The degree and duration of objective tumor response needed to justify that endpoint as a surrogate for clinical benefit in patients with glioblastoma metaforme will be a question before FDA's Oncology Drugs Advisory Committee as it meets March 31 to consider accelerated approval of Genentech's Avastin as a single agent for recalcitrant GBM.

It is unclear whether the objective response rates seen in two single-arm, historically-controlled studies are of sufficient magnitude to support accelerated approval of the supplemental indication for Avastin (bevacizumab), FDA says in its briefing documents. ORR previously has not been used for accelerated approval in GBM, the agency notes.

A problem with using ORR in GBM is the inherent difficulty of measuring GBM tumors, which are morphologically heterogeneous, FDA explains. In the case of bevacizumab, an ORR determination is further complicated because the vascular endothelial growth factor inhibitor reduces radiation necrosis and edema and it is unclear how much of the imaging changes are the result of an anti-tumor effect, FDA points out.

Agency review of radiographic images for responders in the bevacizumab monotherapy arm of AVF3708g, an open-label, multi-center, randomized, non-comparative trial conducted by Genentech, calculated an objective response rate of 25.9 percent, with a median response duration of 4.2 months. The agency disregarded a bevacizumab plus irinotecan arm because of inability to isolate the bevacizumab effect.

The ORR reported by Genentech, based on an independent radiologic facility, was 28.2 percent with a median duration of 5.6 months. The difference in the two sets of figures could be the result of FDA not confirming a response in two patients claimed by Genentech.

The second study referenced in the application, NCI 06-C-0064E, is a single-arm, single-center trial to determine the anti-tumor activity of bevacizumab, with objective tumor response and duration of response determined by independent radiologic review. The results were an ORR of 19.6 percent with a median duration of 3.9 months.

The objective response rates in the studies fall short of what experts at a 2006 brain tumor endpoints workshop suggested would be appropriate (¹ Also see "Progression-Free Survival Is Meaningful Endpoint For Glioma Trials" - Pink Sheet, 6 February, 2006., p. 15). Participants at the meeting agreed that a "response rate of sufficient magnitude (e.g. greater than 30 percent) was likely to be associated with clinical benefit as the magnitude of response rate would outweigh the uncertainties associated with interpreting MRI scans," the FDA briefing documents note.

Genentech originally sought to submit the sBLA on the basis of a six-month progression-free survival rate in AVF3708g. That was 36 percent in the bevacizumab arm and 51 percent in the bevacizumab plus irinotecan arm, according to interim data (² Also see "Genentech Reports Positive Progression-Free Survival Results For Avastin In Brain Cancer Trial" - Pink Sheet, 19 November, 2007.). But FDA nixed that proposal because time-to-event endpoints "must be evaluated in randomized, controlled clinical trials."

-Cathy Dombrowski (c.dombrowski@elsevier.com)

Avastin Advisory Committee To Assess Validity Of Objective Response Rate Endpoint

More from Archive

More from Pink Sheet