Plasma HIV-1 RNA viral load rebound among people who inject drugs receiving antiretroviral therapy (ART) in a Canadian setting: an ethno-epidemiological study

Antiretroviral therapy (ART) inhibits HIV viral replication and suppresses HIV plasma viral load (PVL) to undetectable levels, resulting in reductions in HIV-related morbidity and mortality, as well as reduced potential for onward HIV transmission [1]. Interventions promoting the preventive benefit of ART to reduce the onward transmission of HIV, known as Treatment as Prevention (TasP), have become a cornerstone of global HIV prevention efforts [2]. Due to the high rates of HIV infection among people who inject drugs (PWID), this population is a focus for TasP programs, but PWID often experience sub-optimal ART outcomes [3, 4], including high rates of treatment discontinuation and failure [5, 6]. Poor treatment outcomes among PWID have potential to undermine the effectiveness of TasP in preventing injection-related HIV infections [7], and comprehensive efforts to promote viral suppression in this population are required [4, 8].

While contemporary perspectives on HIV treatment among PWID increasingly emphasize the role of social-structural factors in shaping treatment outcomes [4], relatively few studies have investigated their impact on disease progression [9]. The HIV Risk Environment framework has been utilized to examine how HIV risk behaviours among PWID are shaped by interactions between individuals and social, structural, political, economic influences [10, 11], but has not yet been fully applied to examine the role various environmental forces play in shaping ART outcomes among PWID [9]. The Risk Environment framework has potential to help identify how broad social-level and structural-level influences constrain engagement with public health programs and impede individual ability to adhere to ART regimens [9–11]. While some evidence suggests that socio-economic marginalization, HIV-related stigma and discrimination, the criminalization of drug use, incarceration, and inadequate access to substance use treatment [9, 11], pose barriers to adherence, less is known about how these forces function to produce sub-optimal ART outcomes. Similarly, although drug use has often been identified as a risk factor for suboptimal adherence among injection drug-using populations [12, 13], recent research has suggested that PWID can achieve optimal adherence even in the context of ongoing drug use [14, 15]. This suggests that social-structural forces may play an important role in mediating the relationship between drug use and HIV treatment outcomes [9, 14, 15].

Plasma HIV-1 RNA viral load rebound (hereafter referred to as ‘rebound’), defined as the re-emergence of detectable plasma HIV viral load among individuals who had previously achieved viral suppression through ART, is a form of HIV treatment failure that represents a key challenge to the medical management of HIV [16, 17]. Approximately 10 % of all patients taking ART will experience a rebound episode [18], although it is not clear precisely how quickly viral load rebound emerge may due to missed doses of medication or treatment discontinuation given complex interactions between patient and virus characteristics, treatment histories, and regimen-specific effects. Shorter duration of viral suppression, particular ART medications, and non-adherence to treatment regimens are associated with increased risk of rebound, and viral rebound is a risk factor for the emergence of viral resistance, requiring less tolerable second-line and salvage-level therapies, and the onward transmission of transmission-resistant viruses [16, 18–20]. However, rebound appears to be far more common among PWID than other populations. A study from Vancouver, Canada, indicates that 45 % of PWID receiving ART displayed at least one rebound episode since initiating treatment [17]. In this study, rebound was associated with recent incarceration and involvement in sex work, while methadone maintenance therapy (MMT) was negatively associated with rebound episodes [17]. Importantly, when analyses controlled for adherence to ART, no measure of drug use was found to predict the emergence of rebound [17].

Due to the importance of long-term retention on ART and prolonged viral suppression for the effectiveness of TasP, there is an urgent need for greater understanding of the emergence of virologic failure among PWID [8, 17]. While risk factors associated with rebound among PWID have been identified, research to date has not examined the perspectives of PWID displaying rebound or documented the circumstances surrounding rebound events. Detailed knowledge of the lived experience of drug users can generate understanding of the production of specific health outcomes, and to add a deeper social dimension to conventional quantitative measures [21, 22]. Integrating targeted qualitative research with epidemiological inquiry, known as ‘ethno-epidemiology’ [23–25], has potential to improve knowledge regarding ART treatment outcomes by examining the experiences and perspectives of individuals who display particular outcomes. We conducted an ethno-epidemiological study to investigate the circumstances surrounding the emergence of rebound episodes among PWID in Vancouver, Canada, with a focus on the role of social-structural forces in the production of virologic failure. We examined the experiences and perspectives of PWID who had a recent episode of rebound through targeted qualitative interviews.

