Veronica Miller, PhD, from Frankfurt's Goethe University has been following a cohort of 94 salvage therapy patients on mega-HAART (Lisbon, Abstract 587). Preliminary information from this cohort has been previously discussed in Treatment Issues (April 1999, pages 1-3). When Dr. Miller first began to monitor responses in her cohort, she noticed that a few of the patients who were experiencing the most marked viral load benefit had discontinued their previous regimens and experienced a shift in their viral population to wild type before starting salvage therapy. She decided to analyze the rest of her cohort retrospectively to see whether there was a clear association between structured treatment interruptions (STIs) and subsequent response to treatment. She also examined whether there were any other factors that could explain patients' responses. Knowing that even a brief interruption of therapy could put the patients at risk, she closely monitored and recorded their CD4 cell counts and viral load.
Patients on failing regimens may opt to quit taking the drugs for a number of reasons. Continued exposure to a drug may allow the virus to become more resistant to it or to find more efficient ways to resist it. This could lead to cross-resistance to new compounds that a patient might have access to in the future. Also, many patients choose to not bother with the inconvenience and toxicity of antiretroviral therapy partly because there is a perception that therapy is not working at all when there has been a viral breakthrough. Dr. Miller's findings may show, however, that this is a dubious conclusion.
In the total cohort of 94, she found that using more new drugs, and having virus that was susceptiple to the drugs used were independent predictors of achieving sustained drops in viral load. So were having a low baseline viral load and a high CD4 cell count — findings reported in many other studies. Of more interest, 50 of the 94 patients had chosen to have a treatment interruption of at least two months before mega-HAART. During that period, they experienced a median 0.71 log increase in viral load and a drop in CD4 count of 89 cells, which, as she noted, put many of them at risk of clinical progression. Notably, the ones who experienced a shift to wild type virus had the greatest drops in CD4 cell count and greatest rises in viral load.
We previously reported that 26 out of 39 patients for whom resistance data are available (before and after their STI) experienced a shift to wild type virus and that their virologic response to subsequent treatment was much more profound: by week eight, a 2.9 log versus a 0.78 log drop for those who did not revert to wild type. When these data were reported earlier, the critical question was whether they experienced a rebound in viral load within the first six months. Dr. Miller now can verify that 19 out of 24 have sustained viral loads below 500 copies/ml out to week 24 compared to one out of nine patients whose virus did not shift to wild type.
Even so, the study's findings are anomalous, and raise more questions than answers. There are several reasons to be cautious about these early findings. For one, the risk of going off therapy did not pay off for at least a third of the patients who did not experience a shift toward wild type virus but who did have increasing viral loads and falling CD4 cell counts. Only about half of the patients have regained the CD4 cells lost during their STI, although the time to regaining them has been longer for the patients who shifted to wild type virus (336 days vs. 209 days). Furthermore, many have voiced the criticism that the drug-resistant strains are probably still present, but have yet to re-emerge, and that viral suppression will not prove durable. So are any of these patients better off? Clearly, we need much longer follow-up and prospectively designed studies to answer this.
The data from this study could still use closer inspection as it raises some fascinating questions about the virus, treatment and pathogenesis. What exactly is happening in those patients whose viral populations do not revert to wild type? One can postulate that the time it takes for wild type virus to become predominant is probably a function of both how large a proportion of one's quasi-species it composes and the competitiveness of the resistant strains. Did these patients have mutant strains that grow as well as wild type? This seems unlikely because patients with resistant virus had much lower viral loads after STI than those who reverted to wild type. Or perhaps heavy pretreatment drove their wild type virus from active circulation. If so, it probably survives in low quantities and would grow out eventually, just as the drug-resistant virus in those patients who did "go wild type" might. Even so, it could take more than a few months (the average duration of STI in this study) for wild type to become the predominant strain. But if prospective studies indicate that introducing an STI before mega-HAART is clinically beneficial, it could lead to the following paradox: The faster-replicating wild type virus would be desirable because antiretroviral therapy might be able to tame it.
The flip side of this paradox is that having drug-resistant viral strains could also be a good thing. A few years back, some rather optimistic researchers predicted resistance could wind up virtually crippling the virus. That has not happened, but perhaps they were not far from wrong, given the rapid rise in viral load and fall in CD4 cell counts among the patients who reverted to wild type. It is possible that they might have even progressed faster than those who did not revert to wild type if they had remained off therapy. Granted, we would need to cripple the virus a lot more to put a halt to progression, but still, it is possible that treatment-induced resistance does slow progression down. As alluded to earlier, this could be an important clinical implication for those advanced patients who are considering whether to stop treatment after exhausting their anti-HIV arsenal. Perhaps the drugs are not just firing blanks after all. If you are tolerating your drugs, continued treatment on a failing regimen may yet confer some clinical benefit.