Sage Journals: Discover world-class research

Abstract

Background

The real-world effectiveness of pre-exposure prophylaxis (PrEP) may be influenced by circulating HIV strains resistant to either tenofovir or emtricitabine. Yet, few studies have examined rates of resistance to these drugs in clinical settings.

Methods

We conducted a retrospective cohort study of antiretroviral-naive participants in the Canadian Observational Cohort collaboration who initiated antiretroviral therapy between 2006 and 2014. In separate analyses, we determined the prevalence of pretherapy resistance and cumulative incidence of follow-up resistance to tenofovir and emtricitabine. We used multivariable proportional hazards models to examine associations between baseline variables and the development of resistance.

Results

We studied 6,622 antiretroviral-naive participants initiating therapy, of whom 5,428 (82.0%) had a baseline resistance test. Baseline resistance to tenofovir and emtricitabine was observed in 83 (1.5%) and 21 (0.4%) patients, respectively. Among patients without baseline resistance, the cumulative incidence of resistance to tenofovir and emtricitabine 5 years following treatment initiation was 0.0070 (95% CI 0.0046, 0.0095) and 0.033 (95% CI 0.028, 0.038), respectively. Following multivariable analysis, a baseline viral load ≥100,000 copies/ml was associated with emergence of tenofovir (hazard ratio [HR] 2.88; 95% CI 1.35, 6.15) and emtricitabine (HR 2.27; 95% CI 1.64, 3.15) resistance. Initiating an integrase inhibitor-based regimen and CD4⁺ T-cell count below 200 cells/mm³ were also associated with resistance to each drug.

Conclusions

We observed a low prevalence of baseline resistance and a low incidence of emergence of resistance to tenofovir and emtricitabine among antiretroviral-naive patients in routine clinical care.

Introduction

Pre-exposure prophylaxis (PrEP) with tenofovir disoproxil fumarate and emtricitabine has been shown to reduce the risk of HIV acquisition in populations at high risk for the disease, including men who have sex with men, transgender women, serodiscordant heterosexual couples and people who inject drugs [1–8]. Results from a meta-analysis of published trials and observational studies indicate that PrEP imparts a 51% reduction in the risk of HIV relative to placebo, a finding that was consistent across populations and modes of HIV acquisition [9]. Based on these data, the World Health Organization has recommended that PrEP be offered to all populations at risk of HIV [10]. Adherence has been identified as the most important determinant of PrEP effectiveness, with no protection against HIV demonstrated in trials where fewer than 40% of participants had detectable plasma levels of medication [9,11].

Another potentially important determinant of the effectiveness of PrEP is the prevalence of circulating HIV strains resistant to tenofovir and/or emtricitabine in the community of implementation. The prevalence of virus harbouring the K65R mutation is especially important, given that studies in macaques have correlated this mutation with PrEP failure, while no such effect was observed following exposure to strains containing the 184V mutation [12,13]. In this context, the protective activity of PrEP could be undermined, facilitating the transmission of drug resistant virus. Two recently published cases of infection with multi-drug resistant HIV in individuals with therapeutic levels of emtricitabine and tenofovir while receiving PrEP support this assertion and reinforces the need for ongoing population surveillance of tenofovir- and emtricitabine-resistant HIV [14,15]. Yet, few studies have characterized the incidence of resistance to these drugs in regions where PrEP is being introduced [16]. Consequently, surveillance for circulating strains of HIV resistant to tenofovir and/or emtricitabine has been identified as a research and policy priority as PrEP use increases [17,18]. These data are necessary to inform modelling of the effectiveness of PrEP and support regional demonstration projects designed to optimize its implementation. Accordingly, we examined the incidence of tenofovir- and emtricitabine-resistant HIV in a large cohort of antiretroviral-naive individuals living with HIV in Canada.

Methods

Study design and population

We conducted a retrospective study using data from the Canadian Observational Cohort (CANOC), an interprovincial collaboration of eight cohorts of antiretroviral-naive people with HIV aged 18 years and over who had initiated combination antiretroviral therapy after 1 January 2000 [19]. For this study, we included patients who initiated antiretroviral therapy between 2006 and 2014. We selected 2006 as the start date because it corresponds with the earliest Department of Health and Human Services recommendations for performing genotypic resistance testing on antiretroviral-naive patients and the widespread uptake of tenofovir-based antiretroviral regimens in Canada [20]. We excluded two sites with incomplete data on resistance testing. Data extraction was performed locally at the participating sites and pooled at the coordinating centre in Vancouver, BC. All participating cohorts have received ethical approval from their institutional boards to contribute data to CANOC.

Outcomes

The primary outcomes of the study were the occurrence of baseline and emergence of resistance to tenofovir or emtricitabine among antiretroviral-naive patients. We determined the proportion of patients with baseline resistance to either tenofovir or emtricitabine using all available genotypic resistance tests performed prior to the initiation of treatment. The cumulative incidence of resistance to either tenofovir or emtricitabine during follow-up was computed separately using Kaplan-Meier curves and excluding patients with baseline resistance to the corresponding drug. Patients with baseline resistance to both tenofovir and emtricitabine were not included in time-to-resistance analyses. In all analyses, the resistance interpretation was performed using the virco® TYPE HIV-1 system developed by Virco (now Janssen, Beerse, Belgium). This system makes use of a large database of linked HIV sequences, resistance phenotypes and clinical outcomes to predict a ‘virtual’ resistance phenotype from an input sequence [21,22]. For most drugs (including emtricitabine and tenofovir) the inferred resistance phenotypes are classified as susceptible, low-level or high-level resistance on the basis of clinical cutoffs derived from the analysis of linked clinical outcome data [23]. Accordingly, tenofovir resistance could be primarily defined by the presence of three or more thymidine analogue mutations or the presence of K65R or 69 insertion site mutations, and emtricitabine resistance primarily defined by the presence of M184V/I.

