
III/IV, regardless of whether or not the subject received test article. A subject was considered
randomized when the IxRS provided the test article assignment (ie, completes a randomization
transaction).

• The micro-mITT population consisted of all subjects in the mITT population who had at least 1
causative bacterial pathogen identified at baseline (see above, Outcomes/endpoints), according
to the more stringent criteria for sputum Gram stain of >25 PMNs/LPF and <10 SECs/LPF.

• The expanded micro-mITT population consisted of all subjects in the mITT population who had
at least 1 causative bacterial pathogen identified at baseline (see above, Outcomes/endpoints),
according to the less stringent criteria for sputum Gram stain of >10 PMNs/LPF and <10
SECs/LPF.

• Two CE analysis sets were defined; the CE-EOT and the CE-PTE comprised mITT patients who
received a minimum dose of study drug, completed an outcome assessment, and met all other
evaluability criteria detailed in the SAP. By definition, subjects with an indeterminate clinical
outcome at the EOT or PTE visits were excluded from the relevant CE population.

Multiplicity

No adjustments were made. Confidence intervals for secondary outcomes were descriptive only.

Missing data

For the IACR at the EOT and PTE visits subjects were defined as an indeterminate if the investigator
could not determine whether the subject was a clinical success or failure at the EOT or PTE visits or the
subject had a missing response. By definition, subjects with an indeterminate response were included
in the denominator for analyses in the ITT and microITT populations, and thus, were considered clinical
failures. Subjects with an indeterminate response were excluded from the CE-EOT, CE-PTE, ME-EOT,
and ME-PTE populations.

EMA/595311/2019 Page 71/123

Results (Study PTK0796-CABP-1200)

Figure 6. Participant flow, Study PTK0796-CABP-1200

CE = clinically evaluable, EMA = European Medicines Agency, ITT = intent-to-treat, ME = microbiologically

evaluable, microITT = microbiological intent to treat, PORT = Pneumonia Outcomes Research Team, PTE = post
therapy evaluation. Source: applicant’s Summary of Clinical Efficacy – CABP.

Table 44. Subject disposition, ITT population (Study PTK0796-CABP-1200)

Source: Study PTK0796-CABP-1200 CSR.

• Conduct of the study

Database lock: 24 March 2017

Database unblind: 21 March 2017

EMA/595311/2019 Page 72/123

• Baseline data, subjects with PORT Risk Class III/IV

Table 45. Demographics in the Safety population (Study PTK0796-CABP-1200)

Source: Study PTK0796-CABP-1200 CSR Table 10.

EMA/595311/2019 Page 73/123

 Table 46. Medical history occurring in >10% of Safety population (Study PTK0796-
ABSI-16301)

Source: Study PTK0796-CABP-1200 CSR Table 11.

Two thirds of subjects were in PORT Risk Class III and the remaining third were in PORT Risk Class IV.
The mean PORT score was approximately 87. One third of patients had moderate to severe CABP
based on their CURB-65 score or American Thoracic Society (ATS) criteria.

The proportion of subjects receiving iv treatment with a calculated treatment compliance of ≥ 80% for
the omadacycline and moxifloxacin groups was 98.5% and 99.4%, respectively in the ITT population
and 100% in both groups in the CE-PTE population.

The proportion of subjects receiving po treatment with a calculated treatment compliance of ≥ 80% for
the omadacycline and moxifloxacin groups was 76.0% and 73.7%, respectively in the ITT population
and 79.0% and 75.3%, respectively in the CE-PTE population. Oral compliance may have been lower
due to the administration of study drug in the outpatient setting. It is unlikely that the lower oral
compliance had any clinically meaningful impact since the efficacy at PTE was high and a majority of
patients (~75%) switched to oral therapy.

• Causative organisms

Table 47. CABP Subjects, most prevalent organisms, microITT Population, subjects with
PORT Risk Class III/IV (PTK0796-CABP-1200)

Omadacycline Moxifloxacin
(N=175) (N=156)
Baseline Pathogen n (%) n (%)
Gram-Positive Bacteria (aerobes) 55 (31.4) 51 (32.7)
Streptococcus pneumoniae 39 (22.3) 32 (20.5)
PSSP 24 (13.7) 20 (12.8)
Macrolide Resistant 10 (5.7) 5 (3.2)
MDRSP 7 (4.0) 6 (3.8)
Staphylococcus aureus 10 (5.7) 9 (5.8)
MSSA 10 (5.7) 8 (5.1)
Gram-Negative Bacteria (aerobes) 67 (38.3) 60 (38.5)
Haemophilus influenzae 26 (14.9) 14 (9.0)
Haemophilus parainfluenzae 14 (8.0) 13 (8.3)
Klebsiella pneumoniae 13 (7.4) 11 (7.1)
Escherichia coli 4 (2.3) 6 (3.8)
Pseudomonas aeruginosa 3 (1.7) 3(1.9)

EMA/595311/2019 Page 74/123

 Omadacycline Moxifloxacin
(N=175) (N=156)
Baseline Pathogen n (%) n (%)
Atypical Pathogens 104 (59.4) 91 (58.3)
Mycoplasma pneumoniae 63 (36.0) 50 (32.1)
Legionella pneumophila 31 (17.7) 32 (20.5)
Chlamydia pneumoniae 26 (14.9) 24 (15.4)
MDR = multi-drug resistant, MDRSP = multidrug-resistant S. pneumoniae, microITT = microbiological
intent-to-treat, MRSA = methicillin-resistant Staphylococcus aureus, MSSA = methicillin-susceptible S. aureus,
N = Number of subjects in the microITT population, n = Number of subjects within a specific category,
PORT = Pneumonia Outcomes Research Team, PSSP = penicillin-susceptible S. pneumoniae.
Source: adapted from applicant’s Summary of Clinical Pharmacology Studies – Special Studies – Microbiology.

Table 48. Omadacycline and Moxifloxacin MIC Summary Statistics for Baseline
Pathogens From Blood and/or Respiratory Cultures, microITT population, subjects
with PORT Risk Class III/IV (Study PTK0796-CABP-1200)

Omadacycline Moxifloxacin
(N=175) (N=156)
MIC50/ MIC50/
Range MIC90 Range MIC90
Baseline Pathogen n (µg/mL) (µg/mL) n (µg/mL) (µg/mL)
Gram-positive organisms (aerobes) 42 0.015 - 0.25 0.06/0.12 40 ≤ 0.03-4 0.12/0.25
Streptococcus pneumoniae 26 0.015 - 0.12 0.03/0.06 22 ≤ 0.03 - 0.25 0.12/0.12
MDRSP 7 0.03 - 0.12 NC 6 ≤ 0.03 - 0.25 NC
PSSP 25 0.015 - 0.12 0.03/0.06 20 0.06 - 0.25 0.12/0.12
Macrolide Resistant 10 0.03 - 0.12 0.03/0.06 5 0.06 - 0.25 NC
Staphylococcus aureus 10 0.12 - 0.25 0.12/0.25 9 0.03 – 0.06 NC
MSSA 10 0.12 - 0.25 0.12/0.25 8 0.03 – 0.06 NC
Gram-negative organisms (aerobes) 66 0.12 - > 16 2/8 56 ≤ 0.004 - > 4 0.12/1
Haemophilus influenzae 26 1-2 1/2 13 0.008 - 1 0.03/0.06
Haemophilus parainfluenzae 13 1-4 2/4 12 ≤ 0.004 - 1 0.12/0.25
Klebsiella pneumoniae 12 1 – 16 2/16 11 0.06 - 0.12 0.12/0.12
Escherichia coli 4 0.5 – 4 NC 6 0.03 - > 4 NC
Pseudomonas aeruginosa 3 4 - > 16 NC 3 0.06 - 1 NC
Source: Study PTK0796-CABP-1200 CSR Table 20.

• Numbers analysed

Table 49. Analysis populations Overall and by Geographic Region, subjects with PORT
Risk Class III/IV (Study PTK0796-CABP-1200)

Omadacycline Moxifloxacin All Subjects

Region Population
n (%) n (%) N (%)
Overall ITT 329 331 660
Safety 326 ( 99.1) 331 ( 100) 657 ( 99.5)
micro-mITT 175 ( 53.2) 156 ( 47.1) 331 (50.2)
Expanded micro-mITT 188 (57.1) 169 (51.1) 357 (54.1)
CE-EOT 309 (93.9) 307 (92.7) 616 (93.3)
CE-PTE 295 (89.7) 296 (89.4) 591 (89.5)
ME-EOT 167 (50.8) 151 (45.6) 318 (48.2)
ME-PTE 163 (49.5) 147 (44.4) 310 (47.0)
Western Europe ITT 77 70 147
Safety 76 (98.7) 70 (100.0) 146 (99.3)
micro-mITT 39 (50.6) 28 (40.0) 67 (45.6)
Expanded micro-mITT 40 (51.9) 29 (41.4) 69 (46.9)

EMA/595311/2019 Page 75/123

 CE-EOT 72 (93.5) 62 (88.6) 134 (91.2)
CE-PTE 71 (92.2) 62 (88.6) 133 (90.5)
ME-EOT 35 (45.5) 26 (37.1) 61 (41.5)
ME-PTE 34 (44.2) 26 (37.1) 60 (40.8)
Eastern Europe ITT 209 213 422
Safety 208 ( 99.5) 213 (1 000) 421 ( 99.8)
micro-mITT 116 ( 55.5) 111 ( 52.1) 227 ( 53.8)
Expanded micro-mITT 127 ( 60.8) 122 ( 57.3) 249 ( 59.0)
CE-EOT 196 ( 93.8) 203 ( 95.3) 399 ( 94.5)
CE-PTE 186 ( 89.0) 194 ( 91.1) 380 ( 90.0)
ME-EOT 112 ( 53.6) 108 ( 50.7) 220 ( 52.1)
ME-PTE 109 ( 52.2) 104 ( 48.8) 213 ( 50.5)
North America (USA, Mexico) ITT 0 3 3
Safety 0 3 (100.0) 1 (33.3)
micro-mITT 0 0 0
Expanded micro-mITT 0 0 0
CE-EOT 0 2 (66.7) 2 (66.7)
CE-PTE 0 1 (33.3) 1 (33.3)
ME-EOT 0 0 0
ME-PTE 0 0 0
Rest of World ITT 43 45 88
Safety 42 ( 97.7) 45 (100.0) 87 ( 98.9)
micro-mITT 20 ( 46.5) 17 ( 37.8) 37 ( 42.0)
Expanded micro-mITT 21 ( 48.8) 18 ( 40.0) 39 ( 44.3)
CE-EOT 41 ( 95.3) 40 ( 88.9) 81 ( 92.0)
CE-PTE 38 ( 88.4) 39 ( 86.7) 77 ( 87.5)
ME-EOT 20 ( 46.5) 17 ( 37.8) 37 ( 42.0)
ME-PTE 20 ( 46.5) 17 ( 37.8) 37 ( 42.0)
Table compiled by assessor. Source data: Study PTK0796-CABP-1200 CSR Table 14.1.1.2E.

EMA/595311/2019 Page 76/123

• Outcomes and estimation

Primary efficacy analysis

Table 50. Overall Clinical Response at the PTE in ITT and CE-PTE populations, subjects
with PORT Risk Class III/IV (Study PTK0796-CABP-1200)

Omadacycline Moxifloxacin Difference

Efficacy Outcome n (%) n (%) (97.5% CI)
ITT N = 329 N = 331
Clinical success 291 (88.4) 282 (85.2) 3.3 (-2.7, 9.3)
Clinical failure or indeterminate 38 (11.6) 49 (14.8)
Clinical failure 27 (8.2) 35 (10.6)
Rescue/ alternative antibacterial required for
6 (1.8) 9 (2.7)
CABP
Infectious complication of CABP 2 (0.6) 2 (0.6)
Potentially effective antibacterial received
3 (0.9) 3 (0.9)
for different indication
Died before evaluation 3 (0.9) 3 (0.9)
Clinical failure at EOT 22 (6.7) 30 (9.1)
Rescue/ alternative antibacterial
12 (3.6) 19 (5.7)
required for CABP
Infectious complication of CABP 2 (0.6) 5 (1.5)
AE requiring discontinuation of test
4 (1.2) 8 (2.4)
article
Potentially effective antibacterial
1 (0.3) 1 (0.3)
received for different indication
Died before evaluation 3 (0.9) 1 (0.3)
Indeterminate 11 (3.3) 14 (4.2)
CE-PTE N = 295 N = 296
Clinical success 273 (92.5) 268 (90.5) 2.0 (-3.2, 7.4)
Clinical failure 22 (7.5) 28 (9.5)
CE = clinically evaluable, CI = confidence interval, CSR = clinical study report, EMA = European Medicines
Agency, EOT = end of treatment, ITT = intent-to-treat, PORT = Pneumonia Outcomes Research Team,
PTE = post therapy evaluation.
Source: amalgamated from Study PTK0796-CABP-1200 EMA CSR Tables 14.2.2.1.1E and 14.2.2.2.1E.

Given that the lower limit of the 97.5% CI for the treatment difference between arms in the ITT and CE
populations was above the pre-specified margin of -10%, omadacycline was considered non-inferior to
moxifloxacin. Sensitivity analysis of the primary outcome conducted by constructing an unstratified
97.5% CI yielded results almost identical to the primary analysis in both the ITT (97.5% CI: -2.7, 9.3)
and CE-PTE populations (97.5% CI: -3.2, 7.3).

Of note, the lower limit of a 99% CI was also higher than -10% (-3.6% for ITT population and -4.0%
for CE-PTE population).

Secondary efficacy analyses

Microbiological response (per-subject)

Nearly all cases of microbiological eradication were “presumed” (i.e. there were very few post-baseline
cultures performed) in the micro-mITT.

EMA/595311/2019 Page 77/123

 Table 51. Microbiological response at EOT and PTE in microITT, ME-EOT and ME-PTE
populations, subjects with PORT Risk Class III/IV (Study PTK0796-CABP-1200)

Omadacycline Moxifloxacin
Visit/Outcome n (%) n (%) Difference (97.5% CI)
microITT N = 175 N = 156
Microbiological response at EOT visit
Favourable 160 (91.4) 145 (92.9) -1.5 (-8.5, 5.6)
Eradication 4 (2.3) 8 (5.1)
Presumed eradication 156 (89.1) 137 (87.8)
Unfavourable 13 (7.4) 9 (5.8)
Persistence 1 (0.6) 1 (0.6)
Presumed persistence 12 (6.9) 8 (5.1)
Indeterminate 2 (1.1) 2 (1.3)
Overall microbiological response at PTE visit
Favourable 158 (90.3) 138 (88.5) 1.8 (-5.9, 9.9)
Eradication 2 (1.1) 3 (1.9)
Presumed eradication 156 (89.1) 135 (86.5)
Unfavourable 15 (8.6) 14 (9.0)
Persistence 1 (0.6) 1 (0.6)
Presumed persistence 14 (8.0) 13 (8.3)
Indeterminate 2 (1.1) 4 (2.6)
ME-EOT N = 167 N = 151
Microbiological response at EOT visit
Favourable 156 (93.4) 144 (95.4) -2.0 (-8.3, 4.4)
Eradication 4 (2.4) 8 (5.3)
Presumed eradication 152 (91.0) 136 (90.1)
Unfavourable 11 (6.6) 7 (4.6)
Persistence 1 (0.6) 1 (0.7)
Presumed persistence 10 (6.0) 6 (4.0)
ME-PTE N = 163 N = 147
Overall microbiological response at PTE visit
Favourable 150 (92.0) 134 (91.2) 0.9 (-6.5, 8.5)
Eradication 1 (0.6) 3 (2.0)
Presumed eradication 149 (91.4) 131 (89.1)
Unfavourable 13 (8.0) 13 (8.8)
Persistence 1 (0.6) 1 (0.7)
Presumed persistence 12 (7.4) 12 (8.2)
CABP = community-acquired bacterial pneumonia, CI = confidence interval, CSR = clinical study report, EMA =
European Medicines Agency, EOT = end of treatment, ME = microbiologically evaluable,
microITT = microbiological intent-to-treat, PTE = post therapy evaluation.
Source: Study PTK0796-CABP-1200 EMA CSR.
Microbiological response (per-pathogen)

There was a numerically lower response rate for omadacycline against H. influenzae, when comparing
the overall microbiological response according to pathogen. The absolute numbers are small and
therefore must be interpreted with caution. Of the 4 omadacycline subjects with H. influenzae who had
an unfavourable microbiological response, one was indeterminate and one subject died before
evaluation.

EMA/595311/2019 Page 78/123

 Table 52. Overall Microbiological Favourable Response at PTE Visit by Baseline
Pathogen isolated in ≥ 10 Subjects, microITT Population, subjects with PORT Risk
Class III/IV (Study PTK0796-CABP-1200)

Source: Study PTK0796-CABP-1200 EMA CSR Table 21.

EMA/595311/2019 Page 79/123

Additional efficacy analyses

Mono- and poly-microbial infections

Table 53. Overall Microbiological Response at PTE Visit by Mono-microbial and Poly-
microbial Infection, microITT Population (Study PTK0796-CABP-1200)

Source: Study PTK0796-CABP-1200 EMA CSR Table 14.2.11.1.

