Pharma’s Cutting Edge

In the June 2nd paper issue of BusinessWeek (published online 5/21) the article “Cancer’s Cruel Economics” by Catherine Arnst provides a high-level look at the difficulty some small copmpanies are facing getting their cancer therapies approved in the US. 

The focus is on cancer immunotherapies, particularly Antigenics’ Oncophage.  I last discussed Oncophage in 2006, after the first report of its Phase 3 results.  I have also devoted space in this forum to other cancer therapies mentioned in the BusinessWeek article including Dendreon’s Provenge and Genitope’s MyVax.

I was intrigued by a quote attributed to Richard Pazdur, head of CDER’s Oncology review division:

“[Post hoc subgroup analysis for differential treatment effect] is like shooting an arrow and then painting the bull’s-eye around it,” says Pazdur. “You cannot use subset analysis to salvage a failed trial.”

Pazdur’s concern regarding treatment effect inferences derived from post hoc subgroup (subset) analyses rests on firm grounds, but the quote suggests a black and white attitude towards their utility, without any room for compromise.  That’s too bad, because the rule of thumb Pazdur is apparently using to reject subgroup evidence of efficacy is imperfect, undoubtedly resulting in the rejection of some effective therapies.

I’m not going to write a manuscript-length post describing the many risks inherent in inferential subgroup analyses.  There are many published reviews you can find that do that.  Suffice to say that the risks of both false-negative and false-positive inferences are inflated with subgroup analyses relative to the main analysis (primary hypothesis test), whether the analyses are pre hoc (defined before the trial results accrue) or post hoc (sometimes called retrospective).  Pre hoc analyses are less susceptible to Pazdur’s target drawing, especially when the specifics of the subgroup are rigorously pre-defined than post hoc, and so they are preferred by regulators.

What I’ve found to be less well represented in the literature is a situation in which the weight of evidence presented in the subgroup analysis is sufficient to, as Pazdur says, “rescue a failed study.”  I’ll not focus specifically on Provenge or Oncophage, but the example from the literature I’ll cite is relevant to both.

In order to determine whether any subgroup analyses provide evidence sufficient to warrant drug approval, it is necessary to first know the expected false-positive rate of a subgroup analysis, given a false-positive rate of 5% in the overall (main) analysis.  A 5% rate is chosen, because that rate is usually considered an acceptable one by clinical practitioners and drug regulators.  Thus, the null hypothesis is rejected falsely in 1 in 20 trials.  FDA usually requires two independent experiments (trials) for evidence of efficacy, resulting in an overall false-positive rate of 2.5% (0.05×0.05), though one statistically significant experiment with corroborating evidence from others is sometimes sufficient, particularly for accelerated approvals.

In an important study published in 2001 by the UK’s NHS, Brookes et al used simulations of 100,000 clinical trials each to determine the false-positive (and false-negative) rates of subgroup analyses for different types of study designs.  They simulated two subgroup analysis (ignoring the effects of multiple analyses, which inflate Type 1 error) and tested a variety of relative treatment effect and subgroup sizes.

The simulations showed that when there was in reality no main or subgroup effect of treatment, and the overall (main) analysis of treatment was falsely positive (i.e. null hypothesis was rejected at the nominal p<0.05) then the chance of falsely declaring one subgroup as demonstrating a treatment effect was high.  For a survival study, this chance was 61%.  In other words, with no real main treatment or subgroup effect, when there was a false-positive main effect, one of two subgroups analyzed will appear to have a treatment effect well over half the time. 

However, under the same set of circumstances, when the main effect is not rejected (i.e. a true negative inference is made), then one subgroup will show evidence of a treatment effect much less often, only 6.5% of the time, approaching the overall effect false-positive rate of 5%.  In other words, the probability of falsely rejecting a subgroup-specific null hypothesis in the absence of overall and subgroup effects is reasonably low if the overall effect is correctly negative.

Of course, the above simulation findings aren’t by themselves capable of determining whether a subgroup-specific effect is real or not.  They simply suggest that the regulator need not reject out-of-hand statistical evidence of a subgroup-differential treatment effect when evidence for an overall effect is absent, as Dr. Pazdur’s quote suggests he is willing to do in some cases. 

