• genotypic resistance;
  • protease inhibitor mutations;
  • unscheduled treatment interruptions;
  • viral supression


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References


Although current guidelines recommend resistance testing prior to antiretroviral therapy (ART) reinitiation after treatment interruptions, virological failure of first-line ritonavir-boosted, protease-inhibitor (PI/r)-containing ART is associated with low emergent PI resistance. In patients experiencing unscheduled treatment interruptions (UTrIs) on ritonavir-boosted atazanavir (ATV/r) ART regimens, we hypothesized low emergence of PI mutations conferring resistance to ATV/r.


In a retrospective assessment of HIV-infected patients initiating ATV/r-containing ART, using logistic regression we determined factors associated with UTrI, the prevalence of emergent resistance mutations and virological response after ART reinitiation.


A total of 202 patients [median age 33 years (interquartile range (IQR) 29–40 years); 52% female; median CD4 count 184 cells/μL (IQR 107–280 cells/μL); median HIV RNA 4.6 log10 HIV-1 RNA copies/mL (IQR 3.2–5.1 copies/mL)] initiated ATV/r between 2004 and 2009; 80 (43%) were ART naïve. One hundred and ten patients (55%) underwent 195 UTrIs after a median (IQR) 25 (10–52) weeks on ART, with a median (IQR) UTrI duration of 10 (3–31) weeks. Fifty-four of 110 patients (49%) underwent more than one UTrI. The commonest reasons for UTrI were nonadherence (52.7%) and drug intolerance (20%). Baseline HIV RNA > 100 000 copies\mL [odds ratio (OR) 3.6; 95% confidence interval (CI) 1.3–9.95] and being HCV positive, an injecting drug user or on methadone (OR 2.4; 95% CI 1.3–4.4) were independently associated with UTrI. In 39 patients with at least two resistance assays during UTrIs, 72 new mutations emerged; four nucleoside reverse transcriptase inhibitor (NRTI), two nonnucleoside reverse transcriptase inhibitor (NNRTI) and 66 protease inhibitor (PI) resistance mutations. All emergent PI resistance mutations were minor mutations. At least 65% of patients were re-suppressed on ATV/r reinitiation.


In this PI-treated cohort, UTrIs are common. All emergent PI resistance mutations were minor and ATV/r retained activity and efficacy when reintroduced, even after several UTrIs, raising questions regarding the need for routine genotypic resistance assays in PI/r-treated patients prior to ART reinitiation after UTrI.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Effective combination antiretroviral therapy (ART) has led to a dramatic decrease in HIV-related morbidity and mortality over the last decade [1]. Treatment of HIV-infected patients with ART can result in long-term suppression of HIV viral replication to undetectable levels and substantial CD4 T-cell count increases. However, despite these improvements, patients often discontinue ART for a variety of reasons and for durations ranging from a few days to several months. Common reasons for interrupting therapy include adverse side-effects, poor adherence, drug toxicity and drug resistance [2].

The introduction of the protease inhibitor (PI) atazanavir (ATV) in 2003 allowed HIV-infected patients greater access to once-daily, simplified antiretroviral regimens [3]. ATV combined with ritonavir (ATV/r) has been shown to have comparable efficacy and safety to the twice-daily PI lopinavir/ritonavir in treatment-naïve and treatment-experienced patients as well as having a more favourable lipid profile and fewer gastrointestinal side effects [4-6] Similarly, ART containing ATV/r has also been shown to be noninferior to ART containing the nonnucleoside reverse transcriptase inhibitor (NNRTI) efavirenz [7-9].

Most commonly used PIs have a significantly higher genetic barrier to resistance when compared with both first- and second-generation NNRTIs (including efavirenz, nevirapine and rilpivirine) or integrase inhibitors (such as raltegravir or elvitregravir), for which development of single genotypic mutations can result in high-level phenotypic resistance. Resistance to ATV can develop by two means; rarely as a single major mutation in the protease gene in the form of the I50L mutation [10], which tends to occur in treatment-naïve patients, or more commonly through the accumulation of multiple minor mutations in the protease gene, which is more often observed in PI-experienced patients [11]. In patients who experience virological failure while on their first PI-based regimen, few or no PI resistance mutations are detected at failure [12, 13].

