Biomedical HIV prevention including pre-exposure prophylaxis and opiate agonist therapy for women who inject drugs: State of research and future directions

Abstract

Women who inject drugs are at higher risk of HIV compared to their male counterparts as a result of multiple factors including biological, behavioral and socio-structural, yet comparatively little effort has been invested in testing and delivering prevention methods that directly target this group. In this paper, we discuss the need for expanded prevention interventions for women who inject drugs, focusing on two safe, effective, and approved, yet underutilized biomedical prevention methods: opiate agonist therapy (OAT) and oral pre-exposure prophylaxis (PrEP). While both interventions are well researched they have not been well examined in the context of gender. We discuss the drivers of women injectors’ higher HIV risk, review the effectiveness of OAT and PrEP interventions among women, and explain why these new HIV prevention tools should be prioritized for women who inject drugs. There is substantial potential for impact of OAT and PrEP programs for women who inject drugs in the context of broader gender-responsive HIV prevention initiatives. While awaiting efficacy data on other biomedical approaches in the HIV prevention research ‘pipeline’, we propose that the scale up and implementation of these proven, safe, and effective interventions are needed now.

INTRODUCTION

Preventing HIV infection remains a challenge thirty-two years after the recognition of the virus. New effective prevention strategies are needed, and the implementation of strategies that have been shown to be safe and effective must be accelerated. Globally, injection drug use remains a key route of HIV transmission, and in areas where HIV is driven by injection drug use, women who inject drugs (WWID) play an increasingly important role in the growth and evolution of the HIV epidemic (1–3). Compared to men who inject, WWID have a modest but measurable increased risk of HIV infection (1). This increased risk is likely the result of multiple biological, behavioral and socio-structural factors, yet comparatively little effort has been invested in the evaluation and delivery of effective HIV prevention methods directly targeting this group. In this paper we discuss the need for expanded prevention interventions for women who inject drugs, focusing on two safe, effective, and approved, yet underutilized biomedical prevention methods: opiate agonist therapy (OAT) and oral pre-exposure prophylaxis (PrEP). While both interventions are well researched, they have not been well examined in the context of gender.

We reviewed published data on the efficacy and effectiveness of OAT and PrEP interventions among women and their potential impact on HIV-associated risk behaviors and incidence in this population and examine the reasons why these new HIV prevention tools should be prioritized for women who inject drugs.

Despite widespread knowledge of how HIV is transmitted, as well as inexpensive and effective methods to prevent HIV infection, 2.1 million people worldwide were estimated to have contracted HIV in 2013 (4). Injection drug use contributes to a significant fraction of these new HIV infections; the World Health Organization (WHO) estimates that on average at least 10% of new infections globally result from injection drug use (5), more than 200,000 new HIV infections a year. In 2008, it was estimated that just over two-thirds of the estimated 3 million HIV-infected people who inject drugs (PWID) were in Eastern Europe, East and Southeast Asia, and Latin America (6). These numbers likely underestimate the burden of HIV infections caused by drug use because injecting populations are difficult to enumerate, and HIV surveillance in PWID is less systematic than among other risk groups (6, 7). We are not aware of any global reviews of HIV among PWID that stratify by gender. One study examining trends in the population prevalence of PWID across 96 U.S. metropolitan service areas between 1992 and 2007 examined prevalence of WWID (8). The 2007 estimate of HIV prevalence among WWID (per 10,000) was 74.3 in comparison with an estimated prevalence among men of 132.94, for a male: female ratio of 1.78 to 1. More recently, Lansky et al.(9) estimated that the population proportion of WWID (any lifetime use) in the U.S. was 1.6% (95% CI 1.1, 2.0%) corresponding to 2,059,709 women (range: 1,513,969–2,606,450). In comparison, estimates for men were more than double those for women: prevalence 3.6% (95% CI 2.4, 4.8%), corresponding to 4,532,348 men who had ever injected (range 3,040,447–6,024,250). If we assume that women make up between one-third and a half of all HIV infections among PWID worldwide, then over one million are living with HIV and 82,000 to 125,000 become newly infected each year (based on 2013 estimates). UNAIDS, WHO, the World Bank and other systematic reviews that discuss HIV risk associated with injection drug use do not report gender-specific estimates of population size, and discussion of women’s risk is strictly in terms of their childbearing age, or sexual risk in association with heterosexual sex or transactional sex (4, 6).