Evidence that the apparent subgroup effect in a survival study is real will be strengthened by the following factors:

• The main effect does not contradict the purported subgroup effect 
• The subgroup-specific analysis was defined a priori
• A significant test of interaction between overall treatment effect and the subgroup is in evidence prior to any subgroup-specific test
• The total number of subgroups analyzed is small, and, if not, an inference of treatment effect made on any one subgroup uses an appropriately conservative adjustment of the significance level
• There is strong biological plausibility for the differential subgroup effect
• The size of the subgroup is large relative to the total sample size (i.e. relatively representative of the total population)
• The conduct of the study, particularly the handling of dropouts and non-compliant subjects, creates confidence in the quality of the subgroup data

Finally, as I’ve argued before in the case of Provenge, when the evidence of efficacy is marginal, regulators have a duty to the public they serve to weigh with utmost care and without bias the risk of introducing an ineffective medicine versus the risk of withholding ready availability of an effective medicine from a gravely ill population without other treatment options. 

This is destined to be one of the key HIV papers of this decade, if not longer.

–Robert Gallo

I’m a real fan of unconventional thinking and scientists who are unafraid of venturing outside of their comfort zones.  So, when I read the story of the discovery of new potential drug targets against HIV in Science recently, I was thrilled.  I believe the full article behind the news piece requires a subscription; if you have one it’s worth at least a quick read-through.

Stephen Elledge is a pioneer in the use of “forward genetics” to screen for host factors that influence (suppress and or promote) the malignant phenotype (see this for an example of his group’s creative scientific approach).  In Elledge’s brand of forward genetics applied to cancer target screening, he and his team apply massive pools of interfering RNA to cells of various types to determine whether blocking endogenous gene transcription modulates the malignant cell phenotype. 

In the recent Science paper, Elledge and his collaborators used the same approach with a foreign invader–HIV.  I don’t know whose idea that was, but it was brilliant, and I give Elledge lots of credit for not shying away from a complex, costly, time-consuming HIV study despite his previous lack of experience working with the virus.  In short, the labs grew transformed epithelial cells (HeLa) transfected with some 21,000 pools of siRNAs in pools of 4 siRNAs per gene (i.e. siRNAs against all expressed human genes with transcription-suppresion redundancy to help gaurd against false positives and negatives) and then infected the cells with HIV-1.  When HIV-1 was able to replicate despite host-gene transcription inhibition, the scientists could conclude that the host gene probably wasn’t important for HIV replication, but if HIV replication was disrupted, they could surmise that a host protein involved in HIV replication had been discovered.  As you’ll read, the group found 273 host proteins that HIV relies upon for replication, only 36 of which had been previously discovered, thus explaining how HIV is able to so effectively parasitize its human hosts despite having a genome of just 9000 RNA bases.

As this is a scientific paper, the authors themselves have respectfully avoided aggrandizing this work and their screening approach in general, but this type of study is potentially very clinically relevant in the near-should be a more fruitful strategy to eliminate the virus from the body.  This is because HIV can easily mutate in the face of selection pressure from drugs targeting its own proteins, but it will be much harder for HIV to mutate to circumvent a blockade of a host target (or two) that it requires for reproduction. 

The accompanying news story expresses some skepticism from a Novartis researcher regarding the ease of discovering drugs targeting HDFs and the willingness of pharma to fund such work.  I frankly can’t understand such skepticism, but perhaps the story excerpted the interview in order to provide a bit of contrast to the enthusiasm from Dr. Gallo that opened the piece.  Yes, the HDFs are mostly internal cellular targets, and yes, they probably have some important roles to play in normal host functions, but let’s remember that we already use lots of drugs that fit this description, particularly in oncology, and although side effects are common, so are benefits.  As for pharma funding the work needed to capitalize on this work, let’s remember that it’s one thing to gauge the suits’ interests in the face of established pathway opportunities, and it’s quite another to anticipate their interest in the face of potential therapeutic breakthroughs with large commercial potential.   In my mind, this is a must-do opportunity for any company with virology capabilities and experience.  From a platform perspective, I’d be surprised if most large discovery outfits aren’t already taking advantage of genome-wide functional screening to identify the host factors enabling invaders to live at our expense, whether those invaders be infectious organisms or our own cells gone awry.

Results from ENHANCE are now available.  I don’t think anyone is terribly surprised, given the delay in reporting the results, that addition of ezitimibe to simvasatin (i.e. Vytorin) did not have a sigificant impact on carotid intimal media thickness compared with simvastatin (Zocor) alone in subjects with heterozygous familial hypercholesterolemia (a disorder of LDL receptors that affects about 1 in 500 people globally). 

See my prior post on the implications of ENHANCE for most current Zetia users.  As I indicated previously, these results will likely have very little impact on prescribing of Vytorin or Zetia, despite pronouncements to the contrary yesterday, primarily from Steve Nissen of the Cleveland Clinic.  Dr. Nissen apparently gave lots of interviews yesterday, saying things like:

This wraps it up…That’s all there is. There just isn’t any evidence that adding ezetimibe to simvastatin produces any advantage.

to the Washington Post.