In recent times, performing a resistance assay in patients who undergo unscheduled treatment interruptions (UTrIs) to ascertain if drug resistance has occurred as a result of nonadherence to ART has become part of clinical practice [14]. This practice can be costly, delays reinitiation of ART and, with certain drug classes known to have high genetic barriers to resistance, may be of limited value. We hypothesized that, because of the high genetic barrier to resistance of PIs, the emergence of PI resistance mutations conferring resistance to ATV/r after UTrIs in patients on ATV/r-based ART regimen is low. This study aimed to characterize and determine the frequency, associations and development of resistance in patients undergoing UTrIs after initiating ATV/r-containing ART.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study design and study subjects

We completed a retrospective cohort study of HIV-infected patients attending the Mater Misericordiae University Hospital (MMUH) infectious disease clinic. Subjects were included in the study if they had a documented positive HIV antibody test, initiated ART with ATV/r-containing regimens and were aged ≥18 years. The study was approved by the Mater Misericordiae University Hospital and Mater Private Hospital Research Ethics Committee.

Clinical and medication history

Subjects previously prescribed ATV/r had their case notes reviewed for the following information: demographic characteristics (age, gender and ethnicity) and disease characteristics (CD4 T-cell count, HIV RNA, clinical disease stage, mode of acquisition and viral hepatitis coinfection). Documentation of self-reported or physician-reported UTrI, defined as discontinuation of all ART drugs for any period of time while on ATV/r-containing ART, was recorded. Dates of interruption and reintroduction of ATV/r, documentation of genotypic resistance assays (where available) pre- and post-ATV/r interruption, details of other components of the ART regimen, durations of the treatment interruptions, the number of UTrIs and reasons for interruptions were also collected.

HIV RNA levels were measured using the Amplicor HIV-1 Monitor (Roche Molecular Systems, Branchburg, NJ) and genotypic resistance testing was conducted using the Trugene HIV-1 Genotyping Assay (Bayer, Berkeley, CA). Mutations were classified based on the International AIDS Society−USA drug resistance mutations list [15].

Statistical analysis

Patients' characteristics were described using the median and interquartile range (IQR) for continuous, nonnormal variables and percentages for categorical variables. Covariates with P-values of < 0.2 in univariate analysis were included in multivariate logistic regression to determine factors independently associated with UTrI. In the multivariate analysis, to deal with multicollinearity caused by the high correlation between methadone use, hepatitis C and injecting drug use (IDU), we fitted four separate models; model 1 combined all three covariates while models 2–4 were fitted for these individual covariates. A variance inflation factor (VIF) > 10 was considered an indicator of multicollinearity. The Bayesian information criterion (BIC) was used to determine the model that provided the most information. The backward elimination method was used to derive baseline covariates independently associated with UTrI. The Hosmer−Lemeshow test was used to assess the goodness of fit of the model. The results of the logistic regression analyses are expressed as odds ratios (ORs) and their 95% confidence intervals (CIs).

The prevalences of new emergent resistance mutations were calculated for patients with at least two resistance assays during UTrIs. The resistance profile for each patient was compared against the Stanford University HIV Drug Resistance Database (Version 6.2.0; Stanford University, Stanford, CA). Linear trends in the data were assessed using the Cochran−Armitage trend test. A P-value of < 0.05 was considered to be statistically significant. Analyses were performed using stata 10 (StataCorp, College Station, TX).


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

A total of 202 subjects commencing ATV/r-containing ART between January 2004 and August 2009 had their medical records reviewed for history of UTrIs. Cohort demographics are outlined in Table 1. Most study subjects were female (52%), their median (IQR) age was 33.6 (29–40) years, and most subjects were of European origin. Although the median time since ART initiation was 4.6 (2.8–7.6) years, 80 subjects (43%) were initiating ART for the first time, with 115 (60%) having been on other ART regimens (median time 3.9 years) prior to commencing an ATV/r-containing ART regimen. The median (IQR) CD4 count at ATV/r initiation was 184 (107–280) cells/μL, with a median log10 HIV RNA of 4.6 (IQR 3.2–5.1) HIV-RNA copies/mL. Injecting drug use was the commonest transmission risk group (48%), with all subjects within this risk group on methadone maintenance therapy, and 95 of 97 (97.9%) were hepatitis C virus (HCV) antibody positive.