HIV drug Resistance Testing

The drug resistance testing programme in Quebec is performed on plasma-derived samples on all newly diagnosed patients, first genotyped with physician-designated clinical indication of primary infection (0–6 months after documented seroconversion) or chronic treatment-naive stage infection. Genotyping is also performed following antiretroviral treatment-failure (>100 copies/ml). Sequences (1467 bp) were generated using propriety Virco primers (Virco, BNBA, Mechelen, Belgium), spanning the entire protease gene and reverse transcriptase (RT) codons 1 to 400 [24]. Sequencing of select patients was also performed across the entire integrase gene as previously described [25]. Sequences were compared with HXB2 using Seqscape v2.5 (Applied Biosystems, Foster City, CA, USA) to identify minor or major resistance mutations.

Drug resistance testing for Ontario and Quebec was conducted in British Columbia. RNA extracted from plasma samples with plasma viral load >250 copies/ml underwent nested RT-PCR of HIV protease and reverse transcriptase, followed by standard Sanger sequencing on an ABI instrument. Automated sequence analysis was performed by the custom RECall software package with which nucleotide ‘mixtures’ are called when the secondary peak area exceeded >17.5% that of the primary peak in at least half of the reads [26]. The methodology has been previously described in more detail [27].

Statistical analysis

All analyses were performed using SAS software (version 9.4; SAS Institute Inc., Cary, NC, USA) and R 3.3.1 (R Development Core Team, Vienna, Austria). We conducted separate analyses of baseline resistance and the development of resistance during follow-up. For the first analysis, we examined the association between baseline resistance and clinical and demographic variables using multivariable logistic regression analysis among participants with baseline resistance testing. Next, we conducted separate time to event analyses using multivariable Cox proportional hazards regression to examine the associations between baseline clinical and demographic variables and the emergence of resistance to tenofovir or emtricitabine among those patients who did not have baseline resistance to one or both of these drugs. We adjusted all models for age, sex, ethnicity, HIV acquisition risk factor, baseline viral load and CD4⁺ T-cell count, calendar year of treatment initiation and Canadian province of residence. We also adjusted time-to-resistance models for type of antiretroviral regimen initiated (that is, protease inhibitor, non-nucleoside reverse transcriptase inhibitor, integrase inhibitor). To allow for a non-linear relationship between baseline CD4⁺ T-cell count and time to development of resistance, restricted cubic splines were incorporated into the proportional hazard models with knots at 30, 210, 340 and 680 cells/mm³. Missing values of covariates were imputed using a fully conditional specification method. A discriminant function was used to impute missing values of the categorical ethnicity variable and logistic regression models were used to impute missing values of binary variables ‘injection drug use’ and ‘men who have sex with men.’ All variables contained in the substantive model were used in the imputation models, as well as the binary outcome when imputing for logistic models and the event indicator and Nelson-Aalen estimator of the cumulative hazard when imputing for proportional hazards models [28].

Results

Study population

We studied 6,622 antiretroviral-naive patients initiating antiretroviral therapy during the study period. The majority (n=5,599; 84%) of participants were male and the median age was 40 years (IQR: 32–47; Table 1). The median duration of follow-up after treatment initiation was 3.8 years (IQR 1.9–6.0).

Table 1

Baseline characteristics, overall and by receipt of baseline resistance test

	Total (n=6,622)	No baseline test (n=1,194)	Had baseline resistance test (n=5,428)	P-value
Age at first ARV, years	40 (32–47)	40 (32–47)	40 (32–47)	0.87
Year of first ARV	2010 (2008–2012)	2008 (2007–2011)	2010 (2008–2012)	<0.0001
Gender				<0.0001
Male	5,559 (83.9)	946 (79.2)	4,613 (85.0)
Female	1,063 (16.1)	248 (20.8)	815 (15.0)
Province				<0.0001
BC	3,396 (51.3)	377 (31.6)	3,019 (55.6)
ON	1,848 (27.9)	509 (42.6)	1,339 (24.7)
QC	1,378 (20.8)	308 (25.8)	1,070 (19.7)
Ethnicity				<0.0001
Caucasian	2,073 (31.3)	410 (34.3)	1,663 (30.6)
ACB	572 (8.64)	174 (14.6)	398 (7.33)
Indigenous	447 (6.75)	80 (6.70)	367 (6.76)
Other	734 (11.1)	122 (10.2)	612 (11.3)
Missing	2,796 (42.2)	408 (34.2)	2,388 (44.0)
Risk factor				<0.0001
MSM and IDU	239 (3.61)	34 (2.85)	205 (3.78)
MSM	2,843 (42.9)	507 (42.5)	2,336 (43.0)
IDU	1,066 (16.1)	214 (17.9)	852 (15.7)
Other	981 (14.8)	253 (21.2)	728 (13.4)
Missing	1,493 (22.5)	186 (15.6)	1,307 (24.1)
Baseline CD4⁺ T-cell count/mm³	270 (160–394)	239 (130–350)	280 (170–409)	<0.0001
Baseline CD4⁺ T-cell count <200 cells/mm³	2,113 (31.9)	475 (39.8)	1,638 (30.2)	<0.0001
Baseline viral load, log₁₀ copies/ml	4.79 (4.27–5.15)	4.86 (4.31–5.18)	4.77 (4.26–5.14)	0.03
Baseline VL <100,000 copies/ml	4,117 (62.2)	697 (58.4)	3,420 (63.0)	<0.01
Regimen type				<0.01
Boosted PI	2,871 (43.4)	502 (42.0)	2,369 (43.6)
II	582 (8.79)	114 (9.55)	468 (8.62)
NNRTI	2,873 (43.4)	505 (42.3)	2,368 (43.6)
Other	116 (1.75)	23 (1.93)	93 (1.71)
PI	180 (2.72)	50 (4.19)	130 (2.39)
Number of follow-up resistance tests				<0.01
0	5,054 (76.3)	958 (80.2)	4,096 (75.5)
1	918 (13.9)	140 (11.7)	778 (14.3)
2	307 (4.64)	46 (3.85)	261 (4.81)
3	343 (5.18)	50 (4.19)	293 (5.40)

