Switches to second-line ART occurring faster where viral load testing available

Viral load testing in resource-limited settings results in switching to second-line therapy earlier and at higher CD4 counts when compared to CD4 testing alone, The Antiretroviral Therapy in Lower Income Countries (ART-LINC) collaboration of the International Epidemiological Databases to Evaluate AIDS (leDEA) reports in the September 10^th issue of AIDS.

Low CD4 counts at the start of antiretroviral therapy were predictive of switching in programmes with or without viral load monitoring. This finding lends support to arguments for earlier initiation of antiretroviral therapy to help reduce high mortality during the first few months on ART as well as help preserve first-line regimens.

The practicalities and cost-effectiveness of viral load monitoring in scale-up of antiretroviral therapy in resource-limited settings has been a topic of considerable debate over the past year.

The World Health Organization believes this is useful but not essential in a public health approach to antiretroviral treatment. However evidence of the long-term impact of protracted virological failure on second-line treatment is lacking.

Increased exposure to antiretroviral therapy in resource-limited settings in recent years raises the risks of resistance and ensuing treatment failure and the consequent ever-increasing need to switch patients to alternative second-line regimens.

While viral load monitoring is common practice in the diagnosis of treatment failure and a guide in switching to alternative regimens in industrialised countries, its use in resource-limited settings is restricted due to cost and availability.

Reliance on clinical examination and, when available, CD4 counts to inform treatment decisions is the norm. This may lead to late switching of regimens, increasing the risk of resistance and negatively affecting long-term outcomes. Second-line regimens are the final option for many in resource-limited settings.

Conversely, some studies have found that unnecessary treatment switches are occurring in large numbers of patients who experience Cd4 cell declines without ciological failure. In these cases the lack of viral load testing is leading to switches to more expensive second-line treatment, wasting scarce and diminishing future treatment options.

ART-LINC is a multi-national, multi-centre collaborative research network of clinics and cohorts providing antiretroviral treatment to people living in resource-limited settings in Africa, Asia and Latin America. Data from these programmes are collected on patient characteristics and treatment outcomes.

A multi-cohort study of 17 programmes in 14 countries was undertaken. All sites monitored CD4 cell counts and provided access to second-line antiretroviral regimens. Ten of the sites monitored viral load. Rates of switching, time to switching and determinants of switching were compared between sites that offered viral load monitoring and those that did not.

Out of a database of close to 35,000 patients, data from 20,113 patients were analysed. 31.7% (6,369) were from programmes with routine viral load monitoring and 82.5% (16,591) from sub-Saharan Africa.

Median CD4 counts at the start of antiretroviral therapy were lower in programmes with viral load monitoring than in programmes without (97cells/mm³ compared to 129 cells/mm³). In sites with and without viral load monitoring the four most common first-line regimens were lamivudine (3TC) with stavudine (d4T) or zidovudine (AZT), combined with efavirenz (EFV) or nevirapine (NVP).

576 patients (2.9%) switched to a second-line treatment regimen. The rate of switching was higher in programmes with viral load monitoring than those without: 3.2 switches per 100 patient years (95% CI: 2.8-3.6) compared with 2.0 switches per 100 person-years (95% CI: 1.8-2.3). Switching happened earlier in programmes with access to viral-load monitoring. Median time to switching in programmes with viral-load monitoring was 16.3 months (interquartile range (IQR): 10.1-26.6) compared to 21.8 months (IQR: 14.0-21.8) in those without.

Fewer patients switched in more recent calendar years than in the earlier years of antiretroviral scale-up. The authors note this was not related to increases in CD4 counts at the start of ART but more likely reflect changes in practice due to increases in people starting antiretroviral therapy from the early 2000s.

Switching was more likely in programmes with access to viral load monitoring during months 7-18 following the start of antiretroviral therapy (adjusted HR 1.38; 95% CI 0.97-1.98), similar during months 19-30 (aHR 0.97; 95% CI 0.58-1.60) and less likely during months 31-42 (aHR: 0.29; 95% CI: 0.11-0.79). These results were similar across all first-line regimens.

54.9% (316 patients) who switched changed both nucleoside reverse transcriptase inhibitors (NRTIs).

Median CD4 cell count at the time of switching in programmes with viral load monitoring was 161 cells/mm³ (IQR: 77-265) compared to 102 cells/mm³ IQR: 44-181) in those without.

Low CD4 cell count at the start of antiretroviral therapy, but not clinical stage, was a predictor for switching in all sites.

Of the 576 patients who switched, reasons for switching were available for 42% (241): 74% (179) were due to treatment failure, 10% (25) due to toxicity and the remaining 15% (37) for other reasons. 399 (69.3%) had a CD4 count at the time of switching; 153 (26.6%) had a viral load measurement and 118 (20.5%) had both. Of these patients, 66% or 263 patients had immunological failure, virological failure or both.

The authors point out that the sites in their study represent those with medical record systems, access to CD4 cell counts and second-line regimens and as such are not necessarily representative of all sites in these countries providing antiretroviral therapy.

In addition they note that they did not examine for clinical failure, no information on adherence or drug resistance was available and information about reasons for switching was only available for some patients.

The authors cite a recent multi-country survey by the WHO that revealed considerable variability in rates of switching to second-line regimens. The authors do not believe that significant differences in primary resistance to NRTIs or non-nucleoside reverse transcriptase inhibitors (NNRTIs) is the cause but rather the availability or not of viral load monitoring and differences in clinical practice within and between countries. They cite the example within South Africa in two township programmes in Cape Town, where therapy is switched after two consecutive viral loads above 5000 copies/ml in Khayelitsha, whereas in Gugulethu the threshold is 1000 copies/ml.

As a result of their finding that in programmes with access to viral load monitoring, patients are switched earlier and with higher CD4 cell counts, the authors suggest that future studies “examine what the consequences of earlier or later switching are for clinical outcomes.” Also, “further research is required to determine the optimal frequency of determining CD4 cell counts and of measuring viral load to maximize cost-effectiveness and optimize patient outcomes in different settings in lower income countries,” they conclude.