Report 2
     Written by Jules Levin

It was announced today that Bristol-Myers, showing a commitment to stay in HIV, has purchased DuPont Pharmaceuticals. Thus, BMS gets control of efavirenz to add to their other HIV drugs: BMS-232632 (once-a-day PI in research), ddI, and d4T.

Kaletra Resistance: possible mutations associated with viral failure and using resistance testing to predict response

At yesterday’s Workshop sessions several abstracts were reported on mutations that appear associated with viral failure to Kaletra only for patient with PI experience. Today was another report suggesting similar mutations. This should not be confused with resistance to Kaletra for treatment-naïve individuals, which I discussed in yesterday’s Workshop report: 40 patients who had viral failure to Kaletra in treatment-naïve study did not yet show resistance mutations.

At last year’s resistance Workshop Abbott reported that response to Kaletra in treatment experienced was predictable based on the amount of protease inhibitor phenotypic and genotypic resistance patients had prior to starting Kaletra. The report was based on small numbers of patients but patients with 0-5 mutations did best, while patients with more mutations did not respond as well. Patients with <10 fold phenotypic resistance did better than patients with 20 fold or greater. Calvez looked at 700 patients in the French Kaletra expanded access program. He reported baseline viral load and CD4 were related to how well a patient responded. After 90-129 days 69% of patients with <100,000 copies/ml at baseline had a virologic response (<400 copies/ml and/or a 1 log decrease at any time point) compared to 49% with >100,000 copies/ml. 75% of patients with >150 cd4s had a viral response, compared to 65% with 50-150 cd4s, and 31% with <50 cd4s.

The point of the presentation, however, was that patients with 5 or less genotypic mutations did signifucantly better in response to Kaletra than patients with 6 or more mutations (65% vs 49%). Thus, suggesting that having 5 or less mutations may be a useful in predicting response to Kaletra. In addition, Calvez reported that using a multivariate analysis PI mutations at 54, 10, 82, and 46 were independently associated with viral failure. Using the univariate analysis (supposedly not as good a predictor) mutations at 20, 63, and 33 were associated with viral failure. An Abbott representative went to the microphone to caution that you should use both genotypic and phenotypic testing in trying to predict response to Kaletra, that genotyping alone was not adequate. So, Calvez concluded that in heavily pre-treated patients, virologic response to Kaletra is associated with baseline genotype, and that viral response was higher when patient had 5 or less mutations at baseline.

Military Study Reports Racial Differences in Response to Efavirenz Therapy

This study was to assess the response to therapy comparing African Americans and Caucasians treated with an efavirenz regimen in the military.

Scott Wegner (US Military HIV Research Program) referred to a Retrovirus Conference report that he says reported their were differences in efavirenz trough levels that were predicted by genotypes for CYP450 2D6 and p-glycoprotein, and he said these differences are known to exist at different frequencies between different ethnic groups. Wegner looked back at military personnel receiving efavirenz, indinavir and nelfinavir regimens. They looked time to first virologic failure from the start of therapy. The CERT study looked at the efficacy of resistance testing and was of 450 subjects (228 Caucasian, 170 African American, 40 Hispanic) at 6 US Military HIV evaluation centers.

In the efavirenz group there were 56 African Americans and 43 Caucasians.

The baseline characteristics appeared relatively comparable between the two groups. The Kaplan-Meirer curve showed a median time to failure of 440 days for African Americans and was not yet reached by 1400 days for Caucasians (hazard ratio=2.42, C: 1.35-4.67, p=.0027). Wegner said that for the purpose of this analysis people who stopped taking EFV for reasons other than virologic failure were considered as non-failing and were censored at the day they stopped taking EFV. In looking at the patients taking nelfinavir (n=144: 66 African-Americans, 80 Caucasians) or indinavir (n=100: 39 African-Americans, 61 Caucasians), there were no differences in median time to failure between African-Americans and Caucasians.

They looked at drug level testing that was randomly conducted on about 85 subjects periodically during the study, and Wegner reported EFV blood levels were the same in African-Americans and Caucasians. So he surmised there were no adherence differences between the two ethnic groups, they were all taking their drugs. They compared resistance test results for 2/3s of patients at failure between the two groups and found no difference.

This study created quite a bit of controversy at the presentation. A number of researchers and doctors went to the microphone to say they felt strongly that the differences in time to failure was most likely due to adherence. Wegner did not have adherence data and did not have side effects and toxicity data. In other words, there was no data on how many patients reported CNS side effects in the two arms. All he had was that there were an equal amount of discontinuations due to side effects in the two groups. I asked if the patients knew they were being tested for blood levels when they came in for testing, and Wegner said yes.

In speaking with Wegner after the meeting he feels the differences were due to genetic differences or differences in metabolic processing of the drug. A representative from DuPont was present and said that when they looked at the large 006 EFV study they saw no differences between Blacks and Whites in response to therapy. I agree with the doctors and researchers that the differences were due to adherence and side effects. Many of the doctors in agreement treat patients and felt strongly the differences were due to adherence differences between the two groups.