Follow-on glatiramer acetate

We read with interest the article by Melamed-Gal et al. [1] addressing the question how to establish similarity between an originator and a follow-on drug for non-biological complex drugs, here for glatiramer acetate (GA). They call for further assessment of links between compositional and functional differences of such products and their potential impact on long-term efficacy and safety, including substitution.

We are surprised that the authors dismiss the 2-year clinical evidence coming from the GATE Phase III study [2,3] based on alleged underreporting of injection site reactions (ISRs). They claim that ISRs occurred in less than 24% of patients treated with follow-on GA (FOGA) or branded GA in GATE while ISRs should have been seen in 56–90% [[4], [5], [6]] of patients based on earlier studies with Copaxone®. This claim misrepresents the published information. In the double-blind part of the trial, 80–90% of the patients reported ISRs (Supplemental eFigure) [2] and about 70–80% reported ISRs during the open-label extension (Supplemental Fig. e-2) [3]. During double-blind and open-label parts, patients recorded ISR presence and intensity in diaries for 14 consecutive days at treatment initiation and again during maintenance treatment. Diary-recorded ISRs were not additionally reported as adverse events (AEs) while spontaneously reported ISRs were reported as AEs. Of note, the rates of systemic AEs and serious AEs in the GATE double-blind part were identical to the rates in a contemporary controlled Copaxone® study (GALA) [7].

Melamed-Gal et al. [1] report 15 non-clinical experiments comparing six Copaxone® batches with six batches of the European FOGA for similarity. The authors define “similar” as (1) within Copaxone® specifications, (2) within inherent microheterogeneity ranges of tested Copaxone® batches, or (3) not showing statistically significant differences.

We have the following comments and concerns regarding the reporting and interpretation of the non-clinical assays:

• •The authors do not report on how representative the Copaxone® batches are. The reader should be informed about whether these six batches came from six different drug substance batches or were derived from the same drug substance batch and, thus, being the same.
• •The similarity observed in the TNFa secretion assay (Section 3.2.4) is missing for the count of similar experiments in the paper's abstract and Table 1.
• •The potency for the FOGA in an ex vivo T cell assay was within Copaxone® specifications, yet the authors depreciate that finding by claiming a “significantly higher potency profile” based on statistical methods that are not appropriate in our opinion.
• •Four assays purportedly show dissimilarity: CEX, Viscotec-TDAmax, RPLC 2D-MALLS, IMMS, yet the authors fail to provide the respective specification limits. It may well be that all FOGA batches were within the respective specification limits, i.e., similar.

Further, the gene expression studies illustrate the major short-coming of this paper, that microheterogeneities do not necessarily have biological relevance. While these assays showed dissimilar effects on TNFα signaling, the functional cell-based assay on TNFα secretion demonstrated similarity. Of note, the US FDA identified important methodological flaws in comparable gene expression data submitted by the originator. The FDA's re-analysis of the data found very similar effects on the pathways proposed for GA's mechanism of action [8].

The authors report also a subcutaneous toxicity study in rats using marketed FOGA apparently conducted after human data have been in the public domain for years [2,9]. The minor differences found in 30 rats lack relevance when considering the similarity observed in more than 700 patients exposed to the FOGA [3].

Establishing similarity for two sets of random complex peptide mixtures is not trivial. “Similar” does not equal “identical” as minor differences exist also between batches of the same product with the definition of “minor difference” remaining arguable. What ultimately matters are relevant clinical differences. Melamed-Gal et al. [1] provide no data that any of the microheterogeneities they claim to have demonstrated in the reported assays have any clinical ramification. We believe that “the proof of the pudding is in the eating.” The GATE study and its extension established therapeutic equivalence for the FOGA, including efficacy, safety, tolerability, and immunogenicity [2,3]. Those results are more meaningful than the purported microheterogeneities in non-clinical assays.

Conflict of interest disclosures

Dr. Cohen reports personal compensation for consulting for Alkermes, Biogen, Convelo, EMD Serono, ERT, Gossamer Bio, Mapi, Novartis, Pendopharm, and ProValuate; speaking for Mylan and Synthon; and serving as a Co-Editor of Multiple Sclerosis Journal – Experimental, Translational and Clinical.

Dr. Wolf reports compensation for his organization for consulting from Synthon, Novartis, Teva, BBB, ICON and Desitin; and for speaking from Mylan and Synthon.

Dr. Selmaj reports grants and personal fees from Synthon during the conduct of the GATE study as well as grants from Merck, personal fees from Roche, Biogen, Novartis, Merck, Genzyme, and Celgene.

Dr. Arends is a salaried employee of Synthon.

Funding

No funding was received.

Author contributions

JAC, CW, RA: design, literature search, discussion, first draft, approval; KS: discussion, critical comments, approval.