More Trouble for Vanda's Sleep Disorder Drug

As this story was being published, Vanda announced a follow-on stock offering of 4 million shares.

WASHINGTON, DC ( TheStreet) -- In June, I uncovered a long list of deficiencies and shortcuts taken by Vanda Pharmaceuticals ( VNDA) in the development of tasimelteon, a melatonin receptor agonist, for the treatment of a rare sleep-pattern disorder that affects blind people.

My conclusion: The odds of FDA approving Vanda's tasimelteon for non-24 sleep-wake disorder were exceedingly small.

I based my last analysis primarily on the larger of two phase III studies Vanda conducted on tasimelteon -- the so-called "SET" study. Vanda's second phase III study -- known as "RESET -- is equally troublesome.

Let's take a closer look.

RESET was a randomized withdrawal study, meaning all patients diagnosed with non-24 entered first into a run-in stage in which they were treated with daily tasimelteon for three months. After three months of receiving tasimelteon, patients were randomized to one of two groups: 1) continued tasimelteon treatment for two months; or 2) a switch to placebo for two months.

At the end of the two-month withdrawal phase, patients were measured for response, defined as maintenance of entrainment, or adherence, to a normal 24-hour circadian rhythm.

Vanda declared victory in the RESET study, claiming 90 percent of tasimelteon-treated patients maintained entrainment of their circadian rhythm after two months compared to 20 percent of placebo-treated patients. The difference in entrainment maintenance was statistically significant.

Here's a chart describing the RESET primary endpoint, taken from a poster presented at the Endocrine Society (ENDO) meeting in June.

Was the tasimelteon benefit clinically meaningful or relevant at all for getting the drug approved? I don't believe so.

Vanda chose to measure maintenance of entrainment at a fixed and arbitrary point in time, defined in RESET as two months. A non-24 patient was deemed a responder in RESET if he/she maintained entrainment at two months.

Of the 10 patients treated with tasimelteon, 9 maintained entrainment after two months. Ten patients were treated with placebo but only two maintained entrainment for two months. I'll address the issue of only having 20 patients to analyze in what is supposed to a pivotal, phase III study later.

Why is this claimed success irrelevant? Because it doesn't matter if non-24 patients maintain entrainment for two months if they continue to suffer from sleep problems.

That's exactly what happened to patients in the RESET study.

Here's another, more important chart taken from the RESET poster. This tracks time to relapse in a Kaplan-Meier Curve.

The dotted line represents non-24 patients treated with tasimelteon. Each step down is a patient who relapses, defined as a 45-minute decrease in daily average nighttime sleep over a period of a week. Vanda created this definition of relapse.

Four of the 10 tasimelteon treated patients relapsed before Day 50. A 40 percent relapse rate despite a 90 percent maintenance of entrainment through Day 60, according to Vanda.

Now, look where the dotted line turns horizontal. See those circles? There are six of them (two overlap), each representing a patient censored from the analysis. Censored means Vanda has no data on sleep-time relapse for these patients.

In other words, of the 10 patients treated with tasimelteon, Vanda is missing data on six patients, or 60 percent.

The tasimelteon curve is misleading with all that missing patient data. Vanda claims that median time to relapse for tasimelteon patients was not reached but that's only because 60 percent of the data are missing! In reality, time to relapse for tasimelteon patients might actually look entirely like the time to relapse for placebo patients, depicted in the chart with the solid line.

The p-value calculated for the time to relapse analysis is omitted from the Kaplan-Meier Curve but it's listed elsewhere on the RESET study poster. It's 0.0907, meaning tasimelteon failed to reduce time to sleep relapse compared to placebo, even with all that missing data.

I believe FDA will be much more interested in the failure of tasimelteon to prevent sleep-time relapse in the RESET study than successful maintenance of entrainment at an arbitrarily defined two-month time point.

Vanda has already admitted that FDA will not approve any sleep drug based solely on an entrainment endpoint.

Now, about the 20 patients who made it all the way through the RESET study. The study started with 57 patients screened and entered into the run-in stage, so Vanda was left with just 35 percent of patients to analyze for the primary endpoint (maintenance of entrainment.)

The RESET study lost patients left and right. Here's a rundown from the poster:
-- 58 patients screened, 57 patient eligible for the run-in phase.
-- 48 patients actually entered the run-in phase (9 patients thrown out immediately for missing data)
-- Of the 48 patients treated with tasimelteon for the three months of the run-in stage, 24 patients, or 50 percent, achieved entrainment.
-- If you add back the 9 patients entering the run-in stage but thrown out before treatment began, the tasimelteon entrainment rate was actually 42 percent.
-- The entrainment failure rate was 68 percent.
-- 24 patients exited the three month run-in stage but four patients did not enter the randomized withdrawal stage of the study for a variety of reasons, including adverse events and data quality issues. (Another red flag?)
-- That left 20 patients evaluable for the study's primary endpoint, maintenance of entrainment.

RESET, like the previously discussed SET, are poorly conducted studies which raise lots of troubling questions about tasimelteon. Vanda's chance of gaining FDA approval are slim.

-- Reported by Adam Feuerstein in Boston.

Adam Feuerstein writes regularly for TheStreet. In keeping with company editorial policy, he doesn't own or short individual stocks, although he owns stock in TheStreet. He also doesn't invest in hedge funds or other private investment partnerships. Feuerstein appreciates your feedback; click here to send him an email.

More from Stocks

Jim Cramer: Centene Is in All the Big Medicare Markets

Jim Cramer: Centene Is in All the Big Medicare Markets

Jim Cramer on the Markets: I Struggle With How Negative Things Are

Jim Cramer on the Markets: I Struggle With How Negative Things Are

Video: Jim Cramer on Tariff Worries, Oil, Alphabet and Centene

Video: Jim Cramer on Tariff Worries, Oil, Alphabet and Centene

Emerging Markets Get Pounded by Trade War Concerns

Emerging Markets Get Pounded by Trade War Concerns

Worth a Stunning $6.6 Trillion, Tech Stocks Have Taken Over the Market

Worth a Stunning $6.6 Trillion, Tech Stocks Have Taken Over the Market