Of course that isn’t all there is.  One trial with an iffy surrogate endpoint in a highly selected population does not a definitive treatment recommendation make (said Master Yoda).

Since my last post, I was happy to learn about another trial of Vytorin underway, one with a hard outcome.  The IMPROVE-IT trial, sponsored by Merck and Schering-Plough, is a randomized, active-control, double-blind study of subjects with stabilized high-risk acute coronary syndrome. The primary objective in IMPROVE-IT is to evaluate the clinical benefit of Vytorin 10/40mg compared with Simvastatin 40 mg alone, where clinical benefit is the reduction in the risk of the occurrence of a composite endpoint of CV death, major coronary events, and stroke.  Nice.  Now enroll this sucker.

I’m not stopping my Zetia because of ENHANCE, and I’m confident that most other users of Zetia and Vytorin will not either.  Once the ENHANCE dust settles, it’ll be clear that ezitimibe remains our first best hope for atherosclerosis therapy beyond statins, at least until a definitive study demonstrates otherwise.

Disclosures:  I have no financial ties to Merck or Schering-Plough.

Genitope’s Webcast

A week ago Genitope reported the results of a pivotal Phase 3 trial for its only clinical candidate, the idiotype-specific, personalized active immunotherapy for Non-Hodgkin’s Lymphoma, known as MyVax.  I’ve been interested in this budding therapy since around the time the company went public in early 2004, which roughly coincided with the start of the Phase 3 study.  That was back when I was testing my proclivity for biotech equity analysis, so I began following the company.  

I remember being impressed with the long clinical experience with this form of personalized therapy, which dates back some two decades and which was nurtured by no less than Dr. Ronald Levy of Stanford (who you might know as one of the founders of IDEC Pharmaceuticals and a pioneer of immunotherapy for cancer, lymphoma in particular; see this profile).  Based largely on the Phase 2 experience of Dr. Levy and others, which included a convincing number of long-term remissions with MyVax and MyVax-similar idiotype immunotherapy, I recommended investors buy Genitope stock.  I think the stock was just under $10/share then.

My feelings about the stock and the likelihood of a successful Phase 3 study changed dramatically after a second planned interim analysis of efficacy of the MyVax Phase 3 study failed to demonstrate superiority of the therapeutic vaccine over control measured as progression-free survival, but I’m getting ahead of myself.

Backing up to the beginning of the Phase 3 study, its purpose was to test the ability of MyVax to prolong PFS in patients with newly diagnosed NHL who had received one course of chemotherapy  After a 6-month rest, patients with at least a partial response to the course of chemo were randomly assigned to therapy with either MyVax immunizations (recombinant, cancer-specific idiotype protein conjugated to KLH) + GM-CSF or KLH + GM-CSF.  Multiple immunizations were given and PFS, along with a number of biomarkers, was determined.

As mentioned, a second planned interim analysis for efficacy was perfomed in July 2006.  Based on my own accrual and hazard rate assumptions, I had expected that sufficient evidence of efficacy would be seen to allow stopping the trial at that point.  However, the trial was continued.  I suspected then that the study would miss its primary endpoint.  Full planned follow-up was completed earlier this month and, as you’ll see from the market reaction, the study missed its primary endpoint.

What interests me about this story is less the ostensibly failed study and more the way Genitope spun the news into a positive story about MyVax.  “Simply put,” said CEO Dan Denney Jr., “MyVax works.”  Dr. Denney went on to explain that the primary endpoint was missed because the KLH arm did better than expected, with some patients mounting an immune response that caused prolonged PFS.  However, a secondary endpoint that was “highly statistically significant” showed that MyVax prolonged PFS in a subgroup of the subjects who mounted a “positive immune response” to the therapy when compared with those who did not mount such a response to MyVax. Read the rest of this entry »

Obesity-associated improvements in metabolic profile through expansion of adipose tissue

Obesity is bad for you, right?  Not so quick.  It’s been known for decades that deficiencies of fat in unusual to rare syndromes known as lipodystrophies (aka lipoatrophic diabetes) can cause early-onset severe insulin resistance and eventually diabetes.  Based on these cases, clinical researchers surmised that fat isn’t necessarily bad for you–in fact you need some minimum amount of fat to have normal metabolism–and abnormal patterns of fat accumulation, such as an acquired loss of fat in subcutaneous tissues, can trigger insulin resistance and diabetes. 

