Fishing in the HIV reservoir: proviral HIV DNA levels can predict success of switch to abacavir

French investigators have found that high levels of proviral HIV DNA can predict the success of a switch from a protease inhibitor to abacavir, according to a study published in the 1 January edition of the Journal of Infectious Diseases.

The group also found that the viral reservoir (represented by HIV proviral DNA) harbours drug resistance mutations even when plasma viral load is undetectable, and that these predict the success of subsequent treatment failure in those who switch to abacavir.

Proviral HIV DNA is found in latently infected CD4 cells (cells that are not actively producing infectious virus, but potentially capable of producing virus if the cell is activated).

Between September 2000 and September 2001, French investigators monitored 75 patients who had achieved a viral load below 50 copies/mL on a HAART regimen including two NRTIs and a protease inhibitor. During this period, 55 switched to a triple NRTI regimen because of concerns about side effects, and changed the PI to abacavir.

Baseline tests included CD4 count, viral load, proviral HIV DNA in peripheral blood mononuclear cells (PBMC) and genotypic tests for reverse transcriptase and PI resistance. Patients were monitored at months one, three and six. If a patient’s viral load remained below 50 copies/mL for the study period, they were regarded as virological successes. A blip was defined as a transient increase in viral load to between 50 and 1,000 copies and a subsequent return to undetectable (

Proviral HIV DNA was measured, and tests performed for RT and PI resistance at baseline and six months for successful patients, or at virological failure or blip for other patients. Patients who experienced blips in their viral load received six months of additional follow-up.

At baseline, the overwhelming majority of patients (81.3%) had detectable proviral DNA, and this was found to be strongly associated with previous mono or dual therapy; the number of previous treatment regimens; and, the number of RT or PI resistant mutations. Neither baseline viral load nor CD4 count were found to be significant.

In the 20 patients who remained on PI based therapy, 19 were considered virological successes at six months. In these patients proviral DNA remained stable. At baseline 74% had wild-type virus, with the remaining 26% having only one or two mutations in the reverse transcriptase (RT) gene. At six months, one or two additional mutations had emerged in three patients, whereas RT resistance was no longer detected in the other three. One patient experienced blips at months one and six (61 and 76 copies/mL respectively), but subsequently returned to below 50 copies/mL. No resistance mutations were found.

In the abacavir switch group, six withdrew because of an allergic reaction to the drug or were lost during follow-up. Of the remaining 49, 22 (45%) were successes, 10 (20%) had blips and 17 (35%) failed. Baseline proviral DNA was significantly higher in those considered to have failed or blipped. Prior treatment history was also found to be significant. Only one of the people successfully switching to abacavir-based therapy had been previously treated with suboptimal therapy. Genotypic testing revealed that 18 of the 22 successes had wild-type HIV at baseline, as did fifteen patients at six months. One or two resistance mutations were found in four patients at baseline, and between one and five had emerged in three others by six months.

Amongst the treatment failures, seventeen had prior treatment history and of these six (35%) had a history of earlier treatment with either mono or dual therapy. At baseline, 53% of treatment failures had between one and five RT mutations associated with NRTI resistance.

Ten patients experienced one or more blips in their viral load, of whom two had a history of earlier treatment. Wild type virus was found in eight patients at baseline, the remaining two having between three or four mutations. At the time of the first blip, new resistance mutations were found in two patients who previously had wild-type virus. In one patient, whose viral load blipped to over 500 copies/mL, four new mutations were found, and treatment failure occurred within six months of the blip.

Proviral DNA levels were strongly associated with earlier mono or dual NRTI therapy, and patients with a proviral DNA load of 2.7logcopies/10⁶ or above were more likely to experience virological failure. The authors concluded that proviral DNA testing should be undertaken before “switching to a simplified regimen with the aim of maintaining virus replication suppression.”

The investigators reason that archived resistance mutations could quickly become dominant when treatment was simplified to include three NRTIs, concluding “that switching to a triple NRTI-based protease-sparing regimen may not be as effective in patients who began treatment on suboptimal a(nti) r(etroviral) t(herapy)”.

Further information

Abacavir

Eradicating HIV - discussion of viral reservoirs.