Clinical outcome according to baseline pathogen and/or MIC

Generally, clinical success rates in CABP were high across S. pneumoniae, Enterobacteriacae and
atypical infections, regardless of resistance to other antibiotic classes. Slightly lower success rates
were seen with S. aureus and Haemophilus spp. infections. Clinical success rates were not
demonstrably correlated to omadacycline MIC, indeed clinical success was reported in omadacycline
subjects with baseline pathogens with MIC up to 16 µg/mL. Susceptibility data were not obtained for
atypical pathogens.

EMA/595311/2019 Page 80/123

 Table 54. Clinical Success at PTE by Pathogen and Study Drug Received MIC (µg/mL) in
CABP, microITT population, subjects with PORT Risk Class III/IV (Study PTK0796-
CABP-1200)

Omadacycline Moxifloxacin
(N=175) (N=156)
Baseline Clinical Success Baseline Clinical Success
Baseline Pathogen MIC (µg/mL) n/total (%) MIC (µg/mL) n/total (%)
Gram-positive organisms (aerobes)
Streptococcus pneumoniae 26 22
0.015 2/2 (100.0) ≤0.03 1/1 (100.0)
0.03 12/14 (85.7) 0.06 5/5 (100.0)
0.06 9/9 (100.0) 0.12 14/15 (93.3)
0.12 1/1 (100.0) 0.25 1/1 (100.0)
Penicillin-susceptible S. pneumoniae 25 20
0.015 2/2 (100.0) 0.06 5/5 (100.0)
0.03 12/14 ( 85.7) 0.12 13/14 ( 92.9)
0.06 8/8 (100.0) 0.25 1/1 (100.0)
0.12 1/1 (100.0) 0.5 0
Macrolide Resistant 10 5
0.03 6/6 (100.0) 0.06 2/2 (100.0)
0.06 3/3 (100.0) 0.12 2/2 (100.0)
0.12 1/1 (100.0) 0.25 1/1 (100.0)
Staphylococcus aureus 10 9
0.12 5/8 (62.5) 0.03 6/7 (85.7)
0.25 2/2 (100.0) 0.06 1/2 (50.0)
MSSA 10 8
0.12 5/8 (62.5) 0.03 5/6 (83.3)
0.25 2/2 (100.0) 0.06 1/2 (50.0)
Gram-negative organisms (aerobes)
Haemophilus influenzae 26 13
0.5 0 0.008 1/1 (100.0)
1 13/15 (86.7) 0.015 5/5 (100.0)
2 8/11 (72.7) 0.03 4/4 (100.0)
4 0 0.06 2/2 (100.0)
>4 0 1 1/1 (100.0)
Haemophilus parainfluenzae 13 12
≤0.008 0 ≤0.004 1/1 (100.0)
0.25 0 0.03 1/1 (100.0)
0.5 0 0.06 1/1 (100.0)
1 2/3 (66.7) 0.12 5/5 (100.0)
2 5/7 (71.4) 0.25 2/3 (66.7)
4 3/3 (100.0) 1 0/1 (0.0)
Klebsiella pneumoniae 12 11
1 2/2 (100.0) 0.06 4/4 (100.0)
2 5/7 (71.4) 0.12 5/7 (71.4)
4 0 0.25 0
8 1/1 (100.0) 0.5 0
16 2/2 (100.0) 1 0
Escherichia coli 4 6
0.5 1/1 (100.0) 0.03 1/1 (100.0)
1 1/2 (50.0) 0.06 2/3 (66.7)
2 0 0.12 0/1 (0.0)
4 0/1 (0.0) 0.5 0
>4 0 >4 1/1 (100.0)
Atypical Pathogensa NA NA
Mycoplasma pneumoniae 60/63 (95.2) 45/50 (90.0)
Legionella pneumophila 30/31 (96.8) 31/32 (96.9)
Serology 28/29 (96.6) 29/30 (96.7)

EMA/595311/2019 Page 81/123

 Omadacycline Moxifloxacin
(N=175) (N=156)
Baseline Clinical Success Baseline Clinical Success
Baseline Pathogen MIC (µg/mL) n/total (%) MIC (µg/mL) n/total (%)
Urinary Antigen 5/5 (100.0) 7/7 (100.0)
Chlamydia pneumoniae 23/26 (88.5) 22/24 (91.7)
MIC = minimum inhibitory concentration, microITT = microbiological intent-to-treat,
MSSA = methicillin-susceptible S. aureus, NA=not available.
Source: adapted from applicant’s Summary of Clinical Pharmacology Studies – Special Studies – Microbiology.

New and super-infections

Table 55. New and super-infections (Study PTK0796-CABP-1200)

Subject Treatment New or super-infection

K. pneumoniae (Day 3), super-infection
411-5003 4d iv
P. mirabilis (Days, 4, 5 6), super-infection and then new infection.
74F omadacycline
Note: subject died (Day 25) due to acute respiratory and multi-organ failure.
311-5012 5d iv
E. cloacae and K. oxytoca (Day 6), new infections
64F omadacycline
343-5029 10d iv 3d po
A. baumannii (Day 28), new infection
69F moxifloxacin
405-4020 6d iv 4d po
MRSA (Day 10), new infection
73M moxifloxacin
551-5009 5d iv
M. catarrhalis (Day 5), new infection
66F moxifloxacin
Table compiled by assessor. Source data: Study PTK0796-CABP-1200 EMA CSR 11.5.1.5.8.

Concordance of clinical outcome across visits

The omadacycline arm demonstrated numerically better continuity of cure than moxifloxacin, with
8/270 ECR successes deemed overall failures at PTE in the omadacycline, vs 15/279 ECR successes
deemed overall failures PTE in the moxifloxacin arm. The number of ECR successes later deemed
indeterminate (including subjects not clinically evaluable) was also lower in the omadacycline arm.

Sub-group analysis (PORT Risk Class)

Sub-group analysis revealed, perhaps unsurprisingly, slightly higher rates of clinical success amongst
subjects with PORT Risk Class IV vs III, but the lower limit of the two-sided 97.5% confidence interval
was higher than -10% both in ITT population and CE-PTE populations, for both Class III and IV:

Table 56. Overall Clinical Response at the PTE by PORT Risk Class, ITT and CE-PTE
Populations, subjects with PORT Risk Class III (Study PTK0796-CABP-1200)

Source: Applicant’s responses to D120 LoQ Q210.

EMA/595311/2019 Page 82/123

 Table 57. Overall Clinical Response at the PTE by PORT Risk Class, ITT and CE-PTE
Populations, subjects with PORT Risk Class IV (Study PTK0796-CABP-1200)

Source: Applicant’s responses to D120 LoQ Q210.

Sub-group analysis (prior antibiotic use)

Sub-group analysis revealed slightly lower rates of clinical success amongst subjects who had not
received a single dose of short-acting antibiotic prior to study randomisation, as might be expected,
however the difference was less dramatic in the omadacycline arm than the moxifloxacin arm:

Table 58. Overall Clinical Response at the PTE by PORT Risk Class, ITT Population,
subjects with PORT Risk Class III/IV (Study PTK0796-CABP-1200)

Source: Study PTK0796-CABP-1200 EMA CSR Table 25.

Sub-group analysis (other)

Sub-group analyses revealed no obvious differences between treatment arms according to geographic
region, baseline bacteraemia, baseline CURB-65 score and for subjects who switched to PO treatment
vs subjects who did not switch. Some numerical differences are noted, but small absolute numbers in
some of these groups precludes any robust conclusions.

3.3.5.2.4. Summary of main efficacy results

The following tables summarise the efficacy results from the main studies supporting the present
application. These summaries should be read in conjunction with the discussion on clinical efficacy as
well as the benefit risk assessment (see later sections).

Table 59. Summary of efficacy for pivotal studies

Title: A Phase 3 Randomised, Double-Blind, Multi-Centre Study to Compare the Safety and
Efficacy of Omadacycline IV/PO to Linezolid (Zyvox®) IV/PO for Treating Adult Subjects with
Acute Bacterial Skin and Skin Structure Infection (ABSSSI)

EMA/595311/2019 Page 83/123

 Study identifier PTK0796-ABSI-1108

1:1 randomised, active-controlled, double-blinded Phase 3 non-inferiority

study of omadacycline administered iv and po against comparator Linezolid
Design
administered iv and po in treatment of ABSSSI known or suspected to be
caused by Gram positive pathogens.

Hypothesis Non-inferiority (-10%)

100 mg iv q12h for 2 doses, then 100 mg iv q24h.

Omadacycline
Option to switch to 300 mg po q24h after a minimum
(n=329)
of 3 days.
Treatments Total duration 7-14 days.
groups
600 mg iv q12h.
Option to switch to 600 mg po q12h after a minimum
Linezolid (n=326)
of 3 days.
Total duration 7-14 days.

Overall clinical response at PTE (derived from IACR at

Endpoints Co-Primary endpoint
EOT and PTE) in the mITT and CE-PTE populations

Database lock 09 June 2016

Results and Analysis

Analysis Primary Analysis

Treatment group Omadacycline Linezolid

Number of subjects (mITT) 316 311

Overall clinical success at PTE in

86.1% 83.6%
mITT population

Co-primary 2.5 (95% CI -3.2, 8.1)

endpoint
Number of subjects (CE-PTE) 269 260

Overall clinical success at PTE in

96.3% 93.5%
CE-PTE population

2.8 (95% CI -0.9, 7.1)

Title: A Phase 3 Randomised, Double-Blind, Multi-Centre Study to Compare the Safety and
Efficacy of Oral Omadacycline to Oral Linezolid (Zyvox®) for Treating Adult Subjects with Acute
Bacterial Skin and Skin Structure Infection (ABSSSI)

Study identifier PTK0796-ABSI-16301

EMA/595311/2019 Page 84/123

 1:1 randomised, active-controlled, double-blinded Phase 3 non-inferiority
study of omadacycline administered po against comparator Linezolid
Design
administered po in treatment of ABSSSI known or suspected to be caused by
Gram positive pathogens.

Hypothesis Non-inferiority (-10%)

Omadacycline 450 mg po q24h for 2 doses, followed by 300 mg po

(n=368) q24h.
Treatments
Total duration 7-14 days.
groups

600 mg po q12h.
Linezolid (n=367)
Total duration 7-14 days.

Overall clinical response at PTE (derived from IACR at

Endpoints Co-Primary endpoint
EOT and PTE) in the mITT and CE-PTE populations

Database lock 30 June 2017

Results and Analysis

Analysis Primary Analysis

Treatment group Omadacycline Linezolid

Number of subjects (mITT) 360 360

Overall clinical success at PTE in

84.2% 80.8%
mITT population
Co-primary
3.3 (95% CI -2.2, 9.0)
endpoint

Number of subjects (CE-PTE) 284 292

Overall clinical success at PTE in

97.9% 95.5%
CE-PTE population

2.3 (95% CI -0.5, 5.8)

Title: A Phase 3 Randomised, Double-Blind, Multi-Centre, Non-inferiority Study to Compare the

Safety and Efficacy of Omadacycline IV/PO to Moxifloxacin IV/PO for Treating Adult Subjects with
Community-Acquired Bacterial Pneumonia (CABP)

Study identifier PTK0796-CABP-1200

1:1 randomised, double-blind, active-comparator-controlled, multi-centre

Phase 3 non-inferiority study comparing omadacycline administered iv and po
Design
against moxifloxacin administered iv and po for the treatment of adults with
CABP.

EMA/595311/2019 Page 85/123

 Hypothesis Non-inferiority (-10%)

100 mg iv q12h for 2 doses, then 100 mg iv q24h.

Omadacycline
Option to switch to 300 mg po q24h after a minimum
(n=329)
of 3 days.
Treatments Total duration 7-14 days.
groups
400 mg iv q24h .
Option to switch to 400 mg po q24h after a minimum
Moxifloxacin (n=331)
of 3 days.
Total duration 7-14 days.

Overall clinical response at PTE (derived from IACR at

Endpoints Co-Primary endpoint
EOT and PTE) in the ITT and CE-PTE populations

Database lock 24 March 2017

Results and Analysis

Analysis Primary Analysis

Treatment group Omadacycline Moxifloxacin

Number of subjects (ITT) 329 331

Overall clinical success at PTE in

88.4% 85.2%
ITT population
Co-primary
3.3 (97.5% CI -2.7, 9.3)
endpoint

Number of subjects (CE-PTE) 295 296

Overall clinical success at PTE in

92.5% 90.5%
CE-PTE population

2.0 (97.5% CI -3.2, 7.4)

3.3.6. Analysis performed across trials (pooled analyses)

Causative organisms – pooled Phase 3 ABSSSI studies (PTK0796-ABSI-1108 and PTK0796-

ABSI-16301)

Table 60. ABSSSI Subjects With Pathogens Identified at Baseline, most prevalent
organisms (microITT Population) (Pooled PTK0796-ABSI-1108 and PTK0796-ABSI-
16301)

Omadacycline Linezolid
(N=228) (N=227)
Baseline Pathogen n (%) n (%)
Gram-positive organisms (aerobes) 220 (96.5) 219 (96.5)
Staphylococcus aureus 156 (68.4) 151 (66.5)
MRSA 69 (30.3) 50 (22.0)

EMA/595311/2019 Page 86/123

 Omadacycline Linezolid
(N=228) (N=227)
Baseline Pathogen n (%) n (%)
MSSA 88 (38.6) 102 (44.9)
Streptococcus anginosus group 47 (20.6) 37 (16.3)
Gram-negative organisms (aerobes) 28 (12.3) 23 (10.1)
Other 20 (8.8)a 21 (9.3)b
Klebsiella pneumoniae 6 (2.6) 5 (2.2)
Enterobacter cloacae complex 6 (2.6) 4 (1.8)
Haemophilus parainfluenzae 5 (2.2) 5 (2.2)
Gram-positive organisms (anaerobes) 16 (7.0) 15 (6.6)
Gram-negative organisms (anaerobes) 17 (7.5) 13 (5.7)
ABSSSI = acute bacterial skin and skin structure infections, micro-mITT = microbiological modified intent-to-treat,
MRSA = methicillin-resistant Staphylococcus aureus, MSSA = methicillin-susceptible S. aureus, VSE =
vancomycin-susceptible enterococci.
a
Acinetobacter baumannii (1), A. baumannii complex (1), Eikenella corrodens (2), Enterobacter aerogenes (1), E.
cloacae (3), Escherichia coli (2), Klebsiella oxytoca (2), Morganella morganii (1), Proteus mirabilis (2),
Providencia stuartii (1), Pseudomonas aeruginosa (3), Serratia liquefaciens (1).
b
E. corrodens (2), E. aerogenes (3), E. cloacae (1), E. coli (3), Escherichia hermannii (1), K. oxytoca (1), M.
morganii (3), P. mirabilis (1), P. stuartii (2), P. aeruginosa (2), Stenotrophomonas maltophilia (2).
Source: adapted from applicant’s Summary of Clinical Pharmacology Studies – Special Studies – Microbiology.

Clinical outcome by baseline pathogen– pooled Phase 3 ABSSSI studies (PTK0796-ABSI-

1108 and PTK0796-ABSI-16301)

Generally, clinical success rates in ABSSSI were high regardless of baseline isolate species or
resistance to other antibiotic classes. Notably, clinical success rates with omadacycline were good even
for Gram negative species, although absolute numbers of such isolates are small and adjunctive Gram-
negative therapy, which may have impacted clinical efficacy, was permitted at the discretion of the
investigator in any subject diagnosed with a Gram negative or anaerobic organism (in practice this was
only applied to 9 subjects in study 1108, of which 5 were in the omadacycline group, and no subject in
16301.

Clinical outcome by baseline MIC– pooled Phase 3 ABSSSI studies (PTK0796-ABSI-1108 and
PTK0796-ABSI-16301)

Clinical success rates were not demonstrably correlated to omadacycline MIC, although this apparent
lack of relationship may depend on a high overall clinical success rate and very few isolates identified
with extremely high MIC values. Notably, clinical success rates with omadacycline were good (>80%)
even for species with MIC up to 4 µg/mL (although absolute numbers of isolates are small).

Increased MIC values and emergent resistance in Phase 3 studies

Emergence of resistance, defined as a 4-fold increase in MIC to test article in any post-baseline isolate,
was not observed in any of the three Phase 3 studies (TK0796-PTK0796-CABP-1200, PTK0796-ABSI-
1108 and PTK0796-ABSI-16301), although it should be noted that post-baseline cultures were not
commonly performed in any of the studies, therefore the value of this analysis is very limited.

3.3.7. Clinical studies in special populations

No dedicated clinical studies have been performed in special populations, however older subjects were
included in the main clinical efficacy studies.