This ethno-epidemiological study was nested within the AIDS Care Cohort to Evaluate Exposure to Survival Services (ACCESS), an ongoing prospective cohort of HIV-positive adults (≥18 years of age) with histories of illicit drug use (other than marijuana), which has been described previously [17]. Participants in this community-recruited cohort study provide a blood sample for virologic analysis and complete an interviewer-administered behavioural questionnaire at baseline and every 6 months thereafter. ACCESS data are linked with data collected by the Drug Treatment Programme of the British Columbia Centre for Excellence in HIV/AIDS, a province-wide centralized ART dispensary and clinical monitoring laboratory, providing a complete retrospective and prospective profile of CD4+ cell counts, PVL, and exposure to specific antiretroviral medications for all participants [17]. Universal healthcare is provided in this setting, and all HIV care (including ART) is provided free of charge, permitting examination of treatment outcomes among PWID in a context without financial barriers to access or medications.

In April 2011, we conducted a database query to identify ACCESS participants with recent rebound experiences. Inclusion criteria were all participants who: (i) were exposed to ART since enrolling in ACCESS; (ii) had at least two consecutive measurements indicating PVL suppression (<50 copies/mL) since 2007; and, had at least one measurement indicating elevated PVL (>1000 copies/mL) subsequent to PVL suppression. 43 individuals met these criteria, three of whom were recently deceased. For each participant, we generated a clinical timeline graphic (see Fig. 1) depicting PVL and adherence to ART over time. To estimate adherence, we used a validated measure based on pharmacy refill data [17], defined as the number of days ART was dispensed divided by the number of days since an individual was first dispensed therapy, capped at 180 days.

A semi-structured interview guide was used to facilitate discussion regarding the circumstances surrounding rebound episodes, including HIV care, ART dispensing arrangements, drug use, living circumstances, and other contextual factors. A modified format of timeline interviewing, which is effective in linking individual narratives and experiences to wider social and environmental contexts [26], was used to retrospectively examine HIV treatment experiences in relation to each participant’s drug use career and personal biography, as well as to situate their experiences within their social-structural contexts. A simple timeline was hand-drawn by the interviewer during the interview, noting HIV care milestones and life events with the potential to impact upon ART adherence (e.g., housing transitions, changes in drug use patterns). This approach elicited extensive discussion of social-structural influences shaping adherence, as well as transitions between periods of adherence and non-adherence. The clinical timeline graphic was introduced at the conclusion of the interview to elicit further discussion of treatment events, summarize the interviewer’s interpretation of events, and allow participants to clarify inaccuracies. Some participants displayed more than one rebound episode over the study period, and all were examined during interviews and analysis. A total of 25 participants provided clear accounts of non-adherence related to their viral load rebound, as two participants were unable to provide a narrative regarding circumstances surrounding their rebound event due to problems related to recall.

Analysis focused on examining factors leading to prolonged interruption or discontinuation of ART. Interview transcripts and timelines were examined to understand each participant’s experience and rebound-related circumstances as an individual case, and a coding framework composed of a priori and emergent codes (e.g., housing transitions, drug use transitions, social isolation, understandings of ART) was then used to compare and contrast these cases. We also drew on the Risk Environment framework in our analysis [11], as well as the concept of structural vulnerability to illuminate how social and structural inequalities shaped engagement with care and adherence-related behaviour [27]. The analysis was conducted using word processing software to code, label, and retrieve data segments [28]. Ethics approval for the study was obtained from UBC/PHC Research Ethics Board. We have utilized the consolidated criteria for reporting qualitative research (COREQ) guidelines, to ensure full and accurate description of our study and analytical methods. [29] All names used in presenting the analysis are pseudonyms.

Our findings illustrate how rebound episodes were shaped by the complex interplay between individual-level factors and social, structural, and environmental forces operating within the HIV treatment risk environment. Participants had previously achieved viral suppression by maintaining optimal ART adherence, and their accounts emphasized how they had managed or adjusted their living situations, HIV care, ART dispensing arrangements, and drug use to ensure adherence through the development and maintenance of “stability” or regular “routines”. Rebound episodes were associated with disruption of these stable situations and routines, stemming from housing transitions, difficulty managing co-morbid health conditions, as well as changes in drug use and drug scene involvement. In addition, disruptions occurring due to poor relationships with care providers and misunderstandings regarding ART resulted in intentional treatment discontinuation in several cases.

