Patients receiving antiretroviral therapy (ART) for HIV typically undergo routine blood tests to ensure that the drugs are working and not causing toxic side effects. However, according to a randomized trial conducted in two Sub-Saharan African countries, ART can be dispensed safely without routine testing for toxicity.1 On the other hand, the trial also found that routine lab testing may provide a small improvement in the efficacy of treatment by alerting clinicians that the patient's regimen is losing its effectiveness and that a switch of drugs is warranted. Twenty-one percent of patients assigned to routine laboratory monitoring either died or developed a serious HIV- related condition during five years of ART, compared with 28% of those whose physicians relied on clinical assessments to make therapeutic decisions.
Routine laboratory testing of patients receiving ART is often cost-prohibitive in developing nations; even if such testing is available, the resources it requires might be better used to increase access to ART. For this reason, researchers sought to determine whether routine monitoring of ART efficacy and toxicity has meaningful benefits for patients' long-term clinical outcomes.
Between January 2003 and October 2004, symptomatic HIV-positive adults aged 18 or older who had not begun ART and who had CD4 cell counts of less than 200 per microliter were enrolled from three centers in Uganda and one in Zimbabwe; participants were followed until December 2008. Patients were excluded if they were considered unlikely or unable to follow an ART regimen or attend follow-up sessions, were undergoing chemo-therapy or the intensive phase of antituberculosis therapy, had an acute infection or laboratory results that precluded initiation of ART, or were pregnant or breast-feeding. The final sample of 3,316 participants was randomized into two groups, one that received clinically driven monitoring (i.e., the need to switch medications or test for toxicity was determined by the clinician's assessment) and one that received both clinical monitoring and routine laboratory testing for toxicity and efficacy.
All participants received triple-drug ART. At screening, as well as at four weeks, 12 weeks and every 12 weeks thereafter, participants from both study groups met with a doctor and had routine blood tests. Results for participants in the routine testing group were given to their clinicians, while those for patients in the clinically driven monitoring group were provided only if requested for clinical reasons or if the tests revealed potentially life-threatening toxicity. In both groups, clinicians were free to perform additional tests and treat new diagnoses as needed. Patients who developed severe HIV-related conditions—specifically, those classified as stage 4 conditions by the World Health Organization, such as recurrent severe pneumonia or Karposi's sarcoma—were switched to a different (second-line) ART regimen; in the routine monitoring group, the patients were also switched if their CD4 cell counts fell below 100 per microliter. A nurse used a standard symptom checklist to review the status of all participants every four weeks.
The trial examined rates of two primary patient outcomes. The first was either death or the occurrence of a new stage 4 HIV-related condition; the second was the occurrence of a serious adverse event, which referred not only to HIV-related conditions but to any serious medical condition that was life-threatening or fatal, caused disability or required a hospital admission. The study was a noninferiority trial, which meant that the researchers aimed to show that a new treatment (here, clinically driven monitoring) was not worse than a standard treatment (routine monitoring) by a prespecified, clinically meaningful degree; they determined that the upper 95% confidence limit of the hazard ratio for the measure encompassing death and new stage 4 conditions could not exceed 1.18. Kaplan-Meier plots, log-rank tests and proportional hazard models were used to compare outcomes between groups.
At baseline, characteristics of the two study groups were similar: Some 64–66% of participants were women, their median age was 36 and 99% reported having been infected with HIV through heterosexual contact. Median follow-up was 4.9 years, during which time both groups attended 98% of nurse visits and 99% of doctor visits.
The proportion of participants reporting one or more serious side effects from therapy did not differ between the two groups (16– 17%). Although the blood test results were available upon request to the clinicians of patients in the clinical monitoring group, fewer than 4% of these results were requested.
By the end of the trial, the proportion of patients who had been switched to a second-line ART regimen was higher in the routine monitoring group than in the clinically driven monitoring group (22% vs. 19%). Twenty-eight percent of patients receiving clinically driven monitoring, but only 21% of those receiving routine monitoring, had died or developed a new stage 4 condition, indicating that one death or stage 4 condition was prevented for every 59 patients receiving routine monitoring for one year. This translates to a hazard ratio of 1.31 (95% confidence interval, 1.14–1.51), which means that the two treatment strategies can not be considered equivalent.
A prespecified subgroup analysis showed that outcomes for the two groups were generally similar for the first two years of treatment. However, from the third year on, deaths and stage 4 conditions were roughly twice as frequent among patients receiving clinically driven monitoring as among those receiving routine monitoring. Thirteen percent of patients in the clinically driven monitoring group and 10% of those in the routine monitoring group died, a difference equivalent to one death prevented annually for every 130 patients who received routine instead of clinically driven monitoring.
The authors conclude that although routine laboratory monitoring offers a "small" benefit in reducing disease progression, "good ART outcomes with low mortality can be attained without" such monitoring. They suggest that "a greater public health effect would be gained from widening access to ART for untreated patients with low CD4 cell counts who are at high risk of mortality rather than providing routine laboratory monitoring for people already receiving ART." In light of this, the authors suggest that "funding…be focused on drug procurement, strengthening of diagnostic laboratory services, and training and supervision for health care workers to foster quality clinical monitoring, [in order] to support scale-up of ART rollout to rural Africa where 60% of the HIV-infected population live."
1. DART Trial Team, Routine versus clinically driven laboratory monitoring of HIV antiretroviral therapy in Africa (DART): a randomised non-inferiority trial, Lancet, 2010, 375(9709):123–131.