Secondary Logo

Journal Logo

Use of HIV Resistance Testing After Prolonged Treatment Interruption

Iarikov, Dmitri E MD*; Irizarry-Acosta, Melina MD; Martorell, Claudia MD; Rauch, Carol A MD§; Hoffman, Robert P MD; Skiest, Daniel J MD*

JAIDS Journal of Acquired Immune Deficiency Syndromes: March 1st, 2010 - Volume 53 - Issue 3 - p 333-337
doi: 10.1097/QAI.0b013e3181c79ab0
Clinical Science

Background: HIV-1 genotypic resistance testing is not routinely recommended for patients who have been off antiretroviral therapy (ART) for longer than 4 weeks. We assessed the results and use of resistance testing in patients off ART.

Methods: All HIV resistance genotypes from November 2003 through April 2008 were reviewed from one large teaching hospital and two private HIV practices. Inclusion criterion was having a genotypic resistance test after an ART interruption of at least 2 months. Medical records were reviewed using a standardized data collection sheet.

Results: Sixty-two of 304 treatment-experienced patients with HIV genotypes met the inclusion criteria. Prior cumulative ART class exposure included nucleoside reverse transcriptase inhibitors in 54 patients, nonnucleoside reverse transcriptase inhibitors in 32 patients, and protease inhibitors in 30 patients. Resistance testing was performed at a mean of 12 months (range, 2.5-48 months) after ART interruption. The mean time between ART interruption and resistance testing did not differ for patients with mutations and those without mutations detected. Seventeen of 62 (27.4%) patients were found to have resistance mutations. Eleven patients were found to have mutations to nonnucleoside reverse transcriptase inhibitors, four patients had mutations to nucleoside reverse transcriptase inhibitors, and two patients had protease inhibitor-associated mutations. No patient had multiclass resistance. Among the 17 patients with mutations after treatment interruption, 15 had mutations that were either not present on a prior genotype (n = 2) or did not have a prior genotype (n = 13).

Conclusions: HIV genotypic resistance assays may identify mutations even when performed after a prolonged treatment interruption and may offer clinically significant information. Current guidelines that discourage resistance testing after treatment interruptions of longer than 4 weeks should be re-evaluated.

From the *Division of Infectious Diseases, †Department of Medicine, and §Department of Pathology, Baystate Medical Center, Tufts University School of Medicine, Springfield, MA; ‡The Research Institute, Springfield, MA; and ¶Mercy Medical Center, Springfield, MA.

Received for publication April 7, 2009; accepted August 10, 2009.

Correspondence to: Dmitri E. Iarikov, MD, The Food and Drug Administration, 10903 New Hampshire Avenue, Building 22, Room 6211, Silver Spring, MD 20993 (e-mail:

Back to Top | Article Outline


Currently available highly active antiretroviral therapy usually results in excellent virologic and clinical responses.1,2 Although scheduled long-term antiretroviral therapy (ART) interruptions are not recommended, in practice, patients frequently stop ART. Prior studies indicated that 20% to 78% of patients undergo treatment interruptions.3-5 Reasons for therapy interruption include intolerance, adverse effects, poor virologic response, pill fatigue, incarceration, cost of medications, and ongoing substance abuse. Because treatment interruptions frequently occur without knowledge of the HIV clinician, it is important to know how to manage such patients.

The optimal approach to a patient who has interrupted highly active ART is not clear. Should the most recent ART regimen be restarted or should a resistance assay be performed before reinitiation of therapy? The use of resistance testing in this situation has not been established. It has been recognized that in the absence of selective drug pressure, secondary resistant mutations are rapidly overgrown by wild-type virus.6-10 Thus, some have proposed that HIV-1 resistance testing should only be done when a patient has been off therapy for less than 2 weeks.7 In the current US Department of Health and Human Services ART guidelines, resistance testing is not routinely recommended for patients who have been off ART for longer than 4 weeks.10,11

We hypothesized that because certain resistance mutations decay more slowly in the absence of selective drug pressure, it may be clinically useful to obtain genotypic resistance testing even if patients have stopped ART for longer than 4 weeks.7,8,12-14 If a resistance mutation is present, it is likely to still be clinically significant. We sought to determine the prevalence of drug resistance in treatment-experienced HIV-infected patients who underwent therapy interruption and to determine the use of resistance testing in this group of patients.