Due to concomitant sexual and injecting exposures, PWID are at incrementally higher risk of HIV than other key populations. In a review of 117 studies conducted in 14 countries, Des Jarlais et al., (1) found that female injectors had higher odds of HIV infection compared to male injectors (OR 1.18; 95% CI 1.10, 1.26). Studies have shown that female injectors are more likely to have higher risk behaviors, including sexual and injection risks, than their male counterparts (10–13). Some WWID engage in sex work (14–17), potentially exposing themselves to other sexually transmitted diseases (18), and increasing their risk for HIV infection. Women in overlapping injecting and sexual relationships often have little or no control over partners’ condom use and injecting equipment (19, 20). Other key issues include access to clean needles and syringes, which remains a challenge and is mired in political and moral debates in many parts of the world, including in the U.S. (21). The Joint United Nations Programme on HIV/AIDS (UNAIDS) reports that only “two of 32 reporting countries provide the recommended minimum of at least 200 sterile syringes per year for each person who injects drugs.” (4). While condoms are easily available and widely promoted, women have limited power to control their use. The disproportionate sexual and injection risks experienced by WWID are further exacerbated by stigma/discrimination, and violence (19, 22–24). Women are more likely to conceal their injection risk and as a result, are harder to reach, count and engage in prevention programs(22). Drug users, including both injection and non-injection users, may not be prioritized as a key population in some areas, and women in particular may fall through the HIV prevention ‘cracks’. In many locations, sex work and injection drug use are co-occurring exposures, and prevention programs for these groups need to consider these intersecting risks. Reliable surveillance data on PWID, in general, and WWID in particular is lacking in many settings due to legal barriers preventing access to these populations.

In addition to condom distribution and needle and syringe programs, the current “combination” HIV prevention package proposed for PWID by the WHO includes HIV counseling and testing, linkage to HIV care, access to opiate agonist therapy (OAT), and evidence based psychosocial interventions (25, 26). With persistently high rates of new infection, ongoing risk and limited access to prevention, WWID need access to effective HIV prevention technologies. Biomedical HIV prevention approaches, defined as “strategies that use medical and public health approaches to block infection, decrease infectiousness, and reduce susceptibility” (27), are now available and should be included as a part of combination prevention approaches. Currently, “biomedical HIV prevention” includes effective biomedical strategies including diagnosis and treatment of sexually transmitted infections, male circumcision, pharmacologic therapy for substance abuse treatment (including OAT), and daily oral PrEP (28–30). For women, vaccines and topical protections, including microbicides and vaginal gels such as 1% tenofovir (topical PrEP) are high priority research areas (31). Vaccines remain elusive, and topical 1% vaginal tenofovir, used before and after sex as PrEP, has shown promise in one study but requires confirmatory evidence of efficacy for regulatory approval (53,76,77). Research into long-acting injectable PrEP is progressing, but this research is in the early stages (32, 33). Two important HIV prevention tools that deserve more attention and expanded availability are OAT and PrEP. Both have been shown to be safe and effective and are approved for use – yet both are underutilized.

OAT and HIV prevention in women who inject drugs

Untreated substance use disorders (particularly opioid use disorders) are a major source of HIV transmission, particularly in developing countries (34). Injection drug use is frequently a consequence of untreated opiate addiction, for which there are effective pharmacotherapies. Methadone and buprenorphine have been shown to be effective in reducing opiate use (35), reducing injecting and sexual risk behaviors (36), and preventing HCV and HIV (37–41). Both methadone and buprenorphine have opioid agonist properties, which reduce craving for illicit opiate use. However, buprenorphine, which was approved for use by the U.S. Food and Drug Administration in 2002, is a partial mu-opioid receptor agonist, and therefore has a ceiling that limits side effects such as sedation and respiratory depression. Maintenance therapy with long-acting oral medications that reduce craving can break the cycle of intoxication and withdrawal from heroin use and ongoing injection drug use. Longer courses of OAT in general have been associated with better health and social outcomes (42).