Results are presented as n (%) or median (IQR). ACB, African, Caribbean and Black; ARV, antiretroviral; IDU, intravenous drug use; II, integrase inhibitor; MSM, men who have sex with men; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; VL, viral load.

Baseline resistance

Among 5,428 (82.0%) participants who had undergone baseline genotyping, resistance to tenofovir and emtricitabine was observed in 83 (1.5%) and 21 (0.4%) patients, respectively (see Additional files 1 and 2 for prevalence of individual mutations). Most tenofovir-associated resistance (n=80; 96%) was intermediate level (inferred 1.0- to 2.3-fold decrease in susceptibility). The K65R substitution was found in two (2.4%) isolates with tenofovir resistance, and seven (8.4%) cases were related to the accumulation of three or more thymidine analogue mutations. The majority (n=17; 81%) of emtricitabine resistance was related to the M184I/V mutation. Following multivariable analysis, individuals living in British Columbia (adjusted odds ratio [OR] 0.33; 95 CI 0.20, 0.54) and Quebec (OR 0.10; 95% CI 0.04, 0.29) were less likely to harbour tenofovir resistant virus at baseline relative to individuals living in Ontario (Table 2). Because of small numbers, we could not fit logistic regression models for baseline emtricitabine resistance.

Table 2

Logistic regression model for presence of baseline tenofovir resistance

	Unadjusted models		Adjusted model
	OR (95% CI)	P-value	OR (95% CI)	P-value
Age (per 10 years)	1.00 (0.81, 1.23)	0.99	1.09 (0.88, 1.35)	0.45
Year of cART initiation	1.04 (0.95, 1.14)	0.42	1.02 (0.92, 1.13)	0.76
Gender (ref = female)
Male	1.93 (0.89, 4.21)	0.10	0.81 (0.28, 2.36)	0.70
Province (ref = Ontario)
British Columbia	0.28 (0.18, 0.44)	<0.0001	0.33 (0.2, 0.54)	<0.0001
Quebec	0.10 (0.04, 0.28)	<0.0001	0.1 (0.04, 0.29)	<0.0001
Ethnicity (ref = Caucasian)
African Caribbean Black	0.5 (0.21, 1.22)	0.13	0.81 (0.3, 2.21)	0.68
Indigenous	0.18 (0.03, 0.98)	0.05	0.3 (0.05, 1.8)	0.18
Other	0.69 (0.35, 1.38)	0.29	0.8 (0.4, 1.6)	0.52
HIV transmission risk factor (ref = other)
IDU	0.99 (0.35, 2.8)	0.99	1.64 (0.55, 4.93)	0.38
MSM	2.38 (1.1, 5.17)	0.03	2.3 (0.82, 6.41)	0.11
MSM-IDU	2.36 (0.65, 8.53)	0.19	2.4 (0.55, 10.5)	0.24
Baseline CD4 (ref = 200 cells/mm³)^a
50 cells/mm³	0.97 (0.48, 1.95)	0.93	1.1 (0.53, 2.26)	0.80
100 cells/mm³	0.94 (0.63, 1.40)	0.76	1.02 (0.68, 1.53)	0.93
350 cells/mm³	1.60 (1.02, 2.50)	0.04	1.46 (0.9, 2.38)	0.12
500 cells/mm³	1.86 (0.99, 3.48)	0.05	1.69 (0.84, 3.39)	0.14
800 cells/mm³	1.27 (0.54, 2.98)	0.59	1.25 (0.5, 3.14)	0.63
Baseline viral load (ref ≤100,000 copies/ml)
≥100,000 copies/ml	0.77 (0.49, 1.23)	0.28	0.81 (0.49, 1.34)	0.41

Baseline CD4⁺ T-cell count was fit using restricted cubic splines with knots at 30, 210, 340 and 680 cells/mm³ to allow for a non-linear relationship with baseline resistance. The estimates shown in the table were chosen for their clinical relevance. IDU, injection drug use; MSM, men who have sex with men.