The findings of our study illustrate the role of interactions between individual-level factors and diverse social-structural forces operating within the HIV treatment risk environment in shaping the emergence of rebound episodes among PWID. Rebound episodes emerged when routines that facilitated adherence were disrupted by structural-environmental influences, including housing transitions, changes in drug use patterns and drug scene involvement, and inadequate care for comorbid health conditions. Social-environmental influences similarly disrupted participants’ adherence to ART as poor communication between care providers and patients precipitated non-adherence, and misunderstandings of ART functioned to reduce motivation to adhere to ART and produced intentional treatment discontinuation.

This study is the first to use qualitative and ethno-epidemiological methods to examine viral load rebound among PWID. Our findings generate new knowledge regarding how structural-environmental and social-environmental forces influence ART non-adherence among PWID, and lead to rebound episodes. While being homeless or marginally housed is a known risk factor for poor ART adherence and treatment outcomes among PWID [9, 30, 31], our findings indicate that housing transitions, including relocating to a different neighborhood, may influence treatment outcomes even among HIV-positive PWID who are continuously housed. These dynamics suggest that transitions in housing and relocation to different neighborhoods may be particularly influential among structurally vulnerable PWID, as living and care arrangements that function to enable optimal adherence may be easily disrupted. Changes in housing status and residence may also play a role in driving increases in drug use and involvement in illegal income generation among drug users [32], further underscoring the importance of maintaining housing stability among HIV positive PWID. Interventions designed to improve housing stability and quality may reduce the potential for non-adherence [33], but efforts to improve continuity of HIV care for individuals relocating their residence should focus upon ensuring specialist care and appropriate dispensing arrangements subsequent to housing transitions. Importantly, both supportive housing models and housing utilizing harm reduction supports appear to facilitate adherence among PWID given the documented beneficial effect on viral load [30, 34].

Recent research has challenged the perspective that illicit drug use directly contributes to poor ART outcomes [9], highlighting the importance of ART adherence in mediating the relationship between drug use and treatment outcomes [14, 15]. Our results underscore the importance of understanding how extra-individual forces (including physical and psychological comorbidities) interact with varying patterns of drug use to shape non-adherence [9], and elucidate how changes in drug use and drug scene involvement impede individual ability to adhere by disrupting adherence-related routines. Intense drug scene exposure has previously been associated with poor health outcomes and more severe substance use [35], and our findings suggest that time spent in the drug scene (generating income and using drugs) represents a mechanism through which drug use shapes non-adherence. Given that intensity of drug scene involvement and illegal income generation may mediate the relationship between drug use and ART adherence, the potential of alternative income generation and low-threshold employment interventions for PWID living with HIV should be explored [36, 37]. Improving economic opportunities may represent an effective strategy to reduce or limit drug use and drug scene involvement [32], and employment has been previously documented to positively impact ART adherence [38].

While MMT has been associated with optimal adherence [39–41], our findings illustrate how the benefit of MMT upon ART adherence can be eroded and negated among PWID. Specifically, we found that MMT discontinuation events also coincided with non-adherence and viral rebound. An established literature shows that restrictive programmatic features of MMT can result in suboptimal outcomes and poor treatment retention [42, 43], and additional supports for HIV-positive PWID should be considered, including low threshold MMT, outreach to individuals who discontinue MMT, and alternatives to MMT (including heroin prescription programs and buprenorphine). These supports might help facilitate optimal adherence over the long-term among PWID.

Inadequate support across the continuum of care for PWID living with HIV with comorbid conditions (e.g., mental health conditions) resulted in non-adherence and rebound. While it is recognized that medical and psychiatric comorbidities among HIV-positive PWID (and the concurrent nature of these conditions) contribute to poor ART outcomes [44], appropriate interventions are lacking in most settings. The role comorbid conditions played in disrupting adherence among study participants illustrates the importance of comprehensive care and support for HIV positive PWID in maintaining optimal adherence. There is a need for integrated health services, including multi-component and interdisciplinary approaches, delivering integrated case management to simultaneously improve the management of comorbid conditions and adherence to ART [44–47].

