Year: 2003
Type: letter
Abstract: To the Editor —In his letter [1], Dr. Murray indicates surprise that transmission-fitness calculations were performed only on data on 1997. Space limitations precluded a full exploration of these issues in our article. The year 1997 represented the end of a period of several years of predominantly mono-/di-nucleoside analog therapy. At this time, the low virus suppression levels observed in individuals receiving treatment at the University of California San Diego (UCSD) clinic would have been broadly comparable with the results of individuals receiving therapy from primary care physicians (who are responsible for treating the majority of patients in the area). However, rapid change in the percentage of individuals achieving virus suppression followed as a result of the introduction of potent antiretroviral therapy, and 1999 represented the period of most rapid change in suppression rates in individuals in the UCSD clinic. It is unlikely that the same virus suppression level would have applied to results of therapy received from primary care physicians at that time, which would likely have lagged behind the increase observed at a specialist clinic, but to an unknown degree. We therefore do not consider calculations made using the cross-sectional clinic suppression rates for 1999 to be reliable Dr. Murray suggests that “transmission fitness” could be a misleading term, but we feel that it is clear that the term is being used in a virological sense. Individuals with drug-resistant strains can transmit resistant virus. The issue being addressed is what is the overall rate of transmission of drug-resistant strains in these individuals Dr. Murray then refers to “more-efficient resistance transmitters,” but we do not see this as a useful concept. In the absence of any other information, the probability of any transmission is, for these purposes, best represented by plasma virus load, as we have discussed [2]. The probability of transmission of resistance is, in the absence of resistance assays, likewise best indicated by success or failure of therapy. Although individuals experiencing therapy failure could have been overrepresented in recruitment to triple-drug combination therapy, they would not, by virtue of their prior failure, be expected to be overrepresented among individuals with suppressed virus. However, neither would they “revert to wild type,” because it is not general practice to discontinue therapy just because of the presence of resistant strains Dr. Murray’s fourth point refers to the accumulation of resistance-associated mutations over time. However, for most nonnucleoside reverse-transcriptase inhibitors (NNRTIs) and for lamivudine, a single mutation is sufficient to reduce susceptibility by several hundred–fold, relative to wild type. The study that Dr. Murray cites [3] describes results observed with protease inhibitors (PIs), which require several mutations before substantial reductions in susceptibility are acquired. However, in determining a strain as being drug-resistant, we made no distinction between the presence of 1 or several mutations. As we describe in our study [2], we based this determination on the presence of any primary resistance–associated mutation. The additional suggestion that virus so exposed will acquire enhanced “replicative capacity” is only true in a few cases. In any case, replicative capacity will be incorporated as a contributing component into “transmission fitness,” as we use the term, but, in contrast to what Dr. Murray suggests, Deeks et al. [4] have shown that, in most cases, replication capacity decreases with increasing PI resistance. In our data set, the most commonly transmitted variants in 1997 were partial revertants of T215Y, associated with zidovudine failure, which have been shown to have low fitness costs [5]
DOI: 10.1086/378680
Language: English
Subfield: Infectious Diseases
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4253710249