Pharma’s Cutting Edge

I’m pleased to report that NRDD has published as an advanced online publication the above article that will appear in print in the April edition of the journal. A PDF is also available at the PGS website: http://www.pharmagrowth.com/library.html

As you’ll see, the article is a combination of a brief overview of what is known about market entry order and eventual commercial success, particularly as it relates to pharmaceuticals, and an exploration of additional factors that describe and hopefully illuminate this relationship. As you’ll see, first to market is not always best, at least as far as share of revenues is concerned. Yet that hasn’t stopped most firms from furiously pursuing a pioneer strategy for nearly all of their products. Exactly when firms should pursue the less costly and less risky watch-and-wait (and learn!) strategy remains to be fully fleshed out, but what is clear to me at least is that the decision will rarely be obvious.

The issue described in this article is part of the much broader issue of best practices for strategic thinking for complex, risky, protracted scientifically based innovation. Ill-informed intuition probably still has a role in modern pharmaceutical innovation strategy, but I strongly advocate the idea that science must be applied to the management of scientific innovation if firms are to improve research productivity meaningfully. Measurements (metrics) alone are not enough. Indeed, metrics can be misleading if used to advance a preconceived outcome, as is usual practice. Measurements must be preceded by hypotheses and followed by tests of the hypotheses to be successfully applied. Unsuccessful hypotheses must be rejected. Models incorporating metrics and relationship assumptions derived from successful hypotheses should be created and then used to drive decision-making rather than decision-justification. Lacking adequate quantitative models, strategists might resort to intuition, but at least it should be well-informed.

According to a WSJ article (not linked here but see this story from the Washington Post), investigators believed it was unethical to withhold high-dose statins from the high-risk patients enrolled in this study, so the study was done without a control arm. My question to them: Is it more ethical to conduct a clinical study knowing in advance that its results will do little to advance the state of scientific knowledge significantly? But, you say, the study was published in JAMA and Steve Nissen said this was the first time a statin actually shrunk plaque and raised HDL-C significantly, so it must be a scientific/medical advance. How could it not be an advance? Because, I reply, there was no control group. I don’t care if Osler ran the study and published it in Nature. If a clinical study makes no attempt to control for the introduction of all sorts of biases by comparing the test drug to a concurrent therapy (including the test drug itself, if done properly), its results must be regarded as preliminary, speculative, and probably not very newsworthy.

The proper choice of a control group is increasingly difficult to make as therapeutic advances limit the comparator choices that provide requisite equipoise between groups. But that difficulty isn’t an excuse to abandon a control arm altogether. If readers are interested, I’ll be happy to discuss some ways of designing controlled studies in a future PCE post.

The post your are looking for may now be found here:

 http://pharmaweblog.com/blog/2006/03/10/emea-biosimilars-guidance/

Thank you.

Pharma’s Cutting Edge Vol. 4 Number 3 - March 2006

Review of EMEA Biosimilars Guidelines

Last month EMEA (the European Medicines Agency) released final guidelines containing details of clinical, nonclinical and quality expectations for biosimilar protein therapeutics (aka generic biologics, aka biogenerics, aka follow-on biologics or proteins). These guidelines were themselves as expansion of the general guideline released in September 2005 and of two earlier documents, a note for guidance containing nonclinical and clinical issues (December 2003) and a quality guideline (also December 2003). All of these documents are conveniently located on a single page on the EMEA website.

As of today, two biosimilar proteins have been recommended for approval by the CHMP (Committee for Human Medicinal Products, the human medicines technical component
within EMEA), both somatropins (growth hormones): Omnitrope (Sandoz) and Valtropin (BioPartners). Based on CHMP’s specific guidelines for insulin and G-CSF biosimilars, we can infer that registration dossiers have been submitted for biosimilars in these categories (potentially adversely affecting primarily Novo Nordisk and Hoffman-La Roche, respectively). Other widely used therapeutic proteins that are susceptible to biosimilar competition in Europe today, owing to patent expiration, are erythropoeitin (Eprex, Janssen) and interferon-alpha2b (IntronA, Schering-Plough). We can expect CHMP guidelines for biosimilars for these drugs soon. In addition to Sandoz and BioPartners, companies that stand to gain in the near-term from the EMEA’s actions are listed in this PharmaWeek article.

In the U.S., FDA has been considering its own guidance for follow-on biologics (biosimilar) developers and manufacturers for some time. It’s hinted that a draft guidance will be forthcoming this year. Sandoz isn’t holding its breath waiting for FDA to get its act together on this issue. It sued FDA in September last year, asking the Court to rule that FDA must act on its application for Omnitrope now, one way or another. FDA asked the Court to dismiss the suit in November. So far, no further action has occurred. In the meantime, U.S. observers can learn from the CHMP’s guidelines, as it is likely that scientific guidance from FDA will be similar.