EMA/595311/2019 Page 87/123

 Table 61. Age Groups (Safety Population, SC23 Pooling Group [ABSSSI+cSSI+CABP
Phase 2 and 3 Studies])
e
All
b
Omadacycline c
Linezolid d
Moxifloxacin Subjects
a
Characteristics (N=1252) (N=869) (N=388) (N=2509)
Age Group (years), n
(%)
n 1252 869 388 2509
18-64 1021 (81.5) 803 (92.4) 205 (52.8) 2029 (80.9)
65-74 120 (9.6) 36 (4.1) 95 (24.5) 251 (10.0)
75-84 83 (6.6) 28 (3.2) 70 (18.0) 181 (7.2)
≥85 28 (2.2) 2 (0.2) 18 (4.6) 48 (1.9)
Pooling group SC23 includes: ABSI-1108, ABSI-16301, cSSI-0702, cSSI-0804, CABP-1200.
The denominator for the percentage is the number of subjects who had that parameter assessed.
Source: Table 222.1

The Applicant has provided the requested data. The vast majority of subjects exposed to omadacycline
across the clinical programme were <64 years of age. CABP is a significant problem in elderly patients,
and the study population enrolled in study 1200 was reflective of this, comprising 41.9% >65 years of
age (20.4% >75 years of age). Meanwhile, ABSSSI is a condition predominantly affecting younger
patients, as reflected in the study populations enrolled in studies 1108 and 16301, which make up the
larger part of subject exposure in the Phase 3 programme, and which comprised just 10.5% and 4.6%,
respectively, >65 years of age. Thus, most of the older subjects exposed to omadacycline in the
clinical programme were CABP subjects, which is expected given the epidemiology of the different
conditions. Due to the overall size of the clinical programme, the absolute number of older subjects
exposed to omadacycline is reasonable (231 subjects >65 years old) and, due to the similarity in study
design, these subjects would all have received 7-14 days’ treatment according to protocol.

3.3.8. Supportive studies

Study PTK0796-CSSI-0702 was a 1:1 randomised, active-controlled, single (evaluator)-blinded, multi-

centre US Phase 2 safety and tolerability study of 7-14 days’ omadacycline iv and po against
comparator linezolid iv and po in treatment of adults with cSSSI. Efficacy assessments were a
secondary endpoint. The omadacycline dosing regimen was lower than that tested in the pivotal Phase
3 studies in ABSSSI and was not one of those testing in Monte Carlo simulation. Two thirds of
randomised subjects had a diagnosis of major abscess, but other subjects were enrolled with a
different spectrum of infections to those included in the pivotal Phase 3 studies in ABSSSI, including
infections associated with animal or human bites, second degree burns, removable foreign bodies and
acute lower-extremity ulcers, as well as cellulitis associated with comorbidity (diabetes mellitus,
arterial or venous insufficiency, immunosuppression). The rates of success according to IACR at TOC
were numerically higher for omadacycline both the mITT (98.3% vs 75.6%, difference 13.6 (95% CI
1.3, 26.0)) and CE (98% vs 93.2%, difference 4.8 (95% CI -1.7, 11.4)) populations. However, as this
study was not powered for statistical inference, no formal conclusion of non-inferiority can be made.
Furthermore, the CE and ME populations were smaller for linezolid, on account of a high number of
linezolid subjects who were not clinically evaluable at TOC, which may have driven the difference seen
between treatment arms.

Study PTK0796-CSSI-0804 was a 1:1 randomised, active-controlled, double-blinded, multi-centre US

Phase 3 study of 7-14 days’ omadacycline iv and po against comparator linezolid iv and po in
treatment of adults with cSSSI and evidence of a systemic inflammatory response, or a known,
specified reason for pre-existing immunosuppression. The dosing regimen was one of those tested in

EMA/595311/2019 Page 88/123

Monte Carlo simulation, but not the regimen used in the pivotal Phase 3 studies in ABSSSI. Two thirds
of randomised subjects had a diagnosis of cellulitis, while other major infection types were wound
infection and major abscess (capped at 20%), however, inclusion criteria allowed for infections
associated with animal or human bites, second degree burns and removable foreign bodies. Enrolment
was terminated prematurely by the sponsor following significant changes to the FDA guidance for
cSSSI. Furthermore, several post hoc changes were made to the analyses, including a change in
primary efficacy endpoint from evaluator-defined to Sponsor-defined clinical response. The rates of
Sponsor-defined success at TOC in the ITT and CE populations were high across all analysis
populations (>85%). However, as the study was terminated prematurely, is not powered for its pre-
defined efficacy endpoint.

As these studies applied different inclusion criteria, dose and efficacy endpoints vs the pivotal studies,
they offer very limited support for the overall efficacy of omadacycline in skin infections.

3.3.9. Discussion on clinical efficacy

ABSSSI

Design and conduct of main clinical studies

The pivotal Phase 3 trials for ABSSSI were generally appropriately designed to fulfil the requirements
for a new antibacterial substance according to current EMA guidance. The two dosing regimens used
are not fully supported by the currently presented PTA data , however this is considered mitigated by
the sufficient quantity of clinical data demonstrating good outcomes at the required MIC values.

The choice of linezolid as comparator was appropriate, as this is an antibiotic authorised across the EU
for ABSSSI caused by Gram positive organisms and available in iv and po formulations. The inclusion
criteria were appropriate and major abscess was capped at 30% of subjects, given the significant
therapeutic role of surgical drainage in major abscess. There was good representation of all three
major infection types (wound infection, cellulitis/erysipelas, major abscess). Other infection types,
such as infected leg ulcers, were excluded by the protocol. Only subjects with at least one Gram
positive causative pathogen at baseline were included in the efficacy analysis populations. Subjects
with pre-existing hepatic, renal, immunological and cardiac disease were excluded by the
protocol.Subjects were generally young (<65 years) and the majority of subjects were outpatients who
did not have bacteraemia or SIRS.

The primary objective (non-inferiority) and statistical approach (including NI margin of -10% and alpha
value of 0.05) were appropriate. A total treatment duration of 7-14 days was permitted (despite a
maximum of 10 days’ recommended in CHMP Scientific Advice), to match the authorised posology of
the chosen comparator, and allow flexibility for longer treatment of difficult infections. However, in
practice the majority of patients received 10 days’ or fewer treatment.

Although microbiological analyses were included as secondary efficacy analyses, as requested by

CHMP, the almost negligible proportion of subjects for whom post-baseline cultures were performed
(which is not unusual given that there is little left to sample when ABSSSI is successfully treated)
resulted in the vast majority of microbiological outcomes being “presumed” (i.e. linked to clinical
response) rather than “documented”, limiting the value of these secondary analyses.

Only a minor proportion of subjects in Study PTK0796-ABSI-1108 were enrolled in Europe and
PTK0796-ABSI-16301 was entirely US-based. Sub-group analyses revealed no obvious difference in
infection type or pathogen organism, or the efficacy of omadacycline, by geographic region. Clinical
success with omadacycline was high and comparable to linezolid in both EU and non-EU subjects,

EMA/595311/2019 Page 89/123

although only an approximate numerical comparison is possible. Overall, the results are considered
generalisable to the EU population.

Pregnant women and children <18 years old were also excluded from clinical studies in ABSSSI (a PIP
has previously been agreed) were excluded from the clinical studies in ABSSSI. This is reflected in the
proposed Product Information.

Efficacy data and additional analyses

In studies PTK0796-ABSI-1108 and PTK0796-ABSI-16301, clinical response rates were high across
both treatment arms regardless of baseline pathogen or omadacycline MIC, >80% in the mITT
population and >90% in the CE-PTE population. Both studies demonstrated non-inferiority of
omadacycline compared to linezolid against a pre-defined margin of -10% for the EMA co-primary
efficacy endpoints of overall clinical response at PTE (derived from IACR at EOT and PTE) in the mITT
and CE-PTE populations. The results of all sensitivity analyses, including analysis of the co-primary
endpoint in the all-treated population (as recommended by CHMP Scientific Advice) and analysis
treating indeterminate clinical response as clinical success (to assess the impact of missing data) were
supportive of the primary analysis. Most patients also underwent a protocol-allowed surgical procedure
(which is standard practice in the treatment of ABSSSI). There are no obvious differences in efficacy
trends between subjects who did and did not receive surgical intervention.

CABP

Design and conduct of main clinical studies

Study PTK0796-CABP-1200 was designed as a single pivotal study in CABP. The dosing regimen used
is not supported by the currently presented PTA data.

The study population (ITT=660) comprised moderate two thirds PORT Risk Class III and one third
PORT Risk Class IV in inpatients (mean age >60 years) with moderate to severe CABP and clinically
significant medical comorbidities. The dominant pathogens were M. pneumoniae (one third of
subjects), S. pneumoniae (around 20%), and H. influenzae (around 12%), with over half of subjects
having an atypical causative organism. The majority (>85%) of subjects were recruited in Europe,
making this generally a good representation of EU patients with CABP.

EMA guidance recommends that for studies of IV administration, a minimum of 25% (and ideally 50%)
of patients are in class IV-V, accepting that PORT Risk Class V patients requiring ICU admission may be
excluded. Subjects with PORT Risk Class V were excluded from the single pivotal trial in CABP on the
basis of significant differences in clinical condition and management vs lower PORT Risk Classes,
according to the Applicant.

The choice of moxifloxacin as comparator was appropriate, as this is an antibiotic authorised across the
EU for CABP and available in iv and po formulations. The inclusion criteria were appropriate, requiring
3 symptoms, 2 vital signs, 1 clinical or laboratory sign, and radiographic confirmation of CABP.
Subjects with CABP known to be caused by a pathogen resistant to either test article (including K.
pneumoniae) were excluded by the protocol. Subjects with pre-existing hepatic, renal, immunological
and cardiac disease, and; subjects with pre-existing (and possible pre-disposing) respiratory conditions
(including lung neoplasm, aspiration, CF, active TB, bronchiectasis, COPD) or chronic neurological
disorders affecting airway clearance, were excluded by the protocol. The majority of subjects did not
have bacteraemia (<5%).

The primary objective (non-inferiority) and statistical approach (including NI margin of -10% and alpha
value of 0.025) were appropriate. A total treatment duration of 7-14 days was permitted (despite a

EMA/595311/2019 Page 90/123

maximum of 10 days’ recommended in CHMP Scientific Advice), to match the authorised posology of
the chosen comparator, and allow flexibility for longer treatment of difficult infections. However, in
practice the majority of patients received 10 days’ or fewer treatment.

The vast majority of microbiological outcomes were “presumed” (i.e. linked to clinical response) rather
than “documented”, limiting the value of the microbiological analyses as secondary analyses.

Pregnant women and children <18 years old were also excluded from clinical studies in ABSSSI (a PIP
has previously been agreed) were excluded from the clinical studies in ABSSSI. This is reflected in the
proposed Product Information.

Efficacy data and additional analyses

Clinical response rates were high, >85% in the ITT population and >90% in the CE-PTE population.
The study demonstrated non-inferiority of omadacycline against a pre-defined margin of -10% for the
EMA co-primary efficacy endpoints of overall clinical response at PTE in the ITT and CE-PTE
populations. The results of sensitivity analysis were supportive of the primary analysis.

Sub-group analysis revealed a numerically higher rate of clinical success in subjects with PORT Risk
Class III vs IV, and better rates of clinical success amongst subjects who had not received a single
dose of short-acting antibiotic prior to study randomisation in the omadacycline arm.

Clinical success rates in CABP were high regardless of baseline isolate species or resistance to other
antibiotic classes. Clinical success rates were not demonstrably correlated to omadacycline MIC.
Susceptibility data were not obtained for atypical pathogens.

Conclusions on clinical efficacy

The applicant has presented the results of two pivotal studies in ABSSI in adults and one pivotal study
in CABP in adults, all three of which met their primary objective of demonstrating non-inferiority
against an authorised comparator according to pre-specific primary analysis agreed with CHMP. A
Phase 2 study and a sponsor-terminated Phase 3 study, both in cSSSI, offer only very limited support.

Overall, there is sufficient evidence from two adequately-sized clinical studies for good clinical
outcomes in ABSSSI at suitable MIC values, despite lack of comprehensive PK-PD support for efficacy
of the proposed dose.

However, the evidence for the CAP indication comes solely from a single pivotal study. These
deficiencies mean that Study PTK0796-CABP-1200 has not fulfilled the requirements of a single pivotal
study, the objective of which is to confirm an established hypothesis (MO).

3.3.10. Clinical safety

Safety data were derived from 27 studies in the frame of the phase 1 to 3 programme. The evaluation
of omadacycline safety is specifically based on the safety data available from 5 studies, i.e. 1 phase 2
study (CSSI-0702) and 4 phase 3 studies (CSSI-0804, ABSI-1108, ABSI-16301 and CABP-1200),
including 1252 subjects who received omadacycline. These studies included a total of 705 subjects who
received the intravenous (iv) to oral (po) regimen in each targeted indication, as well as 368 subjects
who received the oral-only regimen for the treatment of ABSSSI.

Pooling strategy

The integrated analysis of omadacycline safety data is based on 5 pooled populations:

1) 2 pivotal Phase 3 studies performed in ABSSSI patients (A3 Pool)

2) Pivotal Phase 3 studies in ABSSSI combined with 2 supportive studies in cSSSI (S23 Pool)

EMA/595311/2019 Page 91/123

3) the pivotal Phase 3 study investigating CABP (C3 Pool)
4) the 3 pivotal Phase 3 studying ABSSSI and CABP (AC3 Pool)
5) all studies pool (SC23 Pool) comprising the 2 supportive studies in cSSSI and the 3 pivotal Phase
3 studies in ABSSSI and CABP
Table S1 Studies incorporated in the 5 study pools used for the integrated safety analysis
Omadacycline Linezolid Moxifloxacin
All Doses 600 mg 400 mg
Analysis Pool (iv + po) (iv + po) (iv + po) Total
Phase 3 ABSSSI studies (A3) 691 689 - 1380
ABSI-1108 323 322 - 645
ABSI-16301 368 367 - 735
Phase 2/3 ABSSSI and cSSSI studies
870 869 - 1739
(S23)
ABSI-1108 323 322 - 645
ABSI-16301 368 367 - 735
CSSI-0702 111 108 - 219
CSSI-0804 68 72 - 140
Phase 3 CABP study (C3) 382 - 388 770
CABP-1200 382 - 388 770
Phase 3 ABSSSI and CABP studies (AC3) 1073 689 388 2150
ABSI-1108 323 322 - 645
ABSI-16301 368 367 - 735
CABP-1200 382 - 388 770
Phase 2/3 ABSSSI, cSSSI, and CABP
1252 869 388 2509
studies (SC23)
ABSI-1108 323 322 - 645
ABSI-16301 368 367 - 735
CSSI-0702 111 108 - 219
CSSI-0804 68 72 - 140
CABP-1200 382 - 388 770
Source: Clinical Safety Summary Table 2 – ISS Table 14.1.1.1

Patient exposure

The mean duration for treatment in the phase 2/3 programme was comparable between treatment
arms (9 days for omadacycline, 8.5 days for linezolid, 9.6 days for moxifloxacin). Similar treatment
durations are observed when stratifying according to the way of application (iv/oral switch therapy in
studies ABSI-1108 and CABP-1200, oral only therapy in ABSI-16301).

Table S2 Extent of exposure in the phase 2/3 program with omadacycline

Category Omada Omada Linezolid Omada Omada Linezolid Moxiflox

SC23 S23 S23/SC23 AC3 A3 A3/AC3 CABP (C3)
n (%) n (%) n (%) n (%) n (%) n(%) n (%)
Treated (n) 1252 870 869 1073 691 689 388
Completed study 1142 (91.2) 786 772 (88.8) 971 (90.5) 615 604 (87.7) 362 (93.3)
(n/%) (90.3) (89.0)

Prematurely 110 (8.8) 84 (9.7) 97 (11.2) 102 (9.5) 76 (11.0) 85 (12.3) 26 (6.7)
discontinued (n/%)
due to
AE 9 (0.7) 2 (0.2) 3 (0.3) 8 (0.7) 1 (0.1) 1 (0.1) 9 (2.3)
Lost to follow-up 51 (4.1) 51 (5.9) 64 (7.4) 48 (4.5) 48 (6.9) 56 (8.1) 3 (0.8)
Withdrawal by 29 (2.3) 22 (2.5) 18 (2.1) 27 (2.5) 20 (2.9) 16 (2.3) 8 (2.1)
subject
Physician decision 1 (0.1) 1 (0.1) 2 (0.2) 1 (0.1) 1 (0.1) 2 (0.3) 1 (0.3)
Death c) 6 (0.5) 0 2 (0.2) 6 (0.6) 0 2 (0.3) 3 (0.8)
Other 14 (1.1) 8 (0.9) 8 (0.9) 12 (1.1) 6 (0.9) 8 (1.2) 2 (0.5)

Mean duration 9.2 (2.99) 9.0 (3.02) 8.9 (3.34) 9.0 (2.87) 8.7 (2.81) 8.5 (2.96) 9 . 6 (2.95)
treatment (SD)
Mean duration iv 5.0 (2.32) 4.4 (2.00) 4.6 (2.66) 5.0 (2.35) 4.3 (1.89) 4.4 (2.10) 5.7 (2.54)
therapy (SD)
Mean duration oral 6.4 (2.59) 6.9 (2.64) 6.8 (2.79) 6.4 (2.66) 7.0 (2.74) 6.8 (2.81) 5.2 (2.04)
therapy (SD)

EMA/595311/2019 Page 92/123

Source: ISS, Table 14.1.3.5.

