30/09/2013

r
Cleaning Standards in
Multi-Product Facilities
Dr. Andreas Flückiger
Chief Occupational Health Officer
F. Hoffmann-La Roche Ltd., Basel
Brussels - 30 September, 2013

• New requirements for cleaning standards

• Toxicological basis

• Challenges in calculating science-based limits

• Conclusions

2
© Andreas Flückiger

The old standards

Conservative but scientific nonsense
Frequently applied (and accepted) cleaning standards

a) Visually clean
b) Not more than 10 ppm of product A in product B
c) Not more than 1/1000 of the smallest daily therapeutic dose of A
in the maximum daily therapeutic dose of B
d) (Not more than 1/50,000 of an LD50 of A in the max. therapeutic dose of B)

Ref. Walsh A.: Cleaning Validation for the 21st Century; Pharm Engin 2011
3
© Andreas Flückiger

1
 30/09/2013

Current regulatory situation

EMA
Draft Chapter 3.6
Risk assessment should include among other parameters a toxicological evaluation
of the products being manufactured (see Guideline …. )

Dedicated facilities are required for manufacturing when a medicinal product

presents a risk:
a) Which cannot be adequately controlled by operational and/ or technical
measures or
b) Scientific data does not support threshold values (e.g. allergenic potential
from highly sensitising materials such as beta lactams) or
c) Threshold values derived from the toxicological evaluation are below the
levels of detection
If you say you cannot set a limit value, you must use dedicated facilities
4
© Andreas Flückiger

Current regulatory situation

EMA

Draft
Start of consultation Jan 8, 2013
End of consultation June 30, 2013
5
© Andreas Flückiger

The new standards say:

Make use of existing data!

APIs are well investigated substances

• In vitro and in silico tests
• Animal tests
• Human experience
• Comparison with other substances

Why ignore all these data??

6
© Andreas Flückiger

2
 30/09/2013

• New requirements for cleaning standards

• Toxicological basis

• Challenges in calculating science-based limits

• Conclusions

7
© Andreas Flückiger

The two key elements of risk assessment:

Hazard and exposure

Risk = Hazard x Exposure

Study of data
-> ADE, PDE* Measurements
Setting limit values

8 * Acceptable/Permitted Daily Exposure

© Andreas Flückiger

Paracelsus
(Theophrastus von Hohenheim, 1493-1541)

• Dose-effect relationship:
“Dosis facit venenum”

…”was ist das nit gifft ist?

alle ding sind gift /
What is not a poison?
und nichts ohn gifft /
Everything is poison,
allein die dosis macht das only the dose makes
ein ding kein gifft ist…” it that a thing is not a
poison…
9
© Andreas Flückiger

3
 30/09/2013

Toxicology in a nutshell
• Every substance is poisonous
• Every substance is also not poisonous – the dose decides
Highly active Not very active

Highly active Not highly active

All sorts of measures No measures

10
© Andreas Flückiger

Wanted: The limit value:

The highest dose that is without an effect
• The maximum dose at which there is no unwanted effect
• in the patient (from cross-contamination)
• in the worker from exposure at work
• Even if exposure is life-long! -
A limit value for chronic exposure!

11
© Andreas Flückiger

The ADE/PDE is such a limit value

Definition

PDE: Permitted Daily Exposure (EMA)

ADE: Acceptable Daily Exposure (ISPE Risk-MaPP, FDA)
„… a substance-specific dose that is unlikely to cause an adverse
effect if an individual is exposed at or below this dose every day for a
lifetime.”
(Definition of the PDE; EMA, abbreviated)
Unit: mg/person/day

12
© Andreas Flückiger

4
 30/09/2013

Setting a limit value: Where to start?

• -> critical questions:

• WHAT
is the effect of the substance?

• AT WHAT DOSE
does this effect set in?

