NIAID SURVEY FINDS WIDE VARIATION IN CD4 CELL COUNTS
NIAID SURVEY FINDS WIDE VARIATION IN CD4 CELL COUNTS indicating that unreliable CD4 lab measurements may be compromising AIDS clinical trial results, Rebecca Gelman, a biostatistician at Harvard University School of Public Health, warned at a Sept. 20 plenary session of the National Institute of Allergy and Infectious Diseases' Community Programs for Clinical Research on AIDS (CPCRA). Gelman presented findings from a recent, 60-lab survey commissioned by NIAID's AIDS Clinical Trials Group (ACTG). The survey showed significant variation in sample CD4 count measurements even within a single laboratory, she said. In some labs, the same patient's blood sample yielded CD4 counts as diverse as 350 and 850. CD4 readings as a percentage of total white blood cells also varied dramatically at many labs, ranging from 24% to 59% for a single sample. Gelman hypothesized that human error, as well as differences in lab protocols, such as the inclusion or exclusion of CD3 and CD8 cells in samples, could account for some of the variation. In some cases, she noted, labs have been known to switch CD4 and CD8 counts on reporting forms. Such errors and discrepancies, Gelman commented, "obviously have serious . . . consequences for patient enrollment and treatment," and bring into question the reliability of past ACTG studies that used CD4 counts as a surrogate marker. In response to Gelman's data, several CPCRA researchers expressed skepticism about the wisdom of continuing to use CD4 counts as a surrogate marker for AIDS. One principal investigator in the audience commented that "the degree of variation" described by Gelman was "horrifying and astounding." The ramifications of the NIAID survey remain to be seen. Just last February, FDA's Antiviral Drugs Advisory Committee supported the choice of CD4 as an acceptable primary endpoint for the evaluation of AIDS therapies. However, the committee's consensus was that CD4 could not be the only surrogate marker but rather a partial one ("The Pink Sheet" Feb. 18, p. 8). To decrease the chances of inaccurate readings, Gelman recommended that CPCRA researchers set up quality control programs to assess "in-lab variation." She also suggested that investigators work consistently with one lab through all stages of AIDS clinical research. Following Gelman's remarks, NIAID biostatistician Jonathan Kagan stated that NIAID's Division of AIDS is currently exploring the possibility of having all CPCRA researchers submit their blood samples to a single regional lab selected by the institute. In combination with rigorous quality control programs, this strategy "could make a big difference" in ensuring more reliable CD4 measurements, he argued.
You may also be interested in...
Newly released Medicare Part D data sheds light on the sales hit that branded pharmaceutical manufacturers will face when the coverage gap discount program gets under way in 2011
FDA appears headed for a showdown with clinicians and the pharmaceutical industry over the proposed new clinical trial endpoints for acute bacterial skin and skin structure infections, the guidance's approach for justifying a non-inferiority margin and proposed changes in the types of patients that should be enrolled in trials
Specialty drug maker Shire has quietly begun scouting deals with a brand-new $50 million venture fund, the latest of several in-house investment arms to launch with their parent company's pipelines, not profits, as the measure of their worth