Table 1. Demographic and clinical characteristics of patients who initiated ritonavir-boosted atazanavir (ATV/r)-containing antiretroviral therapy (ART) regimens between January 2004 and August 2009
 All patients (n = 202)ART treatment interruption
No (n = 92)Yes (n = 110)
  1. Data are n (%), unless otherwise stated.

  2. TDF, tenofovir; FTC, emtricitabine; ABC, abacavir; 3TC, lamuvidine; ZDV, zidovudine; SAg, surface antigen; IDU, injecting drug use; IQR, interquartile range; CDC, Centers for Disease Control and Prevention.

  3. *Age at the commencement of ATV/r. Nine patients missing baseline CD4 counts. Six patients missing HIV RNA. §One patient without HIV transmission risk.

Female104 (51.5)52 (56.5)52 (47.3)
Male98 (48.5)40 (43.5)58 (52.7)
Age (years) [median (IQR)]*33.6 (29.4–39.9)35.7 (29.7–40.7)32.3 (29.2–39.4)
Total years on ART [median (IQR)]4.6 (2.8–7.6)4.7 (2.8–7.5)4.5 (2.8–7.7)
Time since HIV diagnosis (years) [median (IQR)]6 (4–10)6 (4–11)6 (4–10)
European138 (68.3)53 (57.6)85 (77.3)
African/Asian64 (31.7)39 (42.4)25 (22.7)
CDC HIV stage   
A92 (45.5)48 (52.2)44 (40)
B41 (20.3)20 (21.7)21 (19.1)
C69 (34.2)24 (26.1)45 (40.9)
PI ART regimen   
TDF/FTC115 (56.9)46 (50)69 (62.7)
ABC/3TC57 (28.2)31 (33.7)26 (23.6)
ZDV/3TC18 (8.9)13 (14.1)5 (4.6)
Other12 (5.9)2 (2.2)10 (9.1)
Baseline CD4 count   
< 200 cells/μL107 (55.4)46 (50.5)61 (59.8)
200–350 cells/μL53 (27.5)27 (29.7)26 (25.5)
> 350 cells/μL33 (17.1)18 (19.8)15 (14.7)
Baseline HIV RNA   
< 50 copies/mL32 (16.3)23 (25.2)9 (8.6)
50–10000 copies/mL33 (16.9)19 (20.9)14 (13.3)
10000–100000 copies/mL71 (36.2)28 (30.8)43 (41.0)
> 100000 copies/mL60 (30.6)21 (23.1)39 (37.1)
HIV transmission risk§   
IDU97 (48.3)33 (36.3)64 (58.2)
Heterosexual77 (38.3)46 (50.5)31 (28.2)
Homosexual contact16 (8)8 (8.8)8 (7.3)
Needle stick/transfusion11 (5.4)4 (4.4)7 (6.3)
Methadone use   
Yes101 (50)33 (35.9)68 (61.8)
No101 (50)59 (64.1)42 (38.2)
Hepatitis C virus   
Positive97 (48)32 (34.8)65 (59.1)
Negative105 (52)60 (65.2)45 (40.9)
Hepatitis B virus SAg   
Positive9 (4.5)5 (5.4)4 (3.6)
Negative193 (95.5)87 (94.6)106 (96.4)
ART exposure   
Naïve88 (43.6)33 (35.9)55 (50)
Experienced114 (56.4)59 (64.1)55 (50)

Other antiretrovirals initiated alongside a first-line ATV/r-containing ART regimen included tenofovir (TDF)/emtricitabine (FTC) in 115 (56.9%) patients, abacavir (ABC)/lamivudine (3TC) in 57 (28.2%), zidovudine (ZDV)/3TC in 18 (8.9%), and didanosine (ddI)/TDF in seven (3.5%). Other regimens included ATV/r with ddI/3TC (two patients), ddI/nevirapine (NVP) (one patient), ritonavir boosted darunavir (DRV/r) with ZDV/3TC (one patient) and nelfinavir/ritonavir (NFV/r) with ddI/ABC (one patient).

Of the 202 patients, 92 (46%) remained on continuous ART for the duration of the follow-up and 110 (54%) underwent a total of 195 UTrIs after a median of 25 (10–52) weeks spent on ATV/r-containing ART regimens. Fifty-four of 110 patients underwent more than one UTrI, with one patient undergoing a maximum of six interruptions. Within the treatment interruption group, the median (IQR) duration of interruptions and the time between the start of the UTrI and the performance of a genotypic resistance test were 10 (3–31) weeks and 12.1 (6.6–28.3) weeks, respectively. The commonest self-reported reasons for treatment interruptions were poor adherence (52.7%), nausea/vomiting (10%) and other side effects (10%).

