This survey demonstrates that in 4 sites of the Malawi ART program, ≥70% of patients suppressed HIV RNA at 12–15 months after ART initiation, meeting the WHO target of HIVDR prevention. Detected resistance was low at 12–15 months, and HIVDR profiles were as expected for NNRTI-based therapy and highly associated with the presence of baseline resistance. Possible resistance was considerable, mainly determined by high defaulter rates.
Although it is reassuring that HIVDR targets were met, we observed several issues of concern. There was considerable intersite variation of HIVDR prevention and possible HIVDR resistance. The Mzuzu Clinic reported lower viral suppression partly because a number of participants who were alive on first-line ART (and thus included in the denominator) were missed for HIV RNA specimen collection at 12–15 months. A high number of defaulters in QECH and Mzuzu caused relatively low virologic suppression and high possible resistance rates in those sites. The Lighthouse Clinic [24
] and the Thyolo District Hospital have active community tracing programs of patients who miss scheduled ART visits, which prevents defaulting and allows correct categorization of participants as deaths rather than defaulters, thereby resulting in higher viral suppression rates. This and other efforts to retain patients in care are important to improve ART outcomes and to reduce possible HIVDR rates. Men started ART with more advanced HIV disease than women and had a higher mortality rate as a consequence. Although our data did not demonstrate this, it is possible that men also have a higher chance of virologic failure and HIVDR development, as advanced disease at baseline was a risk factor for ART failure in other studies [25
The prevalence of resistance at 12–15 months observed in this survey is comparable to a report from neighboring Mozambique [9
]. The low level of detected resistance in both cohorts may be related to optimal adherence, reflected by a high degree of appointment keeping for drug collection among patients who were alive on first-line ART. The resistance rates in this survey were also similar to those from our retrospective study in 2007 at the same sites [18
], which does not suggest a progressive increase in HIVDR as the ART program matures. However, it will be important to maintain HIVDR vigilance by continuing regular surveys. The predominance of the M184V and K103N mutations was expected given the standard first-line regimens used and was consistent with other studies from Malawi and the region [9
]. The relatively low rate of TAMs we identified at the 12–15-month follow-up point confirms other studies suggesting that TAMs develop relatively slowly compared with M184V- and NNRTI-associated mutations. However, ZDV substitution for d4T toxicity was uncommon (7.1%) in this cohort and TAMs are more likely to develop with ZDV-containing regimens than with d4T-containing regimens [26
]. Similar to other studies on treatment failure of d4T-based regimens, we observed K65R mutations although at lower frequency than when ART failure is detected using clinical or immunological monitoring [4
]. Baseline resistance to NNRTIs was relatively high in this cohort. One explanation could be the widespread use of single-dose NVP in the Malawi PMTCT program, although few women reported its use. However, the mutations occurred in both men and women, suggesting that transmitted resistance or unidentified ART exposure (such as antiretroviral drug sharing) may have occurred.
The 16–24-year age range at ART initiation was a predictor for HIVDR. Adolescents typically struggle with management of chronic diseases and HIV [27
]. Programs addressing the challenges of drug adherence among adolescents may minimize the emergence of drug resistance. The risk of developing HIVDR was also higher in those who missed antiretroviral drugs for a period of 14–31 days compared with those missing drugs for >31 days and those missing <14 days. Wild-type HIV may become the predominant quasi-species when antiretroviral drugs are missed for a longer duration, and therefore drug mutations may have not been detected in those who missed ART for >31 days [28
]. Missing drug pickup by a small number of days may not have been associated with missed doses if patients accumulated tablets at home during ongoing therapy, as clinic visits were scheduled every 28 days, whereas 30-day supplies were dispensed. As expected, NNRTI resistance at baseline was strongly associated with resistance at 12–15 months, although some were able to suppress HIV replication despite the presence of baseline mutations. Baseline NRTI mutations did not prevent viral suppression of 4 patients alive on first-line ART at 12–15 months; however, NRTI resistance at baseline was too infrequent to enable drawing conclusions on its consequences for HIVDR and virologic suppression at 12–15 months.