Was I Wrong, Or Do We Have A Failure to Communicate?

I have been wrong a few times in the sixty-plus years I’ve been on Earth. Was I wrong about the risk of technical failure in biosimilar development in my June26th post on virtual biosimilar development? To be clear, what I mean by technical failure is that the molecule produced and taken through testing turns out to be non-comparable to the original biologic, either because it has less effectiveness or greater risks of adverse events, including immunogenicity. Therecent news about the 40% “first-cycle” approval rate for the first ten biosimilar applications reviewed by the FDA caused me to take another look at the risk of technical failure. I do not believe that the startling low first cycle approval rate indicates that the technical difficulty of creating a reasonably good biosimilar is very high, and that the better explanation for the low initial approval rate may be expressed in one of my very favorite films of all time, “Cool Hand Luke.” “Captain,” the brutal Warden of the chain gang prison to which Cool Hand Luke is sentenced, explains the need to punish Luke by saying “What we’ve got here is a failure to communicate.” While I would never ever suggest any real similarity or parallel between the FDA and that prison, it does seem to me that the best explanation for the low first-pass approval rate for biosimilar applications is indeed a failure to communicate. A closer look at the six biosimilar applications which received Complete Response Letters (CRLs) instead of approval helps to clarify the issue. Here is a list of the biosimilars that received a CRL and the basis for the CRL:

1. Hospira’s Epoetin zeta approval was delayed by manufacturing concerns. There were no concerns about the actual molecules safety or efficacy (the drug has been on the market in Europe for some time).

2. Pfizer’s Epoetin alph was also delayed because of manufacturing concerns; previous requests for more clinical data were satisfied. So this molecule also is approvable by the FDA.

There were three different applications for biosimilars for Pegfilgrastim and all have been delayed.

3. Coherus’ Pegfilgrastim was delayed by the FDA’s request for a reanalysis of a subset of samples with a revised immunogenicity assay and to provide additional manufacturing-related process information. Clinical data on efficacy and adverse reactions apparently were apparently satisfactory to the FDA. So although is possible that the revised immunogenicity assay will disclose some issue with the molecule, it is more likely that the delays in the Pegfilgrastim biosimilar applications indicate that the FDA is simply unsure of exactly how to assess immunogenicity for pegylated biosimilar proteins.

4. Apotex’s Pegfilgrastim­­––Apotex has not disclosed the concerns raised by the FDA in its CRL.

5. Sandoz’s Pegfilgrastim––There also was no disclosure by Sandoz of the concerns expressed by the FDA in its CRL. However, the data from 3 clinical trials, including two trials for safety and efficacy in breast cancer patients would seem to demonstrate that the molecule is safe and effective in treating chemotherapy-induced neutropenia. Here too it appears that the FDA has additional concerns to be addressed for pegylated proteins and is still considering how to best resolve those concerns.

6. Celtrion’s infliximab was approved in the second cycle after the provision of additional data. This proves that the molecule was a reasonable and functional biosimilar but that Celtrion and the FDA had not reached an earlier understanding as to exactly what needed to be included in the biosimilar application.

There are two conclusions that can be drawn from this limited data set of the first ten applications. First, manufacturing issues are always a concern for biologics. Biosimilar manufacturers need to make a good biosimilar, but they also need to make it in an impeccable facility. Second, the FDA has concerns about pegylated proteins (such as Pegfilgrastim) that are still evolving. Manufacturing issues aside, the developers of cytokines and antibody biosimilars have produced approvable molecules. The 29 biosimilars on the market in the EU with very few (if any) involuntary withdrawals for safety are clear evidence of the relative rate technical success. The European experience with biosimilars points to the conclusion that the risk of technical failure in biosimilar development is low. The early going at the FDA represents a learning curve in the data collection, submission, and review process for the biosimilar developers and the FDA. I do not believe I was wrong about the technical difficulty of creating a biosimilar. What we’ve got here is a failure to communicate precisely what needs to be done to obtain the FDA’s approval.

FDA Approves Endari for Sickle Cell Disease

It should not come as a surprise to readers of this blog- Emmaus Medical was given FDA approval for it's l-glutamine treatment for sickle cell disease.  I posted my analysis of the then-pending application back on June 17th: Good News for Sickle-Cell Patients, But How Will Emmaus Make Money? A Look at the Business Side of Pharmaceutical Policy.  As I stated then- it will be very interesting to see whether Emmaus is able to successfully complete with other sources of l-glutamine.  I look forward to Emmaus' quarterly reports.