Factors associated with treatment interruptions

Associations between potential baseline risk factors and having experienced a UTrI were examined using univariable logistic regression (Table 2). The risk of experiencing a UTrI was greater for patients with higher baseline HIV RNA (> 10 000 copies/mL: OR 3.92; 95% CI 1.59–9.71; > 100 000 copies/mL: OR 4.75; 95% CI 1.86–12.1), those of European origin (OR 2.5; 95% CI 1.36–4.60), those using methadone (OR 2.89; 95% CI 1.63–5.14), those who were HCV positive (OR 2.71; 95% CI 1.53–4.80) and those who were injecting drug users (OR 2.88; 95% CI 1.55–5.35). Higher nadir CD4 cell count was marginally protective (OR 0.99; 95% CI 0.99–1.00) for experiencing a UTrI.

Table 2. Univariate logistic regression for factors associated with unscheduled treatment interruption
 Unadjusted odds ratio95% confidence intervalP-value
  1. SAg, surface antigen; IDU, injecting drug use; CDC, Centers for Disease Control and Prevention.

  2. †Age at the commencement of ritonavir-boosted atazanavir (ATV/r).

Age (years)0.970.92–1.000.054
Total years on ART1.000.93–1.080.96
Time since HIV diagnosis0.980.93–1.030.37
CDC HIV stage   
Nadir CD4 count (cells/μL)0.990.99–1.000.05
Baseline CD4 count   
< 200 cells/μL1.590.73–3.490.25
200–350 cells/μL1.160.48–2.760.75
> 350 cells/μL1  
Baseline HIV RNA   
< 50 copies/mL1  
50–10000 copies/mL1.880.67–5.300.23
10000–100000 copies/mL3.921.59–9.710.003
> 100000 copies/mL4.751.86–12.10.001
Transmission risk   
Homosexual contact1.480.50–4.370.47
Needle stick/transfusion2.60 0.15
Methadone use   
Yes2.891.63–5.14< 0.001
Hepatitis C virus   
Hepatitis B virus SAg   

In the multivariate analysis, four models were used to deal with multicollinearity caused by the high correlation between methadone use, hepatitis C and IDU (Table 3). Model 1 combined all three covariates, while models 2–4 were fitted for these individual covariates. In all models, higher baseline HIV RNA was independently associated with experiencing a UTrI along with either use of methadone or being HCV antibody positive, or a combination of the three covariates. Of all the fitted models, the model that combined all three covariates (methadone use, IDU as a transmission risk and being HCV antibody positive; model 1) was determined to be the model that provided the most information according to the BIC.

Table 3. Multivariable logistic regression analysis for factors independently associated with unplanned treatment interruptions
VariableModel 1*Model 2Model 3Model 4§
  1. Only covariates significantly or weakly associated with unscheduled treatment interruption were included in multivariate models and the Bayesian information criterion was used to select the most suitable model.

  2. IDU, injecting drug use; CDC, Centers for Disease Control and Prevention; CI, confidence interval; AOR, adjusted odds ratio.

  3. *Model 1 contains the covariates baseline HIV RNA, CDC HIV stage and the combined covariate methadone use, hepatitis C and IDU. Model 2 contains the covariates baseline HIV RNA, CDC HIV stage and IDU. Model 3 contains the covariates baseline HIV RNA, CDC HIV stage and hepatitis C. §Model 4 contains the covariates baseline HIV RNA, CDC HIV stage and methadone use.

