FDA’s Current Practice and Challenges in the Use of

Dissolution Similarity Testing for Demonstration of

Bioequivalence – Case Studies
Zhen Zhang, Ph.D.
Division of Bioequivalence I
Office of Bioequivalence
Office of Generic Drugs, CDER, FDA

Dissolution Similarity Workshop

University of Maryland School of Pharmacy, May 21-22, 2019
 Disclaimer

The views and opinions presented here represent those

of the speaker and should not be considered to represent
advice or guidance on behalf of the U.S. Food and Drug
Administration

www.fda.gov 2
 Outline

• Definition of Dissolution Similarity

• Dissolution Similarity in Bioequivalence Determination

• Office of Bioequivalence’s Current Practice and Challenges:

Four Case Studies

• Summary/Challenges

www.fda.gov 3
 Definition of Dissolution Similarity

• Dissolution profiles may be considered similar by virtue of overall profile similarity (e.g., f2 ≥ 50)1, 2, 3, 4, 5 and
similarity at every dissolution sample time point (e.g., ≤ 15%) 1, 3, 4. – Profile Comparison

• When both test and reference products dissolve 85 percent or more of the label amount of the drug in 15
minutes using all three dissolution media, the profile comparison with an f2 test is unnecessary 5. – Point
Comparison

• To allow the use of mean data, the coefficient of variation should not be more than 20 percent at the earlier
time points (e.g., 15 minutes), and should not be more than 10 percent at other time points1, 5. – Low
variability

www.fda.gov 4
 Dissolution Similarity in Bioequivalence Determination
- Profile Comparison
Common Medium Sampling time Criteria
‘waiver’ of non-bio QC medium for IR; Sufficient number of intervals to characterize the entire Low variable data:
strength QC + multimedia for MR dissolution profile of drug product Similar if f2 ≥ 50;
Multimedia dissolution pH 1.2, 4.5 and 6.8 Include early sampling times of 1, 2, and 4 hours and Highly variable data:
for MR products buffer continue every 2 hours until at least 80% of the drug is Other methods (e.g.
released bootstrap f2)

Multimedia dissolution Per PSG Example: PSG for Mesalamine DR Tablets, 800 mg
for locally acting drugs strength: 0, 10, 20, 30, 45, 60, 75, 90, 120, 150, 180,
240, 300, and 360 minutes or as needed
QCRT (e.g. Icosapent Develop a discriminatory Early times (e.g. 5, 10, 15, 20, 25 minutes) and as
Ethyl Capsule) QCRT method frequently as possible, until at least 80% of the drug is
released
DESI Drug QC medium Sufficient number of intervals to characterize the entire
dissolution profile of drug product
Half tablets (e.g. scored Same as whole tablets Same sampling time points as whole tablets
ER tablets)
QC: Quality Control; MR: Modified-Release; ER: Extended-Release; IR: Immediate Release
PSG: Product-Specific Guidance; QCRT: Quantitative Capsule Rupture Test; DESI: Drug Efficacy Study Implementation

www.fda.gov 5
 Dissolution Similarity in Bioequivalence Determination
- Point Comparison

Common Medium Sampling Time Criteria

BCS I/III waiver 1) 0.1 N HCl or SGF w/o e.g., 5, 10, 15, 20, and 30 minutes Point Comparison
enzyme; BCS I: ≥ 85% (mean) within 30 min
2) pH 4.5 buffer; BCS III: ≥ 85% (mean) within 15 min
3) pH 6.8 buffer or SIF Profile Comparison
w/o enzyme) similar dissolution profiles (e.g. f2 ≥
50 if applicable)
Alcohol dose 900 mL, 0.1 N HCl, USP Samples of the media are taken once every 15 Comparable % dissolved drug
dumping apparatus 2 at 50 rpm, w/ minutes until 2 hours is reached product
or w/o Alcohol