Follow-up resistance

Among 6,539 patients with no baseline tenofovir resistance, the cumulative incidence of virological failure (viral load >50 copies/ml on two consecutive occasions at least 30 days apart or a viral load >1,000 copies/ml on a single occasion after initial suppression or 16 weeks on therapy) at 1, 3 and 5 years was 0.20 (95% CI 0.19, 0.21), 0.32 (95% CI 0.31, 0.33) and 0.38 (95% CI 0.36, 0.39). The corresponding cumulative incidences of having a resistance test were 0.15 (95% CI 0.14, 0.15), 0.23 (95% CI 0.22, 0.24) and 0.28 (95% CI 0.27, 0.29), respectively. After 1, 3 and 5 years following treatment initiation, the number of patients developing resistance to tenofovir was 16, 29 and 34, and the cumulative incidence of resistance to this drug was 0.0027 (95% CI 0.0014, 0.004), 0.0055 (95% CI 0.0035, 0.0075) and 0.0070 (95% CI 0.0046, 0.0095), respectively. For emtricitabine, the cumulative incidences of virological failure and resistance testing among 6,601 patients without baseline resistance to this drug were identical to those of tenofovir. Following treatment initiation, 74, 133 and 159 patients developed emtricitabine resistance at 1, 3 and 5 years, respectively, and the cumulative incidence of resistance at those time points were 0.012 (95% CI 0.0095, 0.015), 0.025 (95% CI 0.02, 0.029) and 0.033 (95% CI 0.028, 0.038). Tenofovir resistance was most often related to the K65R mutation (n=25; 69%) with the accumulation of multiple thymidine analogue mutations being observed infrequently (n=5; 14%). Resistance to emtricitabine was associated with the M184I/V mutations in nearly all (95%) cases (Additional files 1 and 2).

Following multivariable analysis, baseline viral load ≥100,000 copies/ml was associated with the emergence of resistance to tenofovir (hazard ratio [HR] 2.88; 95% CI 1.35, 6.15) and emtricitabine (HR 2.27; 95% CI 1.64, 3.15). Similarly, initiation of an integrase inhibitor-based regimen relative to a boosted protease inhibitor-based regimen was associated with emergence of resistance to tenofovir (HR 6.25; 95% CI 1.48, 26.29) and emtricitabine (HR 3.11; 95% CI 1.53, 6.35). In contrast, initiation of non-nucleoside reverse transcriptase-based (HR 2.97; 95% CI 1.33, 6.6) or unboosted protease inhibitor-based (HR 6.58; 95% CI 1.64, 26.32) therapy was associated with resistance to tenofovir, but not emtricitabine (Table 3). Demographic variables associated with emtricitabine resistance included Indigenous ethnicity (HR 3.79; 95% CI 2.27, 6.33) and African/Caribbean/Black (HR 2.65; 95% CI 1.38, 5.08) ethnicity (Table 3). In addition, relative to patients with a baseline CD4⁺ T-cell count of 200 cells/mm³, patients with lower CD4⁺ T-cell counts were more likely to develop resistance to either drug, whereas those with higher CD4⁺ T-cell counts were less likely to develop resistance (Figures 1 and 2). However, this relationship reached statistical significance only for emtricitabine (Table 2).

Table 3

Adjusted proportional hazards models of times to tenofovir or emtricitabine resistance

	TDF		FTC
	HR (95% CI)	P-value	HR (95% CI)	P-value
Age (per 10 years)	0.75 (0.52, 1.08)	0.12	0.87 (0.74, 1.01)	0.08
Year of cART initiation	0.96 (0.80, 1.15)	0.66	0.98 (0.90, 1.06)	0.62
Gender (ref = female)
Male	0.74 (0.27, 2.04)	0.56	0.99 (0.68, 1.46)	0.98
Province (ref = Ontario)
British Columbia	1.25 (0.50, 3.14)	0.63	1.53 (0.98, 2.39)	0.06
Quebec	1.28 (0.46, 3.52)	0.64	1.23 (0.74, 2.05)	0.42
Ethnicity (ref = Caucasian)
African Caribbean Black	1.85 (0.50, 6.78)	0.35	2.65 (1.38, 5.08)	<0.01
Indigenous	1.00 (0.20, 5.05)	0.99	3.79 (2.27, 6.33)	<0.0001
Other	1.06 (0.30, 3.72)	0.93	1.43 (0.83, 2.45)	0.20
HIV transmission risk factor (ref = other)
IDU	2.65 (0.70, 9.99)	0.15	1.15 (0.68, 1.95)	0.60
MSM	1.48 (0.43, 5.09)	0.53	0.71 (0.39, 1.30)	0.26
MSM-IDU	2.81 (0.47, 16.76)	0.26	1.05 (0.45, 2.44)	0.91
Baseline CD4 (ref = 200 cells/mm³)^a
50 cells/mm³	1.81 (0.86, 3.78)	0.12	2.47 (1.82, 3.36)	<0.0001
100 cells/mm³	1.48 (0.98, 2.23)	0.06	1.82 (1.53, 2.16)	<0.0001
350 cells/mm³	0.62 (0.25, 1.51)	0.29	0.46 (0.29, 0.74)	<0.01
500 cells/mm³	0.48 (0.14, 1.65)	0.24	0.31 (0.16, 0.60)	<0.001
800 cells/mm³	0.40 (0.04, 4.01)	0.44	0.26 (0.06, 1.04)	0.06
Baseline viral load (ref ≤100,000 copies/ml)
≥100,000 copies/ml	2.88 (1.35, 6.15)	<0.01	2.27 (1.64, 3.15)	<0.0001
Regimen (ref = boosted PI)
II	6.25 (1.48, 26.29)	0.01	3.11 (1.53, 6.35)	<0.01
NNRTI	2.97 (1.33, 6.60)	<0.01	1.22 (0.89, 1.66)	0.22
Other	5.08 (0.60, 42.71)	0.13	2.25 (0.69, 7.36)	0.18
PI	6.58 (1.64, 26.32)	<0.01	1.76 (0.83, 3.75)	0.14

Baseline CD4⁺ T-cell count was fit using restricted cubic splines with knots at 30, 210, 340 and 680 cells/mm³ to allow for a non-linear relationship with time to resistance. The estimates shown in the table were chosen for their clinical relevance. FTC, emtricitabine; HR, hazard ratio; IDU, injection drug use; II, integrase inhibitor; MSM, men who have sex with men; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; TDF, tenofovir disoproxil fumarate.