Methadone use, hepatitis C and IDU    
Yes2.35 (1.25–4.43)   
HIV transmission risk    
Heterosexual 1  
IDU 2.35 (1.20–4.62)  
Homosexual contact 0.90 (0.28–2.88)  
Needle stick/transfusion 3.97 (0.85–18.5)  
Methadone use    
No   1
Yes   2.53 (1.35–4.71)
Hepatitis C virus    
Negative  1 
Positive  2.33 (1.25–4.33) 
Baseline HIV RNA    
< 50 copies/mL1111
50–10000 copies/mL2.23 (0.74–6.71)2.19 (0.73–6.52)2.04 (0.70–5.93)2.12 (0.72–6.21)
10000–1000000 copies/mL3.56 (1.36–9.36)3.58 (1.38–9.27)3.25 (1.28–8.22)3.24 (1.28–8.22)
> 100000 copies/mL3.64 (1.33–9.95)3.55 (1.31–9.60)3.50 (1.32–9.27)3.52 (1.32–9.34)
CDC HIV stage    
B1.12 (0.48–2.58)0.95 (0.43–2.13)0.97 (0.43–2.15)0.94 (0.42–2.11)
C1.95 (0.94–4.04)1.66 (0.81–3.38)1.67 (0.83–3.37)1.64 (0.81–3.32)

Controlled for other baseline factors in model 1, a strong dose−response association between higher HIV RNA and UTrI persisted [adjusted odds ratio (AOR) 3.56; 95% CI 1.36–9.36 for patients with HIV RNA > 10 000 copies/mL; and AOR 3.64; 95% CI 1.33–9.95 for patients with HIV RNA > 100 000 copies/mL]. Injecting drug users on methadone who were HCV antibody positive had greater odds of experiencing one or more UTrIs compared with non-drug users who were HCV antibody negative (AOR 2.35; 95% CI 1.25–4.43). The Hosmer−Lemeshow goodness-of-fit test suggests that the model was a good fit (χ2 = 17.9; P = 0.40).

Genotyping and mutations

A total of 193 genotypic resistance assays were available, 52 (26.9%) of which were baseline assays for patients who did not experience treatment interruptions and 141 (73.1%) of which were genotypic assays for the 88 of 110 patients (80%) who experienced UTrIs. Thirty-nine of these 88 patients (44.3%) had at least two genotypic resistance assays, 36 (40.9%) had baseline assays only and 13 (14.7%) had genotypic assays at one interruption episode only. Among the 141 genotypic resistance assays performed in patients with treatment interruptions, 92 of 141 (65.2%) were in patients with at least two genotypic assays (Fig. 1).


Figure 1. Flow chart for genotypic resistance assays. UTrI, unscheduled treatment interruption.

Download figure to PowerPoint

Analysis of the emergence of new mutations was restricted to patients who had at least two genotypic assays available. In these patients, prior to initiating ATV/r-containing ART regimens, there were 101 baseline mutations, of which three (3%) were nucleoside reverse transcriptase inhibitor (NRTI), six (5.9%) were NNRTI and 92 (89.7%) were PI resistance mutations. During UTrIs, 72 new mutations emerged, four (5.6%) NRTI, two (2.8%) NNRTI and 66 (91.7%) PI resistance mutations. Table 4 shows the various baseline and emergent mutations during treatment interruptions within each class of antiretroviral drug.

Table 4. Baseline and new emerging mutations by drug class
 Baseline mutationsNew emerging mutations
  1. Numbers in brackets represent the frequency of occurrence of the mutation. Mutations are listed in descending order of occurrence.

  2. NRTI, nucleoside reverse transcriptase inhibitor; NNRTI, nonnucleoside reverse transcriptase inhibitor; PI, protease inhibitor.

Primary mutationsM41L (1), MI84V (1), T215Y (1)69Insert (1), M184V (2), K70R (1)
Secondary mutationsNoneNone
Primary mutationsK103N (4)G190A (1), Y181C (1)
Secondary mutationsV90I (1), V179D (1)None
Primary mutationsT74P (1), V82A (1)None
Secondary mutationsL63P (25), M36I (15), H69K (9), A71T (7), I13V (6), L10I (5), L10V (3), I15V (3), K20I (3), L89M (3), G16E (2), I62V (2), A71V (2), K20R (1), K20V (1), G73S (1), I93L (1), I93M (1)I62V (13), H69K (8), I15V (7), V771 (6), I93L (5), K20I (4), I13V (4), L63P (3), A71T (3), L89M (3), L10I (2), L10V (1), G16E (1), M36I (1), A71V (1), A71I (1), I85V (1), G73T (1), D60E (1)

Of the three baseline NRTI resistance mutations, the M41L (1) and T215Y (1) mutations were both detected in patients initiating ART, while M184V (1) was detected in an ART-experienced patient. Four new NRTI resistance mutations emerged during UTrIs, two M184V and a K70R mutation in patients initiating ART and one case of 69Insert in an ART-experienced patient.

