Abstract
Background
Potential bidirectional drug–drug interactions between feminizing hormone therapy (FHT) and antiretroviral therapy (ART) are of concern for trans women with HIV and their healthcare providers. This study aimed to characterize patterns of FHT and ART among trans women with HIV and to compare serum hormone levels to trans women without HIV.
Methods
Charts of trans women were reviewed at seven HIV primary care or endocrinology clinics in Toronto and Montreal from 2018 to 2019. ART regimens, FHT use, serum estradiol, and serum testosterone levels were compared on the basis of HIV status (positive, negative, missing/unknown).
Results
Of 1495 trans women, there were 86 trans women with HIV, of whom 79 (91.8%) were on ART. ART regimens were most commonly integrase inhibitor-based (67.4%), many boosted with ritonavir or cobicistat (45.3%). Fewer (71.8%) trans women with HIV were prescribed FHT, compared to those without HIV (88.4%) and those with missing/unknown status (90.2%, p < 0.001). Among trans women on FHT with recorded serum estradiol (n = 1153), there was no statistical difference in serum estradiol between those with HIV (median: 203 pmol/L, IQR: 95.5, 417.5) and those with negative (200 mol/L [113, 407]) or missing/unknown HIV status (227 pmol/L [127.5, 384.5) (p = 0.633). Serum testosterone concentrations were also similar between groups.
Conclusions
In this cohort, trans women with HIV were prescribed FHT less often than trans women with negative or unknown HIV status. There was no difference in serum estradiol or testosterone levels of trans women on FHT regardless of HIV status, providing reassurance regarding potential drug–drug interactions between FHT and ART.
Introduction
Transgender (trans) women are disproportionally affected by HIV worldwide. Meta-analyses have found trans women have 49 times the odds of acquiring HIV compared to the general population globally, and in pooled US studies the HIV prevalence was 14.11% for trans women1,2 – staggeringly high compared to the national US HIV prevalence of 0.36% 3 and the Canadian HIV prevalence of 0.17%. 4 Thus, engaging and retaining trans women in HIV care is crucially important. However, current evidence suggests that, compared to cisgender (cis) women and men living with HIV, trans women with HIV have lower rates of linkage to care, retention in care, antiretroviral therapy (ART) initiation, ART adherence, and virologic suppression.5–12
Many trans women use feminizing hormone therapy (FHT) to develop physical and emotional changes that affirm their gender. FHT often consists of an oestrogen and an anti-androgen, and is widely accepted as the standard of care for trans women. 13 However, many trans women living with HIV and their medical providers harbour concern that ART and FHT may have clinically significant drug–drug interactions,14,15 on the basis of known ART interactions with oral contraceptives among cis women.16,17 This concern may lead some providers to discourage trans women living with HIV from taking FHT.14,18,19 In response, some trans women prioritize FHT over ART and take their ART and/or FHT differently than prescribed; or seek hormone therapy outside the medical system. 20 In many cases, trans women may not discuss their concerns nor their partial ART adherence with their HIV care providers.10,21 By contrast, co-prescribed FHT with ART is associated with improved engagement in care and virologic suppression. 21
To determine the extent that these tensions affect ART and FHT co-administration among trans women accessing clinical care in two urban cities in Canada, we undertook a retrospective chart review of trans women living with and without HIV to characterize both ART and FHT prescribing practices and the attained serum estradiol and testosterone concentrations. We hypothesized that women living with HIV would access FHT less to their HIV negative peers and may have lower serum estradiol concentrations because of drug–drug interactions or caution in prescribing.
Methods
The Montreal Toronto Trans Study (MTTS) collected sociodemographic and clinical data of trans women aged 16 years and older, to answer questions about social determinants of health, the HIV care continuum, and trans health outcomes. Data were collected in 2018–2019 from electronic medical records (EMR) from one HIV primary care clinic in Montreal, Canada in addition to five HIV primary care clinics and one endocrinology clinic in Toronto, Canada. Some clinics are solo practitioners, whereas others comprise multi-disciplinary teams (with nursing and counselling). Each clinical site created a list of their trans women patients on the basis of their standard EMR documentation procedures and confirmation from relevant care providers. Standardized data collection forms were developed by the team, piloted at two clinics, and then refined. Clinic and linking logs were used to identify and merge duplicate patient records in Toronto. Study data were collected and managed using REDCap electronic data capture tools, a secure, web-based software platform designed to support data capture, auditing, and data integration for research studies.22,23
Charts were retrospectively reviewed for sociodemographic variables (e.g., age), HIV status (positive, negative, unknown/missing), FHT use (yes versus no), current FHT regimen (type of anti-androgen and oestrogen), serum estradiol levels in pmoL/L, and serum testosterone levels in nmol/L. For trans women living with HIV, most recent CD4 count, most recent HIV viral load (classified as suppressed if <200 copies/mL or unsuppressed if ≥200 copies/mL), and current ART regimen (characterized by ART class, and whether unboosted versus boosted with ritonavir or cobicistat) were collected. Current ART and FHT were defined as the most recently prescribed regimens, unless discontinuation was explicitly noted in the chart. Groups on the basis of HIV status (positive, negative, unknown/missing) were compared using chi-square or Fisher’s exact tests for categorical variables (e.g., FHT use, current oestrogen product, and current anti-androgen regimen) and Kruskal–Wallis test for continuous variables (e.g., age, last serum estradiol level). The study was approved by the Women’s College Hospital Ethics Review Board at Women’s College Hospital, Toronto, Canada (#2018–0140-E) and the University of Toronto, Toronto, Canada (#36003).
