Comments on EMA's guidelines for biosimiliar beta interferons

This is the first of two part series regarding biomilars and how they may affect MS (and other diagnoses) in both Europe and the Unites States.

After the EMA issued its draft guidance on biosmilars to beta interferons Dr. Kantor was interviewed by Jennifer C. Smith-Parker of BioPharm Insight -- Biogen, Merck, Bayer's MS drugs safe from EU biosimilars competition for several years due to tough requirements - experts, which was picked up by the Financial Times - Pharma majors’ multiple sclerosis drugs safe from EU competition for several years

In the report below, you will find Dr. Kantor's full responses (blue) to questions (green) raised by an analyst report (black):

Highlights:

· The requirements accept a single, one year trial with MRI primary endpoint in RRMS patients as basis for approval. Further, the results can be extrapolated to related conditions such as clinically isolated syndrome. This is a 'discount' vs. the new agent approval requirements (two trials whose length is two years with clinical efficacy as endpoints). However, the requirements are far from trivial.
I agree. The EMA could have required a non inferiority study with a clinical outcome, but instead they agreed to settle on an equivalency study with a radiologic outcome. The most accepted clinical outcome in MS trials is annualized relapse rate reduction, and you need two large Phase 3 trials that last for 2 years, but MRI studies can be shorter (this is more similar to phase 2 trials used in MS research).

- MRI results are clearly correlated with clinical efficacy, but are far from being a classical 'tight' bio-molecular marker (such as those that exist for already approved Insulin, hGH, EPO or Neupo). Thus there could be significant individual patient variability. EMA's suggested end point is "combined unique active lesions (CUA, defined as new gadolinium-enhancing T1-weighted lesions and new/enlarging T2-weighted lesions without double 170 counting)". However, it notes a baseline may need to be established and potentially a control group may be needed to calibrate the study. We expect that designing these trials will be difficult and they will need to be fairly large to deal with patient variability.

Recent trials, such as Teva's BRAVO study of oral Laquinimod for RRMS have highlighted potential barriers in matching for baseline MRI characteristics. Differences in the baseline characteristics may lead to unexpected disappointing outcomes. Theoretically an equivalence study could be negative, not just because the biosimilar may look worse than the original reference interferon, but also if it looks better than the reference interferon. If the EMA had allowed a non-inferiority study design, a biosimilar could unlikely but theoretically claim that it was superior to the reference interferon.

- Significant pharmacodynamic data and markers of immunogenicity (neutralizing antibodies) collection are required. Each difference between reference product and the biosimilar would need to be justified. The causes and effects of these differences are not always very clear. Again, this is a risk for approval.
- We note that the only company to develop a beta interferon biosimilar so far was Biopartners (Biferonex IF-beta), which failed to gain approval.
My questions: How large, specifically. do you think a control group should be? What beside the patient numbers makes the trial designs difficult and why?

The trial would not have to be as large as the > 1,000 subject phase 3 trials that we have become accustomed to because we are only looking at MRI changes and not primarily clinical factors, but the issue with both equivalency and non-inferiority studies is that there is a temptation to make these trials small because this would make it difficult to find a difference between the two groups, as opposed to a superiority trial, where you want the trial to be as large as you can afford (financially and time-wise) so that you are more likely to show a difference in the two groups. The EMA has tried to solve this problem a little by requiring both the reference interferon and the proposed biosimilar to be compared to placebo, and only if they both look better than placebo can the be compared to each other properly. While the proposed 4 months of placebo exposure is short, it does make it harder to convince patients and investigators to get involved. The EMA partially solves this by allowing the thirs comparison arm to be a lower lesser effective dose of the biosimilar rather than placebo, but it would be more difficult to show that the reference interferon and proposed biosimilar are superior to a lower (and potentially still effective) dose of the biosimilar.

The intricacies and subtleties inherent in an equivalency study make it more difficult to design well, and there isn't an established way of doing such a study in the MS world -- any sponsor considering this would be forging a relatively new scientific and statistical path.

· The commercialization of biosimilar beta interferon will be challenging. Our view is that while hospital products will be relatively easy to commercialize assuming the scientific data is strong, retail-type products will be tougher to commercialize for two reasons (i) Complex (dis)incentive structures in the channel and (ii) need to build a sales force and convince a broad group of physicians. In the MS there is particularly difficult challenge as physicians are risk averse as there is a material question of determining disease progression (i.e. it is tough for a physician to tell if a drug is working in his own patients).

My questions: Can you please explain disincentive structures? I don’t quite get that. Why is there a need to build a sales force for a biosimilar product here? Not sure I quite understand that. Also not sure why particularly challenging in the MS space; what does being risk adverse in this case and determining disease progression have to do w/ biosimilar product acceptance?