Figure 1

Adjusted hazard ratios of development of tenofovir resistance by CD4⁺ T-cell count

Figure 2

Adjusted hazard ratios of development of emtricitabine resistance by CD4⁺ T-cell count

Discussion

In our study of over 6,000 antiretroviral-naive individuals with HIV, we observed a low baseline and follow-up frequency of resistance to tenofovir and emtricitabine during a period of increasing use of these drugs. We found associations between resistance to these drugs and low baseline CD4⁺ T-cell count, high baseline viral load, Indigenous ethnicity and type of antiretroviral regimen initiated.

Several mechanisms likely explain our findings. The association between baseline resistance and province of residence is consistent with findings from the 2012/2013 Canadian HIV Strain and Drug Resistance Surveillance Program, which described the prevalence of drug resistance mutations in specimens obtained for laboratory confirmation of HIV [29]. This report demonstrated an overall higher burden of transmitted resistance in patients diagnosed in Ontario relative to British Columbia (14.7% versus 10.7%), with a greater proportion of specimens in Ontario being resistant to nucleoside analogues alone or with another class (that is, multidrug resistant). Because genotypes in our study are likely to represent samples obtained at a point in time following diagnosis but prior to treatment initiation, a low prevalence of baseline resistance would be expected as conditions would favour the selection of wild-type HIV strains rather than those harbouring nucleoside analogue mutations deleterious to viral fitness. However, this phenomenon should not vary by province, such that inter-provincial differences in the burden of baseline resistance should remain discernible.

The association between baseline viral load and emergence of drug resistance may reflect incomplete virological suppression following the initiation of treatment, and/or the presence of drug-resistant minority variants in CANOC participants living with HIV [30]. The association of emtricitabine resistance with high baseline viral load observed here and elsewhere likely reflects a biological association with reduced likelihood of achieving virological suppression [30]. A possible association between low baseline CD4⁺ T-cell count and antiretroviral resistance, particularly among nucleoside analogues, has been previously described [30,31]. Although a biological mechanism may exist for this finding, CANOC participants with HIV and low baseline CD4⁺ T-cell counts may differ from other patients in ways that could influence treatment outcomes, including social and structural determinants of access to HIV care [32,33]. In addition, findings from qualitative studies have demonstrated how social and structural determinants of health such as housing and food insecurity, institutionalized stigma, systemic racism and trauma can influence health-care access and produce healthcare disparities among Indigenous persons living in Canada [34–36]. Finally, the finding that patients initiating integrase inhibitor-based regimens were at greater risk for developing resistance may reflect the predominant use of raltegravir and elvitegravir during the study period, both of which were used as part of regimens which have a lower genetic barrier to resistance than ritonavir-enhanced protease inhibitor-based regimens. This assertion is supported by findings from comparative clinical trials demonstrating minimal resistance to supporting nucleoside backbones in patients experiencing virological failure with boosted protease inhibitor-containing regimens, whereas such resistance was common when virological failure occurred among patients randomized to either elvitegravir or raltegravir [37,38]. Since dolutegravir-based regimens impose a higher genetic barrier to resistance than elvitegravir and raltegravir [39], it is possible that our observed association between nucleoside resistance and integrase inhibitors would be attenuated with longer follow-up or data from individuals starting since it became available.

Our findings build upon those of earlier studies examining the frequency of resistance to tenofovir and emtricitabine among antiretroviral-naive patients. In one review of 203 studies describing 23,291 patients, the prevalence of mutations known to confer resistance to tenofovir was 0.4% [40]. However, most sequences available for analysis in this study were collected prior to 2004, a period preceding the widespread approval and availability of tenofovir. A more recent review that included 358 studies conducted in low- and middle-income countries and published between January 2012 and August 2017 found that the prevalence of tenofovir resistance was 3% among 56,044 genotypes reviewed [41]. However, this finding includes both antiretroviral-naive patients and pretreatment genotypes in those with a history of prior antiretroviral treatment. A study of 4,717 antiretroviral-naive patients seroconverting before January 2013 found that <1% of patients had high-level resistance to either tenofovir or emtricitabine [42], with similar results reported in a study of 1,318 patients diagnosed with primary HIV infection in France between 2007 and 2012 and 1,946 study participants with baseline resistance testing in the Strategic Timing of Antiretroviral Treatment (START) trial [43,44]. A more recent analysis encompassing the years 2005 to 2014 found the prevalence of tenofovir and emtricitabine mutations to be below 4% among antiretroviral-naive patients in Italy, with a slightly higher frequency of resistance to tenofovir among men who have sex with men (1.4%) relative to women (1.3%) and heterosexual men (0.0%) [16]. Similar findings were obtained from an analysis of 758 patients from the HIV Outpatient Study, in that the prevalence of the K65R and M184V mutations was 0.4% and 0.8%, respectively [45]. However, this finding likely underestimates the prevalence of tenofovir resistance because it does not consider the contribution of thymidine analogue mutations to the occurrence of this phenomenon. Our study addresses this limitation by using an algorithm for defining tenofovir resistance that also includes thymidine analogue mutations.