Back to Top | Article Outline


HIV-infected patients from the HIV clinics affiliated with Baystate Medical Center and two HIV private practices, all in Springfield, MA, were included in the study. Inclusion criteria consisted of receipt of any ART followed by a treatment interruption of at least 2 months. In addition, the patient needed to have at least one genotypic resistance test performed after the treatment interruption. Data on previous ART regimens were collected by reviewing medical records. Patients with genotypic resistance tests performed from November 2003 through April 2008 were included. HIV resistance tests used included HIV GenoSure (LabCorp, Research Triangle Park, NC), HIV-1 Genotyping for Drug Resistance (Quest Diagnostics Nichols Institute, Cambridge, MA), and TRUGENE HIV-1 Genotyping Kit (Visible Genetics, Suwanee, GA) resistance assays.

Mutations were classified based on the International AIDS Society-USA drug resistance mutations list. Only major mutations were considered clinically significant and were included in the analysis.14,23 Only the results of the genotype that was performed after being off ART for more than 2 months were analyzed. Patients with resistance mutations were compared with patients without resistance mutations using a two-tailed t test for continuous variables (Microsoft Excel Version 2008; Microsoft Inc., Redmond, WA) and chi-square test for categorical variables (Statview for Window, Version 5.0.1; SAS Institute, Cary, NC). The Baystate Medical Center Institutional Review Board approved the study.

Back to Top | Article Outline


Five hundred twenty-four genotypes in 403 patients were performed during the study period. Ninety-nine patients were treatment-naïve and 304 were treatment-experienced. Sixty-two of 304 treatment-experienced patients met the inclusion criteria of being off ART for at least 2 months before the resistance test, which constituted 12% of the 524 resistance assays obtained during the study period. There were 32 women and 30 men with a median age of 39 years (range, 15-65 years). Thirty-four patients were Hispanic, 15 patients were black, and 13 were white. HIV risk factors included heterosexual transmission in 29 patients, intravenous drug use in 19 patients, male-male sex in 10 patients, both intravenous drug use and heterosexual contact in three patients, and perinatal transmission in one patient. Twenty-four patients were hepatitis C antibody-positive.

Prior ART included the combination of a protease inhibitor(s) (PI), nonnucleoside reverse transcriptase inhibitor (NNRTI), and nucleoside reverse transcriptase inhibitors (NRTI) in 14 patients; a combination of NNRTI and NRTIs for 18 patients; and a combination of PIs and NRTIs for 16. Six patients reported only having received NRTIs. Eight patients had received prior ART, but records of the specific regimen were not available. Cumulatively, 54 patients had previously been exposed to NRTIs, 32 patients to NNRTIs, and 30 patients to PIs.

The median time between ART interruption and resistance testing was 12 months (range, 2.5-48 months). Resistance mutations were detected in 17 patients (27%). The mean HIV viral load at the time of resistance testing was similar in those with and without mutations (Table 1). The median CD4 count for patients with and without mutations was 203 cells/mm3 and 145 cells/mm3, respectively. The median time between the last receipt of ART and resistance testing did not differ significantly for patients with mutations compared with patients without mutations. The mean time between ART interruption and resistance testing in patients with NNRTI-associated mutations was shorter when compared with patients with NRTI- and PI-associated mutations, 14.4 compared with 29.5 and 36 months, respectively (P = 0.067; Table 2). However, the number of patients with NRTI- or PI-associated mutations was small. Data to estimate the total time on ART were available for 53 patients. The mean time for patients with (n = 14) and without resistance (n = 34) were 55 and 26 months, respectively (P = 0.03).





All detected resistance mutations were confined to a single class of antiretrovirals for each patient. NNRTI mutations were the most common. Eleven patients were found to have a total of 12 major mutations to NNRTIs; one patient had two NNRTI mutations (Table 3). Four patients had mutations to NRTIs and two patients had PI-associated mutations



Among the 17 patients with mutations after treatment interruption, four patients had prior genotypes (Table 3). Two of the four were treatment-naïve at the time of the initial resistance test and did not have resistance mutations on the initial genotype. The other two patients had resistance mutations on prior studies. One patient had mutations at positions 41L, 215V, and 184V on a prior resistance assay; only the 41L was still detected after the patient had been off ART for 38 months. The other patient had mutations at positions 184V and 103N on a previous resistance assay; only the 103N mutation was detected after being off ART for 18 months. Thus, a total of 15 patients had mutations that were either not present or unknown before the current resistance assay.

Patients with prior NNRTI use were more likely to have persistent mutations after treatment interruption. Nine of 32 patients (28%) with NNRTI experience had NNRTI-associated resistance mutations. In contrast, only two of 31 patients (6.4%) with previous PI experience had PI-associated resistance mutations and four of 57 patients (7%) with a history of NRTI experience had NRTI-associated resistance mutations. Both patients who did not know their prior ART regimen were found to have NNRTI-associated mutations; both had K103N.