Results
Demographic and clinical characteristics of trans women in the Montreal Toronto Trans Study (n = 1495).
***statistically significant difference at p < 0.001; computed using Kruskal–Wallis (age) or chi-square/Fisher’s exact (race/ethnicity, legal status).
aStatistically significant between HIV status unknown/missing and HIV positive and HIV negative and HIV positive subgroups.
bOther includes chart-specified ethnicities (e.g., Filipino, Middle Eastern, Muslim, etc.).
cIncludes two undocumented persons.
Current ART use among trans women living with HIV in the Montreal Toronto Trans Study (n = 86).
ART, antiretroviral therapy; INSTI, integrase strand transfer inhibitor; NRTI, nucleoside reverse transcriptase inhibitor, NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor.
Current FHT, type of anti-androgen, type of oestrogen, and estradiol concentration by HIV status among trans in the Montreal Toronto Trans Study (n = 1491). a
*, **, ***Statistically significant difference at p < 0.05, p < 0.01, p < 0.001, respectively; computed using Kruskal–Wallis (serum estradiol) or chi-square/Fisher’s exact (all other variables).
FHT, feminizing hormone therapy.
an = 1491 on the basis of those with complete data on FHT use (yes/no).
Of the trans women on FHT with recorded serum estradiol (n = 1153), there was no statistical difference in serum estradiol between those with HIV (median: 203 pmol/L, IQR: 95.5, 417.5) and those with negative HIV status (200 pmol/L [113, 407]) or those with missing/unknown HIV status (227 pmol/L [127.5, 384.5) (p = 0.633). There was also no statistically significant difference in serum testosterone between those with HIV (median: 0.51 nmol/L, IQR: 0.20, 11.50) and those with HIV negative status (0.70 nmol/L [IQR: 0.30, 4.40] or those with missing/unknown HIV status (0.80 [IQR: 0.40, 5.00]) (p = 0.289).
Discussion
In our study, we found that trans women living with HIV were less frequently prescribed FHT compared to trans women without HIV or of unknown status. Although FHT is considered the standard of care for trans women, the decision of whether and how to take FHT is tailored to the needs, health risks, and social context of each individual. Our analysis does not address whether these differences relate to patient factors independent of HIV status. Trans women with HIV in our study represented a relatively treatment-engaged and treatment-adherent cohort with markers of preserved immune function that were mostly prescribed INSTI-based ART regimens. Serum estradiol and testosterone measurements were not different between trans women living with or without HIV or of unknown HIV status.
Although the trans women with HIV in this study had high rates of clinical care engagement, a notable fraction were not prescribed ART or did not achieve virologic suppression. This analysis did not aim to determine why some women were not prescribed ART. It is possible that some women were reluctant to take ART owing to concern it would interfere with FHT, as reported elsewhere 20 ; that some clinicians declined to prescribe ART to trans women (less likely given the expertise in trans health at participating clinic sites); or that other factors such as housing instability, mental health concerns, and lower prioritization of virologic suppression stalled treatment initiation or adherence.24–26 Medication coverage may also be a barrier for individuals without legal status in Canada, or those in Ontario who are not enrolled in a publicly funded insurance program.
Clinicians may be cautious about co-prescribing ART and FHT owing to concerns of potentially compromising antiretroviral and/or gender-affirming efficacy. 14 Supratherapeutic serum estradiol could increase the risks of FHT such as venous thromboembolism, dyslipidemia, and cardiovascular disease; and lower serum estradiol concentrations could potentially lead to unsatisfactory clinical response to FHT. Evidence from oral contraceptives in cis women living with HIV has demonstrated that certain ART agents alter serum levels of ethinyl estradiol. Specifically, ritonavir and certain NNRTIs (such as efavirenz) decrease ethinyl estradiol, whereas cobicistat increases ethinyl estradiol. 17 However, differences between the metabolism of estradiol and ethinyl estradiol make it unclear whether these observations are relevant to FHT. 16 There is pharmacokinetic evidence that efavirenz, a potent pharmacokinetic inducer, can lower serum estradiol concentrations in trans women taking estradiol or estradiol valerate orally.18,19 However, the clinical importance of this drug–drug interaction is less relevant in the Canadian context because of the current limited use of efavirenz-based regimens. There have been no pharmacokinetic studies of FHT with unboosted INSTI-based ART regimens, which are recommended as first-line by the U.S. Department of Health and Human Services (DHHS) 27 and which more than a third of MTTS trans women living with HIV were taking. On the basis of expert opinion, unboosted INSTIs are not expected to have clinically significant effects on gender-affirming hormonal therapy. 16 In our cohort, an overall difference in serum estradiol among trans women with HIV was not observed.