BCS: Biopharmaceutics Classification System; SGF: Simulated Gastric Fluid; SIF: Simulated Intestinal Fluid;
QC: Quality Control; IR: Immediate Release

www.fda.gov 6
 Case Study #1: Justification of Missing Sampling Time
• The drug product A is an extended release tablet, which has five strengths
• The PSG recommends in vivo bioequivalence (BE) studies on the middle strength.
• As one of the criteria to evaluate the waiver request of non-bio strengths, the PSG recommends
multimedia dissolution testing at pH 1.2, 4.5 and 6.8 buffers including early sampling times of 1, 2 and 4
hours and continue every 2 hours until at least 80% of the drug is release.

www.fda.gov 7
 Justification: 1) low variability; 2) even though non-bio strengths do not have release data at 2-hour
and 4-hour time points, the dissolution data include early, middle, and complete release at different
time points, which is sufficient to capture the whole release profile.

Challenge: What is the appropriate sampling time for immediate release solid oral dosage forms,
especially when the drug release does not reach 85% within 15 min? Include early times (e.g., 5, 10,
15, 20, 25 minutes) and as frequently as possible?

www.fda.gov 8
 Case Study #2: Justification of Similarity When f2 < 50
• Same drug product and dissolution data as in Case Study #1;
• Acceptable in vivo BE studies on the middle strength (bio-strength) and formulation proportionality
across all strengths;
• Per Guidance for Industry: Bioequivalence Studies with Pharmacokinetic Endpoints for Drugs
Submitted Under an ANDA (PK BE Guidance; Dec 2013), we recommend that the drug products
exhibit similar dissolution profiles between the strength on which BE testing was conducted and
other strengths based on the f2 test in at least three dissolution media (e.g., pH 1.2, 4.5, and 6.8).

www.fda.gov 9
 Conclusion: The dissolution data are acceptable to support waiver request.
Note: May not be applicable to other cases.
Challenges: F2 ≥ 50 is used as a criterion to determine the dissolution similarity.
1) If f2 fails within a close margin (e.g., f2 = 49), what is the likelihood of rejecting a ‘good’ drug product if we
strictly follow the criterion of f2 ≥ 50?
2) If f2 fails largely (e.g., f2 = 40), what is an acceptable justification, e.g., clinical relevance, comparable
dissolution profile between test and reference?
www.fda.gov 10
Case Study #3: Bootstrap f2 for Highly Variable Dissolution Data
• Drug product B is a locally acting, extended-release drug;
• The PSG recommends in vitro comparative dissolution testing at different conditions as one of the pivotal
BE studies.
Test Condition #1:

To allow use of mean data, the percent coefficient of variation at the earlier time points (e.g., 15
minutes) should not be more than 20%, and at other time points should not be more than 10%.

Bootstrap f2
Similar T vs. R

www.fda.gov 11
 Test Condition #2:

Bootstrap f2
Different T vs. R

Conclusion: For Test Condition #2, dissolution profiles are not comparable between the test and reference products.
Repeat comparative dissolution testing on the unexpired test product using a larger sample size to provide a better
estimate of the mean difference. The dissolution testing should be conducted on at least 24 units (more if necessary)
of the unexpired test product and at least two lots of unexpired reference product (12 units per lot)
Challenges:
1) When are the dissolution data considered as highly variable for modified release drug product? Do we need to
take the drug release range at early time points into consideration?
2) Are 12-unit data sufficient for the comparison of dissolution profiles for highly variable dissolution data?
www.fda.gov 12
Case Study #4: Comparison of 3 Test Lots and 3 Reference Lots
• Drug C is an immediate release drug product
• The PSG recommends to compare three lots of the test product with three lots of the reference product
using an optimized QCRT method

www.fda.gov 13
 Individual Test Lot vs. Individual Reference Lot (12 T vs. 12 R) – Bootstrap f2

Results: Only Test Lot 1 and Test Lot 2 are

comparable to Reference Lot 3.