Both methadone and buprenorphine have been demonstrated in randomized controlled trials to lead to reductions in illicit opioid use. A 2009 Cochrane review found 11 randomized trials of methadone maintenance compared to non-pharmacological treatment for opioid dependence. Overall, methadone maintenance was shown to be associated with a 44% (RR 0.66; 95% CI 0.56, 0.78) improvement in retention and suppression of heroin use (measured by self-report and biomarkers) (43). A subsequent review of methadone and buprenorphine maintenance therapy which included 31 trials, showed that both of these pharmacological interventions were significantly associated with reductions in illicit opioid use as measured by urine drug tests (37). Additionally, risk behaviors associated with HIV declined, including the number of participants reporting injection drug use, frequency of injection, and sharing of equipment (although these effects were difficult to disentangle from the effects of decreased injecting overall) (38). The reductions seen in transactional sex, including exchanging sex for money or drugs, with OAT (38) may be due to a reduction of the need for income to purchase illicit drugs.

A recent systematic review and meta-analysis of data on 819 incident HIV infections over 23,608 person-years (py) of observation found that OAT was associated with a 54% reduction in the risk of HIV infection among PWID (RR=0.46; 95% CI: 0.32–0.67; p<0.0001). No gender-specific estimates were provided, and most studies included had modest sample sizes with the majority of participants being male, consistent with a male predominance among PWID (8, 9). OAT has also been shown to be effective in preventing HCV (39–41). WWID report higher rates than men for both injecting and sexual HIV risk behaviors (10–12). Thus, the risk of infection with blood-borne viruses including HIV and HCV is likely increased among WWID (1, 12, 44).

Access and retention in substance use treatment that includes OAT should be a priority HIV prevention area for WWID. Women are disproportionately under-represented in inpatient and outpatient substance abuse treatment programs, suggesting a disparity in access to OAT (45), although this may not be true in all settings (46). Women may delay presentation for treatment. For example, a national multi-site study conducted in the U.S. found that among opioid-dependent patients presenting for treatment, women not only demonstrated a more severe clinical profile than men did, as evidenced by higher Addiction Severity Index scores in multiple domains, but also reported higher opioid craving (47) and comorbid mental health diagnoses (48). The data on retention in OAT by gender is mixed; the U.S. Substance Abuse and Mental Health Administration reports that gender is not a good predictor of retention, in part because of the complex interaction of multiple demographic, psychosocial, psychiatric, and health-status factors (49). It should be recognized that not all WWID use opioids, and treatment strategies for other drug dependencies are also urgently needed. To date, pharmacotherapy for non-opiate drug dependence has shown some promise but no evidence of effectiveness in clinical trials (50). Furthermore, the importance of social, cultural and personal barriers in preventing women from accessing drug treatment in general are substantial and should not be overlooked(51). Stigma, fear of inter-partner violence or abandonment, and fear of loss of custody of children are all well-documented barriers that may impede WWID from accessing drug treatment, including OAT (52–54).

Pre-Exposure prophylaxis (PrEP) in women who inject drugs

Seven large randomized controlled trials have been conducted assessing HIV PrEP (28, 55–60). Four have provided compelling evidence that when used, it prevents or substantially reduces the risk of HIV acquisition among women (56, 57, 59, 61). Two of the trials had sufficient data to assess the impact of PrEP on women and showed statistically significant decreases in the risk of HIV infection (54, 55). Trials of oral PrEP have shown reductions in HIV incidence ranging from 49.4% to 78.6% (57, 59, 61). In the CAPRISA study, topical PrEP reduced HIV risk by 38.5% among women (56); however, two other large randomized trials, VOICE (62) and FACTS 001 (63) did not show a reduction in HIV risk among women using topical PrEP. Collectively, these studies enrolled a geographically diverse group of women in different risk settings. The Bangkok Tenofovir Study (BTS) (61) specifically enrolled PWID, and included women. Strong dose–response relationships have been shown between adherence to PrEP pill-taking (52,54,55) or gel-use (53) and HIV protection in the studies where PrEP was effective, whilst no protective effect was demonstrated in studies with overall low study drug adherence (60, 64). Based on the results of these trials, the U.S. Centers for Disease Control and Prevention (CDC) has issued guidance on the use of PrEP to limit sexual and parenteral HIV transmission (65, 66).