Among the commonest baseline NNRTI resistance mutations, four of six (all K103N) were detected in patients initiating ART (presumed to represent transmitted resistance), while two of six (V90I and V179D) were detected in ART-experienced subjects. During UTrIs, two new NNRTI resistance mutations emerged, one case of G190A in an ART-experienced patient previously on an NNRTI-containing ART regimen (considered to be archived resistance) and one case of Y181C developing in a previously ART-naïve patient prescribed ATV/r with TDF/FTC.

Of the 92 baseline PI resistance mutations, 90 (97.8%) were minor mutations and two (2.2%) were major mutations (T74P and V82A, conferring resistance to tipranavir and lopinavir, respectively [15]). Of the secondary PI resistance mutations, L63P (25 of 92), M36I (15 of 92) and H69K (nine of 92) were most common prior to ATV initiation. During treatment interruptions, 66 new PI resistance mutations emerged, all minor mutations, the commonest being I62V (13 of 66), H69K (eight of 66), I15V (seven of 66) and V77I (six of 66), with only three new L63P mutations and one new M36I mutation emerging during treatment interruptions. In five individuals, new I62V mutations emerged in combination with I15V, and in different group of four individuals; I62V emerged in combination with H69K. Another common combination of emerging mutations was I13V+H69K+L98M, which was detected in three individuals. No major PI resistance mutations emerged during UTrIs.

Of the patients with at least two genotypic assays performed, one of 39 (2.6%) developed enough minor mutations to confer intermediate resistance to ATV/r, against the Stanford HIV Drug Resistance Database (Version 6.2.0).

Viral suppression after unscheduled treatment interruption

Of the 110 patients who experienced a first UTrI episode, 74 (67.3%) reinitiated ART containing ATV/r (36 were either lost to follow-up or reinitiated ART not containing ATV/r). Figure 2 presents the variation in viral response on reintroducing AVT/r ART regimens after a UTrI. Of those reinitiating ATV/r-containing ART, 52 of 74 (70.3%) attained HIV RNA < 50 copies/mL during the subsequent treatment period. Overall, at least 65% of the patients achieved HIV RNA < 50 copies/mL during different subsequent treatment episodes (with 80% achieving HIV RNA < 50 copies/mL after their fourth UTrI) on reintroducing ATV/r-containing ART regimens after a UTrI. The Cochrane−Armitage test for trend (χ2 = 1.34; P = 0.25) demonstrated no downward trend in rate of virological suppression with increased number of UTrI episodes.


Figure 2. Viral response on ATV/r reintroduction after UTrI. UTrI, unscheduled treatment interruption.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This is the first study examining the emergence of resistance with multiple UTrIs to boosted PI-containing ART. Although other studies have extensively described the characteristics of patients undergoing treatment interruptions [16-18], none has focused on the emergence of genotypic resistance after multiple interruptions of ritonavir-boosted PI-containing ART.

Our data demonstrate that unscheduled interruption of ART is a relatively common phenomenon in this PI-treated cohort of patients, with nearly 54% of individuals having at least one UTrI in a median of 25 weeks on ART. A strength of this cohort is that all major risk groups for HIV infection are broadly represented, with a high proportion of women and injecting drug users within the cohort. The relatively high representation of women in this study reflects the general demographic profile of the HIV-infected cohort within our clinic, as well as the historical preferential use of boosted PIs over NNRTIs in women of child-bearing potential.

Although there is a high proportion of injecting drug users within the cohort, 42% of the UTrIs occurred in the non-IDU risk group. Poor adherence and drug intolerance were the most frequently self-reported reasons for UTrI. Findings from this study are consistent with other studies that have demonstrated as association between UTrI and risk factors; high baseline HIV RNA [19, 20], being HCV antibody positive [18], being an injecting drug user or being on methadone [19]. The finding of an association between high baseline HIV RNA and UTrI may seem surprising, but we would speculate that higher baseline HIV RNA may reflect more advanced and perhaps symptomatic HIV disease, which has been associated with poor adherence in a number of studies [21-23].

We did not observe emergence of major PI resistance mutations to ATV/r despite some patients undergoing up to six UTrIs, which is consistent with the high prevalence of suppression upon reinitiation of an ATV/r-containing antiretroviral regimen. Specifically, we did not identify the emergence of the major ATV/r mutation I50L and a few significant minor mutations were observed.

