Illicit Fentanyl Use Independently Predicts HCV Seroconversion Among a Cohort of People Who Inject Drugs in Tijuana and San Diego

Introduction

Illicitly manufactured fentanyls (IMF) have transformed the risk environment for people who use drugs (PWUD) in North America [1]. Having outcompeted heroin for dominance in the illicit opioid market, IMF have caused dramatic increases in overdose mortality over the past decade [2]. Given their high potency, seemingly small fluctuations in product quality can result in large increases in overdose risk [1].

IMF also have additional properties that may elevate infectious disease transmission risk, including shorter half-lives and more powerful euphoric effects relative to heroin and other opioids [3]. This has been associated with increased injection frequency to prevent withdrawal symptoms and achieve sustained effects [4]. IMF use has consequently been associated with higher-risk injection practices compared to heroin use, such as sharing syringes [5–7]. Literature describing implications of IMF use, including drug preparation and administration practices [8] and behavioral aspects (including the pursuit of euphoria) [9], suggest that there may be an important—albeit not currently described—link between IMF use and human immunodeficiency virus (HIV) and hepatitis C virus (HCV). Sharply rising HCV incidence among young people over the past two decades has been linked to increasing rates of injection drug use nationally, although the particular role of fentanyl has not been determined [10–12].

We examined whether IMF use predicts HIV and HCV incidence among a longitudinal cohort of people who inject drugs (PWID) in the US-Mexico border region of Tijuana, Mexico, and San Diego, California, a region with a large population of vulnerable PWID who experience a high burden of HCV, and where IMF has broadly overtaken the drug supply.

Methods

Results

Participant characteristics, stratified by HCV seroconversion status, are shown in Table 1 for the 398 participants in the analytic sample (i.e., n=102 individuals who seroconverted and n=296 individuals who remained HCV seronegative during follow-up). At baseline, 67.3% resided in San Diego; 69.8% were male; median age was 42.5 years; 66.8% identified as Hispanic, Latinx or Mexican; and median years of schooling was 10.5 (i.e., some secondary school/high school) [Interquartile range [IQR]: 7.0-12.0]. Baseline characteristics revealed a highly vulnerable population, with 41.5% experiencing homelessness in the past 6 months.

Polysubstance use was common; the most commonly used substances were methamphetamine (83.7%), heroin (80.2%), cannabis (53.8%) and fentanyl (25.4%). Forty-two percent reported receptive needle sharing at baseline. The median number of average injections per day was 2.5 (IQR: 0.3-4.0). Addiction-related healthcare engagement was poor, with only 6.3% of participants reporting current enrollment in a methadone, buprenorphine or other addiction treatment program. Participants contributed a median of six semi-annual study visits (IQR:4-6).

We observed 10 HIV seroconversions during the study period (Table 2), resulting in an incidence rate of 1.29 per 100 person-years (95%CI: 1.00-2.28). Among people reporting fentanyl use, HIV incidence was 2.28 per 100 person-years (95% confidence interval[CI]: 0.05-4.50) compared to 1.00 (95%CI: 0.20-1.80) among those not reporting fentanyl use. The univariable HR for the association between fentanyl use and HIV seroconversion was 2.53 (95%CI: 0.7-9.15).

HCV incidence over the 36-month study period was 14.26 per 100 person-years (95% CI: 11.49-17.02). Among individuals who reported using fentanyl at baseline, HCV incidence was 23.7 per 100 person-years (95%CI: 15.9-31.6) compared to 11.8 per 100 person-years (95%CI: 8.97-14.63) among those not reporting fentanyl use (Figure 1). Fentanyl use, both at baseline and as a time-varying predictor (lagged by 12 months) was significantly associated with HCV seroconversion. As a time-varying predictor, fentanyl use had a univariable hazard ratio (HR) of 1.68 (95%CI: 1.12-2.53) (Figure 2) and remained independently associated with HCV seroconversion after controlling for receptive needle sharing, sex, and age (adjHR1.54; 95%CI:1.01-2.33). Fentanyl use at baseline was also significantly associated with HCV seroconversion, with a similar magnitude to the time-varying variable (see supplemental table 5). Adjusting for city of residence or cross-border mobility did not appreciably affect parameter estimates.

• Download figure
• Open in new tab

Figure 1. HCV Incidence Rates Stratified by Key Characteristics of Interest

HCV incidence density shown for a 36-month period per 100 person-years by key selected characteristics. Seen supplemental table 3 for exact values.