The BTS assessed the safety and efficacy of oral PrEP in HIV-uninfected men and women, aged 20 to 60 years, who reported injecting drugs in the year before enrollment; trial results and a review of participant risk behavior have been published (61, 67). The trial was conducted at 17 Bangkok Metropolitan Administration (BMA) drug treatment clinics. Participants were not required to have attended the BMA clinics prior to the study. Thailand’s narcotics law prohibits the distribution of needles to inject illicit drugs and needles are not provided in the clinics. However, sterile needles and syringes are available to the public over the counter at low cost in pharmacies in Bangkok. A total of 2413 participants were randomly assigned in a 1:1 ratio to receive daily oral tenofovir 300 mg or placebo.

We analyzed BTS data limited to women participants using chi-square for baseline data and generalized estimating equations. Table 1 provides the baseline characteristics of the 489 WWID enrolled in the BTS trial: median age 30 (Interquartile range [IQR] 26, 36) years, 260 (53.2%) with primary (6 years) education or less, and 101 (20.7%) with self-report of being in police custody in the 12 weeks prior to enrollment. There were no differences between the group receiving Tenofovir compared to placebo with respect to reported drug use or sexual behavior risks, with 62 (12.7%) women enrolled in methadone programs, and over half (54.4%, n=266) reported injecting drugs in the 12 weeks before enrollment. Women were followed for a total of 1990 person-years (py) (maximum 6.9 years). Among the 489 women, the 11 who became HIV infected included 9 in the placebo group (incidence, 0.92/100 py) and 2 in the Tenofovir group (incidence, 0.20/100 py) indicating a 78.6% reduction in risk (95% CI, 16.8, 96.7; p=0.03). This protective effect was almost twice that seen in men (37.6%), though the 95% CIs overlap. The number and proportion of women reporting injection drug use, and needle sharing during the previous 3 months decreased significantly during follow-up (Table 2). Reported sexual risk behaviors also decreased, including the proportion reporting sex with more than one partner during the previous 3 months and sex with casual partners (Table 2). Controlling for age, adherence (average/mean of days based on study diaries) was better in women (median 95.6%, IQR 81.1, 98.9) compared with men (median 93.8%; IQR 78.8, 98.7; p=0.04). In summary, these results support the efficacy of PrEP in WWID. Along with a good safety profile, the protective effect of PrEP was high, and participant reports of drug use and needle sharing decreased, demonstrating that with appropriate risk-reduction counseling and access to methadone maintenance support, risk compensation can be limited or avoided. All of these findings suggest that PrEP should be scaled up among WWID.

Table 1.

Baseline characteristics of women participating in the Bangkok Tenofovir Study, 2005–2012 (n=489).

Characteristics	Tenofovir N=246	Placebo N=243	Total N=489	p-value
Age
Median (IQR*)	30 (26–35)	30 (26–38)	30 (26–36)	0.18
Education level
Primary or less (<6 years)	127 (51.6)	133 (54.7)	260 (53.2)
Secondary or more	119 (52.0)	110 (48.0)	229 (46.8)	0.49
Risk factors
Incarceration
In police holding cell past 12 weeks	52 (21.1)	49 (20.2)	101 (20.7)	0.79
In prison in the past 12 weeks	37 (15.0)	30 (12.4)	67 (13.7)	0.39
Drug use
Currently in methadone program	28 (11.4)	34 (14.0)	62 (12.7)	0.39
Injected drugs in past 12 weeks	132 (53.7)	134 (55.1)	266 (54.4)	0.74
Heroin	28 (11.4)	32 (13.2)	60 (12.3)	0.55
Methamphetamine	87 (35.4)	77 (31.7)	164 (33.5)	0.39
Midazolam	35 (14.2)	33 (13.6)	68 (13.9)	0.84
Other	9 (3.7)	15 (6.2)	24 (4.9)	0.20
Injection frequency in past 12 weeks
Every day	16 (6.5)	11 (4.5)	27 (5.5)	0.34
Every week	45 (18.3)	49 (20.2)	94 (19.2)	0.60
Less frequent than every week	71 (28.9)	74 (30.5)	145 (29.7)	0.70
Shared needles in past 12 weeks	38 (15.5)	40 (16.5)	78(16.0)	0.76
Sexual behaviors
Number of opposite sex sexual partners in past 12 weeks
0	79 (32.1)	67 (27.6)	146 (29.9)	0.27
≥1	167 (67.9)	176 (72.4)	343 (70.1)	0.27
Reported sexual intercourse with live-in partner in past 12 weeks	134 (54.5)	142 (58.4)	276 (56.4)	0.38
Reported sexual intercourse with casual partner in past 12 weeks	48 (19.5)	58 (23.9)	106 (21.7)	0.24

