Surreptitious Antiretroviral Use as an Explanation for Apparent Elite HIV Control

Dear Editor,

Representing less than 1% of all human immunodeficiency virus (HIV)-infected individuals, elite controllers are characterized by their ability to maintain viral loads<50 copies/mL without antiretroviral (ARV) therapy. ^1,2 Due to implications for vaccine design, elite controllers are the focus of much attention.

We encountered a patient whom we initially thought was an elite controller. This 41-year-old woman originally from Burundi was seen in April 2009 for HIV infection that was detected during immigration screening 1 year earlier. She was asymptomatic and did not have any significant past medical history. She reported not taking any medications and denied taking any ARVs at any time in the past. Her CD4 count at the time of her first clinic visit was 470 cells/μL (17%) and her CD4:CD8 ratio was 0.29. As her viral load was<50 copies/mL, it appeared as though she was an elite controller.

The patient continued to do well clinically and maintained a CD4 count between 594–995 cells/μL on quarterly visits, with undetectable viral loads as measured by The Quantiplex HIV-1 RNA 3.0 Assay (Chiron®). In July 2010, on screening blood work for enrollment in a trial involving elite controllers and long-term nonprogressors, the patient's viral load was found to be 31,810 copies/mL and her CD4 count had declined to 287 cells/μL (9.3%). This was in clear contrast to results from March 2010, which had revealed a viral load<50 copies/mL and CD4 count of 735 (25.1%).

Due to this dramatic change, blood tests were repeated in August 2010, revealing a viral load of 106,543 copies/mL and CD4 count of 384 (7.7%). Other than for mild fatigue, the patient felt well. Furthermore, there was no history of any acute preceding illness to explain this change or reported risk of acquiring superinfection. The patient was started on efavirenz/tenofovir/emtricitabine (Atripla®) in October 2010.

Surprisingly, genotyping of the virus using the vircoTYPE HIV-1® resistance assay from the blood sample taken in July 2010 demonstrated CCR5-tropic Clade C virus with the following drug resistance mutations: reverse transcriptase: 67wt/N, 70wt/R, 98wt/G, 181 wt/C, and 219 wt/Q; protease: 10F, 15V, 20wt/R, 24wt/I, 36I, 54V, 62wt/V, 63P, 69K, 82wt/A, 89M, and 93wt/L. These mutations conferred resistance or reduced susceptibility to zidovudine, lamivudine, didanosine, tenofovir, nevirapine, efavirenz, etravirine, indinavir, nelfinavir, saquinavir, fosamprenavir, lopinavir, and atazanavir. There were no drug resistance-associated mutations detected in the integrase gene.

To confirm this unexpected result, resistance testing was performed on the blood sample from August 2010. Results revealed again the following mutations: reverse transcriptase 70wt/R, 98wt/G, 103wt/N, 181wt/C, 219wt/Q, and 181wt/C; protease: 10wt/F, 15V, 20wt/R, 36I, 54wt/V, 63 wt/P, 69K, 70wt/R, 89M, and 93L. The patient's ARV regimen was subsequently changed from Atripla® to tenofovir/emtricitabine (Truvada®), ritonavir (Norvir®), and darunavir (Prezista®). Within 3 months, her viral load became undetectable (<40 copies/mL) and her CD4 count rose. Her most recent CD4 count in October 2011 was 604 cells/μL (17.3%), her viral load remains undetectable and she continues to be clinically well.

Blood that had been cryopreserved from March 2010, at which time her viral load was undetectable without the apparent use of ARV therapy, was analyzed for suspected ARV use using a validated high performance liquid chromatography tandem mass spectrometry assay that was modified from a previously published method by our group. ³ This revealed the presence of raltegravir at 0.23 mg/L, etravirine at 0.35 mg/L, and ritonavir at 0.35 mg/L.

Many elite controllers either maintain relatively stable CD4 T cell counts over time or experience declines in their CD4 T cell counts. ^{4 –6} Therefore, a possible clue to surreptitious ARV use in our patient may have been her rising CD4 T cell count. Furthermore, the resistance profile for this patient's HIV isolate suggested exposure to several ARVs, as multi-class resistance in HAART-naïve individuals is uncommon, and particularly to the degree seen here. In a study by Chilton et al., patterns of HIV resistance by clade were explored in a review of resistance test results from the United Kingdom HIV Drug Resistance Database between 2001 and 2006 for treatment-naïve adults migrating to the United Kingdom. Among individuals who had resistant virus, defined as the presence of>1 surveillance drug resistance mutation in each drug class, triple class ARV resistance occurred in 1% (1/128), 3% (2/68), and 4% (30/745) of individuals with clades C, non-B non-C, and B virus, respectively. ⁷ Interestingly, within the NRTI class, the prevalence of M184V was higher, and the prevalence of 215 variants was lower, than anticipated. As M184V is rarely detected in primary infection cohorts, ⁸ and as 215 variants are more common in untreated—rather than treated—individuals, ^9,10 the authors suggest that some “HAART-naïve” individuals in their study on individuals with clade B virus might actually be HAART-experienced. ⁷ The extent or duration of ARV drug use in our patient is unknown and, therefore, the pattern of drug resistance is unpredictable.

Although likely uncommon, we present the case of a woman who was erroneously thought to be an elite controller. Her reason for denying the use of ARV therapy has never been clearly articulated. The frequency with which this occurs in unknown, but this case raises the issue of whether elite controllers should be screened for surreptitious ARV use. Unmonitored use of ARVs carries significant safety risks to patients and may provide fertile ground for the development of drug resistance. Screening elite controllers for surreptitious ARV use may be of particular value if these individuals are to be enrolled in a clinical trial. Inclusion of such individuals in clinical trials and cohort studies of elite controllers may potentially skew outcomes observed in these studies, especially given that studies involving elite controllers include low numbers of subjects in the first place. This may, in turn, obscure our understanding of the immunological mechanisms at play in this very select and important group of individuals. Given the potential implications, screening of elite controllers for surreptitious ARV use may be merited and would seem prudent, if not necessary, prior to enrollment of such individuals into clinical trials.