Forty-nine of 62 patients restarted ART of whom 28 failed therapy (defined as viral load greater than 50 copies/mL after at least 8 months on ART). Eleven stopped taking their ART and four patients were lost to follow up. Nine of 49 patients had repeated genotypic assays after virologic failure after they restarted ART. Two of nine patients were found to have persistence of previously detected NNRTI-associated resistance mutations despite being switched to a PI-based regimen. One patient had persistence of the 103N mutation at 21 months and the other patient had persistence of the 108I mutation at 46 months despite the absence of NNRTI selective pressure. The third patient developed mutations at 103N and 190S on a repeated genotype while receiving an NNRTI-based regimen, was switched to a PI-based regimen, and still had 103N and 190S on a subsequent genotypic assay after 20 months off NNRTIs. When three patients with persistent NNRTI-associated mutations while on a non-NNRTI-based regimen were combined with the already mentioned 11 patients, who had detectable NNRTI-associated mutations while off ART, the median duration of persistence of NNRTI-associated mutations was 15 months (range, 2.5-48 months).

Patients who tested positive for antibodies to hepatitis C were less likely to have persistent mutations after treatment interruption. Three of 17 (17.6%) patients with resistance mutations were hepatitis C (HCV)-positive. In comparison, 21 of 45 (46.6%) patients without resistance mutations were HCV-positive (P = 0.04). There was no difference in time to resistance testing between HC- positive patients (median, 19 months; range, 3-60 months) and HCV-negative patients (median, 13 months; range, 2-156 months) (P = 0.79).

Back to Top | Article Outline


Although studies of structured treatment interruptions have demonstrated the potential for negative clinical consequences, in practice, nonstructured treatment interruptions are common. A recent study from Baltimore demonstrated that 78% of patients underwent at least one treatment interruption of longer than 6 months.5

The best approach to the patient after a treatment interruption has not been determined. Although information on prior regimens and resistance assays should always be sought, this information is frequently not available or unreliable. This is especially true for patients with a history of poor adherence who move from one location to another. Even when previous ART is known, it is not always clear if the ART taken in the past has selected for drug resistance mutations. The use of resistance testing in these cases remains to be determined. One potential strategy is to restart the most recent ART regimen and assess virologic response 4 to 6 weeks later. If the virologic response is not adequate, a resistance test can be obtained at that point. An alternative strategy is to obtain a resistance test before resumption of the prior regimen to select the best combination ART. If clinically significant mutations are present, the patient is unlikely to benefit from use of the specific antiretrovirals. The absence of mutations in this setting is not as helpful, however, because archived mutations may still be present but are not adequately sampled in the absence of ongoing selection pressure.

Despite the fact that current guidelines do not recommend performing resistance testing after treatment interruption of longer than 4 weeks, in practice, physicians do obtain resistance assays under those circumstances.10 In our series, 12% of all resistance tests were done after an ART interruption of at least 2 months.

Our study shows that clinically important mutations may be detected in a significant number of patients after prolonged ART interruption. Twenty-seven percent of our patients had detectable major resistance mutations after being off antiretroviral therapy a median of 12 months and as long as 48 months. In addition, three patients who had NNRTI-associated mutations on prior assays and were subsequently restarted on a regimen not containing an NNRTI continued to have detectable NNRTI-associated mutations for up to 46 months in the absence of selective drug pressure. If we had tested more patients sooner after treatment interruption, it is likely we would have detected even more resistance mutations. Our data are consistent with another study in which 21% of resistance tests showed key mutations in patients who were off therapy for more than 12 months.14

We did not find significant differences between patients with and without mutations in terms of demographics, HIV RNA viral load, or CD4 count. As expected, the total time on antiretrovirals was significantly greater in patients with resistance mutations in comparison to those without resistance mutations. However, the range of exposure was large in both groups; therefore, the time on ART does not appear to be a reliable estimate of the use of resistance testing in an individual patient. Patients with resistance mutations were also less likely to have HCV coinfection. The reasons for this are not clear, but this finding deserves further investigation.

Resistance testing after treatment interruption may be particularly helpful in cases of poorly documented ART history. The K103N mutation was present in two of eight patients (8 and 48 months after ART cessation, respectively) whose ART history was not well documented. Having this information available before restarting highly active ART can be very valuable, because although a first-generation NNRTI-based regimen is attractive in many cases as a result of a low pill burden and a favorable side effect profile, it is unlikely to be efficacious in a patient with NNRTI mutations. Resistance testing may also be valuable in cases of exposure to multiple classes of antiretrovirals. Six of 14 patients who were exposed to all three classes of antiretrovirals were found to have single-class resistance mutations. The detection of NRTI-associated mutations may help with selection of the NRTI component of ART. Two of our patients were found to have a 41L mutation, which would make thymidine analogs a suboptimal treatment option. The finding of major PI-associated mutations as it occurred to two of our patients may prompt the clinician to select a regimen not including PI or to use a second-generation PI such as darunavir.