Our study is strengthened by the large sample size of antiretroviral-naive patients and the comprehensive ascertainment of resistance to tenofovir and emtricitabine from several clinical cohorts across Canada. However, our study has several limitations. First, we did not have data on the time between infection with HIV and baseline genotypic resistance testing; it is therefore possible that mutations conferring resistance to tenofovir or emtricitabine could have become archived in some individuals, underestimating the prevalence of resistance to one or both drugs. This is important when viewed in light of recent findings indicating that mutations associated with the component drugs of PrEP are no longer detected after 6 months of seroconversion [46]. Similarly, not all patients underwent genotyping prior to initiating therapy, thereby limiting our ability to accurately measure the frequency of drug resistance. However, these limitations are common to all clinical cohort studies undertaking surveillance for drug resistance. We fit non-parsimonious regression models to examine associations between demographic and clinical characteristics with the development of resistance. These findings should be interpreted cautiously and considered hypothesis generating given the small number of events, particularly when interpreting associations between resistance and regimen type. Finally, we are unable to directly address the impact of the level of resistance documented on the effectiveness of PrEP. Although our results provide a measure of reassurance that the reliability of PrEP is unlikely to be undermined by the frequency of resistant HIV observed in this analysis, further research is required to conclusively address this question and examine whether the impact of resistance varies among different sub-populations of persons with HIV.

In conclusion, we observed a low incidence of tenofovir and emtricitabine resistance among antiretroviral-naive patients in a large Canadian cohort of patients receiving routine care. As PrEP becomes increasingly used, ongoing surveillance of the frequency of drug resistance is required to inform delivery of this intervention and model its effectiveness among populations at high risk for HIV.

Footnotes

Acknowledgements

The authors would like to thank all of the CANOC participants for their valued information, Valerie Nicholson (BC Centre for Excellence in HIV/AIDS) for her review of the manuscript, Chanson Brumme and Conan Woods of the BC Centre for Excellence in HIV/AIDS for their assistance with assembling the mutation data and Bluma Brenner of Lady Davis Institute at McGill AIDS Centre for providing information about the assay for detecting drug mutations in Quebec.

The CANOC principal investigators consist of Centre investigator Robert Hogg and Centre site investigators, Curtis Cooper, Deborah Kelly, Marina Klein, Mona Loutfy, Nima Machouf, Julio Montaner, Stephen Sanche, Alexander Wong, Sharon Walmsley, Abigail Kroch, Rejean Thomas. CANOC co-investigators include Ann N Burchell, Tony Antoniou, Ahmed Bayoumi, Mark Hull, Bohdan Nosyk, Sean B Rourke, Chris Tsoukas, Angela Cescon, Michelle Cotterchio, Charlie Goldsmith, Silvia Guillemi, P Richard Harrigan, Marianne Harris, Sean Hosein, Sharon Johnston, Claire Kendall, Clare Liddy, Viviane Lima, David Moore, Alexis Palmer, Sophie Patterson, Peter Phillips, Nisha Andany, Hasina Samji, Marek Smieja.

CANOC is funded by the Canadian Institutes of Health Research (CIHR) through a Centres Grant (Centres for HIV/AIDS Population Health and Health Services Research [CIHR #02684]); 2 Operating Grants (HIV/AIDS Priority Announcement [CIHR #134047]; a Population and Public Health Grant [CIHR #136882]); a Foundation Grant (Expansion of Antiretroviral Therapy and its Impact on Vulnerable Populations in Canada and Global Settings [CIHR #143342]); and in collaboration with the CIHR Canadian HIV Trials Network [CTN #242]. AMB is supported by the Fondation Baxter & Alma Ricard Chair in Inner City Health at St. Michael's Hospital and the University of Toronto. TA is supported by a CIHR New Investigator and University of Toronto Department of Family and Community Medicine Clinician Investigator Award. MBK is supported by a Chercheur National Career Award from the Fonds de recherche du Québec - Santé (FRQ-S). JM is supported by the British Columbia Ministry of Health and by the US National Institutes of Health (R01DA036307). SW is supported through an Applied Health Research Chair from the OHTN.

The funders had no role in the design of the study, data collection, analysis, interpretation of data and the decision to publish the manuscript.

CC has served on advisory boards for AbbVie and Gilead Sciences. MBK reports grants from Merck and ViiV Healthcare and personal fees for consultancy from ViiV Healthcare, Bristol-Myers Squibb, and Merck. ML has served on advisory boards and spoken at CME events for ViiV Healthcare, AbbVie, Merck Canada Inc. and Gilead Sciences. JM is supported with grants paid to his institution by the British Columbia Ministry of Health and by the US National Institutes of Health (R01DA036307). He has also received limited unrestricted funding, paid to his institution, from AbbVie, Bristol-Myers Squibb, Gilead Sciences, Janssen, Merck and ViiV Healthcare. JR is co-investigator on four projects outside the submitted work with in-kind contributions or financial support from Merck and Gilead Sciences. SW has served on advisory boards and spoken at CME events for ViiV, AbbVie, Merck, Gilead, Janssen and Bristol-Myers Squibb. RH reports personal fees from Gilead Sciences. For the remaining authors no conflicts of interest were declared.

Additional file 1: A table showing prevalence of individual mutations associated with emtricitabine resistance can be found at

Additional file 2: A table showing prevalence of individual mutations associated with tenofovir resistance can be found at

References

Grant

R.M.

, Lama

J.R.

, Anderson

P.L.

Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med 2010; 363: 2587–2599

Choopanya

, Martin

, Suntharasamai

Antiretroviral prophylaxis for HIV infection in injecting drug users in Bangkok, Thailand (the Bangkok Tenofovir Study): a randomised, double-blind, placebo-controlled Phase 3 trial. Lancet 2013; 381: 2083–2090

Baeten

J.M.

, Donnell

, Ndase

Antiretroviral prophylaxis for HIV prevention in heterosexual men and women. N Engl J Med 2012; 367: 399–410

Deutsch

M.B.

, Glidden

D.V.