• Download figure
• Open in new tab

Figure 2. Hazard Ratios From Time Dependent Cox Models for HCV Seroconversion

Adjusted and unadjusted (univariable) model results are shown using a log scale on the x-axis. See supplemental table 4 for exact values. *predictor in the past 6 months

Individuals who practiced receptive needle sharing at baseline had an HCV incidence rate of 15.9 per 100 person-years (95%CI: 11.5-20.4) compared to 13.0 per 100 person-years among individuals who did not practice receptive needle sharing (95%CI: 9.5-16.5). In univariable models, as a time-varying predictor, receptive needle sharing was associated with a HR of 1.76 (95%CI: 1.17, 2.67), which remained significant in the multivariable model (adjHR: 1.75; 95%CI: 1.15, 2.65).

Among individuals experiencing homelessness at baseline, HCV incidence was 18.5 per 100 person-years (95%CI: 13.3-23.7) compared to 11.76 per 100-person years among those not experiencing homelessness (95%CI: 8.6-14.93). Unhoused status was associated with a univariable HR of 1.5 (95%CI: 1.0, 2.2) but its marginal association precluded us from including it in multivariable models.

Males consistently showed a higher risk of HCV seroconversion than females, with an incidence rate of 16.47 per 100 person-years (95%CI: 12.8-20.1) compared to 9.75 for females (95%CI: 5.77-13.74). Male sex had an unadjusted HR of 1.68 (95%CI: 1.05-2.68) and an adjusted hazard ratio of 1.66 (95%CI: 1.04-2.66) for HCV seroconversion. No significant differences were found for age or city of residence.

Discussion

We document a novel association between IMF use and HCV seroconversion among PWID in Tijuana-San Diego. Although IMF have drastically transformed health risks for PWID, most research has focused on increases in fatal overdoses [1]. Less is known about the effects of the broad transition to IMF on HIV and HCV transmission. Leveraging a prospective cohort design and employing street outreach techniques to reach clandestine populations, we found a robust association between IMF use and HCV seroconversion. These findings suggest that dramatic increases in IMF use in North America since 2013 may represent a key, albeit understudied driver of contemporaneous increases in HCV incidence [11]. Given the recent White House HCV elimination plan [16], our findings suggest IMF-specific risk patterns may warrant further study to guide the implementation of the nation’s renewed efforts to reduce the burden of HCV. Although the direction and magnitude of the association between IMF use and HIV seroconversion was similar to that of HCV, the small number of HIV seroconversions precluded our ability to examine this relationship controlling for other covariates.

More research is needed to understand the mechanisms that may be underpinning the relationship between IMF use and infectious disease transmission. The shorter half-life of most IMF analogues is likely playing a key role. Qualitative and ethnographic data can help us understand the impact of this shift. Compared to the longer half-life of heroin, IMF require much more frequent dosing to prevent withdrawal symptoms, which decreases individuals’ ability to work a complete shift at their place of employment, sleep through the night, or conduct other aspects of life without continuous interruptions to use IMF [4]. This is also a frequently stated reason why PWUD seek to augment IMF with other substances, such as methamphetamine, benzodiazepines, non-IMF opioids such as nitazenes, cannabinoids, and xylazine to extend the subjective duration of effect [4,17]. These properties may account for the strong association between IMF and increased injection frequency, as well as syringe sharing that has been reported previously [5–7]. However, IMF remained significantly associated with HCV seroconversion even after accounting for receptive syringe sharing, suggesting that there may be additional effects warranting consideration in understanding the causal mechanism. It is possible that IMF’s overall destabilizing effect on the lives of PWID and associated polysubstance use may induce turbulent life circumstances, decreasing their engagement in health and preventative services and predisposing them to behaviors that increase their vulnerability to HCV infection.

An important contextual factor is the extreme vulnerability of PWID in the study region, especially in Tijuana where HCV prevalence among PWID is higher than in San Diego[18]. Social vulnerabilities represent critical drivers of HCV and HIV risk[19]. Previous studies have shown that a critical factor driving PWID to share syringes and preparation materials is a limited ability to carry paraphernalia without risking police surveillance, violence, and incarceration [20]. Municipal police have historically created challenging circumstances in Tijuana that limit PWID in their ability to carry sterile syringes [20,21]. Similar problems have also been noted in parts of California [22]. There is also a dearth of harm reduction and other services for PWID available in Tijuana, especially given recent withdrawals of governmental funding for the civil sector [21]. Harm reduction services in San Diego were also relatively limited earlier in the study period, due to the COVID-19 pandemic [28]; however, we found no difference in the weekly rates of interviews conducted inside vs. outside COVID-19 restrictive periods.

