Resistance & Resistance Testing in Chronic Treatment-Naïve & Acute HIV

These studies highlights the reason why resistance testing may not be helpful in detecting resistance to drugs that patients have not been taking for a while and in patients with chronic infection starting their first regimen. It also suggests to me that resistance in newly infected patients can be detectable if testing occurs immediately, but after a period of time resistance is not detectable and the reason for lack of response to therapy may not be identifiable.

A Pavia from the University of Utah reported on the ability of baseline genotype testing to predict outcome in chronically infected treatment-naïve patients receiving indinavir and two nukes in two trials (abstract 439). Patients were recruited during '97-'98 at 30 centers when resistance in newly infected patients was likely at lower levels of incidence than today. Baseline genotypes were complete in 268 of 409 patients, but incidence of resistance was low: at least 1 significant mutation was present in 5.2%. Neither baseline genotypic resistance score nor the number of PI or NRTI mutations was predictive of failure at week 24 or 48 or of time of failure. Pavia concluded that this data do not support routine, non-selective use of baseline genotypic testing for chronically infected treatment-naïve patients.

In an interesting study by E Daar, he reports on 1 patient who was identified shortly after HIV infection. He had 7 days of fever, rash, myalgia, swollen lymph glands and oral ulcers with an evolving HIV Western Blot. Baseline cd4 was 396 and viral load 5.48 log (over 100,000 copies/ml), with major NRTI mutations at M41L, T69SST, K103N, Y188C, T215F, and protease changes at L10I, M46L, V77I, L90M. He had >2.5-fold phenotypic resistance to all nukes except ddI and ddC, all NNRTIs, and all protease inhibitors except saquinavir. Therapy was deferred. Over the next 4 months cd4 and viral load remained between 2.91-3.05 log and 500-629 cd4s. In month 5 cd4s started to decline and viral load started to increase. By month 7, cd4s were 451 and viral load was 4.51 log. At month 5 genotyping no longer showed any of the major nuke, non-nuke, or PI mutations seen at baseline. Daar concluded that in the absence of drug pressure from being on therapy resistant virus was replaced by phenotypically more susceptible viruses (the ViroLogic Phenosense phenotypic test was used). The question remains if the resistant virus will affect response to therapy. The program abstract did not report response to subsequent therapy.

Clinical Impact of Baseline Genotypic Resistance

S.A. Tasker from the US Military HIV program reported on 21 military personnel who had intermediate or resistant virus according to Vircogen-I (abstract 436). Baseline median cd4 and viral load were similar in patients with resistant or sensitive virus (cd4s- R- 498, S- 463; viral load- R- 4.7 log, S- 4.5 log), but 6 months resistant patients had greater cd4 increases (230 vs 122) and greater decreases in viral load (3 log vs 2.1 log). There were no differences in HIV therapy between the two groups. However, for patients on 3 or more drug regimens with no primary resistance mutations the cd4 increase was 269 vs 170 for patients with sensitive virus (p<0.05). The difference in changes in viral load between the two groups were not significant (R- 3.2 log vs S- 2.7 log, p=0.14). For patients with primary resistance mutations, the mean cd4 increase was 332 (p=0.02 vs S group), and mean viral load decrease was 3.4 log (sill not significant, p=0.19 vs S group, at 6 months. The authors concluded that the mutations causing resistance may have reduced viral fitness, or caused hypersensitivity to other drugs.