Adverse events

In the All Studies Pool (SC23) approximately 50% of subjects reported AEs. Drug related TEAEs
occurred in approximately 30% of the subjects in the ABSSSI+cSSSI Phase 2 and 3 Studies (S23 pool)
vs approximately 10 % in the CABP study (C3 pool). Serious TAES and TEAEs leading to
discontinuation of the treatment occurred more often in individuals treated for pneumonia compared to
subjects treated for ABSSSI/cSSSI. Of note, although overall low in frequency, TEAEs leading to death
are more often observed in the omadacycline arm of the CABP study compared to active comparator.

Table S5A Overview of Treatment-Emergent Adverse Events in the different safety pools

Category Omada Omada Linezolid Omada Omada Linezolid Moxiflox

SC23 S23 SC23/S23 AC3 A3 A3 N=388
N=1252 N=870 N=869 N=1073 N=691 N=689 n (%)
n (%) n (%) n (%) n (%) n (%) n(%)
Subjects with 611 (48.8) 454 (52.2) 397 (45.7) 510 (47.5) 353 (51.1) 284 (41.2) 188 (48.5)
any TEAE
n (%) with:
Drug-related 301 (24.0) 262 (30.1) 185 (21.3) 236 (22.0) 197 (28.5) 111 (16.1) 69 (17.8)
TEAE
Serious TEAE 43 (3.4) 20 (2.3) 16 (1.8) 39 (3.6) 16 (2.3) 13 (1.9) 26 (6.7)
Drug-related 2 (0.2) 0 1 (0.1) 2 (0.2) 0 1 (0.1) 2 (0.5)
serious TEAE
TEAE leading to 9 (0.7) 1 (0.1) 3 (0.3) 8 (0.7) 0 3 (0.4) 4 (1.0)
death
TEAE leading to 35 (2.8) 14 (1.6) 12 (1.4) 33 (3.1) 12 (1.7) 10 (1.5) 27 (7.0)
premature
discontinuation of
the article
TEAE leading to 3 (0.2) 3 (0.3) 0 2 (0.2) 2 (0.3) 0 0
dose interruption
of test article
Serious TEAEs 17 (1.4) 7 (0.8) 5 (0.6) 16 (1.5) 6 (0.9) 5 (0.7) 11 (2.8)
leading to
premature
discontinuation of
test article

Source: Table 7 Clinical summary, ISS Table 14.3.1.1.1-5

Table S5B Overview of Treatment-Emergent Adverse Events in the CABP phase III study

Omadacycline Moxifloxacin
Number of Patients (%) with: N = 382 N = 388
Any TEAE 157 (41.1) 188 (48.5)
Drug-related TEAE 39 (10.2) 69 (17.8)
Serious TEAE 23 (6.0) 26 (6.7)
Drug-related serious TEAE 2 (0.5) 2 (0.5)
TEAE leading to premature discontinuation of test article 21 (5.5) 27 (7.0)
TEAE leading to dose interruption of test article 0 0
Serious TEAEs leading to premature discontinuation of test article 10 (2.6) 11 (2.8)
Patients who died 8 (2.1) 4 (1.0)

Source: Clinical Overview, Table 15

The most common adverse events associated with omadacycline are nausea (15.3%), vomiting
(8.0%), headache (4.2%), ALT increased (3.7%), AST increased (3.0%), diarrhoea (2.6%), cellulitis
(2.4%), wound infection (2.4%), constipation (2.2%) and infusion site extravasation (2.2%) (SC23
pool). Elevations of ALT, AST and GGT were common observations in the S23 and C3 study pool with
omadacycline. Infusion site reactions are seen in very low frequency with omadacycline.

EMA/595311/2019 Page 93/123

Heart rate related TEAEs including tachycardia (0.7%, SC23 pool), atrial fibrillation (0.2%, SC23 pool)
and palpitations (0.5%, SC23 pool) are seen in omadacycline treated subjects. Those events occurred
in low frequency and were only slightly more often observed in the omadacycline group compared to
comparator treated subjects (frequencies of 0.5%/0.1%/0%, respectively, in the Linezolid arm of the
SC23 pool).

Allergic reactions (pruritus, rash) were rarely seen with omadacycline and in a comparable frequency
to linezolid or moxifloxacin.

Class side effects of tetracyclines, such as phototoxicity or pseudotumor cerebri, are not observed with
omadacycline. No cases of C. difficile were identified with omadacycline. Furthermore, lipase (and
amylase) elevations are observed with omadacycline (included in the SmPC).

Table S7A. Summary of Treatment-Emergent Adverse Events (TEAEs) ≥2% by Preferred Term in the
Safety Population S23 ABSSSI Pooling Group (ABSSSI+cSSSI Phase 2 and 3 Studies)

Preferred Term (PT) Omadacycline Linezolid Risk Difference

(n=870) (N=869) Omadacycline – Linezolid
n (%) n (%) (95% CI)
Subjects with at least one TEAE 454 (52.2) 397 (45.7) 6.7 (2.05, 11.29)
Nausea 182 (20.9) 87 (10.0) 11.0 (7.62, 14.34)
Vomiting 90 (10.3) 42 (4.8) 5.5 (3.06, 8.03)
Headache 45 (5.2) 34 (3.9) 1.3 (-0.64,3.25)
Alanine aminotransferase increased 32 (3.7) 36 (4.1) -0.5 (-2.29,1.36)
Wound infection 30 (3.4) 23 (2.6) 0.8 (-0.81,2.40)
Aspartate aminotransferase increased 29 (3.3) 31 (3.6) -0.2 (-1.96,1.48)
Cellulitis 29 (3.3) 26 (3.0) 0.3 (-1.29,1.99)
Diarrhoea 28 (3.2) 40 (4.6) -1.4 (-3.17,0.47)
Infusion site extravasation 28 (3.2) 20 (2.3) 0.9 (-0.59,2.42)
Subcutaneous abscess 23 (2.6) 27 (3.1) -0.5 (-2.03,1.09)
Blood creatine phosphokinase increased 19 (2.2) 9 (1.0) 1.2 (-0.01,2.35)
Constipation 18 (2.1) 9 (1.0) 1.0 (-0.11,2.20)
Dizziness 18 (2.1) 17 (2.0) 0.1 (-1.16,1.44)
Source: ISS Table 14.3.1.4.2

Table S7B. Summary of Treatment-Emergent Adverse Events (TEAEs) ≥2% by Preferred Term in the
Safety Population C3 CABP (Phase 3 Study)

PT Omadacycline Moxifloxacin Risk Difference

N = 382 N = 388 Omada - Moxifloxacin
n (%) n (%) (95% CI)
Subjects with at least one TEAE 157 (41.1) 188 (48.5) -7.4 (-14.36, -0.35)
ALT increased 14 (3.7) 18 (4.6) -1.0 (-3.79, 1.84)
GGT increased 13 (3.4) 11 (2.8) 0.6 (-1.89, 3.02)
Insomnia 10 (2.6) 8 (2.1) 0.6 (-1.58, 2.69)
Vomiting 10 (2.6) 8 (2.1) 0.6 (-1.58, 2.69)
Constipation 9 (2.4) 6 (1.5) 1.1 (-0.95, 2.09)
Nausea 9 (2.4) 21 (5.4) 0.8 (-1.15, 2.76)
AST increased 8 (2.1) 14 (3.6) -3.1 (-5.77, -0.34)
Headache 7 (1.8) 3 (0.8) -1.5 (-3.86, 0.83)
Anaemia 7 (1.8) 3 (0.8) 0.8 (-1.02, 2.63)

Source: ISS Table 14.3.1.4.3

In the phase 1 studies nausea, vomiting, diarrhoea and ALAT increased were primarily observed.
Furthermore transient dose-dependent increases of HR were seen in relation with omadacycline with
resolved within 6 hours.

AE by severity
Mild and moderate TEAEs occurred slightly more often in the omadacycline arm of the ABSSSI studies
(32.3% omadacycline, 26.0% linezolid and 17.1% omadacycline, 12.8% linezolid). Moderate TEAEs

EMA/595311/2019 Page 94/123

 occurred less often in omadacycline treated subjects compared to moxifloxacin in the CABP study
(11.0% omadacycline, 17.0% moxifloxacin). Severe TEAEs were equally distributed between test drug
and control arm in the ABSSSI studies (1.7% of subjects in the omadacycline group and 2.5% of
subjects in the linezolid group) and the CABP study (6.5% of subjects in the omadacycline group and
6.7% of subjects in the moxifloxacin group).

Drug related TEAEs

The most frequently reported drug-related TEAEs in the A3 ABSSSI Pooling Group (ABSSSI Phase 3
Studies) in the omadacycline arm were nausea (18.5%) and vomiting (9.6%), ALT increased (3.2%),
AST increased (2.7%) and diarrhoea (2.3%).
In the ABSSSI Phase 3 studies, the frequency of drug-related TEAEs in the omadacycline group was
higher (28.5%) compared to the omadacycline group in Study PTK0796-CABP-1200 (10.2%). The
most frequently reported drug-related TEAEs (≥ 1% for any group) by PT that occurred at a higher
percentage (≥ 1% difference) in the omadacycline group were nausea (1.8% CABP, 18.5% ABSSSI)
and vomiting (0% CABP, 9.6% ABSSSI). All other drug-related TEAEs (≥ 1% for any group) were
similar between treatment groups for both indications. The higher drug related TEAEs for the C3 safety
pool was due to greater drug-related nausea and vomiting in the PTK0796 ABSI-16301 study.
Furthermore rates of liver enzyme elevations were also higher in the A3 pool compared to CABP (ALT
3.2% vs 1.8%; AST: 2.7% vs. 1.3%)

AEs by route of administration

Nausea and vomiting occur more often in cases where omadacycline is given po rather than iv.
Gastrointestinal events are known and generally manageable tetracycline class effects. Based on the
food interaction PK studies revealing decreased exposure to omadacycline when administered following
a meal the Applicant proposed the administration of this medicinal product under fasting condition
which is agreed. The proposed precaution and listed AEs related to the gastrointestinal disorders in
SmPC are considered sufficient at this time. Finally, the Applicant claims that the impact of a light meal
on the tolerability is currently evaluated and therefore is asked to provide such safety data (LoQI).

Table S9. Number (%) of Subjects With the Most Frequent TEAEs (≥ 2% for Any Group) Overall and by PT
(Studies ABSI-1108 and CABP-1200)
Omadacycline Linezolid Moxifloxacin
iv po Overall iv po Overall iv po Overall
N = 705 N = 581 N = 705 N = 322 N = 283 N = 322 N = 388 N = 294 N = 388
PT n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%)
Subjects
226 165 313 100 147 143 188
with at least 77 (27.2) 74 (25.2)
(32.1) (28.4) (44.4) (31.1) (45.7) (36.9) (48.5)
1 TEAE
Infusion site
28 (4.0) 0 28 (4.0) 19 (5.9) 0 19 (5.9) 0 0 0
extravasation
Nausea 16 (2.3) 33 (5.7) 49 (7.0) 22 (6.8) 11 (3.9) 32 (9.9) 10 (2.6) 11 (3.7) 21 (5.4)
Hypertension 15 (2.1) 5 (0.9) 19 (2.7) 3 (0.9) 2 (0.7) 4 (1.2) 8 (2.1) 3 (1.0) 11 (2.8)
ALT increased 13 (1.8) 10 (1.7) 23 (3.3) 1 (0.3) 13 (4.6) 14 (4.3) 17 (4.4) 1 (0.3) 18 (4.6)
Headache 12 (1.7) 7 (1.2) 18 (2.6) 8 (2.5) 5 (1.8) 13 (4.0) 5 (1.3) 0 5 (1.3)
Vomiting 11 (1.6) 16 (2.8) 27 (3.8) 12 (3.7) 6 (2.1) 16 (5.0) 2 (0.5) 5 (1.7) 6 (1.5)
AST increased 10 (1.4) 6 (1.0) 16 (2.3) 3 (0.9) 10 (3.5) 12 (3.7) 13 (3.4) 1 (0.3) 14 (3.6)
Insomnia 8 (1.1) 5 (0.9) 13 (1.8) 4 (1.2) 0 4 (1.2) 8 (2.1) 0 8 (2.1)
Subcutaneous
8 (1.1) 10 (1.7) 17 (2.4) 14 (4.3) 5 (1.8) 19 (5.9) 0 0 0
abscess
Cellulitis 6 (0.9) 10 (1.7) 16 (2.3) 5 (1.6) 10 (3.5) 15 (4.7) 0 0 0
Diarrhoea 6 (0.9) 6 (1.0) 11 (1.6) 5 (1.6) 5 (1.8) 10 (3.1) 18 (4.6) 14 (4.8) 31 (8.0)
Source: ISS Tables 14.3.1.9.1.1 and Table 14.3.1.9.1.2

EMA/595311/2019 Page 95/123

Table S10. Number (%) of Subjects With the Most Frequent TEAEs (≥ 2% for Any Group) Overall and by PT
(Study ABSI-16301 (oral only administration))
Omadacycline Linezolid
(N = 368) (N = 367)
Body System n (%) n (%)
Subjects with at least 1 TEAE 197 (53.5) 137 (37.3)
Nausea 111 (30.2) 28 (7.6)
Vomiting 62 (16.8) 11 (3.0)
Wound infection 22 (6.0) 17 (4.6)
ALT increased 19 (5.2) 11 (3.0)
AST increased 17 (4.6) 12 (3.3)
Diarrhoea 15 (4.1) 10 (2.7)
Headache 13 (3.5) 8 (2.2)
Cellulitis 12 (3.3) 9 (2.5)
Abdominal pain upper 10 (2.7) 4 (1.1)
Subcutaneous abscess 6 (1.6) 8 (2.2)
Source: Study ABSI-16301, Table 14.3.1.1.2.1

AEs of special interest

1. Hepatic events

Liver enzyme and bilirubin elevations are typical class effects of tetracyclines and were expected to be
seen. Elevations of AST, ALT, GGT and Bilirubin were observed with omadacycline. The frequencies of
these events were roughly comparable with frequencies seen in the comparator arms in the different
study pools. The great majority of these events were mild in intensity. Only in one case, where
omadacycline was evaluated as causative for liver enzyme elevations, the treatment with the test drug
had to be discontinued. No cases of liver failure or cases fulfilling Hy´s law were detected as being
caused by omadacycline. The Applicant has included elevations of GGT, AST, ALT (common) and
bilirubin and ALP (uncommon) in the SmPC.

Table S13. Summary of Hepatic Events of Interest by PT (AC3 Pool)

Omadacycline Linezolid Moxifloxacin

Category N = 1073 N = 689 N = 388
PT n (%) n (%) n (%)
Liver-related investigations, signs, and
58 (5.4) 34 (4.9) 28 (7.2)
symptoms
ALT increased 42 (3.9) 25 (3.6) 18 (4.6)
AST increased 33 (3.1) 24 (3.5) 14 (3.6)
GGT increased 15 (1.4) 8 (1.2) 8 (2.1)
Blood bilirubin increased 5 (0.5) 1 (0.1) 3 (0.8)
Blood ALP increased 3 (0.3) 1 (0.1) 4 (1.0)
Hypoalbuminemia 2 (0.2) 0 3 (0.8)
Hepatic enzyme increased 1 (0.1) 0 0
Hepatic congestion 0 0 1 (0.3)
Biliary system-related investigations,
7 (0.7) 2 (0.3) 6 (1.5)
signs, and symptoms
Blood bilirubin increased 5 (0.5) 1 (0.1) 3 (0.8)
Blood ALP increased 3 (0.3) 1 (0.1) 4 (1.0)
Source: ISS Table 14.3.1.12.4.

2. Gastrointestinal Events

Nausea and vomiting were the most abundant side effects with omadacycline, and occurred especially
during the oral application of the drug. The events were of mild or moderate intensity.

The applicant has included nausea (very common), vomiting (common), diarrhoea (common) in the
SmPC, which is endorsed. Other GI events that have been observed with omadacycline include
constipation (2.1%), abdominal pain (1.3%, included in the SmPC as common), and dyspepsia (0.8%,
included in the SmPC as uncommon) which occurred in similar frequencies as with the comparator.