Pooled Test Data vs. Pooled Reference Data (36 T vs. 36 R) – Bootstrap f2

Results: The QCRT data are not comparable

between the test and reference products

Exploratory: Mahalanobis distance (M-distance)-based approach also shows that the QCRT data are not
comparable between the test and reference products
Challenge: What is an appropriate approach to compare multiple T vs. multiple R? Individual or pooled data?
www.fda.gov 14
 Summary

• Dissolution is a critical tool for the evaluation of generic drug products;

• Low variable dissolution data

o Similar if f2 ≥ 50;

• Highly variable dissolution data

o Bootstrap f2 method;
o Other methods with sufficient justification are also acceptable.

• If f2 fails to meet the acceptance criteria, justification is welcome.

• Challenges

www.fda.gov 15
 Challenges
• What is the appropriate sampling interval for immediate release solid oral dosage forms, especially when
the drug release does not reach 85% within 15 min? Include early times (e.g., 5, 10, 15, 20, 25 minutes) and
as frequently as possible?

• If f2 fails within a close margin (e.g., f2 = 49), what is the likelihood of rejecting a ‘good’ drug product if we
strictly follow the criterion of f2 ≥ 50?

• If f2 fails largely (e.g., f2 = 40), what is an acceptable justification, e.g., clinical relevance, comparable
dissolution profile between test and reference?

• When are the dissolution data considered as highly variable for modified release drug products? Do we need
to take the drug release range at early time points into consideration?

• Are 12-unit data sufficient for the comparison of dissolution profiles for highly variable dissolution data?

• What is an appropriate approach to compare multiple T vs. multiple R? Individual or pooled data?

www.fda.gov 16
 Acknowledgements

Division of Bioequivalence I
Bing V. Li
Utpal Munshi
Qing Liu
Rong Wang

Division of Bioequivalence II
Hongling Zhang

Division of Bioequivalence III

Nilufer Tampal
April Braddy

Division of Biopharmaceutics
Sandra Suarez
www.fda.gov 17
 References
1. Dissolution Testing of Immediate Release Solid Oral Dosage Forms (Aug 1997): Dissolution profiles may be considered similar by
virtue of (1) overall profile similarity and (2) similarity at every dissolution sample time point… Generally, f1 values up to 15 (0-15)
and f2 values greater than 50 (50-100) ensure sameness or equivalence of the two curves… To allow use of mean data, the percent
coefficient of variation at the earlier time points (e.g., 15 minutes) should not be more than 20%, and at other time points should
not be more than 10%.
2. SUPAC-IR: Immediate-Release Solid Oral Dosage Forms: Scale-Up and Post-Approval Changes: Chemistry, Manufacturing and
Controls, In Vitro Dissolution Testing, and In Vivo Bioequivalence Documentation (Nov 1995): An f2 value between 50 and 100
suggests the two dissolution profiles are similar.
3. Extended Release Oral Dosage Forms: Development, Evaluation, and Application of In Vitro/In Vivo Correlations (Sep 1997): A
model independent approach using a similarity factor, and comparison criteria are described in SUPAC-MR.
4. SUPAC-MR: Modified Release Solid Oral Dosage Forms Scale-Up and Postapproval Changes: Chemistry, Manufacturing, and
Controls; In Vitro Dissolution Testing and In Vivo Bioequivalence Documentation (Sep 1997): An f2 value between 50 and 100
suggests the two dissolution profiles are similar. Also, the average difference at any dissolution sampling time point should not be
greater than 15% between the changed drug product and the biobatch or marketed batch (unchanged drug product) dissolution
profiles. An f2 value less than 50 does not necessarily indicate lack of similarity.
5. Waiver of In Vivo Bioavailability and Bioequivalence Studies for Immediate-Release Solid Oral Dosage Forms Based on a
Biopharmaceutics Classification System. Guidance for Industry (Dec 2017): Two dissolution profiles are considered similar when
the f2 value is ≥ 50. To allow the use of mean data, the coefficient of variation should not be more than 20 percent at the earlier
time points (e.g., 15 minutes), and should not be more than 10 percent at other time points. Only one measurement should be
considered after 85 percent dissolution of both products. In addition, when both test and reference products dissolve 85 percent or
more of the label amount of the drug in 15 minutes using all three dissolution media recommended above, the profile comparison
with an f2 test is unnecessary.
www.fda.gov 18