The relationship between IMF and HCV raises important implications for treatment and prevention efforts among PWID to prevent further infectious disease transmission. For instance, tailoring dosing of medications for opioid use disorder (MOUD), to respond to the unique pressures of IMF, will be important for reducing HCV transmission risks, as MOUD is known to reduce transmission risk[23]. The lipophilic nature of IMF has complicated buprenorphine induction, making precipitated withdrawal much more likely. However, novel approaches including microinduction (small buprenorphine doses scaled up as IMF are tapered), macroinduction (large starting doses of buprenorphine), and bridging to buprenorphine with short-acting opioids, all offer promising avenues to stabilize patients withdrawing from IMF [24–26]. Treatment modalities for people using IMF will require flexibility to promote MOUD uptake and adherence. For instance, mobile treatment units may connect PWID to healthcare and facilitate MOUD initiation and HCV and HIV testing[27]. Telehealth approaches may also be helpful for PWID who own electronic devices but find travelling for care difficult[28]. Long-acting MOUD also represent a powerful option to increase retention and decrease the negative health effects of return to IMF use [29]. In the inpatient setting, adequate pain control (using short-acting opioids) is increasingly required to manage pain and withdrawal symptoms related to IMF use and prevent against-medical-advice (AMA) discharges, which are currently on the rise [24,30]. Other kinds of infections affecting PWUD—such as skin and soft tissue infections—have also increased in recent years, which requires further study for treatment optimization in the fentanyl era [31,32]. Finally, the rise in IMF adds further evidence of the need to increase point-of-care testing and HCV treatment among PWUD. Even despite ongoing substance use, re-infection rates among PWID have been shown to be uncommon [33], and HCV treatment is cost-effective [34].

The IMF-HCV relationship also reinforces the need for increasing preventative and primary care services among PWUD. Research comparing the transmission of HCV and HIV suggests that HCV is the first infection to be acquired once young people start injecting, and there is generally a brief window of opportunity to prevent subsequent HIV infection [35]. Reaching individuals in this window can be accomplished via increasing access to low-barrier services that PWID need to remain safe (such as sterile syringes) alongside HIV and HCV testing [36]. On the West Coast of the U.S., there are also early signs of a broad transition from injecting to smoking IMF, which raises important considerations for blood-borne disease prevention efforts [37,38]. This shift may be helpful for reducing HCV transmission if smoking supplies are less likely than syringes to transfer HCV, although evidence on this topic is mixed [39]. Regardless, investing in safe smoking supplies is warranted to reduce the need for sharing equipment among the growing segment of PWUD smoking IMF.

This study has several limitations. Although the HCV and HIV assays employed here were highly sensitive and specific [13,14], the reliance on self-reported fentanyl use represents a limitation. Some PWUD may not know they were using IMF, given limited availability of drug checking services during the study period[40]. This could have caused misclassification, with a tendency to attenuate associations between IMF and HIV/HCV seroconversion towards the null. Follow-up studies using biological markers of fentanyl use should be conducted for confirmation. Additionally, given the known shift from injecting to smoking fentanyl occurring during the follow-up period [38], statistical power to detect an association between injection of IMF and HCV seroconversion may have been more limited. Additionally, we did not have a sufficient sample size of HIV seroconverters to assess the link between IMF and HIV, although this remains an important area for investigation. The results of this study apply to a highly vulnerable population of PWID on the US-Mexico border, and further study is needed to assess the generalizability to the wider North American context of IMF use.

Conclusions

The broad shift among PWUD to using synthetic opioids such as IMF represents a profound change for the health risks of a broad swath of vulnerable individuals in North America. Although most research has focused on the implications for fatal overdose rates, we provide novel evidence that IMF use is associated with increased HCV incidence. This raises important considerations for treatment and prevention efforts, including access to point-of-care HCV testing and treatment, MOUD modalities and dosing, provision of harm reduction supplies and interventions to prevent subsequent HIV acquisition. More research is needed to understand potential mechanisms underpinning this association, as well as the role of IMF use on other infectious diseases, such as HIV and soft-tissue infections.