EMA/595311/2019 Page 96/123

Table S14. Summary of GI Events of Interest by PT (AC3 Pool)

Omadacycline Linezolid Moxifloxacin

N = 1073 N = 689 N = 388
PT n (%) n (%) n (%)
Nausea 160 (14.9) 60 (8.7) 21 (5.4)
Vomiting 89 (8.3) 27 (3.9) 6 (1.5)
Diarrhoea 26 (2.4) 20 (2.9) 31 (8.0)
Constipation 16 (1.5) 5 (0.7) 6 (1.5)
Abdominal pain upper 15 (1.4) 6 (0.9) 2 (0.5)
Abdominal pain 9 (0.8) 5 (0.7) 2 (0.5)
Dyspepsia 7 (0.7) 7 (1.0) 2 (0.5)
Dry mouth 3 (0.3) 6 (0.9) 0
Toothache 3 (0.3) 0 0
Pooling group AC3 includes: Studies ABSI-1108, ABSI-16301, and CABP-1200. PT, for each PT term;
Source: Summary of clinical safety, Table 40

3. Heart Rate and Cardiac events

The frequency of subjects with at least 1 TEAE within the Cardiac disorders SOC in the pooled AC3
pool, were highest in the moxifloxacin group (5.2%), followed by 2.1% in the omadacycline group and
0.4% in the linezolid group. The Applicant has investigated HR and cardiac events of interest by PT
(see Table S15). Only slight numeric differences are observed for PTs belonging to the category cardiac
arrest and tachyarrhythmias in the CABP study (C3 pool, table below, updated/completed table to be
found in the answer to question 233). The Applicant has given a detailed description on
cerebrovascular events in the different study pools compared to comparator arm. That is: 3 subjects
had a cerebrovascular event in the omadacycline arm, 2 in the CABP study (discussed in question
230), in the A3 pool and 1 event of transient dysarthria together with transient hemiparesis was
detected in the phase 2 ABSSSI studies. These events were probably a result of transient progression
of Parkinson’s disease (discussed in question 230). No event was detected in the comparator arms,
leading to a slightly higher overall frequency of 2/vs 0/ in the SC3 pool. Those two cases are
characterised by older subjects with underlying cardiovascular multi-morbidity, which predisposed to
the strokes the occurred under study run. A clear causative relation to study drug treatment can
therefore not be drawn.

Table S15. Summary of HR and Cardiac Events of Interest by PT (A3 and C3 Pool)
A3 (ABSSSI) C3 (CABP)
Omadacycline Linezolid Omadacycline Moxifloxacin Risk Difference
Category N = 691 N = 689 N = 382 N = 388 Omada (CABP)-
PT n (%) n (%) n (%) n (%) Moxi (95%)
Cardiac arrest 0 1 (0.1) 4 (1.0) 1 (0.3) 0.8 (-0.35, 1.93)
Cardiogenic shock 1 (0.1) 2 (0.5) 1 (0.3)
Cardiac arrest ng 1 (0.3) 0
Cardio-respiratory ng 1 (0.3) 0
arrest
Cardiac failure 0 2 (0.3) 6 (1.6) 5 (1.3) 0.3 (-1.4, 1.96)
Cardiac failure 1 (0.1) 3 (0.8) 3 (0.8)
Cardiogenic shock 1 (0.1) 2 (0.5) 1 (0.3)
Acute pulmonary ng 1 (0.3) 0
edema
Cardiac failure ng 0 1 (0.3)
congestive
Hepatic ng 0 1 (0.3)
congestion
Right ventricular ng 0 1 (0.3)
failure
Ischemic heart disease 0 1 (0.1) 3 (0.8) 4 (1.0) -0.2 (-1.58, 1.09)

EMA/595311/2019 Page 97/123

Table S15. Summary of HR and Cardiac Events of Interest by PT (A3 and C3 Pool)
A3 (ABSSSI) C3 (CABP)
Omadacycline Linezolid Omadacycline Moxifloxacin Risk Difference
Category N = 691 N = 689 N = 382 N = 388 Omada (CABP)-
PT n (%) n (%) n (%) n (%) Moxi (95%)
Acute myocardial ng 2 (0.5) 0
infarction
Myocardial 1 (0.1) 1 (0.3) 2 (0.5)
ischemia
Angina pectoris ng 0 2 (0.5)
Tachiarrhythmias 3 (0.4) 0 3 (0.8) 1 (0.3) 0.5 (-0.49, 1.55)
Atrial fibrillation 2 (0.3) 0 3 (0.8) 1 (0.3)
Atrial flutter Ng ng 1 (0.3) 0
Sinus tachycardia 1 (0.1) 0 Ng ng
ng = value not displayed in the ISS table
Source: ISS Table 14.3.1.12.1. and 14.3.1.12.3.

The applicant has added a 3 endpoint MACE analysis for both the AC3 and the SC23 pool. Results
confirm preliminary observations of numeric imbalances for the events cerebrovascular events and
cardiac infarction events on an overall low frequency level and show balanced counts in cardiac
induced deaths. Events occurred in patients often showing several cardiovascular risk factors.
Numerical imbalances for MI almost disappear when additionally taking cases of elevated CPK into
account (1.2% (O) vs 0.9% (L) vs 1.0% (M), AC3 group). At this stage, MACE analysis does not
provide further explanation for imbalances in mortality seen with omadacycline.

Post-hoc table 1: Summary of Cerebrovascular, Myocardial Infarction Treatment-Emergent

Adverse Events (TEAEs), and Cardiac Deaths AC3 Pooling Group (ABSSSI+CABP Phase 3
Studies) – Safety Population

4. Tetracycline Class Related Events

• anti-anabolic events as represented by blood urea nitrogen (BUN) increased and azotemia
• central nervous system side effects including light-headedness, vertigo or dizziness
• hypersensitivity
• photosensitivity
• pseudotumor cerebri
• acute pancreatitis
• fungal infections, in particular vulvovaginal fungal infections.
• Pill esophagitis is noted with tetracyclines, in particular doxycycline

EMA/595311/2019 Page 98/123

Table S16. Summary of Tetracycline Class Events of Interest by PT (AC3 Pool)
Omadacycline Linezolid Moxifloxacin
Category N = 1073 N = 689 N = 388
PT n (%) n (%) n (%)
Hypersensitivity reactions 20 (1.9) 12 (1.7) 10 (2.6)
Pruritus 8 (0.7) 1 (0.1) 1 (0.3)
Rash 6 (0.6) 3 (0.4) 5 (1.3)
Urticaria 3 (0.3) 0 0
Dermatitis 1 (0.1) 1 (0.1) 0
Hypersensitivity 1 (0.1) 1 (0.1) 4 (1.0)
Rash pustular 1 (0.1) 1 (0.1) 0
Swelling face 1 (0.1) 1 (0.1) 0
Angioedema 0 1 (0.1) 0
Bronchospasm 0 0 1 (0.3)
Drug eruption 0 1 (0.1) 0
Infusion site urticaria 0 0 1 (0.3)
Pruritus generalized 0 1 (0.1) 0
Rash generalized 0 1 (0.1) 0
Fungal infections 11 (1.0) 6 (0.9) 5 (1.3)
Oral candidiasis 5 (0.5) 1 (0.1) 1 (0.3)
Vulvovaginal mycotic 3 (0.3) 5 (0.7) 0
infection
Fungal skin infection 1 (0.1) 0 0
Oesophageal candidiasis 1 (0.1) 0 1 (0.3)
Vulvovaginal candidiasis 1 (0.1) 0 0
Oral fungal infection 0 0 1 (0.3)
Respiratory moniliasis 0 0 2 (0.5)
Vestibular disorders 9 (0.8) 6 (0.9) 4 (1.0)
Dizziness 7 (0.7) 6 (0.9) 4 (1.0)
Vertigo 2 (0.2) 0 0
Blood urea increased 1 (0.1) 0 0
Blood urea increased 1 (0.1) 0 0
Oesophageal disorders 1 (0.1) 0 0
Esophagitis 1 (0.1) 0 0
Pancreatitis 1 (0.1) 0 1 (0.3)
Pancreatitis chronic 1 (0.1) 0 0
Pancreatic pseudocyst 0 0 1 (0.3)
C. difficile infection 0 0 8 (2.1)
C. difficile colitis 0 0 1 (0.3)
C. difficile infection 0 0 6 (1.5)
Pseudomembranous colitis 0 0 1 (0.3)
Pooling group AC3 includes: Studies ABSI-1108, ABSI-16301, and CABP-1200.
Source: ISS Table 14.3.1.12.4.

• Anti-anabolic events as represented by blood urea nitrogen (BUN) increased and azotemia
The TEAE of blood urea increased occurred in 1 (0.1%) omadacycline subject in the AC3 pool (Subject
1200-555 5003; mild, unrelated to test article, resolved) and no linezolid or moxifloxacin subject.

• Central nervous system side effects including light-headedness, vertigo or dizziness

Vertigo and dizziness occurred in < 2% of subjects and were comparable between treatment groups

• Hypersensitivity
Hypersensitivity reactions occurred in 1.9% of omadacycline subjects, 1.7% of linezolid subjects, and
2.6% of moxifloxacin subjects, none of them serious in the omadacycline arm.

A total of 9 subjects (1 omadacycline, 4 linezolid, 4 moxifloxacin) had a hypersensitivity event that led
to test article discontinuation. Pruritus and rash have been included in the SmPC as uncommon events
under “skin reactions”. hypersensitivity reaction has been included in the SmPC (uncommon), which is
in principal endorsed.

In the pivotal phase 3 trials (AC3 pool) the preferred term “erythema” was uncommon; treatment
emergent adverse events of infusion site erythema, application site erythema, and penile erythema
were not included in determining the frequency of this adverse event. “Erythema” occurred in 6 (0.6%)

EMA/595311/2019 Page 99/123

of OMC patients, 3 (0.4% linezolid patients, and in 0 moxifloxacin patients. In only 1 omadacycline
patient was the erythema considered related to the study drug however the number of cases was
double in the omadacycline treatment group relative to the comparators and skin reactions are
plausible and not infrequent following antibiotic administration. The Applicant has suggested including
Erythema as an uncommon ADR in the SmPC, which is endorsed

• Photosensitivity
No omadacycline subjects reported photosensitivity. Study subjects had a certain amount of sun
exposure as studies were run in sun-rich areas such as e.g. California. Photosensitivity has been
included as a warning in the SmPC (class effect), which is endorsed.

• Pancreatitis

One subject (Subject 1200-373-5004) in the omadacycline group had a TEAE of chronic pancreatitis on
Day 5 that was considered mild in severity and not related to test article. No action was taken with
test article and the event was considered ongoing.

• Fungal infections, in particular vulvovaginal fungal infections

The incidence of fungal infections was low, occurring in 11 (1.0%) omadacycline subjects, 6 (0.9%)
linezolid subjects, and 5 (1.3%) moxifloxacin subjects. These events were considered mild or moderate
in severity and did not lead to test article discontinuation or dose interruption.

• Pill esophagitis is noted with tetracyclines, in particular doxycycline

Esophagitis was reported in only 1 (0.1%) omadacycline subject (Subject 1200-414-5001) during iv
treatment. The event was mild, not related, and resolved. Oesophageal ulceration was not reported.

• C. difficile infection
C. difficile infection cases occurred only in the moxifloxacin group (2.1%).

Serious adverse events and deaths

An imbalance in the mortality rate is noted in the CABP study between omadacycline (2.1%) and
moxifloxacin (1.0%) (95% CI -0.7 to 2.8). The point estimate for the omadacycline and moxifloxacin
arms in CABP-1200 study is within the reported range of other CAP/CAPB studies, however this is not
entirely reassuring, as the very reason for a randomised comparator arm is to calibrate the experiment
and provide a reference for the numerical findings in the experimental arm. Overall, there were 17
deaths (10 omadacycline, 3 linezolid, 4 moxifloxacin) reported in the five Phase 2 and 3 studies.
Across the omadacycline clinical development programme, there were 10 (0.8%) omadacycline and 7
(0.6%) comparator subjects who died. Among the comparator deaths, 4 (1.0%) occurred on
moxifloxacin and 3 (0.4%) occurred on linezolid.

Table S17. Listing of Deaths

Last Study Date of Relationship to
Age/ Study Medication Death Study Medication
Subject ID Gender Medication PT (Study Day) (Study Day) (study physician)
Study CSSI-0804
25-MAR-2010 19-APR-2010
804-811-4135 51/M Omadacycline Pleural effusion Not related
(Day 10) (Day 35)
Lung cancer 25-MAR-2010 19-APR-2010
Not related
metastatic (Day 10) (Day 35)
Study ABSI-1108
08-AUG-
07-AUG-2015
1108-262-0035 60/M Omadacycline Overdose 2015 Not related
(Day 1)
(Day 2)

EMA/595311/2019 Page 100/123

Table S17. Listing of Deaths
Last Study Date of Relationship to
Age/ Study Medication Death Study Medication
Subject ID Gender Medication PT (Study Day) (Study Day) (study physician)
26-MAR-
21-MAR-2016
1108-124-0005 88/M Linezolid Cardiac failure 2016 Not related
(Day 7)
(Day 12)
16-JAN-2016 18-JAN-2016
1108-253-0018 43/M Linezolid Cardiac arrest Not related
(Day 7) (Day 9)
Study ABSI-16301
16301-604- 10-JAN-2017 APR-2017
62/F Linezolid Death Not related
3056 (Day 10) (unknown)

Study CABP-
1200
Cerebrovascular 25-OCT-2016 31-OCT-2016
1200-241-5019 68/M Omadacycline Not related
accident (Day 7) (Day 13)
Aortic 29-AUG-
28-AUG-2016
1200-244-5008 72/M Omadacycline aneurysm 2016 Not related
(Day 8)
rupture (Day 9)
01-FEB-2016 02-FEB-2016
1200-311-5005 67/M Omadacycline Septic shock Related
(Day 1) (Day 2)
22-JAN-2016 12-FEB-2016
1200-334-5001 86/F Omadacycline Pneumonia Not related
(Day 9) (Day 30)
Acute
respiratory 22-JAN-2016 12-FEB-2016
Not related
distress (Day 9) (Day 30)
syndrome
Cardiogenic 13-APR-2016 20-APR-2016
1200-343-5001 90/F Omadacycline Not related
shock (Day 13) (Day 20)
Cardio-
21-FEB-2016 21-FEB-2016
1200-405-5015 76/M Omadacycline respiratory Related
(Day 2) (Day 2)
arrest
Acute
05-DEC-2016 26-DEC-2016
1200-411-5003 74/F Omadacycline respiratory Not related
(Day 4) (Day 25)
failure
Multi-organ 05-DEC-2016 26-DEC-2016
Not related
failure (Day 4) (Day 25)
Acute
16-OCT-2016 16-OCT-2016
1200-551-5007 66/M Omadacycline myocardial Not related
(Day 2) (Day 2)
infarction
31-DEC-2015 31-DEC-2015
1200-307-5006 83/M Moxifloxacin Cardiac failure Not related
(Day 9) (Day 9)
Acute
28-APR-2016 04-MAY-2016
1200-334-5005 85/F Moxifloxacin respiratory Not related
(Day 3) (Day 9)
failure
05-DEC-2016 11-DEC-2016
1200-354-5001 82/M Moxifloxacin Lung neoplasm Not related
(Day 14) (Day 20)
Pancreatic 24-NOV-2016 26-JAN-2017
1200-355-5003 72/F Moxifloxacin Not related
carcinoma (Day 8) (Day 71)
Source: ISS Listing 16.2.7.3.

EMA/595311/2019 Page 101/123

Mortality

Figure 7. Forest Plot of Mortality in the SC23 Pool

Source: Clinical Overview, Figure 1

Mortality in ABSSSI

In all Phase 2 and 3 studies of skin infections (S23 Pool), the mortality rate was 0.2% for
omadacycline versus 0.3% for linezolid (Figure 7).

Mortality in CABP

The demographic and baseline factors for the omadacycline and moxifloxacin-treated subjects who
died generally describe a subject with CABP and an elevated moderate risk of mortality (age over 65
years, PORT Risk Class IV, with underlying cardiovascular and pulmonary comorbidities); 6 of 8
omadacycline subjects who died, and 2 of 4 moxifloxacin subjects who died, were PORT Risk Class IV.

To determine if there was a numerical trend for specific mortality-associated TEAEs (TEAEs with the
outcome of death) that could represent a potential biologically plausible relationship to treatment,
“mortality-associated TEAEs by PTs” were examined amongst subjects who survived Study CABP-1200.

Table S18 Summary of Mortality-associated TEAEs by PT in Subjects Who Did Not Die in Study CABP-
1200 (C3 Pool, Safety Population)

Omadacycline Moxifloxacin
PT (N = 374) (N = 384)
Acute myocardial infarction 1 0
Cardiogenic shock 1 1
Cardiac failure 3 2
Cardio-respiratory arrest 0 0
Pneumonia 3 7
Septic shock 0 2
Acute respiratory distress syndrome 1 0
Acute respiratory failure 1 2
Aortic aneurysm rupture 0 0
Multi-organ failure 0 0
Cerebrovascular accident 1 0
Lung neoplasm 7 2
Pancreatic carcinoma 0 0
Source: Clinical Overview, Appendix, Table 21

Similar numbers of events were observed between the omadacycline and moxifloxacin groups for all of
the mortality associated TEAE PTs, including terms related to myocardial ischemia, heart failure, and
progression of pneumonia, including respiratory failure.

EMA/595311/2019 Page 102/123

An additional analysis was conducted to determine if the omadacycline subjects who died in Study
CABP 1200 had significant changes in HR. Changes in HR for subjects who died were of small
magnitude (< 10 bpm) and well within the observed population based median values for HR change in
subjects treated with either omadacycline (who survived) or moxifloxacin (who survived or died). QT
interval corrected for HR using Fridericia’s formula (QTcF) was not increased in omadacycline subjects
across the ABSSSI and CABP populations studied (SCS, Table 30). No imbalances were seen between
treatment arms in cardiac TEAE (myocardial infarction, myocardial ischemia, heart failure, and
tachyarrhythmias) in Study CABP 1200 (ISS, Table 14.3.1.12.3, TEAEs of interest, Safety table S15) or
across the integrated safety database (SCS, Table 20).