This cohort is a challenging group to study, with the majority of patients having acquired HIV infection through IDU. This predominance of IDU mainly reflects the preferential use of boosted PIs in these patients to avoid clinically meaningful drug-to-drug interactions with methadone replacement therapy. However, this study also suggests that these patients may also benefit from the selection of an ATV/r-based ART regimen on the basis that it retains efficacy when reintroduced in those undergoing a UTrI.

The high prevalence of UTrIs points to advantages of the use of PI-containing ART, as drug resistance to most PIs requires multiple mutations and seldom develops after early virological failure, particularly with the use of ritonavir boosting. With NNRTIs, partial resistance to efavirenz/nevirapine/rilpivirine is conferred by a single mutation and develops rapidly after virological failure, while integrase inhibitors such as raltegravir also have a lower genetic barrier to resistance than PI-based regimens [24].

In this study, of patients who had at least two resistance assays preformed, no major PI mutation was recorded. Several minor mutations did emerge that may confer advantage to the virus, although further work is required to elucidate the clinical relevance, if any, of the emergence of these resistance mutations.

This study has a number of limitations. First, as it was a retrospective study, it was prone to recall bias surrounding the occurrence, duration and frequency of UTrIs reported by the patient and limited by physician documentation of the same. Secondly, although most patients had resistance assays performed when necessary, not everyone had, raising the possibility of underestimate the emergence of mutations. Thirdly, only patients on ATV/r-based regimes were included in this study, and thereby it is not fully generalized to all PI-based regimens. However, clinical trial data would suggest that other boosted PIs (LPV and DRV) retain genotypic characteristics similar to those we have demonstrated for ATV/r. Finally, there are no phenotypic data to determine the relevance of specific groups of mutations. A prospective study would be required to determine this.

