Pediatric Highly Active Antiretroviral Therapy in Africa: Potential Benefits of a Family‐Centered Model

Abstract

To the Editor-We read with interest the article by Goicoechea et al. on the role of antiretroviral medications and tenofovirassociated kidney dysfunction. The authors concludedthatregimensbasedonritonavirboosted protease inhibitor (PI/r) were associated with a greater decline in kidney function than regimens based on nonnucleoside reverse-transcriptase inhibitors (NNRTIs) among patients receiving tenofovir. However the majority (75%) of patients in their analysis who were receiving a PI/r-based regimen were receiving lopinavir/ritonavir and a similar proportion (79%) of patients who were receiving an NNRTI-based regimen were receiving efavirenz. Thus we suggest that their findings may relate more to the specific effect of lopinavir/ ritonavir regimens versus that of efavirenz- based regimens on tenofovirassociated kidney dysfunction rather than to PIs or NNRTIs as a class. We performed similar analyses in an observational cohort with a larger sample size and found that amprenavir-based regimens were associated with a greater decline in kidney function than efavirenzbased regimens among 445 patients who initiated tenofovir treatment. Our findings also suggested that ritonavir boosting did not account for the differences between agents. In addition the concurrent use of didanosine with tenofovir was associated with an increased risk of kidney dysfunction. A second issue in the article by Goicoechea and colleagues is the decision to only include patients who received tenofovir for 48 weeks. The authors stated that 48 (24%) of 199 patients did not start or discontinued tenofovir before week 40 and were excluded. It is not clear whether patients who discontinued use of tenofovir before 48 weeks did so because of kidney dysfunction. In our cohort severe declines in kidney function although uncommon occurred on average 3-4 months after initiation of tenofovir treatment (range 3-9 months) whereas more moderate declines generally occurred after 6 months (mean 7 months [range 1-33 months]). In conclusion the article by Goicoechea et al. [1] provides interesting information on risk factors for tenofovirassociated kidney dysfunction. Future studies of larger cohorts of patients are needed to examine the roles of all individual medications rather than the roles of classes of medications in tenofovirassociated kidney dysfunction. (full text)