Data are n/N (%) or n (%);

^*Interquartile range (IQR)

Table 2.

Risk behaviors reported by female participants in the Bangkok Tenofovir Study by study visit, Thailand, 2005–2012^a.

	Enrollment n=489 n (%)	Month 12 n=392 n (%)	Month 24 n=358 n (%)	Month 36 n=317 n (%)	Month 48 n=275 n(%)	Month 60 n=213 n (%)	Month 72 n=147 n (%)	P value
Drug injection practices in past 3 months
Injected drugs	266 (54.4)	49 (12.5)	44 (12.3)	30 (9.5)	26 (9.5)	17 (8.0)	12 (8.2)	<0.001
Shared needles	78 (16.0)	7 (1.8)	7 (2.0)	3 (1.0)	1 (0.4)	1 (0.5)	1 (0.7)	<0.001
Injected daily	27 (5.5)	8 (2.0)	11 (3.1)	8 (2.5)	6 (2.2)	8 (3.8)	8 (5.4)	0.85
Drugs injected in past 3 months
Methamphetamine	164 (33.5)	20 (5.1)	14 (3.9)	13 (4.1)	9 (3.3)	7 (3.3)	2 (1.4)	<0.001
Midazolam	68 (13.9)	32 (8.2)	23 (6.4)	21 (6.6)	16 (5.8)	11 (5.2)	11 (7.5)	<0.001
Heroin	60 (12.3)	12 (3.1)	8 (2.2)	9 (2.8)	6 (2.2)	6 (2.8)	3 (2.0)	<0.001
Incarceration in past 3 months
In jail	101 (20.7)	32 (8.2)	29 (8.1)	29 (9.2)	21 (7.6)	23 (10.8)	9 (6.1)	0.005
In prison	67 (13.7)	33 (8.4)	28 (7.8)	35 (11.0)	22 (8.0)	25 (11.7)	8 (5.4)	0.71
Sexual activity in past 3 months
More than one sexual partner	44 (9.0)	17 (4.3)	15 (4.2)	15 (4.7)	15 (5.5)	10 (4.7)	3 (2.0)	0.04
Sex with casual partner	106 (21.7)	46 (11.7)	50 (14.0)	37 (11.7)	24 (8.7)	22 (10.3)	12 (8.2)	<0.001

^aRisk behaviors reported at annual visits are shown in the table; data from all 3-monthly visits through month 72 were used in the generalized estimating equations logistic regression analysis.