13
© Andreas Flückiger

Targets of pharmacological effects and toxicity

Skin
Thymus Liver

Embryo Kidney
Substance

Eye Blood

DNA Heart
Brain Lung

14

© Andreas Flückiger

The lead effect or critical effect

Effects of an API – wanted and unwanted effects:

Lowering blood pressure

Lower cholesterol Allergies
Lowering blood sugar Malformations
Sleep induction Genetic damage
Killing cancer cells Cancer
Anticoagulation etc
etc

Both can be the lead effect

15
© Andreas Flückiger

5
 30/09/2013

The lead effect or critical effect

• The first effect that will occur in an exposed person
• Other effects will only occur at higher doses
• If the dose is low enough to avoid the lead effect, no
biological effects will occur at all

Disturbed function
Structural organ damage
Reversible
Irreversible
16
© Andreas Flückiger

Setting a limit value: Where to start?

• -> critical questions:

• WHAT
is the effect of the substance?

• AT WHAT DOSE
does this effect set in?

17
© Andreas Flückiger

Wanted: The limit value

What we need ideally:
The NOEL for life-long exposure in
patients (and workers [for OELs])

Is almost never directly available

Must be derived from other data that are available

-> One takes the best available data set …

and starts calculating.
18
© Andreas Flückiger

6
 30/09/2013

Wanted: The limit value

Off-label use of data:
Studies made for the registration of a drug
• Dosage much too high and often too short
• Animal studies instead of studies in man
• Interpretation takes into account benefits for the patient
(what is „adverse“?)
• Data are missing: No reprotox and cancer study data for
certain types of drugs
The full therapeutic dose is known, but what about the
NOEL? Therapeutic dose may not correlate with lead
effect. -> Adjustments are necessary!

19
© Andreas Flückiger

NOEL ADE/PDE Formula 50/60 kg body

NOAEL Basis: ICH Q3C (2011) weight
LOEL
For the critical effect

Reference dose x Correction for body weight

ADE =
F1 x F2 x F3 x F4 x F5

Correction for
Differences reference dose
animal/man
Variability Severity of
man/man effect
Length of
study
20
20
© Andreas Flückiger

Starting point: the Reference Dose

Ideally
• NOEL: No (Observed) Effect Level

otherwise
• NOAEL: No (Observed) Adverse Effect Level
• LOEL: Lowest (Observed) Effect Level

21
© Andreas Flückiger

7
 30/09/2013

A given substance can have many

NO(A)ELs - every study leads to a NO(A)EL

• The NO(A)EL is the result of a study with different dose levels

• Each study has a clear objective, e.g.:
• Investigate the desired effect
• Investigate side effects
• Organ damage, disturbed organ function
• Special effects such as ability to cause cancer or
disturbances of fertility/malformations

22
© Andreas Flückiger

The NOEL and the NOAEL - Problems

• -> Expertise is needed to select or derive the NO(A)EL of
the most relevant study
• Pick the right starting point of the calculation
• -> Expertise is necessary to
select the right adjustment factors
These factors multiply!

23
© Andreas Flückiger

And because things have not been

complicated enough…

24
© Andreas Flückiger

8
 30/09/2013

Bio(logical) availability
• What is the percentage of a substance that is taken up via
the different routes of exposure?
• by injection (parenteral)
• if breathed in (inhalation)
• if swallowed (oral)
• via the skin (dermal)

-> If a substance is poorly absorbed via a given route,

then a higher exposure via that route can be tolerated.

25
© Andreas Flückiger

Bio(logical) availability
At the workplace: Inhalation is most important
• Inhalation studies are rare
• Need to extrapolate from oral or parenteral studies
• If no other data are available:
Assumption of 100% bioavailability by inhalation

GMP: Route of administration of the contaminant „A“

• The route of administration of „product B“ is relevant,
since „product B“ carries „product A“ as a contaminant.
• „A“ has not always been investigated for the route of
administration of „B“.
26
© Andreas Flückiger

ADEs for occupational toxicology and for GMP

Review of all
data ADE
ADE

27
© Andreas Flückiger * ADE = Acceptable Daily Exposure

9
 30/09/2013

ADEs for occupational toxicology and for GMP

Review of all For the best

data ADE
ADE investigated route
of exposure

Inhalation Oral Parenteral

ADE ADE ADE

: 10 m3

OEL GMP Cleaning

Standard

* ADE = Acceptable Daily Exposure

28

© Andreas Flückiger

Calculation of cleaning limits:

Wipes of product contact surfaces
Limit (mg/wipe) = PDE* x BS x TS x RR
MDD x TCS
PDE = Permitted Daily Exposure of A (mg/day)
BS = Batch size of B (mg)
TS = Test Surface – wiped surface (cm2)
RR = Recovery Rate
MDD = Maximum Daily Dose of B (mg)
TCS = Total Contact Surface (cm2)

• Use the PDE adapted to the route of administration of B

• A = contaminant; B = contaminated product

29

Calculation of cleaning limits:

Rinsate of product contact surfaces
Limit (mg/liter) = PDE* x BS x TS
V x MDD x TCS
PDE = Permitted Daily Exposure of A (mg/day)
BS = Batch size of B (mg)
TS = Test Surface – rinsed surface (cm2)
V = Volume of rinsate
MDD = Maximum Daily Dose of B (mg)
TCS = Total Contact Surface (cm2)

• Use the PDE adapted to the route of administration of B

• A = contaminant; B = contaminated product

30

10
 30/09/2013

• New requirements for cleaning standards

• Toxicological basis

• Challenges in calculating science-based limits

• Challenging EMA proposals
• Challenging real-life situations

• Conclusions

31
© Andreas Flückiger

EMA guideline proposal: challenges

1. TTC of 0.15 μg/day for genotoxic contaminants coming
from cross-contamination
– 1.5 μg/day for genotoxic contaminants coming from the
synthesis process itself (EMEA 2006)
– 1 case of cancer in 1 million human lives
Increase of individual risk from 30% to 30.0003%
2. Covering reprotox risk
with no or few data
– TTC (what TTC?)
– Add safety factor to
general tox data?
3. More weight to clinical data

32
© Andreas Flückiger

Special cases

1. Genotoxic contaminants
2. Missing reprotox data
3. Potent sensitisers (allergens)
4. Large molecules
5. Clinical trial material
6. Sensitisers

33
© Andreas Flückiger

11
 30/09/2013

Special cases
1. Genotoxic substances
The threshold of many has not been determined.
Alternative: Use the TTC concept (Threshold of
Toxicological Concern):
Industry proposal:
Use the same limit as for genotoxic contaminants that
come from the synthesis: 1.5 μg/day.

34
© Andreas Flückiger

Special cases
2. Missing reprotox data
Always the case in early development
Never generated for antineoplastic agents

EMA proposal: no scientific basis for a limit value

-> dedicated facility

Alternative: Science-based “generic” alternative

 Add a safety factor to the NOAEL of a general tox
study (Factor 10 covers reprotox risk very well)
 “TTC” for reprotox effects (under investigation)

35
© Andreas Flückiger

Special cases

3. “Large molecules” (Proteins, peptides)

Same approach as for small molecules: no special case
Oral bioavailability often low or nil.
Very large molecules have a low bioavailability by
inhalation (<5%)

Special precautions:
Antibody-Drug-
Conjugates (ADCs)

36
© Andreas Flückiger

12
 30/09/2013

Special cases
4. Clinical trial material
This is GMP material
Same science-based risk assessments for multi-product
facilities as for commercial product
ADE/PDE concept needs to be applied as well
No fundamental difference to commercial production

Only limited data available?

What to do?
Banding approach?

37
© Andreas Flückiger

Special cases

4. Clinical trial material

A well established and proven solution
from occupational toxicology:
OEBs = Occupational Exposure Bands (categories)

• Substance is placed into a band/category instead

of setting a “precise” limit value
• Categorisation/banding is possible with fewer
data
• Usually includes additional conservatism

38
© Andreas Flückiger

Special cases

5. Potent sensitisers
ADE/PDE concept is not applicable
Dedicated facility compulsory
Example: Betalactams
a dedicated facility for each of the 5 sub-families

Necessary because it is not possible to set a science-

based limit value (?)
FAO ADI in food items for penicillin 30 μg/d
for cephalosporin 50 μg/d

39
© Andreas Flückiger

13
 30/09/2013

• New requirements for cleaning standards

• Toxicological basis

• Challenges in calculating science-based limits

• Conclusions

40
© Andreas Flückiger

Conclusions

Using existing data to establish cleaning limits is common

sense and a necessity
Expertise and experience is needed
Occupational toxicology is best suited as a background
Setting science-based limits is time-consuming
Switching from the old cleaning limits to the new ones
takes time

41
© Andreas Flückiger

14