Since a lack of efficacy could result in increased mortality, additional analyses were conducted to
determine whether efficacy was decreased in important subgroups: 1) higher probability of mortality
(eg, PORT Risk Class); 2) higher severity (eg, SMART-COP); 3) subjects meeting sepsis criteria, and 4)
baseline bacteraemia). The results of these subgroup population analyses demonstrated similar
efficacy between omadacycline and moxifloxacin.

To assess whether early failure to either antibiotic therapy could result in the observed numerical
difference in mortality, achievement of early clinical stability criteria was analysed in the omadacycline
and moxifloxacin treatment groups. These data demonstrated a high and similar percentage of
subjects in both treatment groups (89%) who successfully achieved early clinical stability, as defined
by meeting all 5 of the stabilization criteria and did not vary appreciably between PORT Risk Class III
and IV. The high percentages are consistent with the high level of efficacy observed at the ECR
assessment.

Finally, inspection of individual microbiology data revealed no pathogen consistently associated with
the progression of the incident pneumonia to death.

Serious AEs
The incidence of SAEs was overall low in the omadacycline and the comparator arms. A systematic
occurrence of a specific type of SAEs is not observed in the omadacycline group.

3 cases of cerebrovascular accidents/hemiparesis plus dysarthria were observed in the omadacycline

group in the SC23 pool. Scrutiny of submitted narratives allowed the conclusion that those events are
not clearly associated with study drug treatment (1 case of Parkinson’s disease and strokes in two
older subjects with underlying cardiovascular-morbidity)

Table S21 Patients with at least one serious TEAE – SC23 safety pool
System Organ Class (SOC) Omada Linezolid Moxifloxacin
Preferred Term (PT) (N=1252) (N=869) (N=388)
n (%) n (%) n (%)

Subjects with at Least One Serious TEAE 43 (3.4) 16 (1.8) 26 (6.7)

Cardiac disorders 5(0.4) 2 (0.2) 2(0.5)
Acute myocardial infarction 2(0.2) 0 0
Cardiogenic shock 2(0.2) 0 1(0.3)
Cardiac arrest 1(0.1) 1 (0.1) 0
Cardiac failure 1(0.1) 1 (0.1) 1(0.3)
Cardio-respiratory arrest 1(0.1) 0 0
Tachycardia 1(0.1) 0 0
Pericardial effusion 0 0 1(0.3)
Right ventricular failure 0 0 1(0.3)
Gastrointestinal disorders 1(0.1) 0 1(0.3)
Small intestinal obstruction 1(0.1) 0 0
Colitis 0 0 1(0.3)

EMA/595311/2019 Page 103/123

General disorders and a dministration site 3(0.2) 1 (0.1) 0
conditions
Drug withdrawal syndrome 1(0.1) 0 0
Multi-organ failure 1(0.1) 0 0
Non-cardiac chest pain 1(0.1) 0 0
Death 0 1 (0.1) 0
Hepatobiliary disorders 3(0.2) 0 1(0.3)
Cholecystitis acute 2(0.2) 0 0
Hepatic failure 1(0.1) 0 0
Hepatic congestion 0 0 1(0.3)
Infections and infestations 21(1.7) 8 (0.9) 16(4.1)
Cellulitis 3(0.2) 3 (0.3) 0
Influenza 3(0.2) 0 0
Subcutaneous abscess 3(0.2) 0 0
Wound infection 3(0.2) 2 (0.2) 0
Pneumonia 2(0.2) 0 6(1.5)
Bacteraemia 1 (0.1) 0 0
Gastroenteritis 1 (0.1) 0 0
Gastroenteritis rotavirus 1 (0.1) 0 0
Hepatitis C 1 (0.1) 0 0
Infectious pleural effusion 1 (0.1) 0 1(0.3)
Septic shock 1 (0.1) 0 2(0.5)
Staphylococcal bacteraemia 1 (0.1) 0 0
Atypical mycobacterial pneumonia 0 0 1(0.3)
Clostridium difficile colitis 0 0 1(0.3)
Clostridium difficile infection 0 0 2(0.5)
HIV infection 0 0 1(0.3)
Infective exacerbation of bronchiectasis 0 0 1(0.3)
Lung abscess 0 0 1(0.3)
Pneumonia viral 0 0 1(0.3)
Sepsis 0 3 (0.3) 0
Skin infection 0 1 (0.1) 0
Injury, poisoning and procedural
2(0.2) 1 (0.1) 1 (0.3)
complications
Joint dislocation 1(0.1) 0 0
Overdose 1(0.1) 1 (0.1) 0
Bladder injury 0 0 1(0.3)
Metabolism and nutrition disorders 1(0.1) 0 0
Decreased appetite 1(0.1) 0 0
Musculoskeletal and connective tissue
1(0.1) 0 0
disorders
Back pain 1(0.1) 0 0
Neoplasms benign, malignant and 3(0.2) 0 6 (1.5)
unspecified (incl c y s t s a n d polyps)
Lung neoplasm 2(0.2) 0 2 (0.5)
Lung neoplasm 2 (0.2) 0 2 (0.5)
Lung cancer metastatic 1(0.1) 0 0
Adenocarcinoma 0 0 11 (0.3)
Chronic lymphocytic leukaemia 0 0 1 (0.3)
Colon cancer metastatic 0 0 1 (0.3)
Pancreatic carcinoma 0 0 1 (0.3)
Nervous system disorders 3(0.2) 0 0
Cerebrovascular accident 2(0.2) 0 0
Hemiparesis 1(0.1) 0 0
Psychiatric disorders 3(0.2) 1 (0.1) 0
Anxiety 1(0.1) 0 0
Confusional state 1(0.1) 0 0
Depression 1(0.1) 0 0
Drug abuse 0 1 (0.1) 0
Renal and urinary disorders 0 0 2 (0.5)
Renal failure acute 0 0 2 (0.5)

EMA/595311/2019 Page 104/123

Respiratory, thoracic and 9(0.7) 2 (0.2) 3 (0.8)
mediastinal disorders
Acute respiratory failure 3(0.2) 0 3 (0.8)
Pleural effusion 3(0.2) 0 0
Acute respiratory distress syndrome 2(0.2) 0 0
distress syndrome
Acute p u l m o n a r y o e d e m a 1(0.1) 0 0
Chronic obstructive pulmonary disease 0 1 (0.1) 0
Pulmonary embolism 0 1 (0.1) 0
Skin and subcutaneous tissue disorders 0 1 (0.1) 0
tissue disorders
Angioedema 0 1 (0.1) 0
Vascular disorders 1(0.1) 0 1 (0.3)
Aortic aneurysm rupture 1(0.1) 0 0
Peripheral ischemia 0 0 1 (0.3)

Source: ISS Table 14.3.1.7.5

Laboratory findings

Haematology
In general, no clinically meaningful changes from baseline were observed which were related in a
negative way to omadacycline treatment. Leukocyte and neutrophil counts, which decreased after
baseline in both treatment groups, and platelets which increased after baseline in the omadacycline
and moxifloxacin treatment groups are consistent with clinical improvement during treatment.
However, in several cases lower neutrophil counts were observed with omadacycline (>1% with a >=
2 Grade decrease). Five omadacycline subjects had neutrophil counts below 0.5. Adverse events
associated with infection were limited to a single case. Cases of neutropenia were partly caused of
underlying infection, in one case abundant at baseline, very short/ transient of nature and occurred in
two cases at PTE. No clear association with treatment can be drawn.

The investigator reported 1% of cases of anaemia as TEAE in omadacycline subjects. The Applicant has
included anaemia as common event in the SmPC. Thrombocytes did not decrease significantly in
omadacycline treated subjects and suggested including thrombocytosis as an uncommon event in the
SmPC. As requested, the Applicant further analysed the frequency of thrombocytosis in the AC3 pool.
Data presented demonstrate that 1% of cases show at least one event of moderate platelet increase
under treatment with omadacycline. The frequency is higher compared to linezolid (0.1%) and lower
compared to moxifloxacin (2.1%). It is noted, the moxifloxacin lists thrombocythemia as a common
occurring event in the ADR list section 4.8. Furthermore, a clear treatment to event relationship
regarding shifts in platelet counts is seen with omadacycline in four cases experiencing a TEAE of
thrombocytosis. Therefore, it is suggested to include thrombocytosis as often occurring event in the
ADR list, section 4.8 of the SmPC.

Table S22 Summary of TEAEs by SOC and PT in the Phase 3 ABSSSI and CABP Studies (AC3 Pool)

Omadacycline Linezolid Moxifloxacin

SOC (N = 1073) (N = 689) (N = 388)
PT n (%) n (%) n (%)
Subjects with at Least 1 TEAE 510 (47.5) 284 (41.2) 188 (48.5)
Blood and lymphatic system 20 (1.9) 11 (1.6) 7 (1.8)
disorders
Anaemia 12 (1.1) 4 (0.6) 3 (0.8)
Thrombocytosis 4 (0.4) 1 (0.1) 0
Source: Summary Table 40

EMA/595311/2019 Page 105/123

Clinical Chemistry
Meaningful changes of lipase (5.9%) and amylase (2.1%) are seen with omadacycline. Frequencies
are roughly comparable with those seen with linezolid or moxifloxacin. No case of pancreatitis was
observed with omadacycline. The applicant has included lipase elevations as uncommon event in the
SmPC. Urea nitrogen elevations (class effect of tetracyclines, 4 cases showing a shift ≥ 2 in the
omadacycline arm of the SC23 pool) and creatinine elevations have been rarely observed with
omadacycline and were comparable to the rate seen with comparator treatment. No significant shifts in
electrolytes are observed with omadacycline.
Table S23. Chemistry Results: Subjects With Incidence of at Least 2 Grade Change From Baseline
(AC3 Pool)
Omadacycline Linezolid Moxifloxacin
N = 1073 N = 689 N = 388
Number
Number of Number of of
Subjects Subjects Subjects
With a With a With a
Subjects With at Least a Grade 2 or Grade 2 or Grade 2
2 Grade Change From Less at Less at or Less at
Baseline in: n (%) Baseline n (%) Baseline n (%) Baseline
Renal 4 (0.4) 1049 3 (0.5) 666 5 (1.3) 382
Urea nitrogen 1 (0.1) 1042 0 663 2 (0.5) 380
Creatinine 4 (0.4) 1049 3 (0.5) 666 5 (1.3) 382
Liver 65 (6.2) 1049 42 (6.3) 666 41 (10.7) 382
ALP 2 (0.2) 1039 1 (0.2) 663 3 (0.8) 377
ALT 41 (3.9) 1049 24 (3.6) 665 21 (5.5) 381
AST 32 (3.1) 1049 24 (3.6) 665 16 (4.2) 381
GGT 25 (2.5) 1019 15 (2.3) 653 18 (4.9) 366
Total bilirubin 7 (0.7) 1047 2 (0.3) 666 6 (1.6) 381
Electrolytes 25 (2.4) 1044 16 (2.4) 665 23 (6.0) 381
Calcium hypocalcaemia
1 (0.1) 1039 0 663 3 (0.8) 377
(corrected for albumin)
Calcium hypercalcaemia
0 1038 0 663 0 377
(corrected for albumin)
Magnesium
5 (0.5) 1041 4 (0.6) 664 2 (0.5) 380
(hypomagnesaemia)
Potassium (hypokalaemia) 2 (0.2) 1034 2 (0.3) 651 8 (2.1) 376
Potassium
8 (0.8) 1034 2 (0.3) 651 8 (2.1) 376
(hyperkalaemia)
Sodium (hyponatraemia) 4 (0.4) 1041 5 (0.8) 662 2 (0.5) 377
Sodium (hypernatraemia) 7 (0.7) 1041 3 (0.5) 662 6 (1.6) 377
Other 149 (14.3) 1043 77 (11.6) 665 63 (16.5) 381
Amylase 22 (2.1) 1037 9 (1.4) 659 6 (1.6) 377
Blood glucose
11 (1.1) 1036 7 (1.1) 655 3 (0.8) 377
(Hypoglycaemia)
Blood glucose
55 (5.5) 1003 29 (4.6) 635 20 (5.7) 353
(hyperglycaemia)
Lipase 61 (5.9) 1037 27 (4.1) 658 22 (5.8) 379
Phosphate
14 (1.4) 1026 12 (1.8) 654 19 (5.1) 374
(hypophosphatemia)
Uric Acid
8 (0.8) 1036 2 (0.3) 661 8 (2.1) 378
(hyperuricaemia)
Pooling group AC3 includes: Studies ABSI-1108, ABSI-16301, and CABP-1200.
Source: ISS Table 14.3.3.6.4.

Liver values
Five omadacycline subjects (Subjects 1108-120-0004, 1200-307-5018, 16301-601-3038, 16301-604-
3055, and 16301-608-3095; no subject in the comparator arms) met Hy’s Law laboratory criteria
assessments (defined as an ALT or AST > 3 × ULN) and total bilirubin > 2 × ULN and ALP < 2 × ULN)
at any time post-Baseline in the AC3 pool. Of these, 2 (0.2%) omadacycline subjects
(Subjects 16301-604-3055 and 16301-608-3095) met the Hy’s law laboratory criteria assessment at
the same visit post-Baseline. These patients had a medical history including concomitant medication
able to induce liver enzyme elevations, ALT elevations evaluated as not related to test drug by the
physician (1), drug abuse and hepatitis B infection (1), drug abuse and hepatitis C infection (2), liver

EMA/595311/2019 Page 106/123

enzyme elevations already at screening (1). Taking this into consideration, the view of the Applicant
that these cases do not represent drug induced liver injury seems plausible.

Changes in AST and ALT > x10 ULN occurred rarely in and were comparable between the different
arms. Increases in total bilirubin and ALP were comparable between treatment groups. Increases in
ALT, AST, GGT and bilirubin have been included in the SmPC which is endorsed.

Laboratory related TEAEs including ALT and AST increased and total bilirubin were comparable between
treatment groups. No laboratory-related TEAEs were considered serious. Treatment emergent AEs that
led to discontinuation of test article occurred in 2 omadacycline subjects (Subjects 1200-332-5002 and
1200-376-5046) and 2 moxifloxacin subjects (Subjects 1200-311-5010 and 1200-420-5004) who
discontinued due to TEAEs of ALT and/or AST increased.
Table S24. Liver Chemistry Elevations for Subjects with Normal Baseline Values (AC3 Pool)
Omadacycline Linezolid Moxifloxacin
N = 1073 N = 689 N = 388
Lab Parameter (SI unit) Parameter n (%) n (%) n (%)
ALT (U/L)
Normal at Baseline, N1 772 537 295
Meeting criterion at post-Baseline > 3 × ULN 13 (1.7) 18 (3.4) 11 (3.7)
> 5 × ULN 7 (0.9) 4 (0.7) 1 (0.3)
> 10 × ULN 5 (0.6) 3 (0.6) 0
AST (U/L)
Normal at Baseline, N1 858 561 328
Meeting criterion at post-Baseline > 3 × ULN 13 (1.5) 16 (2.9) 5 (1.5)
> 5 × ULN 8 (0.9) 6 (1.1) 1 (0.3)
> 10 × ULN 3 (0.3) 1 (0.2) 0
ALT or AST (U/L)
Normal at Baseline, N1 718 494 271
Meeting criterion at post-Baseline > 3 × ULN 15 (2.1) 19 (3.8) 12 (4.4)
> 5 × ULN 9 (1.3) 6 (1.2) 1 (0.4)
> 10 × ULN 5 (0.7) 3 (0.6) 0
Total bilirubin (μmol/L)
Normal at Baseline, N1 937 593 364
Meeting criterion at post-Baseline > 1.5 × ULN 7 (0.7) 2 (0.3) 6 (1.6)
> 2 × ULN 5 (0.5) 1 (0.2) 4 (1.1)
ALP (U/L)
Normal at Baseline, N1 819 523 311
Meeting criterion at post-Baseline > 2 × ULN 5 (0.6) 1 (0.2) 6 (1.9)
ALT or AST and total bilirubin
assessed at same visit post-
Baseline
Normal at Baseline, N1 639 434 254
Meeting criterion at post-Baseline ALT or AST > 3 × ULN, total
2 (0.3) 0 0
bilirubin > 2 × ULN
ALT or AST > 5 × ULN, total
2 (0.3) 0 0
bilirubin > 2 × ULN
ALT or AST > 10 × ULN, total
2 (0.3) 0 0
bilirubin > 2 × ULN
ALT or AST and total bilirubin
assessed at any time post-
Baseline
Normal at Baseline, N1 639 434 254
Meeting criterion at post-Baseline ALT or AST > 3 × ULN, total
2 (0.3) 0 0
bilirubin > 2 × ULN
ALT or AST > 5 × ULN, total
2 (0.3) 0 0
bilirubin > 2 × ULN
ALT or AST > 10 × ULN, total
2 (0.3) 0 0
bilirubin > 2 × ULN
ALP and total bilirubin assessed
at same visit post-Baseline
Normal at Baseline, N1 739 484 300
Meeting criterion at post-Baseline ALP > 2 × ULN, total bilirubin
1 (0.1) 0 1 (0.3)
> 2 × ULN
Source: ISS Table 14.3.3.8.4.2.