I think that the operative word is "complex;" it sounds like they are referring to the heavy involvement of Pharma in the MS world and the clear dissatisfaction that current interferon manufacturers would express to prescribers (in a number of subtle, and potentially, not so subtle ways). A biosimilar product is not the same product -- Nike and Reebok shoes are similar products, but they still need their own competing sales forces. Physicians have also been used to the brands of the reference interferons for almost two decades, and any new (biosimilar) interferons on the market may not be well recognized or understood by many prescribers. Unless the prescriber was forced by the insurance carrier to choose the biosimilar over the reference interferon, it is unclear why a prscriber would want to take the chance that the two are not really exactly the same.

In many other disease states, the patient and the doctor know whether a treatment is working for them. In MS disease modification, we are using the medications as insurance for the future -- taking the medication is supposed to prevent future relapses, not fix problems already present. This is one of the greatest difficulties in maintaining patient adherence to therapy -- if you don't know that you are feeling better from a treatment, you may not simply continue it because it may protect you in the future. When physicians treat hypertension, we know in real-time whether the blood pressure is lower and better controlled than it was before initiating treatment. When physicians treat MS, we can only opine retroactively whether the treatment was beneficial or not. Additionally, because of the highly individualized and variable course of MS, a patient may do well because they were "destined" to do well with or without a particular treatment, and converesely, some patients will worsen no matter which treatment (or no treatment) is chosen. This is why NeuroFunctional Enhancers (NFEs) are so empowering to our patients -- these are treatments that enhance the neurologic functioning of our patients in real-time, and patient either respond to them or they don't, but that response is noticed within the first 2 months of treatment (examples include Acorda's Ampyra and Avanir's Nuedexta).

· We think several products will be commercialized, but not as many as in the antibody field for three reasons: (i) These products are generally commercially smaller than most antibody products targeted in the first wave of biosimilars (and will get smaller as the new generation of MS products gain dominance); (ii) Beta interferon's require largely dedicated manufacturing equipment, having limited overlap with antibody manufacturing; (iii) As discussed above, trial risk is relatively high and will require relatively larger investment in marketing and sales. Interestingly, some of the biosimilar leaders (Novartis, Teva) have mixed incentive structures; they have innovative products in the market and do not necessarily want to see price attrition in the beta interferon category (although this may change for Teva).
My questions: What products in the MS field do you think will be commercialized? Have heard Rituxan from one doc already. Why do beta interferons require dedicated manufacturing equipment?
Rituxan will definitely NOT be commercialized -- Roche is moving forward in two phase 3 trials for RRMS (OPERA 1 and OPERA 2) on the fully humanized version, Ocrelizumab). New treatments closer to commercialization are BG12, Lemtrada and Teriflunomide. Manufacturing beta interferon is labor intensive because there are many steps:

1) Obtaining an interferon beta gene by extracting the gene coding for interferon beta from human fibroblast cells.
2) Modifying the gene
3) Making a recombinant DNA molecule
4) Adding the recombinant DNA molecule to a bacterium
5) Producing interferon beta-1b by bacterial fermentation6) Extracting interferon beta-1b from the ferment
7) Adding inactive ingredients

The machines to do all this are different than those used to make other biologics.

· The cross-read to Copaxone. We understand Copaxone was initially included in the scope of the beta interferon guidance and eventually excluded. Synthon's clinical trial of Copaxone and commentary from market participants suggest to us EMA requirements for Copaxone are similar to beta interferon, but the approval would be as a generic, not a biosimilar. We also understand some companies are challenging this guidance.

My question: Can you shed some light on why some companies are challenging this guidance? What significance, if any, is there for you that EMA is not considering Copaxone as a biosim but as a generic?

Synthon's protocol for their trial of GTR is more stringent EMA's draft guideline on similar biological medicinal products containing interferon beta because GTR is being compared to Copaxone for 24 months (instead of 12) and the placebo arm lasts 9 months (instead of 4).

Current manufacturers of Beta interferons will challenge the EMA's draft guidelines because it opens a route for biosimilars, and thus threatens their market, just as Teva opposes a route for biosimilar Glatiramer Acetate. The definition of a generic vs. a biosimilar will be meaningful to prescribers and patients, as genetics are often prejudiced as being not as good as the original branded products. On the other hand, the term "generic" is one that the Public is well aware of, and is sometimes equated with "the same but cheaper."


The second report in this series regarding biosimilars may be found here.


- Dr. Daniel Kantor, MD
Medical Director
Neurologique

info@neurologique.org
www.neurologique.org