My goal in this issue is to provide an overview to these recent guidelines, focusing on the key issues. I’ll discuss them in the order of general guideline, nonclinical, clinical and quality and then touch on the specific product guidelines. I realize that this is a long post, but it’s far shorter than the original guideline documents, and it isn’t every year that a completely new category of drug marketing authorization comes to fruition.

General Guideline

Scope

• Theoretically covers all biologics, including vaccines, blood-derived products, antibodies, etc.

Application of “similar biological medicinal products approach:”

• differs from the standard generics approach as it is based on a comparability exercise rather than a demonstration of bioequivalence
• is more likely to be applied (and to be successful with) highly purified biologicals, such as recombinant proteins
• must fulfill quality, safety and efficacy requirements (discussed later)
• results in products that are similar but not in generic equivalents, necessitating the need for clear product identification to facilitate pharmacovigilance.

Choice of reference product

• Active substances must be similar (structure and function).
• Reference product must be approved in the Community (meaning the countries embodied by the EMEA).
• Pharmaceutical form, strength, and route should be the same; differences will have to be justified.

Note on immunologicals (e.g. vaccines)

• Vaccines and allergens are complex and unlikely to be well characterized molecularly.
• They will therefore consider them on a case-by-case basis.

Note on blood or plasma-derived products

• Such products have complex and variable physico-chemical, biological and functional characteristics.
• Therefore, such products must satisfy safety and efficacy requirements as found in BPWG “new products” guidance note and related documents.

Other biologicals

• Cell and gene therapy products will be considered in the future.

My take away: Although the similar biological approach is potentially relevant to all biological medications, in fact, the current guidelines are intended primarily for development of biosimilars referenced to well-characterized recombinant protein therapeutics, including monocloncal antibodies.

Nonclinical Guideline

General (applicable to clinical as well)

• Within a biosimilar MAA (marketing authorization application) dossier, each indication for use will be considered separately.
• Efficacy and safety for each indication will either have to be demonstrated or an extrapolation from one indication to another justified.

Scope (applicable to clinical as well)

• Covers biosimilar applications for recombinant proteins only and only when no changes are introduced in the manufacturing process.

Nonclinical study program

• Studies should be done prior to clinical studies.
• Should be designed to detect response differences between the reference and biosimilar, not just responses per se.
• In vitro receptor-binding or cell-based [binding] assays should normally be done.
• In vivo studies should measure the pharmacodynamics relevant to clinical use.
• At least one repeat-dose toxicity study should be conducted.
• Toxicokinetics should include antibody titers, cross reactivity and neutralizing capacity
• Other routine types of tox studies are NOT normally required.

Clinical Guideline

General

• Test product should be studied clinically using the final manufacturing process, unless justified and supported by additional data.
• Clinical comparability is done in stages, much like a traditional program.

Pharmacokinetics

• Clinical comparability pharmacokinetic (PK) studies should NOT necessarily mimic a standard comparability design that focuses on absorption and bioavailability. Rather, such studies should also include exploration of clearance and elimination.
• Choice of design must be justified and careful attention should be paid to issues such as elimination half-life.
• The acceptance range to conclude clinical comparability should be defined and justified.

Pharmacodynamics

• The pharmacodynamic (PD) effects should be compared in a population where “possible differences can best be observed.”

When will PK/PD studies alone suffice to determine clinical comparability:

• when PK of the reference product is well characterized
• when PD properties (e.g. binding to receptor) of the reference product are well characterized
• when the PK/PD relationship of the reference is well characterized
• when a PD marker exists and the PK/PD relationship using the marker is well known.
• when the PD marker is a well characterized surrogate of the clinical outcome of interest
• when the dose range is chosen to demonstrate assay sensitivity (see ICH E10 for a discussion of assay sensitivity)
• when the PK and PD margins defining clinical comparability are defined and justified a priori. Read the rest of this entry »
  Sphere: Related Content
  
  

Anyone who has ever had a frank discussion with an office-based oncologist or rheumatologist is well aware that reimbursement structures for office-infused drugs sway the choice of drugs used. Pharmaceutical and biotechnology companies are certainly well aware of that fact. Now a study in Health Affairs has apparently put the practice into the peer-reviewed literature. Medicare’s new method of reimbursing for office infusions should help diminish the incentives for doctors to prescribe preferentially well-reimbursed drugs. It would be worthwhile for CMS to study the issue itself to determine if its policies are working. For those who are interested in this issue, take a look also at this article published last year describing how differences in Parts D and B of Medicare reimbursement create perverse prescribing incentives.