Biovest A Risky Bet

Nastech, now known as MDRNA (MRNA), is a good example of how tough biotech investing can be. Some smart folks invested in the company, based upon claims by its CEO of a revolutionary drug delivery system that big pharma just couldn’t wait to get its hands on. The claims seemed to be backed up by deeds, in that big pharma did sign deals to investigate the delivery systems.

That approach failed miserably, as the company has essentially abandoned the nasal delivery technology in favor of pursuing an RNA interference strategy. They demoted their CEO, who was the reason I avoided the company, and hired the former CEO of Sirna Therapeutics – which sold itself to Merck (MRK) for a billion dollars a couple of years back.

That brings some instant credibility to the claims MDRNA has legitimate RNAi technology, but critics point to the fact that companies in the RNAi field live and die by patents and MDRNA seems to be a little short in that area. I’m not sure MDRNA needs to have core RNAi patents, however, as they appear to be focusing more on how to deliver RNAi drugs to patients. The issue of delivery is the #1 problem in the RNAi space. If MDRNA has a solution, they could see very significant value growth.

Of course, this is a story we’ve heard already from the company. Encapsulating drugs for easy delivery by nasal spray was the company’s claim to fame – several product failures and two partnership failures ago. I would expect investors to be a little gun-shy this time around, waiting for major investments in MDRNA until the company proves its largely theoretical RNAi delivery system in human trials.

I’ll reiterate my prediction from before – MDRNA’s corporate presentations will be at least as much about the $1B buyout of Sirna and the recent $1B-plus deal RNAi deal between Japan’s Takeda and Alnylam (ALNY) as it will about their own technology. Smart investors will realize there is a big gap between those deals – which involved patent-protected RNAi core technologies – and MDRNA, which is looking only at drug delivery.

When a company blows up investors like Nastech/MDRNA has, it’s always best to make the management team prove they have something in hand instead of investing on the promise they might develop something later.

EGFR Inhibitors

June is a big month for the EGFR inhibitor space. EGFR inhibitors are used to treat cancer. The two prominent players in the space are ImClone’s (IMCL) Erbitux and Amgen’s (AMGN) Vectibix.

At the recent American Society of Clinical Oncology (ASCO) meeting, we learned all EGFR drugs have an Achilles’ heel in that they do not work in a 20-40% portion of the target disease population. The cancers in these patients have a genetic shift called a “k-ras mutation” that prevents the EGFR class of drug from working. The FDA is likely to change the label on these drugs very soon, and there should be FDA-approved tests for k-ras mutations on the market by the end of the year.

There is actually a third EGFR drug approved for patient use, though approvals are largely limited to second and third-world countries. The drug is called nimotuzumab, and has been licensed from Cuba for marketing in the U.S. and Europe by Canadian biotech YM Biosciences (YMI). Nimotuzumab is controversial because the drug does not cause the rash that is thought to be a sign of efficacy for patients given either Erbitux or Vectibix. Nimotuzumab’s pedigree as a Cuban-developed drug hasn’t helped, either.

By the end of the month, however, YM will release data from a Phase II trial that mimics the BOND-1 trial used to approve Erbitux. YM made the dumb choice not to randomize this Phase II trial, so direct comparisons will be difficult. Nevertheless, if nimotuzumab can approximate the results seen with the Erbitux arm in this trial it will go a long way towards legitimizing the drug in the minds of the scientific and investment communities. The smart money is looking for an overall response rate (percentage of patients whose tumors shrink over a set percentage) above 20% and a continued lack of rash.

The stock trades around $1. If the data are bad, then it will go much lower (> 50% decline). There are January 2009 options available for pennies for those with the expertise to play options. I think the drug works, but I’m not sure if the current trial will show it clearly. The trial was designed and enrolled before proof these drugs don’t work in patients with k-ras mutations. If the trial enrolled many of these patients by dumb bad luck, then figuring out the results will be difficult.
 

 And I Thought I’d Seen It All

Biovest (BVTI), which is apparently part of Accentia Biopharma (ABPI), lit up the biotech "expert" IM and e-mail lines earlier this week when they released “blinded” disease-free survival Kaplan-Meier curves for their trial of BiovaxID, a vaccine for non-Hodgkin’s lymphoma (NHL).

We’ve seen blinded data released before, most notoriously what Cell Therapeutics (CTIC) did with their drug Xyotax. More smart biotech investors were hooked by that nonsense argument than any in recent memory. Biovest is the first time either I or any of the other biotech folks I converse with regularly can ever remember seeing blinded, paired Kaplan-Meier curves.

A Kaplan-Meier curve (“KM curve”) is simply a way to represent the results of a trial. The y-axis is the percentage of patients who have not yet achieved some event. The x-axis is time. The results of patients in both the study and the control arm are plotted on the same graph.

A KM curve is a handy visual representation of the results of a trial. Where the curves cross at the 0.50 (50%) mark tells you the median value (usually survival) for either arm of the trial. The gap between the two lines tells you the magnitude of difference in effect of the treatments used in either arm of the trial. The steepness of the curves can tell you things about the magnitude of benefit of the drugs studied, especially in oncology studies.

KM curves are useless unless you know which line belongs to what arm. That’s why Biovest’s release of “blinded” KM curves was so unusual. Both curves are there, but investors are left to guess which one is BiovaxID and which is the control arm. Management wants you to assume the upper curve is BiovaxID, of course, since that is the better curve on the graph.

BiovaxID supporters think that’s the case, judging from how fast they bid up the company’s shares. They say their enthusiasm is supported by a release in April where the company said the independent Board (DSMB) monitoring the trial looked at these unblinded results and recommended presenting the data to regulatory agencies. That release also caused a round of head scratching among the crew of biotech folks I chat with, because legitimate DSMBs never do that sort of thing.

Smart investors will read that April release very carefully. The press release headline suggests the DSMB said to apply for marketing approval based on the unblinded BiovaxID data they reviewed. The quote from the head of the DSMB says nothing of the sort. He says the companies should present the data to regulatory agencies as part of the regulatory pathway for the drug. Since all data has to be presented to regulatory authorities – whether for approval or to design a confirmatory pivotal trial – that isn’t saying much.

The KM chart that accompanied the most recent release has some fine print worth reading, too.  

Note the x-axis is labeled “months from 6 months post randomization.” What happened in the first six months of the trial? That’s not a trivial question since the FDA and EMEA will certainly want to know. Moreover, the FDA will not allow the company to just ignore it. An overriding principle of data presentation at the FDA is to measure from randomization – especially in oncology trials. My guess is the first six months were excluded to try and get around a central problem in vaccines that require bits of a tumor to create – manufacturing failures.

My guess is the trial, as measured from the point of randomization (i.e. with the first six months put back into the graph), will not be statistically significant in favor of BiovaxID. There might be an apparent late effect to taking the drug, but data collection by this small company will not be robust enough to determine whether the late separation in the curve is due to BiovaxID or due to subsequent medications given to the patients. In any case, the FDA will call the data “hypothesis generating” and require a new trial. I would be really surprised if the EMEA didn’t do the same.

With two oddball news releases in two months, the story will be worth watching for entertainment value at the least. Unless you have a history of successful investing in biotech stories whose chief benefit is entertainment, I’d be careful on this one.