DISCUSSION/CONCLUSIONS

Several commentaries have been published about the need to integrate HIV biomedical prevention into practice (30, 68, 69), particularly with interventions for high-risk women (29). These discussions are important because they note the difficulties associated with behavior change and the limitations of current mainstream (non-biomedical) prevention approaches. However, few reviews address the needs of PWID, and little attention is given to WWID despite their high risk of HIV infection (23, 70). There is an important need to better quantify the population of WWID globally. Gender-specific population estimates could greatly inform planning and HIV prevention activities. While there is agreement within the HIV prevention field on the importance of implementing HIV biomedical interventions as part of a ‘combined’ and strengthened response to the HIV epidemic that includes integration of behavioral, structural and community-based efforts (68, 71–74), there has been limited focus on the needs of women who inject drugs. The CDC has recommended PrEP use for PWID and has issued guidance (66); however, the WHO has not (25). Globally, it is widely acknowledged that existing interventions for PWID, including needle and syringe programs and OAT, need to be scaled up (21). Despite support for OAT by many countries there is a significant unmet need for this intervention, with an estimated 90% of opiate users lacking access(4). A recent review, Degenhardt et al., (75) notes progress in access to OAT in China and some Southeast Asian countries, but little progress in the U.S. or Russia, two countries with the largest numbers of PWID. There are also gaps in needle and syringe coverage, access to antiretroviral therapy, and ongoing adverse effects of stigmatization and incarceration of PWID. In summary, OAT has been demonstrated to be effective in preventing HIV among men and women who inject drugs. WWID may be at increased risk for acquiring HIV because of higher levels of risk taking associated with sexual activity and drug injection; therefore, OAT may be particularly effective among women, as it has been linked to both improved sex and drug risk behaviors. However, gender-specific estimates of the impact of OAT on HIV incidence are not known. Further research is needed to assess barriers women face accessing and consistently using OAT. There is also a significant need to address non-opioid injection drug use among women.

The results of the BTS study provide important information for WWID. Although the study was not powered to assess efficacy by subgroup, the results among women showed high levels of protection against incident HIV infection and significant reductions in drug use. These combined effects are important. The availability of directly observed therapy (DOT) delivery likely facilitated adherence, and ongoing counseling and access to OAT likely contributed to reduced risk behaviors. Although the study did not supply syringes and needles, participants could purchase them at low cost without a prescription from pharmacies in Bangkok. These factors suggest that oral PrEP, currently recommended as daily FTC/TDF (61), can have a substantial impact on the risk of HIV infection among women who inject.

The evidence reviewed here supports the expansion of programs to make OAT and PrEP more accessible to WWID in the context of HIV prevention using gender-responsive approaches to enhance impact. It is well recognized that women benefit from women-centered services. Several “gender-based” HIV prevention interventions focused on women have been successfully developed with the aim of reducing sexual risks in women, including increasing condom use and reducing sex client and partner violence, promoting gender empowerment, community mobilization and micro-enterprise (76, 77). Gender-specific services, including ‘women only’ needle and syringe programs (60) have been shown to appeal to WWID. Gender-based approaches can reduce HIV incidence, are cost-effective, and, in some cases cost-saving (78). The U.S.-based Center for Substance Abuse and Mental Health Services Administration has published a report assessing the needs of women in drug treatment that is a useful guide for gender-based biomedical HIV prevention (49), and the United Nations Development Programme has issued new guidance stressing the value and need for synergistic investment in integrating gender-responsive HIV programming (79). While the latter report (79) is more focused on promoting gender equality and violence reduction in general, than on specific HIV prevention strategies among women, it is encouraging that UN agencies are recognizing the importance of gender-specific programming. In a recent commentary on scaling up PrEP, Katz (69) asks pertinent questions regarding ‘who’ and ‘how many’ receive PrEP. WWID represent a highly impacted and large proportion of the population at risk for HIV that is often overlooked. There are important opportunities to access this hard-to-reach population, including working with community-based groups, syringe and needle exchange programs, OAT programs, and possibly HCV treatment clinics. Providers, including clinicians providing primary care and reproductive health services, and prevention educators should be aware of the findings and recommendations of these two important prevention approaches (62). Further evidence including modeling and cost-effectiveness studies of the impact of expanded access to OAT and PrEP is needed to inform policy-makers and HIV programs globally. Programs designed to provide social and behavioral support for women using OAT and PrEP are needed to maintain good adherence and retention. Considerable energy is currently being focused on new technologies for women at risk of HIV (53,76,77). However, it is important to not lose sight of the fact that two highly effective interventions for preventing HIV (OAT and PrEP) exist. Successfully implementing these safe and effective HIV prevention tools will limit new HIV infections among women who inject drugs and help inform subsequent HIV prevention efforts.