Editor's pick: Originally published March 11, 2016
I devote a good portion of this week's Biotech Stock Mailbag to a post-mortem on the Celldex Therapeutics (CLDX) Rintega blow-up, but first, a look ahead at the next important data catalyst from Biomarin Pharmaceuticals (BMRN) .
Andrew O. writes, "Adam, I found your preview of the Biomarin Batten disease drug trial very helpful. I was hoping you'd do something similar for the Peg-Pal results which should be coming soon as well."
Next up for Biomarin is the release of top-line results from a phase III study of pegvaliase, also known as Peg Pal, in adults with phenylketonuria, or PKU. Biomarin has guided to an announcement of study results in March. If positive, the pegvaliase data will form the basis of a U.S. approval filing.
People with phenylketonuria have a genetic deficiency in an enzyme which prevents the breakdown of phenylalanine, or Phe, in the blood. Phe is present in protein-rich foods and is an essential amino acid in the human diet, but excess levels, if not lowered, can lead to intellectual disability and other serious neurological problems.
There is no cure for PKU, but high Phe levels can be managed with a special, restrictive diet. One of Biomarin's currently approved products, Kuvan, when used in combination with a restrictive diet, lowers Phe to safe levels in about 40% to 50% of patients. Kuvan is a pill. Sales in 2015 increased 18% to $239 million. The company forecasts 2016 Kuvan sales in the range of $320 million to $350 million.
Pegvaliase is a daily injection which also works by lowering Phe levels, although it's designed to be more potent than Kuvan and obviate the need for a restricted diet. Biomarin believes pegvaliase will be used by PKU patients, particularly adults, with more severe disease than those who use Kuvan.
The phase III study uses a randomized, discontinuation design. All enrolled PKU patients start on pegvaliase injections. Those who respond, defined as a minimum 20% Phe reduction, are then randomized to either continue with pegvaliase or switch to a placebo. The primary endpoint measures the difference in Phe levels between the two patient arms after eight weeks.
Based on prior data and interim results from this phase III study, investors are most definitely expecting the Phe-lowering efficacy of pegvaliase to be confirmed when Biomarin reports results later this month.
More controversial is the ability of pegvaliase to improve neurocognitive symptoms associated with PKU. Key secondary efficacy endpoints in the study measure hyperactivity and inattentiveness. Biomarin has said FDA wants to see a statistically significant reduction in Phe levels and an improvement trend in neurocognitive function to satisfy requirements for full approval.
Pegvaliase is not an easy drug for PKU patients to tolerate. Almost every patient injected with the drug suffers from hypersensitivity reactions, which forced Biomarin into a prolonged clinical development program to come up with better, safer dosing procedures. As such, the safety profile from the phase III study will be important to watch.
The market potential for pegvaliase is also up for debate. Biomarin says peak commercial sales could reach $1 billion. Investors I spoke with are more cautious. Daily injections are not something PKU patients are accustomed to dealing with, particularly if the shots are not well tolerated. Uptake of pegvaliase will depend on the amount of Phe lowering and the extent to which the drug improves neurocognitive function.
Tobias F. asks, "I was surprised by the Celldex failure and was hoping you could offer some tips on how to avoid a situation like this in the future. Thank you."
I didn't predict the outright success of the Rintega phase III study but I was certainly bullish on Celldex largely because I, like you, believed Rintega had a decent shot at success in the front-line glioblastoma phase III study. And like you, I was wrong.
I'm not surprised by the premature stop to the Rintega study because failure in biotech happens. You have to plan for it.
I was surprised, however, at the drug's total lack of efficacy. Not to get too jargony on you, but the reported overall survival hazard ratio of 0.99 at the second interim analysis means Rintega had essentially no beneficial effect at all compared to the control arm.
It's often the case that an experimental cancer drug prolongs survival but not enough to reach statistical significance in a randomized clinical trial. Celldex's Rintega, on the other hand, was a complete whiff.
Investors choosing to buy and hold speculative biotech stocks as the companies go into binary events like phase III clinical trials can never protect themselves 100% against blow-ups. But the Rintega failure is a good reminder that using historical survival data to predict the outcome of an experimental drug and clinical trial is often no better than just throwing darts blindfolded.
Celldex conducted multiple phase II studies of Rintega in front-line glioblastoma patients which generally showed overall median survival in the range of 21 to 24 months. None of these studies included a control arm, so Celldex relied on data from a previously conducted study which showed similar, or "matched," patients showing overall survival in the range of 14 to 16 months, depending on how the historical comparisons were made.
Celldex believed replicating this difference in survival in its phase III study would yield a positive result and lead to Rintega' approval. That assumption turned out wrong.
In the phase III study, Rintega's median overall survival was 20.4 months, a bit lower than expected but generally within expectations. However, the control arm patients reported median overall survival of 21.1 months, well above Celldex's assumptions.
The positive efficacy signal for Rintega that Celldex thought it saw in the phase II studies against historical control turned out to be a mirage.
Investor pro tip: Consider historical control data very carefully and skeptically.
Lori B. writes, "How did you get Celldex so wrong? You should be embarrassed by your Celldex pumping. It also shows how wrong you are with the constant bashing of Northwest Biotherapeutics (NWBO) , which is now going to dominate the market when approved. At this point, you should think about finding a new job, perhaps dog catcher since you like to post so many dog pics on Twitter."
The dog catcher insult was a real zinger.
I did get Celldex wrong, but not because I fell for the overly optimistic historical controls mentioned above. Instead, my mistake was having too much confidence in the survival benefit reported in the previously conducted REACT study, which compared Rintega and Avastin against Avastin alone in recurrent GBM patients. I believed the positive efficacy from the REACT study in more advanced, harder-to-treat GBM patients was a confidence-boosting signal for the front-line study.
Others, including the FDA, didn't have great confidence in the REACT data, which turned out to be the right call.
I also need to tip my cap to the fund manager who walked me through a very detailed (and slightly over my head) scientific explanation for why Rintega was a total failure. He dug deep into old Rintega data presentations and published papers to figure out that Rintega's immune response was too weak and too transient to have significant impact on brain tumors. On this basis, he was short Celldex.
I bring this up only to counter the popular idea that all biotech short sellers are devious stock manipulators. No, they're often really smart guys and gals who use their Ph.D.s and M.D.s to identify dislocations between stock market value and scientific reality.As for Northwest Bio, the obvious take-home message from the Rintega blow up is that the prospects for DCVax, the company's already-troubled brain tumor vaccine, have dimmed even more.