, Sevelius

HIV pre-exposure prophylaxis in transgender women: a subgroup analysis of the iPrEx trial. Lancet HIV 2015; 2: e512–e519

McCormack

, Dunn

D.T.

, Desai

Pre-exposure prophylaxis to prevent the acquisition of HIV-1 infection (PROUD): effectiveness results from the pilot phase of a pragmatic open-label randomised trial. Lancet 2016; 387: 53–60

Thigpen

M.C.

, Kebaabetswe

P.M.

, Paxton

L.A.

Antiretroviral preexposure prophylaxis for heterosexual HIV transmission in Botswana. N Engl J Med 2012; 367: 423–434

Molina

J.M.

, Capitant

, Spire

On-demand preexposure prophylaxis in men at high risk for HIV-1 infection. N Engl J Med 2015; 373: 2237–2246

Molina

J.M.

, Charreau

, Spire

Efficacy, safety, and effect on sexual behaviour of on-demand pre-exposure prophylaxis for HIV in men who have sex with men: an observational cohort study. Lancet HIV 2017; 4: e402–e410

Fonner

V.A.

, Dalglish

S.L.

, Kennedy

C.E.

Effectiveness and safety of oral HIV preexposure prophylaxis for all populations. AIDS 2016; 30: 1973–1983

10.

World Health Organization. WHO expands recommendation on oral pre-exposure prophylaxis of HIV infection (PrEP). (Accessed 5 October 2018.) Available from http://www.who.int/hiv/pub/prep/policy-brief-prep-2015/en/

11.

Donnell

, Baeten

J.M.

, Bumpus

N.N.

HIV protective efficacy and correlates of tenofovir blood concentrations in a clinical trial of PrEP for HIV prevention. J Acquir Immune Defic Syndr 2014; 66: 340–348

12.

Cong

M.E.

, Mitchell

, Sweeney

Prophylactic efficacy of oral emtricitabine and tenofovir disoproxil fumarate combination therapy against a tenofovir-resistant simian/human immunodeficiency virus containing the K65R mutation in macaques. J Infect Dis 2013; 208: 463–467

13.

Cong

M.E.

, Youngpairoj

A.S.

, Zheng

Protection against rectal transmission of an emtricitabine-resistant simian/human immunodeficiency virus SHIV162p3M184V mutant by intermittent prophylaxis with Truvada. J Virol 2011; 85: 7933–7936

14.

Knox

D.C.

, Anderson

P.L.

, Harrigan

P.R.

, Tan

D.H.S.

Multidrug-resistant HIV-1 infection despite preexposure prophylaxis. N Engl J Med 2017; 376: 501–502

15.

Markowitz

, Grossman

, Anderson

P.L.

Newly acquired infection with multidrug-resistant HIV-1 in a patient adherent to preexposure prophylaxis. J Acquir Immune Defic Syndr 2017; 76: e104–e106

16.

Rossotti

, Borghi

, Callegaro

A.P.

Impact of circulating resistance-associated mutations on HIV pre-exposure prophylaxis (PrEP) efficacy: modeling from antiretroviral resistance cohort analysis (ARCA) national database. J Clin Virol 2016; 83: 48–53

17.

van Tienen

, van de Vijver

, Noori

, Sönnerborg

, Boucher

Letter to the editor: pre-exposure prophylaxis for HIV in Europe: the need for resistance surveillance. Euro Surveill 2017; 22: pii: 30483

18.

Dimitrov

, Boily

M.C.

, Brown

E.R.

, Hallett

T.B.

Analytic review of modeling studies of ARV based PrEP interventions reveals strong influence of drug-resistance assumptions on the population-level effectiveness. PLoS One 2013; 8: e80927

19.

Palmer

A.K.

, Klein

, Raboud

Cohort profile: the Canadian Observational Cohort Collaboration. Int J Epidemiol 2011; 40: 25–32

20.

Department of Health and Human Services. 2006 Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. (Accessed 5 October 2018.) Available from http://aidsinfo.nih.gov/guidelines/archive/adult-and-adolescent-guidelines

21.

Vermeiren

, Van Craenenbroeck

, Alen

Prediction of HIV-1 drug susceptibility phenotype from the viral genotype using linear regression modeling. J Virol Methods 2007; 145: 47–55

22.

Pattery

, Verlinden

, De Wolf

Development and performance of conventional HIV-1 phenotyping (Antivirogram®) and genotype-based calculated phenotyping assay (virco®TYPE HIV-1) on protease and reverse transcriptase genes to evaluate drug resistance. Intervirology 2012; 55: 138–146

23.

Winters

, Montaner

, Harrigan

P.R.

Determination of clinically relevant cutoffs for HIV-1 phenotypic resistance estimates through a combined analysis of clinical trial and cohort data. J Acquir Immune Defic Syndr 2008; 48: 26–34

24.

Charest

, Doualla-Bell

, Cantin

A significant reduction in the frequency of HIV-1 drug resistance in Quebec from 2001 to 2011 is associated with a decrease in the monitored viral load. PLoS One 2014; 9: e109420

25.

Brenner

B.G.

, Lowe

, Moisi

Subtype diversity associated with the development of HIV-1 resistance to integrase inhibitors. J Med Virol 2011; 83: 751–759

26.

Woods

C.K.

, Brumme

C.J.

, Liu

T.F.

Automating HIV drug resistance genotyping with RECall, a freely accessible sequence analysis tool. J Clin Microbiol 2012; 50: 1936–1942

27.

Swenson

L.C.

, Min

J.E.

, Woods

C.K.

HIV drug resistance detected during low-level viraemia is associated with subsequent virologic failure. AIDS 2014; 28: 1125–1134

28.

