Notes from AASLD: Apples and Oranges and Null Responders

The liver disease community – if not the investment community – has largely moved on from the novelty of comparing Vertex Pharmaceuticals' telaprevir and Merck's boceprevir, the two direct-acting antivirals that together form the threshold to a new era in hepatitis treatments if the buzz at the American Association for the Study of Liver Diseases is any indication.

Instead, physicians are celebrating the fact that two therapies may soon be available that can help patients achieve success rates that handily best the current 50% success rate from the standard of care interferon and ribavirin therapy, which is described as 48 weeks of constant flu-like symptoms and PMS.

The sense of promise is palpable at AASLD, now under way in Boston, where many of the sessions are standing room only.

A lot of the excitement now is around the IL28b genetic marker and its implication for better cure rates, and the lure of a still-years-away all-oral therapeutic regimen.

But first, there will be protease inhibitors. Merck and Vertex, who are jockeying to be first-to-market with a direct-acting antiviral for hepatitis C, plan to complete FDA submissions by the end of the year, with approval possible in mid-2011.

Based on overall profile, the odds-on favorite for best-in-class in this initial class of two seems to be telaprevir, but boceprevir may find a top-rung niche in experienced patients.

However, comparisons can be tough. Vertex had no short-course option in the Phase III study of telaprevir in experienced patients, REALIZE. Final results in that study have not yet been reported, but top line data showed 65% of experienced patients treated with telaprevir achieved SVR compared to 17% in the control arm.

Meanwhile, Merck released data at AASLD showing their response-guided therapy plan, which shortens the treatment period for patients who respond early, can work in prior treatment failures.

Another question frequently asked of presenters this year at AASLD concerns the definition of null responder, that is, patients who have had the poorest results with standard of care.

The definition of null responder used by Merck in RESPOND-2 is those who achieved less than 1 log decrease in viral load after the four-week lead-in period with standard of care. According to the abstract on the trial presented at AASLD, 33% of null responders (15/46) in the response-guided arm achieved SVR, a statistically significant improvement over the control arm, in which none of 12 patients had a cure. Null responders in the 44-week triple therapy arm had a 34% cure rate (15/44), also statistically significant.

That null-responder definition, however, appears to be at odds with the FDA guidance. In the document, FDA describes that population as people with "less than a 2 log10 reduction in HCV RNA at week 12" of standard of care therapy, the point at which the therapy is typically dropped for futility.

A footnote in the draft indicates that "other definitions for null response have been proposed, such as less than 1 log decline in HCV RNA at week four of treatment. However, failure to achieve a greater than 2 log decline … at week 12 has typically been used as a treatment futility criterion," and use of the 1 log decline definition "causes a gap in classification for individuals with a viral load reduction" that falls between the two.

Vertex has used that definition in its REALIZE study in experienced patients. In an interview, Robert Kauffman, chief medical officer at Vertex, reasoned that using what he called the "standard definition" prospectively to identify patients at the time of enrollment, rather than "on treatment," ensured that all patients were in the appropriate group.

To test the difference in the two definitions, Vertex used a four-week induction arm in REALIZE, and looked at the correlations between less than a 1 log drop at week four of the delayed start and the standard definition, Kauffman explained.

The outcome of the analysis was "a clear difference," he said. The two groups had different responses to the triple therapy. In addition, he said, the trial showed that both groups can have a "very, very good response."

A subanalysis of REALIZE reported at AASLD showed that among combined partial responder and relapser patients in the lead-in arm, 18% (31/171) fit the less than 1 log reduction definition at the end of the lead-in with standard of care. Of those patients 58% (18/31) went on to achieve SVR compared to 31% (46/147) patients prospectively-defined as prior null responders using the "standard" definition.

– Shirley Haley

flickr image by e g g used under a creative commons license.