Another important question is the use of etravarine in patients with a history of NNRTI exposure. In certain instances, genotypic resistance testing in patients off therapy may help in the assessment of clinical use of etravarine. One of our patients was found to have a 100I in addition to 103N mutation. Although 103N mutation does not appear to affect the response to etravirine, the presence of the 181C/I/V, 101P, and 100I mutations leads to reduced activity of the drug.15 In addition, the presence of any NNRTI-associated mutation may suggest existence of other archived mutations from the same drug class.

Transmitted drug resistance mutations are usually detected for a longer period of time in comparison to acquired resistance mutations.11,16,17 It has been suggested that patients with transmitted drug-resistant virus do not have a reservoir of drug-susceptible virus and that transmitted virus can only change to wild-type by back mutation.16 In the case of acquired drug resistance, archived wild-type virus usually overgrows resistant quasispecies within weeks of withdrawal of the selective pressure of ART.6-9,11,13,18 However, the time at which resistance mutations become undetectable on current assays varies depends on the class of antiretroviral medication, the individual drug, and specific mutation(s). For example, certain NRTI mutations, especially the M184V mutation, disappear quickly in the absence of selective drug pressure, whereas other NRTI mutations persist longer.13,19 In a small treatment interruption study, 68% of patients had detectable NRTI mutations at 8 weeks or longer after cessation of ART compared with 21% and 26% of patients with detectable mutations to NNRTIs and PIs, respectively.13 NNRTI mutations, especially the K103N mutation, often persists for long periods of time, which is thought to be partly the result of the fact that this mutation does not result in a decrease in viral replication capacity. Recent studies have demonstrated prolonged persistence of NNRTI mutations using highly sensitive assays.20,21 Four of six patients showed persistence of the 103N variants from 6 months to 5 years after cessation of NNRTI therapy when the allele-specific real-time reverse transcriptase-polymerase chain reaction assay was used for the detection of the mutation.21 The clinical use of highly sensitive assays in treatment-experienced HIV-infected patients needs further investigation.

There are certain mutations that are viewed as reversion mutations. In particular, 215C, 215D, and 215S may represent reversions of 215Y/F thymidine analog-associated mutation. Their detection could imply the presence of additional resistant mutations. One of our patients had a 215C/N/S/Y mutations, which may reflect a reversion to a wild-type virus.22

Considering the relatively low cost of genotypic resistance assays when compared with the cost of 1 or more months of potentially ineffective ART, and the relatively high prevalence of resistance after treatment interruption, it may be cost-effective to obtain a resistance assay in patients off ART for longer than 4 weeks. Although the absence of resistance mutations in this setting does not rule out archived mutations, the presence of mutations, especially NNRTI-associated mutations, can have a significant impact on the choice of subsequent ART.

There are certain limitations to our findings. The study was retrospective and the decision to obtain the resistance tests might have been influenced by knowledge of previous treatment failures, history of noncompliance, or other factors that could lead to selection bias. In addition, resistance assays were done at random frequencies and were not repeated until detected mutations had disappeared. Thus, it is difficult to quantitatively assess the duration of persistence of resistance mutations. Although we cannot rule out transmitted resistance, it is unlikely to explain the persistence of mutations because the rate of transmitted resistance in our population is low (5%) (unpublished observations). Finally, we only evaluated genotypic resistance assays and thus cannot make any conclusions regarding the use of phenotypic assays.

In summary, we found resistance mutations in 27% of patients after prolonged treatment interruption. Although it remains preferable to obtain resistance testing while on a failing antiretroviral regimen, to maximize the sensitivity of detection of relevant mutations, in practice, this is not always possible. Our data suggest a role for resistance testing in patients off ART and a possible need to revisit current guidelines.

