Let's go the slides. First, let's take a look at the obvious co-primary endpoint, which assessed Xiaflex's ability to straighten those crooked penises following a year of treatment.
Both studies hit this primary endpoint with strong statistical significance. In the IMPRESS-1 study, men in the Xiaflex arm had 49 degrees of bend in their penis at baseline. After 52 weeks, the bend was reduced to 31 degrees, or a 38 percent improvement. Men in the sham injection arm entered the study with penises that curved, on average, by 49 degrees. At study end, that curve was 39 degrees, or a 21 percent improvement. Results from the IMPRESS-2 study were similar, as displayed in the slide.
Does statistical significance translate into a clinically meaningful benefit? That's a tough question here given the high placebo response rate. On placebo-adjusted basis, Xiaflex only reduced penile curvature by 8 degrees. That's not much of an improvement, particularly when you factor in side effects, which I'll discuss in a bit.
Auxilium argues it's more relevant to look at individual patient results compared to their own baseline and not against the placebo injections. Fine, but even looking at just the Xiaflex arm, is reducing the bend of a penis, on average, from 49 degrees to 31 degrees, a robust-enough result to warrant approval? There are no "straight" penises at the end of this study, unlike Xiaflex's ability to fully straighten fingers in patients with Dupuytren's Contracture. Does that matter?
Next slide. This one graphs the second co-primary endpoint of the studies -- the reduction in a Peyronie's "PDQ bother" score. Auxilium developed the PDQ questionnaire to assess the overall impact on quality of life that Peyronie's has on men with the disease. Using 15 questions, the PDQ measures 1) the "bothersomness" of Peyronie's; 2) psychological and physical symptoms of Peyronie's; and 3) penile pain.
For the phase III studies co-primary endpoint, Auxilium only used responses to four of 15 questions that measure the "bother" portion of the PDQ. Here are the results:
Looking at IMPRESS-1 results, you see a 44 percent reduction in the bother score for men treated with Xiaflex compared to a 27 percent reduction for men treated with sham injections. The placebo-adjusted difference is small but still statistically significant with a p value of 0.0451. A p value greater than 0.05 would have been a negative result so this study barely passed.