White

I.R.

, Royston

Imputing missing covariate values for the Cox model. Stat Med 2009; 28: 1982–1998

29.

Public Health Agency of Canada. Update on HIV-1 Strain and Transmitted Drug Resistance in Canada: 2012–2013. Centre for Communicable Diseases and Infection Control, Public Health Agency of Canada, 2017. (Accessed 5 October 2018.) Available from https://www.canada.ca/en/public-health/services/publications/diseases-conditions/hiv-canadian-strain-drug-resistance-surveillance-2012-2013.html

30.

Harrigan

P.R.

, Hogg

R.S.

, Dong

W.W.

Predictors of HIV drug-resistance mutations in a large antiretroviral-naive cohort initiating triple antiretroviral therapy. J Infect Dis 2005; 191: 339–347

31.

, Armon

, Buchacz

, Wood

, Brooks

J.T.

Initiation of HAART at higher CD4 cell counts is associated with a lower frequency of antiretroviral drug resistance mutations at virologic failure. J Acquir Immune Defic Syndr 2009; 51: 450–453

32.

Gesesew

H.A.

, Tesfay Gebremedhin

, Demissie

T.D.

Significant association between perceived HIV related stigma and late presentation for HIV/AIDS care in low and middle-income countries: a systematic review and meta-analysis. PLoS One 2017; 12: e0173928

33.

Metsch

, Philbin

M.M.

, Parish

, Shiu

, Frimpong

J.A.

, Giang le

HIV testing, care, and treatment among women who use drugs from a global perspective: progress and challenges. J Acquir Immune Defic Syndr 2015; 69 Suppl 2: S162–S168

34.

Woodgate

R.L.

, Zurba

, Tennent

, Cochrane

, Payne

, Mignone

“People try and label me as someone I'm not”: the social ecology of Indigenous people living with HIV, stigma, and discrimination in Manitoba, Canada. Soc Sci Med 2017; 194: 17–24

35.

Woodgate

R.L.

, Zurba

, Tennent

, Cochrane

, Payne

, Mignone

A qualitative study on the intersectional social determinants for indigenous people who become infected with HIV in their youth. Int J Equity Health 2017; 16: 132

36.

Goodman

, Fleming

, Markwick

“They treated me like crap and I know it was because I was Native”: the healthcare experiences of Aboriginal peoples living in Vancouver's inner city. Soc Sci Med 2017; 178: 87–94

37.

Lennox

J.L.

, Landovitz

R.J.

, Ribaudo

H.J.

A Phase III comparative study of the efficacy and tolerability of three non-nucleoside reverse transcriptase inhibitor-sparing antiretroviral regimens for treatment-naive HIV-1-infected volunteers: a randomized, controlled trial. Ann Intern Med 2014; 161: 461–471

38.

DeJesus

, Rockstroh

J.K.

, Henry

Coformulated elvitegravir, cobicistat, emtricitabine, and tenofovir disoproxil fumarate versus ritonavir-boosted atazanavir plus co-formulated emtricitabine and tenofovir disoproxil fumarate for initial treatment of HIV-1 infection: a randomised, double-blind, Phase 3, non-inferiority trial. Lancet 2012; 379: 2429–2438

39.

Llibre

J.M.

, Pulido

, Garcia

, Deltoro

M.G.

, Blanco

J.L.

, Delgado

Genetic barrier to resistance for dolutegravir. AIDS Rev 2015; 17: 56–64

40.

Chan

P.A.

, Huang

, Kantor

Low prevalence of transmitted K65R and other tenofovir resistance mutations across different HIV-1 subtypes: implications for pre-exposure prophylaxis. J Int AIDS Soc 2012; 15: 17701

41.

Gupta

R.K.

, Gregson

, Parkin

HIV-1 drug resistance before initiation or re-initiation of first-line antiretroviral therapy in low-income and middle-income countries: a systematic review and meta-regression analysis. Lancet Infect Dis 2018; 18: 346–355

42.

Olson

, Bannert

, Sönnerborg

Temporal tends of transmitted HIV drug resistance in a multinational seroconversion cohort. AIDS 2018; 32: 161–169

43.

Frange

, Assoumou

, Descamps

HIV-1 subtype B-infected MSM may have driven the spread of transmitted resistant strains in France in 2007-12: impact on susceptibility to first-line strategies. J Antimicrob Chemother 2015; 70: 2084–2089

44.

Baxter

J.D.

, Dunn

, White

Global HIV-1 transmitted drug resistance in the INSIGHT Strategic Timing of AntiRetroviral Treatment trial. HIV Med 2015; 16: 77–87

45.

Buchacz

, Young

, Palella

F.J.

, Armon

Brooks JT on behalf of the HOPS investigators. Trends in the use of genotypic resistance testing and frequency of major drug resistance among antiretroviral-naive persons in the HIV Outpatient Study, 1999-2011. J Antimicrob Chemother 2015; 70: 2337–2346

46.

Weis

J.F.

, Baeten

J.M.

, McCoy

C.O.

Preexposure prophylaxis-selected drug resistance decays rapidly after drug cessation. AIDS 2016; 30: 31–35

Tenofovir and Emtricitabine Resistance among Antiretroviral-Naive Patients in the Canadian Observational Cohort Collaboration: Implications for PrEP

Abstract

Background

Methods

Results

Conclusions

Introduction

Methods

Study design and population

Outcomes

HIV drug Resistance Testing

Statistical analysis

Results

Study population

Baseline resistance

Follow-up resistance

Discussion

Footnotes

Acknowledgements

References