CROI: 70% still on first regimen after three years in Khayelitsha, but lactic acidosis a major challenge

Lactic acidosis is occurring at an unusually high frequency in South African women receiving d4T-based therapy and is the main reason for toxicity-related treatment switches, a South African research group reported on Monday at the Thirteenth Conference on Retroviruses and Opportunistic Infections, taking place this week in Denver.

Dr Andrew Boulle of the University of Cape Town was presenting a three-year follow-up on the MSF treatment programme in Khayelitsha, a township on the outskirts of Cape Town. The programme pioneered the delivery of antiretroviral therapy in resource-limited settings and has been an important model for other treatment programmes.

The treatment programme in Khayelitsha began in May 2001; by the end of 2004, more than 1,700 previously untreated HIV-positive adults had started antiretroviral therapy. The initial standard regimen consisted of AZT (zidovudine, Retrovir), plus 3TC (lamivudine, Epivir), plus either nevirapine (Viramune) or efavirenz (Sustiva). Later in the study period, the standard first-line regimen was changed to d4T (stavudine, Zerit) plus 3TC plus nevirapine.

In this prospective cohort study, data were collected through the end of June 2005. The researchers analysed the time to virological failure (defined as two consecutive viral load measurements above 5,000 copies/ml), the time to switching drugs, and the reasons for treatment changes.

In a preliminary analysis of data collected through the end of 2004, about one in ten patients (11.9%; 95% confidence interval [CI] 7.6-18.4) had switched to a second-line regimen by 36 months; fewer, the researchers said, than would be expected based on viral load results. Overall, 59% of patients remained on their initial regimen at 36 months.

By month 24, 7.2% had switched from their initial regimen due to treatment failure, and by month 36, 10.8% had switched due to failure.

After 24 months on therapy, similar proportions of patients had switched from d4T, AZT, and nevirapine due to toxicity (8.5%, 8.7%, and 8.9%, respectively); in contrast, only 1.7% switched from 3TC. Most switches (36 of 44, or about 82%) were attributable to anaemia in patients on AZT.

Among participants taking d4T, the rates of switching due to symptomatic hyperlactatemia/lactic acidosis and peripheral neuropathy were 15 per 1,000 patient-years (PY) (95% CI 9-31) and 17 per 1,000 PY (95% CI 8-29), respectively.

However, when cases of hyperlactatemia were analysed according to weight and gender, the risk was found to be substantially higher in some subgroups. Women weighing 75kg or more had an incidence of hyperlactatemia of 316 cases per year 100 patient years of follow-up, compared to an incidence close to the population average for women weighing less than 75kg.

Multivariate analysis found that women weighing 75kg or more had an adjusted hazard ratio of 25 for lactic acidosis when compared with males, while women weighing between 60 and 75kg had an AHR of 5.6 for lactic acidosis. Weight gain of 5kg or more by month 3 of treatment carried an AHR of 2.5 for lactic acidosis.(all p

Discussion following the presentation could not offer a biological explanation for this phenomenon, but questions from the floor revealed that a similar trend has been seen in the Botswana treatment programme.

After d4T was substituted for AZT in the standard first-line regimen, fewer patients changed therapy due to anaemia or neutropenia, but more switched due to lactic acidosis or peripheral neuropathy. Among participants taking nevirapine, switching was more often due to elevated transaminases (liver enzymes) than skin rash.

Key messages from Khayelitsha

• Seven out of ten patient remained on their initial regimen after three years of treatment.
• One in ten patients were taking a second-line regimen by 36 months.
• Switches from AZT and nevirapine occurred largely during the first six months of treatment.
• Switches from d4T occurred largely beyond month 6 of treatment, predominantly due to lactic acidosis and peripheral neuropathy rather than lipoatrophy.
• Haematological toxicity related to AZT was greater in patients with CD4 cell counts below 50 cells/mm³ at baseline.

Is it necessary to review the regimen being used in South Africa?

Dr Boulle highlighted some of the problems facing South African doctors as they grapple with the potential toxicity of some of the drugs being used in the national treatment programme.

Dropping d4T and using AZT instead as the preferred drug in the nucleoside analogue backbone for all patients would require more intensive monitoring, he pointed out. Monthly haematological monitoring is required for patients taking AZT during the first six months of treatment.

Using tenofovir instead of d4T is attractive but unaffordable, he noted. Even at the `access` price offered by manufacturer Gilead the drug costs ten times as much as d4T, and has to be supplied on a named patient basis because it is not yet registered in South Africa.

Reducing the dose of d4T for all patients to 30mg twice daily might limit toxicity, but this is speculative. A stratified approach based on the risk factors for lactic acidosis identified in this analysis may be the best approach at present, he suggested, with women of high body mass index or weight above 75kg started on treatment with AZT rather than d4T.