Jules Levin, NATAP

This was a NIH study and results at ICAAC were reported by Mark Dybul of the NIH. This small pilot study was undertaken to evaluate the antiviral activity of efavirenz in lymph nodes. Although efavirenz effectively reduces viral load in plasma what about residual replication in lymph nodes. There has been some concern that PI therapy may be more effective at reducing HIV in lymph tissue than non-PI regimens and that there could be clinical differences as a result. In addition, the Public Health Service Guidelines were recently modified to include efavirenz as a possible substitute for a PI in initial therapy. Although the suppressive effect on plasma viremia by efavirenz containing regimens supports this recommendation, the absence of detectable plasma virus does not reflect the lack of virus replication. Follow-up data from study # 006 out to 72 weeks was presented at ICAAC showing the same results seen at at earlier time points. Taken together, the results of this lymph node study and the data from 006--Dybul and a few other doctors I spoke to at ICAAC feel comfortable as clinicians in using either a PI or efavirenz as initial therapy

Nine treatment naïve individuals received efavirenz with d4T+3TC. An additional 4 treatment-naïve patients received indinavir or nelfinavir with AZT+3TC or d4T+3TC. All patients achieved and maintained plasma HIV < 50 copies/ml. At 7-8 months they conducted excisional lymp node biopsies. The mean baseline viral load in the 9 patients receiving the EFV regimen was about 354,000 copies/ml (range 14,000 to > 1 million) and about 48,000 copies/ml in the patients receiving a PI regimen. Actually, 1 patient receiving IDV+d4T+3TC had a baseline viral load of 180,000 but the viral load of the other 3 were <14,000 copies/ml.

They used several assays to look for HIV. Using an in situ hybridization method the investigators said "they found almost nothing" and no differences between the PI and EFV arms. They found no HIV in germinal centers using a less sensitive method but using a sensitive method looking at multiple sections in the lymph tissue in all patients they found evidence of replication in only 2 cells—1 in 1 patient in each group.

Looking at LNMC –lymph node mononuclear cells—(1 million cells per patient) using NASBA quantitative assay, 6 of 8 patients receiving EFV and 3 of 4 receiving PI had undetectable HIV-RNA (<100 copies/ml). The person receiving PI had about 800 copies/ml and the 2 receiving EFV had 49,000 and about 2,500 copies/ml, respectively. They also looked at the presence of replication competent HIV. Investigators reported finding "relatively low levels of replication competent virus in both the EFV and PI groups"; 5 of 8 in the EFV group and 2 of 3 in the PI group had < 1 infectious unit per 1 million cells. One EFV patient had about 30 infectious units per million cells (IUPM), a second EFV patient had about 8 IUPM, the third EFV patient had about 2 IUPM, and the 1 PI patient with > 1 IUPM had about 5 IUPM. In speaking with Dybul, he feels these differences in the numbers between the EFV and PI arms referred to just above are not meaningful, because the study is small. I spoke with a few doctors at ICAAC who feel this data satisfies their concerns about using EFV as first-line therapy. However, future studies will continue to look at this question.

Using Virco phenotypic resistance testing the investigators reported no resistance was seen to EFV or protease inhibitors patients were taking, but they reported low level phenotypic resistance to other drugs. Looking at Virco genotyping results they found 2 patients in the EFV group had NNRTI mutations at 98 and 101. But also 2 PI patients who had never seen an NNRTI had these NNRTI mutations. Taken together this resistance data along with recently presented data (at Resistance Workshop) on transmission of resistant virus suggests the resistance detected by both genotypic and phenotypic assays were present prior to therapy.