Back to Top | Article Outline


1. Moore RD, Keruly JC, Gebo KA, et al. An improvement in virologic response to highly active antiretroviral therapy in clinical practice from 1996 through 2002. J Acquir Immune Defic Syndr. 2005;39:195-198.
2. Bartlett JA, Fath MJ, Demasi R, et al. An updated systematic overview of triple combination therapy in antiretroviral-naive HIV-infected adults. AIDS. 2006;20:2051-2064.
3. Touloumi G, Pantazis N, Antoniou A, et al; CASCADE Collaboration. Highly active antiretroviral therapy interruption: predictors and virological and immunologic consequences. J Acquir Immune Defic Syndr. 2006;42:554-561.
4. Holkmann Olsen C, Mocroft A, Vella S, et al. Interruption of combination antiretroviral therapy and risk of clinical disease progression to AIDS or death. HIV Med. 2007;8:96-104.
5. Kavasery R, Galai N, Astemborski J, Lucas GM, et al. Nonstructured treatment interruptions among injection drug users in Baltimore, MD. J Acquir Immune Defic Syndr. 2009;50:360-366.
6. Verhofstede C, Wanzeeleb FV, Van Der Guchtb B, et al. Interruption of reverse transcriptase inhibitors or a 7. switch from reverse transcriptase to protease inhibitors resulted in a fast reappearance of virus strains with a reverse transcriptase inhibitor-sensitive genotype. AIDS. 1999;13:2541-2546.
7. Devereux HL, Youlea M, Margaret A, et al. Rapid decline in detectability of HIV-1 drug resistance mutations after stopping therapy. AIDS. 1999;13:F123-F127.
8. Deeks SG, Wrin T, Liegler T, et al. Virologic and immunologic consequences of discontinuing combination antiretroviral-drug therapy in HIV-infected patients with detectable viremia. N Engl J Med. 2001;344:472-480.
9. Lawrence J, Mayers DL, Hullsiek KH, et al. Structured treatment interruption in patients with multidrug-resistant human immunodeficiency virus. N Engl J Med. 2003;349:837-846.
10. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents. Department of Health and Human Services. November 3, 2008:1-139. Available at: Accessed January 27, 2009. Table 4. Recommendations for Using Drug Resistance Assays, p 12.
11. Hirsch MS, Gunthard HF, Schapiro JM, et al. Antiretroviral drug resistance testing in adult HIV-1 infection: 2008 recommendations of an International AIDS Society-USA Panel. Clin Infect Dis. 2008;47:266-285.
12. Albert J, Wahlberg J, Lundeberg J, et al. Persistence of azidothymidine-resistant human immunodeficiency virus type 1 RNA genotypes in posttreatment sera. J Virol. 1992;66:5627-5630.
13. Wirden M, Delaugerre C, Marcelin AG, et al. Comparison of the dynamics of resistance-associated mutations to nucleoside reverse transcriptase inhibitors, nonnucleoside reverse transcriptase inhibitors, and protease inhibitors after cessation of antiretroviral combination therapy. Antimicrob Agents Chemother. 2004;48:644-647.
14. Chilton D, Dervisevic S, Pillay D, et al. Determinants of HIV drug resistance mutations in plasma virus after treatment interruption. AIDS. 2005;19:2174-2175.
15. Etravirine [package insert]. Bridgewater, NJ: Tibotec Therapeutics; 2008.
16. Little SJ, Frost SD, Wong JK, et al. The persistence of transmitted drug resistance among subjects with primary HIV infection. J Virol. 2008;82:5510-5518.
17. Barbour JD, Hecht FM, Wrin T, et al. Persistence of primary drug resistance among recently HIV-1 infected adults. AIDS. 2004;18:1683-1689.
18. Birk M, Svedhem V, Sönnerborg A. Kinetics of HIV-1 RNA and resistance-associated mutations after cessation of antiretroviral combination therapy. AIDS. 2001;15:1359-1368.
19. Deeks SG, Hoh R, Neilands TB, et al. Interruption of treatment with individual therapeutic drug classes in adults with multidrug-resistant HIV-1 infection. J Infect Dis. 2005;192:1537-1544.
20. Flys T, Nissley DV, Claasen CW, et al. Sensitive drug-resistance assays reveal long-term persistence of HIV-1 variants with the K103N nevirapine (NVP) resistance mutation in some women and infants after the administration of single-dose NVP: HIVNET 012. J Infect Dis. 2005;192:24-29.
21. Palmer S, Boltz V, Maldarelli F, et al. Selection and persistence of non-nucleoside reverse transcriptase inhibitor-resistant HIV-1 in patients starting and stopping non-nucleoside therapy. AIDS. 2006;20:701-710.
22. Garcia-Lerma JG. Diversity of thymidine analogue resistance genotypes among newly diagnosed HIV-1 infected persons. J Antimicrob Chemother. 2005;56:265-269.
23. Johnson VA, Brun-Vezinet F, Clotet B, et al. Update of the drug resistance mutations in HIV-1. Top HIV Med. 2008;16:138-145.

HIV-1 genotypic resistance testing; resistance testing in treatment-experienced patients; persistence of resistance mutations

© 2010 Lippincott Williams & Wilkins, Inc.