Biotech Values

[DewDiligence]

Latest QLTI CC:

Between the World Series and the presidential election, I’ve fallen behind in conference-call listening during the just-ended peak season for third-quarter CC’s. I finally heard the QLTI call from 10/21/04, and there was a cautionary note for QLTI longs (a group which does not include me).

As previously reported, the VIO (Visudyne in Occult) trial did not achieve statistical significance relative to placebo at one year (#msg-4300583). However, the above statement is an oversimplification. In actuality, the VIO trial missed two distinct visual-acuity endpoints at one year: 1) Stable vision, defined as an avoidance of a 3+ line loss; and 2) Avoidance of severe vision loss of 6+ lines.

The VIO trial will continue for a second year, and will have two, two-year endpoints similar to the ones at one year described above.

The problematic aspect of all this for QLTI is that the FDA allocated a cumulative 0.05 of p-value to QLTI to be divided equally among all four endpoints: the two at one year and the two at two years. Thus, each individual endpoint will have to meet a stringent p-value threshold of only 0.0125.

QLTI would not disclose by how much the trial missed on the two one-year endpoints, but I felt from the tone of the CC discussion that it was not an especially close miss. This makes it seem doubtful that p-values of 0.0125 can be achieved at two years.

Moreover, the effect of missing all of the one-year and two-year endpoints in the VIO trial could be reduced Visudyne sales in Europe, where certain countries have stated their intention to revisit reimbursement policy for Visudyne in the occult subgroup of AMD. However, in the U.S., where Visudyne is reimbursable for only a portion of the occult subgroup, CMS does not plan to revisit its policies based on the two-year data.