Vital signs

EMA/595311/2019 Page 107/123

No significant impact of omadacycline on SBP, DBP and ECG parameters PR, RR, QRS and QT values
was noted with omadacycline in the phase 2/3 program. Analyses of the mean change over time in HR
showed that HR tended to decline slightly in all categories at both the EOT and PTE visits, although the
decline over time was less rapid for omadacycline subjects compared to moxifloxacin subjects.

Cardiac Safety report (phase 1 study)

PTK 0796-TQTC-0803 was a randomised, placebo-controlled, double blind, double-dummy TQTc cross-
over study conducted in 64 healthy subjects who received single iv doses of PTK 0796 100 mg, 300
mg, placebo, and 400 mg moxifloxacin po in separate treatment periods.

Effect on Heart Rate

A single iv dose of 100 mg and 300 mg PTK 0796 resulted in a clear effect on heart rate with the
largest mean placebo-corrected ΔHR observed at early time-points (16.8 bpm at 35 minutes post-dose
after the 100 mg dose and 21.6 bpm at 50 minutes after dosing with 300 mg). Mean ΔΔHR thereafter
declined but remained above 5 bpm during 6 hours after the omadacycline 100 mg dose and for the
full observation period up to 22 hours after the 300 mg dose.

Omadacycline did not have an effect on cardiac conduction as shown for PR, QRS and ATcS intervals.

The thorough cardiac safety and QTc study TR701 115 evaluated omadacycline for its potential to
induce QT interval prolongations in healthy subjects and indicated no concern.

Figure 8. Placebo-Corrected ΔHR Across Post-Dosing Timepoints, PTK0796-TQTC-0803

Source: Figure 7.2.4.1 and 7.2.5.1 Cardiac Safety report

EMA/595311/2019 Page 108/123

 Figure 9. Panel B: Placebo-Corrected ΔQTcS (ΔΔQTcS) Across Post-Dosing Timepoints
in PTK0796-TQTC-0803

Figure 7.2.4.6 and 7.2.5.6 Cardiac Safety report

Safety in special populations

Age

All age groups, < 65 years, > 65 to 75 years and > 75 years of age had a similar frequency of TEAEs
in the omadacycline arm (48%, 49.6% and 48%, respectively) with gastrointestinal side effects as the
most abundant events. The age group < 65 years of age showed the highest rate of nausea and
vomiting. The Applicant has provided a more detailed table for the different type and severity stages of
AEs for the age groups < 65 years, 65-74 years, 75-84 years and >= 85 years which did not reveal
any safety concerns

Diabetes

Omadacycline subjects who reported at least 1 TEAE was comparable in subjects with a history of
diabetes compared to subjects without a history of diabetes. The most notable event between
omadacycline subjects with and without a history of diabetes (based on a ≥ 5% difference) was
nausea, which occurred in a higher percentage of omadacycline subjects without a history of diabetes.
In addition, TEAEs of diarrhoea were reported by a higher percentage of subjects with a history of
diabetes (10.5%) than those without diabetes (2.8%) in the omadacycline group in the A3 Pool. No
differences in diarrhoea were observed in the C3 Pool.

COPD/asthma

The percentage of subjects who had at least 1 TEAE was higher in subjects with a history of
asthma/COPD (56.5% omadacycline, 55.3% moxifloxacin) compared to subjects without a history of
asthma/COPD (36.7% omadacycline, 46.8% moxifloxacin). The few notable events (based on a ≥ 5%
difference) were as follows:

• Constipation events occurred in a higher percentage of subjects with a history of asthma/COPD

(8.2% omadacycline, 3.9% moxifloxacin) compared to subjects without a history of asthma/COPD
(0.7% omadacycline, 1.0% moxifloxacin).

EMA/595311/2019 Page 109/123

• Nausea events occurred in a higher percentage of subjects with a history of asthma/COPD (7.1%
omadacycline, 6.6% moxifloxacin) than subjects without a history of asthma/COPD (1.0%
omadacycline, 5.1% moxifloxacin).

• Vomiting events occurred in a higher percentage of omadacycline subjects with a history of

asthma/COPD (7.1% omadacycline, 1.3% moxifloxacin). Subjects without a history of
asthma/COPD had similar rates of vomiting in the omadacycline and moxifloxacin groups.

Hepatic impairment

Study CPTK796A-2201 investigated subjects with varying degrees of hepatic impairment (mild,
moderate and severe). The primarily affected SOCs were nervous system disorder, gastrointestinal
disorders, general disorders and administration site conditions. Headache (13.3%), nausea (6.7%),
infusion site pain (6.7%), contusion (6.7%) and dizziness (6.7%) were the most commonly
experienced AEs. No death occurred in this study. One subject experienced SAEs of severe intensity
related to alcohol poisoning (intoxication), angina pectoris, hypocalcaemia, hypotension and
rhabdomyolysis. None of the events were suspected to be related to the study medication. Another
subject was discontinued from the study due to an AE of mild rash, which was considered possibly
related to the study medication. In addition, no clinically relevant changes in clinical laboratory tests or
physical examination findings including heart rate were reported.

Renal impairment

Study PTK0796-RENL-15102 investigated subjects with end-stage renal disease (ESRD) on

hemodialysis compared to healthy adults. Overall, 5 of 16 subjects experienced a total of 8 TEAEs
during the study. The TEAEs included upper respiratory infection (2), viral upper respiratory infection,
dizziness, headache, infusion site erythema, bronchospasm, and rash papular. One additional subject
had an AE of injection site hematoma that was not treatment-emergent. Only the dizziness and rash
were considered related to the study drug. There were no SAEs or deaths reported. No subjects
withdrew from the study due to an AE, and there were no AEs resulting in study drug discontinuation
or interruption. Although individual changes from baseline were observed in chemistry, haematology,
and urinalysis values, no apparent clinically relevant trends were observed.

Paediatric use

There have been no studies of omadacycline in children. All tetracyclines form a stable calcium
complex in any bone-forming tissue. A decrease in the fibula growth rate has been observed in
premature human infants given po tetracycline. This reaction was shown to be reversible when the
drug was discontinued. Accordingly, children less than 8 years of age should not receive omadacycline.

Use in pregnancy and lactation

Omadacycline may cause fetal harm when administered to a pregnant woman. Reproductive toxicity
studies suggest that omadacycline may affect the developing fetus during the organogenesis period,
resulting in abortion or reduced fetal body weight. Pregnant women should not receive omadacycline.

It is not known whether omadacycline is excreted in human milk. During tooth development (last half
of pregnancy, infancy, and childhood to the age of 8 years) omadacycline, like all tetracyclines, may
cause permanent discoloration of the teeth. Until more is known, nursing women should not receive
omadacycline.

EMA/595311/2019 Page 110/123

Safety related to drug-drug interactions and other interactions

No safety relevant aspects have been detected in studies investigating the potential for drug-drug
interactions of omadacycline.

Discontinuation due to AES

Discontinuation rates due to TEAES were overall low and comparable between omadacycline and the
comparator arms. A systematic occurrence of a specific type of TEAE leading to discontinuation of
treatment is not observed in the different treatment arms.

Immunological events - N/A

Post marketing experience – N/A

3.3.11. Discussion on clinical safety

Overall, 1947 subjects received at least 1 dose of omadacycline. Of these 1252 subjects have been
exposed to omadacycline in the frame of the phase 2/3 study programme obtaining 100 mg iv /
300mg orally of the test drug once daily for eight to nine days in average which roughly corresponds to
the recommended dosage (100 mg iv/300mg oral once daily for 7 to 14 days).

The most common adverse events associated with omadacycline in phase 2 and 3 studies were nausea
(15.3%), vomiting (8.0%), headache (4.2%), ALT increased (3.7%), AST increased (3.0%), diarrhoea
(2.6%), cellulitis (2.4%), wound infection (2.4%), constipation (2.2%) and infusion site extravasation
(2.2%). The Applicant has provided an adequate rationale for terms included in the SmPC as outlined
in the answer to question 230. The SmPC will be adjusted according to the strategy presented.

The most frequent type of AEs across all pools was gastrointestinal events, with higher frequencies
under omadacycline compared to linezolid or moxifloxacin. The highest rates of nausea and vomiting
events were reported in the Phase 3 ABSSSI oral-only study, the majority of these TEAEs occurred
during the “loading dose” phase on Days 1 and 2 of the treatment. These TEAEs were also more
common during the po than iv administration in studies, which used both ways of application.
Concomitant anti-emetic therapy was used in almost half of the subjects experiencing GIT AEs.
Gastrointestinal events are known and generally manageable tetracycline class effects. Based on the
food interaction PK studies revealing decreased exposure to omadacycline when administered following
a meal the Applicant proposed the administration of this medicinal product under fasting condition
which is agreed. The proposed precaution and listed AEs related to the gastrointestinal disorders in
SmPC are considered sufficient at this time. The effect of a light meal on tolerability is being evaluated
in an ongoing study. Further information on the outcome of the study will be provided, when the study
is completed and reported.

Across the omadacycline clinical development programme, there were 10 (0.8%) omadacycline and 7
(0.6%) comparator subjects who died. An imbalance in the mortality rate is noted in the CABP study
between omadacycline (2.1%) and moxifloxacin (1.0%), (95% CI -0.7 to 2.8). Narrative evaluation by
the Rapporteur revealed that five of the fatal cases, comprising vascular (1), cardiac (3) and
cerebrovascular (1) events, may be a result of the underlying cardiovascular and pulmonary
comorbidities presented by the patients. In one case, the patient was re-admitted for a new (likely
hospital-acquired) pneumonia episode caused by non-susceptible organisms. In two cases, patients
experienced a severe rapid deteriorating CABP disease, secondary in at least one case to known non-
susceptible organisms. Overall, the relatedness of the imbalance to the treatments received is unclear.
Only a larger study population would permit more reliable evaluation of any true difference.

EMA/595311/2019 Page 111/123

It is noted that FDA has included a short statement regarding the outcome of mortality rates in the
CABP-1200 study in the US prescribing information in section Warnings/Precautions, and furthermore
included as a post marketing requirement that a trial similar to the CABP-1200 study design, should be
performed to further evaluate the safety signal of the mortality imbalance in the CABP trial
(https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/209816_209817lbl.pdf). The Applicant
has specified the planned additional study to be performed in compliance with study design
suggestions made by the FDA. The Applicant intends to perform an additional study with the primary
endpoint of efficacy (early clinical response) including 450 to 670 POST Risk Class III and IV subjects.
A Data Monitoring Committee will review safety data, including deaths throughout the trial. The study
will be completed November 2022 (reporting April 2023 to EMA and FDA).

Allergic reactions were not frequently observed with omadacycline and included the PTs pruritus, rash,
urticaria and hypersensitivity reactions. Pruritus and rash have been included in the SmPC as
uncommon events under “skin reactions”. Erythema has been included as an uncommon event in the
SmPC. Furthermore, hypersensitivity reaction has been included in the SmPC as uncommon event
which is endorsed.

In principal, no clinically meaningful changes of haematology parameters and electrolytes from the
baseline were observed which were related in a negative way to omadacycline treatment. Of note, in
several cases lower neutrophil counts were observed with omadacycline (>1% with a >= 2 Grade
decrease). The investigation of cases narratives (5) did not reveal a causative association with drug
treatment. Thrombocytes did not decrease significantly in omadacycline treated subjects. Based on
further analyses demonstrating that 1% of individuals show at least one episode of moderate
thrombocytosis post-baseline and a clear treatment-event relationship in several individuals
experiencing a TEAE of thrombocytosis, it is suggested to include thrombocytosis as often occurring
event in the SmPC.

No subject had a TEAE of acute pancreatitis under treatment with omadacycline. Of note, grade 2
increase of amylase and lipase values was observed in 22 (2,1%) and 61 (5,9%) omadacycline
subjects with isolated cases of > 10 time increase of amylase. Out of 22 cases, only 3 subjects
showed amylase values over 3 xULN. One of them showed in parallel increased lipase levels without
nausea, vomiting or abdominal pain. TEAEs of nausea, vomiting or abdominal pain were neither
reported for the remaining 2 cases. 11 omadacycline subjects (out of 61 with the grade 2 lipase
increase) showed nausea, vomiting or abdominal pain without more serious clinical patterns
potentially indicating symptoms of acute pancreatitis. Out of 5 TEAEs of increased amylase or lipase,
only one subject experienced increase in both values (amylase 1,2xULN, lipase 2,2xULN) without
concomitant symptoms. No safety issue was identified based on the provided data.

No clinically significant impact of omadacycline on ECG parameters (PR, RR, QRS, QT) is noted in the
phase 2/3 pool with the exception of an increased heart rate. Analyses of the mean change over time
in HR showed that HR tended to decline slightly in all categories at both the EOT and PTE visits and the
decline was less rapid over treatment time for omadacycline subjects compared to moxifloxacin
subjects. Post-baseline elevations of HR were seen in 1.4% of omadacycline treated subjects and were
slightly less common compared to comparators. It is suggested to include tachycardia as common
event in the SmPC.

The overall incidence of TEAEs showed no major differences with respect to subpopulations, i.e., age,
sex, race, and BMI, and underlying disease characteristics relating to renal function, hepatic function,
and diabetes, was analysed. No major shifts in the AE profile have been identified in relation to these
parameters.

EMA/595311/2019 Page 112/123

3.3.12. Conclusions on clinical safety

The safety profile of omadacycline appears generally acceptable. An imbalance in the mortality rate is
noted in the CABP study between omadacycline (2.1%) and moxifloxacin (1.0%), for which the reason
is not known. This poses an uncertainty with regard to unfavourable effects of the test drug.

3.4. Risk management plan

3.4.1. Safety Specification

Summary of safety concerns

The following changes to the list of safety concerns have been implemented in the RMP Version 0.3:

Summary of safety concerns

Important identified risks None

Important potential risks None

Missing information None

3.4.2. Discussion on safety specification

The implemented changes to the safety concerns are endorsed, based on the argumentation
summarized below.

3.4.3. Conclusions on the safety specification

In the first round it was proposed including “Dental enamel hypoplasia, teeth discoloration and
reversible bone growth depression in infants exposed in utero or through maternal milk” as
an important potential risk: Omadacycline may cause foetal harm when administered to a pregnant
woman. Reproductive toxicity studies in animals suggest that omadacycline may affect the developing
foetus during the organogenesis period, resulting in abortion or reduced foetal body weight. It is not
known whether omadacycline is excreted in human milk. During tooth development (last half of
pregnancy, infancy, and childhood to the age of 8 years) omadacycline, like all tetracyclines, may
cause permanent discoloration of the teeth. The applicant suggests routine pharmacovigilance
activities for a follow up of these events (PSURs) which is endorsed. A special warning is included in
the SmPC to avoid using the drug during pregnancy and in young individuals. No additional PV
activities will take place. Therefore, the item was removed from the list of safety concerns.

‘Immunosuppressed individuals, resulting in a loss in efficacy’: Following EMA

recommendations, it is expected that the Applicant will monitor this aspect as an efficacy issue in the
PSURs. The item has been removed from the list of safety concerns, which is endorsed.

‘Prolonged treatment for more than 14 days’: The absence of data itself does not automatically
constitute a safety concern. The Applicant has removed this item from the list of safety concerns and
will monitor this aspect through PSURs, which is endorsed.

Regarding “Bacterial resistance” see comments in 3.4.4.

EMA/595311/2019 Page 113/123

3.4.4. Pharmacovigilance plan

Summary of planned additional PhV activities from RMP

There are no ongoing or planned omadacycline category 1 to 3 studies.

The Applicant proposed bacterial resistance as important potential risk during the first round.
Following EMA recommendations, it is expected that the applicant will monitor this as an efficacy issue
in the PSURs. The applicant has consecutively deleted this item from the list of safety concerns, which
is endorsed.

During the third round the Applicant has updated the list of safety concerns (no issues listed) and the
section regarding activities in the frame of the pharmacovigilance plan, accordingly.

The safety concerns to be discussed in the PSURs include:

- Bacterial resistance

- Treatment in immunosuppressed individuals, resulting in a loss in efficacy

- Prolonged treatment for more than 14 days

- Safety in pregnancy and lactation

- Phototoxicity

- Dental enamel hypoplasia, teeth discoloration and reversible bone growth depression in infants
exposed in utero or through maternal milk

3.4.5. Risk minimisation measures

The paragraph describing risk minimisation contains the following information in the RMP Version 0.3:

V.1. Routine Risk Minimisation Measures

Routine risk minimization measures are risk communications through product information, labeling, or
packaging. The risks associated with the use of omadacycline are discussed under the appropriate
sections of the SmPC.

V.2. Additional Risk Minimisation Measures

Routine risk minimisation measures as described in Part V.1 are considered sufficient for omadacycline.
Therefore, no additional risk minimisation measures are proposed.

