Dodgy data analysis sees XBiotech shares plummet

XBiotech ($XBIT) tried to talk up a questionable colon cancer trial as being positive, but investors were having none of it as its shares tanked by nearly 50% yesterday.

The Texas biotech released new data from its Phase III study which focuses on Xilonix, its antibody designed to neutralize biological activity of interleukin-1 alpha.

XBiotech said its data in certain colon cancer patients were “positive,” but this has been ridiculed by some analysts and put its future European approval--which the data were to be used toward--under question.

The company’s data only looked at responders, i.e., those both on its drug and those in the control arm (who were given a placebo) and said that “clinical response was associated with a 2.7-fold increase in overall survival (11.5 versus 4.2 months in responders vs. non-responders, respectively).”

But this only shows that those who responded to either treatment saw a greater OS--not just those being given Xilonix. The data do not give a direct comparison of OS in the Xilonix arm versus placebo, so it’s impossible to tell if the drug was in fact better than a dummy treatment in improving survival. 

The biotech said that further overall survival, the golden endpoint in any cancer trial, was also “not compared” between treatment arms because after 8 weeks, “all patients were eligible to receive study drug.”

The study’s primary endpoint was in fact a measurement called “clinical response rate” (CRR), what the biotech deems a “novel measure of anti-cancer activity,” although the traditionally measured responses, such as anti-tumor activity, progression-free survival (PFS) and objective response rates (ORR), were not given.

In its own CRR findings, of the 309 patients taking the drug after 8 weeks of therapy compared to placebo, 33% on Xilonix achieved a CRR versus 19% in the control arm.

CRR was not set up to assess survival or its anti-tumor properties, but rather to determine lean body mass, as well as being based on a standard quality of life questionnaire to see how patients feel about fatigue, pain and anorexia during two months of treatment.

Further details of how these CRR measurements were met were not, however, shared by the company. The only information given was that the “responders gained more lean body mass compared to non-responders (p<0.0007), had reduced fatigue and pain (p<0.001) and improved appetite (p<0.001),” but this is essentially meaningless as no details have been given on each CRR measurement, with the ambiguity of the “responders” still not dealt with by the biotech.

This endpoint was in fact set up with the EMA, but the results’ efficacy leave many questions unanswered, and even though the biotech said it expects a decision on an approval by Q4, the regulator may not be impressed. A second late-stage U.S. study is also ongoing for Xilonix, with this European trial specifically designed for an EU green light.

This comes after the biotech’s shares have climbed ever higher due to positive noises coming from its testing, and it once had a market cap of nearly $1 billion. 

However, after the data were posted, the company’s shares shot downward as analysts were left deeply unimpressed. It still, however, commands a market cap of $646 million as of yesterday. 

And it’s not the first time things have gone wrong, as late last year some early data from its Phase III study was found to be riddled with practical errors, including a number of patients being given the wrong drug, with 33 also not getting the X-ray required to prove it they had improved lean body mass--one of the primary endpoints of the trial.

- check out the release

Related Articles:
XBiotech lines up for a $76M IPO to fund its antibody R&D
XBiotech flubs a Phase III trial with its lead cancer drug