BIO 2017: Part II ––Virtual Development Is A Reality

In my June 26th post on BIO 2017 I discussed how developing a biosimilar could now be done by a virtual company, with minimal staff overseeing the project and contracting with outside vendors for every step in the process from cell-line development through regulatory approval. The giant Exhibition Hall for BIO 2017 provided a showcase for a number of vendors who can perform the necessary parts of the biosimilar development process.  I concluded that post by noting that although a virtual company could oversee all of the development of, for example, a biosimilar Remicade/infliximab there would certainly be some level of risk in at least 3 areas. First is the straightforward risk of clinical failure. Clinical testing of the biosimilar could reveal that the product either is not comparably effective or is immunogenic, despite pre-clinical analysis of comparability. There have now been a sufficient number of companies entering into the development of biosimilars and a sufficient number of biosimilars approved in Europe (35 as of June 16, 2017) for the risk of clinical failure to be at least roughly estimated. A more complete analysis of companies abandoning particular biosimilar development efforts or failing to win approval is beyond the scope of this post, but it is now reasonably clear that while the risk of failure is relatively low.  There is some risk of clinical failure but it is certainly much lower than is the case for new small molecules or pioneer drugs. The EMA and FDA have responded to very few biosimilars applications either with an outright rejection of an application or a demand for further clinical trials.

The second risk is the competition risk. With multiple biosimilars for the same original biologic (as is would be the case for our hypothetical Remicade biosimilar), selling sufficient quantities of the biosimilar will require extensive marketing efforts and possibly price discounting to persuade doctors and payers to prefer our biosimilar over the others. Unlike the risk of technical failure, the extent of the market risk is both difficult to estimate, but, in my view, completely manageable. The solution to the market risk for a virtual company is simply to enter into an agreement with a major buyer at the very beginning of the effort. For example, under an alternative business model, the virtual development would undertake procure a supply contract with a major payer such as United Healthcare, at a pre-agreed deep discount, before beginning to incur significant development expenses. This approach eliminates market risk and most marketing expenses.

In addition to the relatively low risk of technical failure and the manageable market competition risk, there is one more significant risk to be considered, which is the risk presented by expensive patent infringement litigation with no certainty concerning the eventual outcome. The current thicket of large numbers of process patents surrounding most pioneer biologics will provide far too much fuel for patent litigation and unpredictability to the process. Although I am generally wary of proposing legislative fixes for problems, it seems to me that one may be needed here, in the form of an outer limit on the duration of patent-protection and market exclusivity for the maker of a reference biologic. One of the most hotly debated provisions of the Biosimilar Competition and Innovation Act (BCPIA) was Section 7(A), which provides a twelve year of period of exclusivity during which no biosimilar applicant may be approved simply on the basis of its comparability to the reference, or pioneer, biologic. Without rehashing that debate, it is interesting to speculate on at least one likely explanation for that 12-year exclusivity period: it is remarkably similar to the average period of market exclusivity that is enjoyed by branded small molecule drugs with reasonably large markets. According to a study by Grabowski, Long, and Mortimer the average period of market exclusivity for the branded small-molecule drugs through 2013-2014 was 12.5 years. However, it is increasingly clear that the BCPIA, with its complex “patent dance” provisions, leaves the development of biosimilars open to significant risks in the actual infringement litigation. If 12 years is not the outer limit of market exclusivity, then a slightly longer period, for example 15 years after market entry, certainly is more than long enough for any biologics reference drug maker to have been rewarded very handsomely for its pioneering efforts. Amgen is currently suing Hospira for infringement of process patents that Amgen holds for the making of erythropoietin, which it markets under the brand name Epogen. Amgen’s Epogen, its original blockbuster drug, was first approved by the FDA in 1989. Although some of Epogen’s patents may have been filed under the former system of 17 years from date of issue rather than 20 years from filing, it is close to stunning that twenty-eight years later Amgen is still seeking to defend its market exclusivity by claiming patent infringement. While I am certain that my proposed 15-year limit after market entry on the right to bring infringement actions against any biosimilar applicant will be debated as too long or too short, can anyone seriously defend 28 years, or even 20? Virtual biosimilar development is now possible, and could offer very significant efficiencies and savings. The technical failure risk and the market competition risk are ordinary business risks that can be factored into a business plan and dealt with. However, the risk of costly and uncertain patent infringement litigation for a biosimilar drug needs more certainty than the BCPIA provides. Fifteen years is enough for pioneers and would give the healthcare system a real chance at significant and desperately needed savings.