Treatment resistance after sequential interruption of a non-nucleoside reverse transcriptase inhibitor-based regimen

Abstract

Stopping HIV therapy may reduce costs and side effects, but carries the risk of increased immune suppression and the emergence of resistance. To investigate this issue, the Strategies for Management of Anti-Retroviral Therapy (SMART) randomized study was conducted to compare the long-term consequences of two antiretroviral management strategies: continuous therapy versus scheduled treatment interruption. The study was interrupted for harm [1]. Non-nucleoside reverse transcriptase inhibitors (NNRTI), namely efavirenz and nevirapine, are frequently part of recommended combinations for the treatment of HIV-1 infection. NNRTI, however, have a low genetic barrier to the selection of resistance, and a single key mutation in the NNRTI-specific pocket site or in the surrounding domain of reverse transcriptase poses a major therapeutic problem for this class [2]. If administered as monotherapy, NNRTI selects for resistant mutants within one week, most commonly those harbouring the K103N or Y181C mutation, which confer cross-resistance to other approved drugs in the NNRTI class. The K103N mutation appears to have little effect on the replicative capacity of HIV-1, allowing variants to persist long after nevirapine therapy has been stopped. Resistance mutations have been observed after exposure to a single dose of nevirapine [3]. For the reasons described above, the SMART protocol recommended that NNRTI should be discontinued 7 days before nucleoside reverse transcriptase inhibitor (NRTI) backbone. Here we report the case of a patient who was included in the SMART study and who developed resistance after sequentially stopping a nevirapine-containing regimen. Case report A 44-year-old Caucasian man, diagnosed HIV-1 positive in 1994, was enrolled in SMART in March 2005 in the scheduled interruption treatment group. At this time, the CD4 lymphocyte count was 896 cells/μl (36%, CD4/CD8 0.87) and RNA HIV was undetectable (< 50 copies/ml). The nadir CD4 cell count was 2 cells/μl in 1995. He had no history of opportunistic infections. Lamivudine (300 mg once a day) was initiated in 1995. In April 1996, the patient received indinavir (2400 mg/day), lamivudine (300 mg/day) and didanosine (400 mg/day). In March 1998, indinavir was stopped and switched to ritonavir (1000 mg/day). NNRTI with efavirenz (600 mg/day) substituted ritonavir in May 1999 for simplification. Efavirenz was switched to nevirapine (400 mg/day) in August 2003. Our case had had undetectable RNA HIV of less than 400 copies/ml since 1996 and less than 50 copies/ml since 1999 in all 3-month routine follow-ups until May 2005. According to the protocol, he stopped nevirapine on 12 April 2005 and stopped didanosine and lamivudine on 19 April 2005. His CD4 lymphocyte count decreased (Table 1).Table 1: Immunovirological history according to antiretroviral therapy.In October 2005, the CD4 lymphocyte count was 129 cells/μl and according to the SMART protocol, the treatment was reinitiated with cotrimoxazole for the prevention of opportunistic infections. We chose to restart the same treatment with didanosine (400 mg), lamivudine (300 mg) and nevirapine (200 mg once a day for 14 days and then 200 mg twice a day). An HIV-1 genotypic resistance test for each drug was performed in March 2006, and revealed Y181C and M184I reverse transcriptase mutations, which confer drug resistance to NNRTI and lamivudine, respectively. The NRTI backbone was changed and lopinavir–ritonavir replaced nevirapine with virological control. NNRTI is an attractive class of treatment for HIV-1 infection because of its good bioavailability, long half-life and convenient administration. It should be stressed that the necessity to stop NNRTI is not uncommon in clinical practice, because of adverse events, concomitant conditions or other patient non-medical related factors. Two factors have been associated with the emergence of NNRTI resistance: previous suboptimal treatment [4] and treatment interruptions [5,6]. Both were present in our case, and the emergence of resistance occurred despite sequential treatment interruption. Even with relatively long-term virological control, antiretroviral therapy cessation clearly triggered resistance in this case. Our report has potential important implications when stopping an NNRTI, in particular in patients with previous NRTI suboptimal therapy. The current protocol for NNRTI cessation requires that two NRTI alone are sufficient to control RNA-HIV replication during the 7 days of NNRTI clearance, and thus avoid HIV replication in the presence of subtherapeutic NNRTI levels [7]. This may not be possible in patients with previous suboptimal NRTI treatment, suggesting the presence of archived resistant viruses, even in the case of long-term virological control. We suggest that in this setting, protease inhibitors should replace NNRTI in addition to NRTI, during the delay of NNRTI clearance (which remains to be determined) in order to preserve this potent antiretroviral option. Acknowledgement We wish to dedicate our work to our SMART volunteer who consented to make a sacrifice of himself for the good of others. The authors also thank Pascale Goubin for data management.