Since those early clinical studies of lipoatrophic diabetes, many clinical and animal studies have demonstrated that the above is consistently true, and have recently led to a hypothesis that visceral fat (i.e. the fat surrounding internal organs, particularly in the abdomen) is harmful, whereas subcutaneous fat is not and may even help protect against diabetes.  The prevailing theory devised to account for the visceral/subcutaneous fat dichotomy is that triglyceride accumulation within subcutaneous fat serves as a kind of sink that shields vital organs from the sometimes harmful metabolic consequences of intracellular triglycerides (I say sometimes because elite athletes accumulate intracellular fat in muscle tissue without apparent harmful consequences).  In contrast, visceral fat accumulation leads to harmful metabolic consequences because venous drainage of visceral fat feeds fatty acids and glycerol to the liver, where they are reformed in hepatocytes into triglycerides.  Hepatic steatosis results, leading to insulin resistance.  For some unknown reason, says the lipotoxicity theory of diabetes, some people who consume an excess of energy relative to their energy expenditure, are able to accumulate visceral fat more effectively than subcutaneous fat and develop insulin resistance and beta cell failure (as a consequence of fat accumulation in the pancreas), leading to diabetes.  Aside from the “unknown reason” bit, it’s an elegant theory supported by a large body of observational and experimental data in multiple species, including humans.  What has been lacking, however, is an experimental model demonstrating that unconstrained subcutaneous fat expansion allows for development of obesity uncoupled from harmful metabolic consequences (i.e. insulin resistance, beta cell failure and diabetes).

In yesterday’s JCI (link above), Kim et al from Philpp Scherer’s group at UT Southwestern have published detailed descriprions of just such a model.  They created the unusual obesity model by breeding a transgenic mouse that constitutively overexpresses the hormone adiponectin onto an ob/ob mouse background.  The ob/ob mouse is a well described model of congenital leptin deficiency that results in morbid obesity and insulin-resistant diabetes.  Adiponectin is one of the hormones whose expression increases in response to PPAR gamma agonists, like the TZDs pioglitazone (Actos) and rosiglitazone (Avandia), drugs that increase fat mass, primarily subcutaneously, and that also improve insulin resistance.  As it turns out overexpressing adiponectin in leptin-deficient mice leads to diminished food intake (relative to body size) and diminished calorie expenditure (lower activity and lower body temperature) relative to leptin-deficiency alone.  It also leads to fat accumulation and even more weight gain, resulting in massively obese mice.  The balance of energy expenditure and fat accumulation without a change in fat clearance rate suggests enhanced efficiency of fat generation as an important mechanism for obesity in the mice.  Despite the weight gain and fat accumulation, the mice were protected from insulin resistance, beta cell failure and diabetes.  Furthermore, the triglyceride accumulation seen in the organs of leptin-deficient animals was absent in the transgenic animals, as was the associated inflammation (thought to be related to the lipotoxicity of intracellular fat).

Of interest clinically, the transgenic animals, including those bred on a WT background, also developed cardiomyopathy without fat accumulation in heart muscle.  As you’ll recall, as of August, the TZDs now have black-box warnings in their US labels regarding an increased risk of heart failure.  The animal finding suggests that adiponectin might be mechanistically related to the observed heart failure risk.  Keep in mind, though, that Kim et al also found that adiponectin feeds forward to enhance synthesis of PPAR gamma, so it’s not possible to make a TZD-adiponectin link to CHF based on these observations alone.

Although this model probably doesn’t offer any direct uses to drug hunters, it does offer an example of the rescue of a mouse model of obesity-related diabetes through fat accumulation.  In that regard, it serves two important purposes relevant to drug discovery.  First, it adds substantial weight to the lipotoxicity theory of type 2 diabetes, providing a basis for continued drug research in that area.  Many questions remain:  Is it possible to promote subcutaneous lipogenesis without simultaneously promoting morbid obesity?  Can these experimental findings be replicated by manipulating the downstream effectors of adiponectin action in fat; can it be done without promoting cardiomyopathy?  What exactly are the mediators of the toxic effects of triglycerides in liver, muscle and pancreas, and can they be safely interrupted?  Second it provides an animal-model benchmark for drugs intended to disrupt the obesity-diabetes link.  As it happens, the mark for prevention is set pretty high; the mice were completely protected against insulin resistance, beta cell dysfunction, and diabetes resulting from increased food intake.  Of course, the cure also further worsened their already morbid obesity, so there’s still lots of room for improvement. 

In some future post, I’ll discuss what has already been done and what might be done in the future to further test the lipotoxicity/fat deficiency model of diabetes in humans.