[Minutes, Advisory Committee to the Director GP-160 Meeting (1992: Bethesda, Maryland)]

133

1 won't go into.
2 Even if you had an agent that did what you
3 wanted it to do and you could see it through the
4 surrogate, that was mediated through the surrogate, it
5 might be doing something else that you weren't
6 measuring through the surrogate that was ultimately
7 going to outweigh the benefit.
8 Next slide. There are some specific
9 problems that are worth mentioning with CD4 as a
10 surrogate endpoint in these trials. First of all, the
11 data that are available suggest that relatively little
12 of the clinical effect of AZT, which has been the
13 treatment that is most widely studied, is mediated
14 through the CD4 count.
15 At late-stage patients, where we have the
16 most data, almost none of the effect seems to be
17 predictable by the CD4 count. There have been some
18 preliminary looks at CD4 in mid-stage patients,
19 patients with CD4 counts, say, between 200 and 500,
20 that suggest that more of the effect may be mediated
21 there, but it's still small.
22 We've got no data on correlations between
NEAL R. GROSS
COURT REPORTERS AND TRANSCRISERS
1323 RHODE ISLAND AVENUE, N.W.
[phone] WASHINGTON, D.C. 20005 [phone]