Vexed Virologists: Study After Study, Antiretroviral Analyses Confound More Than They Clarify

First there was Concorde, the now notorious Anglo-French study which last summer sent shock waves across the Atlantic as it cast doubt on the wisdom of early AZT. Equally sobering to those who looked more closely, however, was the observation that AZT-induced rises in CD4+ T cell counts, at 3 years, showed no correlation with clinical outcome; that is, just because AZT made someone's T cells go up didn't mean that person would live longer. While in a longitudinal sense, an HIV-infected person's T cell counts are still believed useful in estimating when clinical deterioration or death might strike, their utility in helping to quickly gauge response to therapy has been forever compromised.

If Concorde could be dismissed as a crazy transcontinental aberration, then all the fuss would be for naught. But as more U.S. studies are completed and analyzed, Concorde's findings are more often corroborated than challenged. Don Abrams' CPCRA study (002) reported observations similar to Concorde's as early as last spring. In gauging a patient's response to therapy, its authors observed a disturbing dichotomy between changes in CD4 T cell counts and clinical outcome: in the group with the larger CD4 count increase (8.6 cells/mm3), the investigators observed shorter survival (65.4% alive at 12 months). In the other group, where CD4 counts fell an average of 5.9 cells/mm3 throughout the follow-up period, longer survival was observed, averaging 71.2% at 12 months.

All along, more circumspect investigators had questioned the wisdom of CD4 T cell counts as surrogate markers. Biologically, it required quite a leap of faith to think that the introduction of an antiretroviral agent could so quickly improve the state of one's immune system that a T cell count rise would be observed as early as 2-3 weeks later. Rather, it was more likely that the nucleoside therapy simply produced a "re-trafficking" of the lymphocytes: T cells that had once been sequestered within tissue or lymphoid organs moved into the circulating blood and were then hailed by unwitting investigators (and salivating drug company study monitors) as evidence that the intervention had proven successful.

The biggest blow to the surrogate marker orthodoxy, however, may turn out to be a trial dubbed ACTG 116B/117. Last year in Berlin (for those delegates who stayed for the hastily composed addendum to the program), Alabama virologist Vicky Johnson showed how resistance at baseline could not explain why patients who switched from AZT to ddI appeared to do better; nor could the presence of the much balleyhooed syncytia inducing ("SI") virus. Patients who were switched over to ddI did better clinically whether their time on prior AZT had been 2 weeks or 2 years.

Subsequent analyses of the 116B/117 data by University of Washington virologist Robert Coombs and Brooks Jackson of Case Western Reserve in Cleveland found that not even the quantity of virus at baseline (nor changes in the levels of virus throughout the study) could explain differences in response to therapy -- or clinical outcome. "We were unable to show [with 116B/117] that changes in RNA levels were predictive of clinical outcome," Dr. Jackson explained. And patients with the highest viral titers at baseline showed, surprisingly, the greatest CD4 count rises in response to therapy.

Like the alluring SI/NSI appellation before it, the high-tech resistance-conferring 215 mutation's contribution to clinical failure is also turning out to be a bust: some 25% of those who developed the threonine mutation at codon 215 continued to do well clinically, despite what erstwhile Gang-of-Five investigator Dr. Tom Merigan and others would have one accept as irrefutable proof that the virus is no longer sensitive to AZT. In spite of all this, the ACTG will pour millions of its dwindling resources into his pet-project (ACTG #244) whose very scientific underpinnings have now been shown to be misunderstood -- if not out-right flawed.

Huddled away off-site at the recent ACTG retreat, far from the mundane proceedings of lesser committees, presentation after presentation before the Primary Infection committee evoked alternating hushes and whispers from the roomful of bewildered virologists. UC-San Diego's Doug Richman may have heralded the naked Emperor most succintly with his poker-faced pronouncement: "There are basic virological processes going on here that we simply do not understand."