In the context of inadequate linkage with mental health services, existing and emergent mental health conditions affected adherence, illustrating the importance of enhanced care and support for mental health conditions among HIV-positive PWID. Interventions to manage mental health conditions among PWID include community-based integrated mental health services programs, social supports, peer-support programs, and housing-based interventions [48, 49]. Unfortunately, in many settings (including Vancouver), biomedical and legal approaches to the management of mental illness have become increasingly dominant, detracting attention from social and structural forces driving suboptimal treatment of mental health conditions, and eroding support for community-based programs [48]. Detainment of individuals diagnosed with mental illness produces negative effects on adherence and treatment outcomes [4]. The utilization of peer-based interventions [50, 51], including peer-counselling, peer navigators, and peer-support programs, could contribute to the effective management of mental illness and simultaneously support optimal adherence among PWID [52–54]. Local drug user groups may be a vehicle for mobilizing community resources to promote social support and engagement with care [55].

Our findings also indicate that poor communication and interactions between care providers and patients precipitated discontinuation of ART, particularly in relation to physician decisions regarding the necessity of MAT/DOT programs to ensure adherence among PWID. While MAT/DOT programs are associated with improved adherence and increased likelihood of viral suppression [56], treatment decisions should be jointly made by patients and physicians [57], and with the understanding that some participants perceive observed therapy to be paternalistic. While our findings are consistent with past work showing that incarceration may shape non-adherence through treatment interruptions stemming from delays in obtaining medications while in custody [6, 58], poor relationships with institutional health care providers may also lead PWID to discontinue ART while incarcerated. These dynamics have previously been documented in our setting [59], and PWID routinely re-initiate treatment upon return to the community, although some do experience challenges obtaining medications and connecting with physicians post-release. As noted previously, correctional services have a duty to provide essential health care in a timely manner, including ART medications, and such efforts are essential to ensuring continuity of HIV care for PWID who experience incarceration [59]. Improving relationships and communication between PWID and physicians may benefit adherence [60], and efforts to improve patient-provider interactions may involve better training for physicians who care for PWID, as well as education regarding addiction medicine and substitution therapies for opioid dependence [61, 62].

Understandings of HIV treatment played a role in shaping adherence to ART among study participants. Our findings illustrate how numerous years of ART may accumulate to make individuals “tired” of pills given a lack of appreciation that managing HIV with ART requires lifelong adherence. This illustrates how treatment fatigue plays a role in shaping intentional treatment discontinuation, as qualitative examination of these issues has been lacking to date [63]. While increasingly simple regimens involving fewer pills may reduce levels of fatigue, ensuring long-term retention of PWID will require further interventions to prevent intentional discontinuation and that care providers emphasize the lifelong nature of ART when communicating with patients [64]. In addition, individuals may want to have a sense of control over their HIV treatment and medications, which may entail stopping or modifying treatment regimens (particularly in the context of long-term ART), and these dynamics may further complicate communication between patients and care providers. Our findings document how particular beliefs regarding resistance and disease progression, as well as anachronistic ideas regarding best practices in HIV treatment, played a role in producing intentional treatment discontinuation. This is particularly significant given emerging clinical research suggesting that rates of antiretroviral resistance among PWID are less of a concern than previously thought [65], highlighting the importance of care providers delivering ongoing education and accurate information about HIV treatment [66]. While improving patient-provider communication is an important strategy for correcting misconceptions regarding ART, social models of education and information provision (similar to peer-led safer injection education) might also improve understandings of ART [55], correct outdated beliefs, and address common concerns regarding long term retention in HIV treatment [67].

The current study has several limitations. Our findings are specific to the experience of individuals participating in the study, and other PWID displaying viral rebound may have different experiences and perspectives. Some participants were not able to describe circumstances surrounding the emergence of rebound episodes, and our analysis has placed emphasis upon cases where participants were able to provide a complete account of the emergence of rebound episodes. While the design of our study allowed for direct examination of the experiences of PWID who recently displayed rebound, further research regarding circumstances surrounding rebound is needed.

In conclusion, this study identified pathways leading to viral rebound episodes among PWID who had achieved viral suppression, and illustrates the impact of environmental forces in shaping non-adherence and treatment failure. Our findings highlight the need for interventions and supports that address the social-structural forces that create vulnerability to ART non-adherence among PWID.