Nevertheless, this is an important study which demonstrates that UTrIs are common even within non-IDU populations. Furthermore, only minor PI mutations, which for the vast majority did not confer resistance to ATV/r, emerged after UTrI, and ATV/r retained its activity and efficacy when reintroduced, even after several UTrIs. These data raise questions as to the utility of routine genotypic testing prior to ART reinitiation after UTrI in patients on PI therapy and would suggest that this expensive assay may be reserved for those who reinitiate therapy and experience suboptimal response.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors would like to thank the participants attending the Mater Misericordiae University Hospital (MMUH) infectious diseases clinic, Dublin and the National Virus Reference Laboratory (University College Dublin) for the genotypic resistance assays.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Lima VD, Harrigan PR, Bangsberg DR et al. The combined effect of modern highly active antiretroviral therapy regimens and adherence on mortality over time. J Acquir Immune Defic Syndr 2009; 50: 529536.
  • 2
    Iarikov DE, Irizarry-Acosta M, Martorell C et al. Use of HIV resistance testing after prolonged treatment interruption. J Acquir Immune Defic Syndr 2010; 53: 333337.
  • 3
    Le Tiec C, Barrail A, Goujard C, Taburet AM. Clinical pharmacokinetics and summary of efficacy and tolerability of atazanavir. Clin Pharmacokinet 2005; 44: 10351050.
  • 4
    Johnson M, Grinsztejn B, Rodriguez C et al. 96-week comparison of once-daily atazanavir/ritonavir and twice-daily lopinavir/ritonavir in patients with multiple virologic failures. AIDS 2006; 20: 711718.
  • 5
    Molina JM, Andrade-Villanueva J, Echevarria J et al. Once-daily atazanavir/ritonavir compared with twice-daily lopinavir/ritonavir, each in combination with tenofovir and emtricitabine, for management of antiretroviral-naive HIV-1-infected patients: 96-week efficacy and safety results of the CASTLE study. J Acquir Immune Defic Syndr 2010; 53: 323332.
  • 6
    Molina JM, Andrade-Villanueva J, Echevarria J et al. Once-daily atazanavir/ritonavir versus twice-daily lopinavir/ritonavir, each in combination with tenofovir and emtricitabine, for management of antiretroviral-naive HIV-1-infected patients: 48 week efficacy and safety results of the CASTLE study. Lancet 2008; 372: 646655.
  • 7
    Puls RL, Srasuebkul P, Petoumenos K et al. Efavirenz versus boosted atazanavir or zidovudine and abacavir in antiretroviral treatment-naive, HIV-infected subjects: week 48 data from the Altair study. Clin Infect Dis 2010; 51: 855864.
  • 8
    Squires K, Lazzarin A, Gatell JM et al. Comparison of once-daily atazanavir with efavirenz, each in combination with fixed-dose zidovudine and lamivudine, as initial therapy for patients infected with HIV. J Acquir Immune Defic Syndr 2004; 36: 10111019.
  • 9
    Daar ES, Tierney C, Fischl MA et al. Atazanavir plus Ritonavir or Efavirenz as part of a 3-drug regimen for initial treatment of HIV-1A randomized trial. Ann Intern Med 2011; 154: 445456.
  • 10
    Alcaro S, Artese A, Ceccherini-Silberstein F et al. Molecular dynamics and free energy studies on the wild-type and mutated HIV-1 protease complexed with four approved drugs: mechanism of binding and drug Resistance. J Chem Inf Model 2009; 49: 17511761.
  • 11
    Rivas P, Morello J, Garrido C, Rodríguez-Nóvoa S, Soriano V. Role of atazanavir in the treatment of HIV infection. Ther Clin Risk Manag 2009; 5: 99116.
  • 12
    Lathouwers E, De Meyer S, Dierynck I et al. Virological characterization of patients failing darunavir/ritonavir or lopinavir/ritonavir treatment in the ARTEMIS study: 96-week analysis. Antivir Ther 2011; 16: 99108.
  • 13
    Soriano V, Arasteh K, Migrone H et al. Nevirapine versus atazanavir/ritonavir, each combined with tenofovir disoproxil fumarate/emtricitabine, in antiretroviral-naive HIV-1 patients: the ARTEN Trial. Antivir Ther 2011; 16: 339348.
  • 14
    Vandamme AM, Camacho RJ, Ceccherini-Silberstein F et al. European HIV drug resistance guidelines panel. European recommendations for the clinical use of HIV drug resistance testing: 2011 update. AIDS Rev 2011; 13: 77108.
  • 15
    Johnson VA, Calvez V, Gunthard HF et al. Update of the drug resistance mutations in HIV-1: March 2013. Top Antivir Med 2013; 21: 614.
  • 16
    Fox Z, Philips A, Cohen C et al. Viral re-suppression and emergence of drug resistance following interruption of a suppressive NNRTI-containing regimen in the SMART study. AIDS 2008; 22: 22792289.
  • 17
    Holkmann Olsen C, Mocroft A, Kirk O et al. Interruption of combination antiretroviral therapy and risk of clinical disease progression to AIDS or death. HIV Med 2007; 8: 96104.
  • 18
    Moore DM, Zhang W, Yip B et al. Non-medically supervised treatment interruptions among participants in a universally accessible antiretroviral therapy programme. HIV Med 2009; 11: 299307.
  • 19
    Touloumi G, Pantazis N, Antoniou A et al. Highly active antiretroviral therapy interruption: predictors and virological and immunologic consequences. J Acquir Immune Defic Syndr 2006; 42: 554561 510.
  • 20
    Taffé P, Rickenbach M, Hirschel B et al. Impact of occasional short interruptions of HAART on the progression of HIV infection: results from a cohort study. AIDS 2002; 16: 747755.
  • 21
    Trotta MP, Ammassari A, Cozzi-Lepri A et al. Adherence to highly active antiretroviral therapy is better in patients receiving non-nucleoside reverse transcriptase inhibitor-containing regimens than in those receiving protease inhibitor-containing regimens. AIDS 2003; 17: 10991102.
  • 22
    Holzemer WL, Corless IB, Nokes KM et al. Predictors of self-reported adherence in persons living with HIV disease. AIDS Patient Care STDS 1999; 13: 185197.
  • 23
    Wagner GJ, Kanouse DE, Koegel P, Sullivan G. Correlates of HIV antiretroviral adherence in persons with serious mental illness. AIDS Care 2004; 16: 501506.
  • 24
    Rockstroh JK, Lennox JL, Dejesus E et al. Long-term treatment with raltegravir or efavirenz combined with Tenofovir/Emtricitabine for treatment-naïve human immunodeficiency virus-1-infected patients: 156-week results from STARTMRK. Clin Infect Dis 2011; 53: 807816.