V.3. Summary of Risk Minimisation Measures

Routine risk minimization measures and routine pharmacovigilance activities are considered sufficient
for omadacycline.

Overall conclusions on risk minimisation measures

The PRAC Rapporteur having considered the data submitted was of the opinion that:

The proposed risk minimisation measures are sufficient to minimise the risks of the product in the
proposed indication(s).

EMA/595311/2019 Page 114/123

3.4.6. Conclusion on the RMP

The CHMP and PRAC considered that the risk management plan version 0.3, dated 31-07-19, is
acceptable.

3.5. Pharmacovigilance system

It is considered that the pharmacovigilance system summary submitted by the applicant fulfils the
requirements of Article 8(3) of Directive 2001/83/EC.

Periodic Safety Update Reports submission requirements

The active substance is not included in the EURD list and a new entry will be required. The new EURD
list entry uses the European Birth Date (EBD) or international birth date (IBD) to determine the
forthcoming Data Lock Points. The requirements for submission of periodic safety update reports for
this medicinal product are set out in the Annex II, Section C of the CHMP Opinion.

The applicant should indicate if they wish to align the PSUR cycle with the EBD or the IBD.

4. Benefit risk assessment

4.1. Therapeutic Context

4.1.1. Disease or condition

The indications applied for are treatment of adults with:

• Acute bacterial skin and skin structure infections (ABSSSI)

• Community-acquired pneumonia (CAP)

Acute bacterial skin and skin structure infections (ABSSSI) are generally caused by Gram positive
pathogens including S. aureus and Streptococcus spp.. They can be serious, limb- or life-threatening
conditions, often requiring systemic antibiotic therapy and possibly surgical management and
hospitalisation.

Community-acquired pneumonia is caused most frequently by S. pneumoniae, Haemophilus influenzae,

and some atypical pathogens such as Mycoplasma pneumoniae and Legionella pneumophila, but can
also involve rarer atypicals and Gram negative organisms. CABP is classified separately from Hospital-
acquired pneumonia (HAP) (developing ≥48 hours after hospital admission), Ventilator-acquired
pneumonia (VAP) or Health-care associated pneumonia (HCAP), which are caused by a different range
of pathogens. Management includes systemic antibiotic therapy and supportive care.

4.1.2. Available therapies and unmet medical need

The most commonly used classes of antibiotics for ABSSSI and CAP include beta-lactams, lipopeptides,
oxazolidinones, glycopeptides and tetracyclines, and for CAP additionally macrolides, fluoroquinolones
are sometimes used, for coverage of atypical pathogens. Omadacycline does not address an unmet
need, but would provide a further antibiotic option.

EMA/595311/2019 Page 115/123

4.1.3. Main clinical studies

The clinical programme for omadacycline comprises two pivotal Phase 3 clinical studies in ABSSSI and
one single pivotal Phase 3 study in CAP.

Studies PTK0796-ABSI-1108 (n=655) and PTK0796-ABSI-16301 (n=735) were similarly designed, 1:1
randomised, active-controlled, double-blinded Phase 3 non-inferiority studies of 7-14 days’ iv-to-po or
po-only (respectively) treatment with omadacycline, against authorised comparator Linezolid, in
treatment of adults with ABSSSI (comprising cellulitis/erysipelas, major abscess, wound infection)
known or suspected to be caused by Gram positive pathogens, associated with signs of a systemic
inflammatory response.

Study PTK0796-CABP-1200 (n=660) was a 1:1 randomised, double-blind, active-comparator-

controlled, multi-centre Phase 3 non-inferiority study comparing 7-14 days’ treatment with an iv-to-po
regimen of omadacycline, against an iv-to-po regimen of authorised comparator moxifloxacin, for the
treatment of adults with CABP, in PORT Risk Class III or IV and not requiring intensive care.

4.2. Favourable effects

ABSSSI

Study PTK0796-ABSI-1108 demonstrated high rates of overall clinical response at PTE (the EU primary
efficacy endpoint) in both treatment arms in both the mITT population (86.1% omadacycline, 83.6%
linezolid) and CE-PTE population (96.3% omadacycline, 93.5% linezolid). The difference between
treatment arms was 2.5 (95% CI -3.2, 8.1) in the mITT population and 2.8 (95% CI -0.9, 7.1) in the
CE-PTE population.

Study PTK0796-ABSI-16301 also demonstrated high rates of overall clinical response at PTE (the EU
primary efficacy endpoint) in both treatment arms in both the mITT population (84.2% omadacycline,
80.8% linezolid) and CE-PTE population (97.9% omadacycline, 95.5% linezolid). The difference
between treatment arms was 3.3 (95% CI -2.2, 9.0) in the mITT population and 2.3 (95% CI -0.5,
5.8) in the CE-PTE population.

Given that the lower limit of the 95% CI for the treatment difference between arms in the mITT and CE
populations was above the pre-specified margin of -10%, omadacycline was demonstrated non-inferior
to linezolid in both studies, according to the EMA statistical analysis plan.

CAP

The single pivotal trial PTK0796-CABP-1200 demonstrated high rates of overall clinical response at PTE
(the EU primary efficacy endpoint) in both treatment arms in both the ITT population (84.2%
omadacycline, 80.8% moxifloxacin) and CE-PTE population (97.9% omadacycline, 95.5%
moxifloxacin). The difference between treatment arms was 3.3 (97.5% CI -2.2, 9.0) in the ITT
population and 2.3 (97.5% CI -0.5, 5.8) in the CE-PTE population. Given that the lower limit of the
97.5% CI for the treatment difference between arms in the ITT and CE populations was above the pre-
specified margin of -10%, iv/po omadacycline was demonstrated non-inferior to iv/po moxifloxacin,
according to the EMA statistical analysis plan, using a 1-sided alpha level of 0.0125.

Sub-group analysis revealed, perhaps unsurprisingly, numerically higher rates of clinical success
amongst subjects with PORT Risk Class III vs IV (90.7% vs 83.3% for omadacycline, 88% vs 80% for
moxifloxacin).

EMA/595311/2019 Page 116/123

4.3. Uncertainties and limitations about favourable effects

In the absence of a traditional dose finding study, PK/PD and PTA data provide important alternative
support for the dose selected for Phase 3 studies. Several deficiencies in the non-clinical PK/PD
package and PTA analysis mean that support in this application for dose selection is not robust.
Success rates were high in the clinical studies even for pathogens with MIC values several dilutions
above the highest MIC expected to be covered according to PTA analysis, and clinical success rates did
not appear to correlate to species or MIC. The lack of comprehensive PK-PD support for efficacy of the
proposed dose is mitigated for the ABSSSI infection by a sufficient quantity of clinical outcome data.
However, this is not true for CAP, where there is a clear lack of support for all proposed dosing
regimens to cover key pathogens, and only a single clinical study has been conducted. The Applicant
considers that nonclinical PK-PD PTA data may not be as predictive for certain classes of antibiotics
(tetracyclines for instance). If this is so, this makes it more difficult to justify an indication based on a
single pivotal trial; non-predictivity of PK/PD does not in any way strengthen the evidence that can be
derived from a single pivotal trial.

Clinical success rates were very similar between the two ABSSSI studies, i.e. regardless of iv-to-po or
po dosing regimen and are generally consistent with those seen in previous studies. For example, the
rates of clinical success based on IACR at PTE were 88% in the ITT population and 96% in the CE
population for linezolid, in a recent iv to po clinical trial in ABSSSI subjects (Moran, et al., 2014).
Sensitivity analyses, including analysis of the co-primary endpoint in the all-treated population and
analysis assuming subjects with indeterminate clinical outcomes to be clinical success, were supportive
of the primary analysis.

Clinical success rates in the treatment of CABP were numerically lower for omadacycline compared to
moxifloxacin in eradication of some gram-negative bacteria, namely Haemophilus influenzae,
Haemophilus parainfluenzae, and Klebsiella pneumoniae. This aligns with the observed less potent in
vitro activity of omadacycline against Gram negative species.

4.4. Unfavourable effects

The most common events in omadacycline subjects included side effects at the gastrointestinal tract of
mild to moderate intensity (see effects table).

An imbalance in the mortality rate is noted in the CABP study between omadacycline and moxifloxacin
(2.1% for omadacycline (8 cases) versus 1.0% for moxifloxacin (4 cases)), with a rate difference of
1.1% (95%CI -0.7 to 2.8). Based on a request from the FDA, the Applicant plans to perform an
additional post-marketing study of CABP to further investigate the observed imbalance.

While no impact of omadacycline on heart conduction was observed, the heart rate was significantly
increased after a single iv omadacycline dose (100 mg, 300 mg) in the TQTc study. The peak is
approximately within the first 60 minutes after administration and the increases in heart rate are dose
related. The thorough cardiac safety and QTc study evaluated omadacycline for the potential for QT
interval prolongation in healthy subjects indicated no concern. No clinically significant impact of
omadacycline was noted on SBP and DBP and ECG parameters (PR, RR, QRS, QT) in the phase 2/3
pool. Analyses of the mean change over time in HR showed a slower decline in the omadacycline arm
over treatment time compared to comparator arms. Comparable elevations in HR (> 120bpm or
elevation by > 15bpm from baseline) post baseline where observed in 1.4% omadacycline and 2.2%
linezolid treated subjects respectively.

EMA/595311/2019 Page 117/123

4.5. Uncertainties and limitations about unfavourable effects

A numerical difference of eight (2.1%, omadacycline) and four cases of death (1,0%, moxifloxacin
arm) have been observed in the CABP study, the reason for which is unknown.

4.6. Effects Table

Effects Table for Omadacycline

Uncertainties/
Short
Effect Unit Treatment Control Strength of References
Description
evidence
Favourable Effects
ABSSSI
Omadacycline Linezolid
(95% CI -3.2, 8.1)
mITT mITT
Non-inferior
86.1 83.6
PTK0796-
Difference=2.5
ABSI-1108
CE-PTE
CE-PTE (95% CI -0.9, 7.1)
Overall 96.3
Based on IACR 93.5 Non-inferior
clinical Difference=2.8
at TOC and PTE %
response at Omadacycline Linezolid
visits
PTE visit mITT mITT (95% -2.2, 9.0)
84.2 80.8 Non-inferior
PTK0796-
Difference=3.3
ABSI-16301
CE-PTE (95% -0.5, 5.8)
CE-PTE
97.9 Non-inferior
95.5
2.3
CABP
Omadacycline Moxifloxacin Single pivotal study,
mITT mITT no PKPD support
Overall 88.4 85.2
Based on IACR
clinical Difference=3.3 (97.5% CI -2.7, 9.3) PTK0796-
at TOC and PTE %
response at Non-inferior CABP-1200
visits CE-PTE
PTE visit CE-PTE
92.5
90.5 (97.5% CI -3.2, 7.4)
Difference=2.0
Non-inferior
Unfavourable Effects

Based on the Omadacycline Linezolid/Moxifl.

Nausea % Strong evidence SC23 pool
SC23 pool 15.3 10.0/5.4

Based on the Omadacycline Linezolid/Moxifl.

Vomiting % Strong evidence SC23 pool
SC23 pool 8.0 4.8/1.5

Based on the Omadacycline Linezolid/Moxifl.

Diarrhoea % Strong evidence SC23 pool
SC23 pool 2.6 4.6/8.0
Non-negligible
numeric difference
Within the omadacycline Moxifloxacin No clear trend PTK0796-
Deaths %
CABP study 2.1 1.0 towards therapy CABP-1200
failure or specific side
effects
Abbreviations: SC23 pool (comprises all 5 phase 2/3 studies), CABP study (includes patients with
pneumonia)

4.7. Benefit-risk assessment and discussion

4.7.1. Importance of favourable and unfavourable effects

The clinical success rates for treatment of both ABSSSI and CABP were high across the pivotal Phase 3
studies and aligned with those reported in recent studies of authorised comparators. Omadacycline
demonstrated non-inferiority to authorised comparators used in these indications and the statistical

EMA/595311/2019 Page 118/123

results were robust and supported by sensitivity analyses. Clinical outcome did not appear correlated
to MIC, and organisms with omadacycline MIC values up to 16 µg/mL were successfully treated
according to overall clinical response at PTE.

The tolerability of omadacycline is good.

A numeric imbalance of mortality has been observed in the CABP study. It is possible that this is a
chance finding, but this non-negligible difference does create a further caveat with regard to the
appropriateness of approving omadacycline for CABP based on a single pivotal study.

4.7.2. Balance of benefits and risks

ABSSSI

Omadacycline is considered to have demonstrated good safety and tolerability and high clinical success
rates, comparable to an authorised comparator and recently published studies, in two Phase 3 studies
of ABSSSI, and this suggests that omadacycline could offer an alternative to current antibiotic options,
with generally acceptable safety and tolerability. An advantage in comparison to several already
authorised tetracyclines is preserved activity in the presence of many common tetracycline-specific
resistance mechanisms. Omadacycline is available in both iv and po formulations, albeit with fasting
requirements when administered po.

Considering all findings, the favourable effects of omadacycline are considered to outweigh the
unfavourable effects for the treatment of ABSSSI.

CAP

Evidence for clinical efficacy in CAP comes from a single pivotal study, in which a non-negligible
numeric imbalance in mortality was observed. A causative association could not be identified from
careful examination of the individual cases but cannot be ruled out. The applicant plans to perform an
additional post-marketing study in CAP at the request of FDA.

At present, study PTK0796-CABP-1200 has not fulfilled the requirements of a single pivotal study, the
objective of which is to confirm an already established hypothesis. Additional efficacy and safety data
from the planned further study in CAP will be required before a decision can be made on the B/R
balance for omadacycline in CAP, for the purposes of authorisation for this indication (MO B/R).

4.8. Conclusions

The overall B/R of omadacycline is negative.

EMA/595311/2019 Page 119/123

5. QRD checklist for the review of user testing results

PRODUCT INFORMATION

Name of the medicinal product: Nuzyra

Name and address of the applicant: Paratek Ireland Limited, Dublin, Ireland

Name of company which has performed NDA Regulatory Science Ltd, UK

the user testing:

Type of Marketing Authorisation New application

Application:

Active substance: Omadacycline tosylate

Pharmaco-therapeutic group J01AA

(ATC Code):

Therapeutic indication(s): for the treatment of adults with the following infections
(see sections 4.4 and 5.1):
- Community-acquired bacterial pneumonia (CABP)
- Acute bacterial skin and skin structure infections
(ABSSSI)

Orphan designation yes no

Rapporteur/CoRapporteur SE/CZ

- Full user testing report provided yes no

- Focus test report provided yes no

- Bridging form provided 1 yes no

- In case bridging form1 has been provided, please perform the assessment in the bridging form and
state the overall conclusion/recommendations below:

A bridging report covering the IV formulation has been provided, referring to the full test for
the oral formulation. The bridging is acceptable.

- Is the justification for bridging acceptable? yes no

1
QRD form for submission and assessment of user testing bridging proposals [EMA/355722/2014]

EMA/595311/2019 Page 120/123

1. Technical assessment

1.1 Recruitment

• Is the interviewed population acceptable? yes no

no information
Comments/further details:

1.2 Questionnaire

• Is the number of questions ______ sufficient? yes no

no information

• Questions cover significant (safety) issues for the PL concerned? yes no

no information
Comments/further details:

1.3 Time aspects

• Is the time given to answer acceptable? yes no

no information

• Is the length of interview acceptable? yes no

no information
Comments/further details:

1.4 Procedural aspects

• Rounds of testing including pilot ______ yes no

no information
Comments/further details:

1.5 Interview aspects

• Was the interview conducted in well structured/organised manner? yes no

no information
Comments/further details:

2. Evaluation of responses

EMA/595311/2019 Page 121/123

2.1 Evaluation system

• Is the qualitative evaluation of responses acceptable? yes no

no information

• Does the evaluation methodology satisfy the minimum prerequisites? yes no

no information
Comments/further details:

2.2 Question rating system

• Is the quantitative evaluation of responses acceptable? yes no

no information
Comments/further details:

3. Data processing

• Are data well recorded and documented? yes no

no information
Comments/further details:

4. Quality aspects

4.1 Evaluation of diagnostic questions

• Does the methodology follow Readability guideline Annex? yes no

no information

•Overall, each and every question meets criterion of 81% correct answers (e.g. 16 out of 20
participants) yes no
no information
Comments/further details:

4.2 Evaluation of layout and design

• Follows general design principles of Readability guideline yes no

• Language includes patient friendly descriptions yes no

• Layout navigable yes no

• Use of diagrams acceptable yes no

Comments/further details:

EMA/595311/2019 Page 122/123

5. Diagnostic quality/evaluation

• Have any weaknesses of the PL been identified? yes no

• Have these weaknesses been addressed in the appropriate way? yes no

Comments/further details:

Both language and layout have been adjusted according to the result of the user consultation.

6. Conclusions

• Have the main objectives of the user testing been achieved? yes no

• Is the conclusion of applicant accurate? yes no

• Overall impression of methodology positive negative

• Overall impressions of leaflet structure positive negative

CONCLUSION/OVERVIEW
The user consultation is found acceptable for both the content and layout.

EMA/595311/2019 Page 123/123