There are also sparse data that suggest FHT can affect concentrations of some antiretrovirals, which could affect ART efficacy. 17 Pharmacokinetic studies in trans women living with HIV have demonstrated that FHT (comprising oral estradiol valerate and cyproterone) lowered serum levels of efavirenz and tenofovir, with 90% of participants nonetheless achieving virologic suppression within 12 weeks of ART initiation. 18 The same FHT regimen was shown to lower serum tenofovir levels when it was being used as part of HIV pre-exposure prophylaxis (PrEP). 18 FHT was also associated with lower concentration of tenofovir active metabolites in rectal tissue among trans women taking PrEP compared with cis men. 28 The clinical significance of these reductions in antiretroviral exposures is unclear. Studies of trans women living with HIV who were receiving FHT were more likely to be taking ART 29 and to achieve virologic suppression 21 compared to those not receiving FHT. In our chart review, the majority of trans women achieved virologic suppression.
The large number of trans women with no documented HIV test is a major limitation of our study because of the possibility of undiagnosed or unrecognized HIV in the cohort. There are many possible factors for this; the most likely being that one participating site was an endocrinology office, which would not routinely order sexual health screening, but would rely on the patient profile reported by the referring provider. In the primary care sites, there may have been some trans women who were not offered HIV testing or declined because of a lack of perceived behavioural risk factors. A respondent-driven sampling survey in Ontario found heterogeneous sexual activity and sexual orientation among transfeminine individuals: 8.3% reported no lifetime sexual activity, and 42.6% of those with lifetime sexual activity reported no sexual activity in the past year. The same study found 14.9% of participants reported no attraction to anyone, and 26.8% were attracted primarily to female-identified partners. 30 Thus, trans women with only female partners or no partners may not have requested or been offered HIV testing. In our cohort, an HIV infection unrecognized by the trans woman and her provider would not influence FHT prescribing, but some trans women could have potentially concealed their HIV status from their care providers to access FHT.
Additionally, our dataset lacked data on PrEP prescription. It is possible that those taking PrEP were included in the HIV negative subgroup, whom we have assumed are not taking ART. More robust analyses might have included those HIV positive, HIV negative on PrEP, HIV negative not on PrEP, and missing.
Our dataset also included a fraction of trans women living with HIV who did not have viral load testing on file. Some women living with HIV were also followed by an external HIV specialist and would not have HIV monitoring tests documented in their study site’s chart. By contrast, all sites had providers with expertise in gender-affirming care; therefore, we would expect that all charts reviewed would accurately represent FHT and monitoring for the participants. The familiarity of the clinicians with FHT suggests our findings may actually overrepresent rates of FHT and ART co-prescription amongst most trans women living with HIV.
The Endocrine Society guidelines on gender-affirming hormone therapy, 13 cited in the WPATH Standards of Care version 8, 31 recommend serum estradiol not to exceed 367–734 pmol/L, and the Sherbourne Guidelines used in Ontario suggest a target estradiol level of 200–500 pmol/L is clinically appropriate. 32 The serum estradiol levels across the MTTS cohort were at the lower range of local guideline recommendations. Our analysis of estradiol levels has limited precision with wide interquartile ranges. Serum estradiol concentration varies with timing of lab work relative to administration, especially for sublingual and intramuscular estradiol. However, timing of lab work is not standardized in clinical practice. This variability could confound subtle differences between the groups. Furthermore, owing to the complexity of ART drug–drug interactions, subgroup analysis of estradiol levels stratified or controlled by ART type would provide valuable granular detail. However, because of a relatively small sample size and clinician-selected ART regimens, this was not possible.
Conclusions
In this Canadian multicentre retrospective chart review, trans women living with HIV were less frequently prescribed FHT compared to trans women without HIV. Trans women with HIV were mostly prescribed INSTI-based ART regimens. Serum estradiol measurements were not different between trans women living with or without HIV or of unknown HIV status, although wide variability was observed because of non-standardized timing of measurement and different hormone therapy regimens. Further research is needed to definitively address concerns about drug–drug interactions between first-line ART and FHT and to characterize barriers to gender-affirming care, to ensure trans women living with HIV have equitable access to health and healthcare.
Footnotes
Acknowledgements
We would like to thank all the team members who offered their insight into these analyses, all of the community members involved in the Trans Women HIV Research Initiative, as well as the research coordinators and assistants, partnering clinics, and funders who made this study possible. We would also like to thank Chris Battiston, the Women’s College Hospital REDCap Database Administrator, for his time and effort with data management.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by a Canadian Institutes of Health Research (CIHR) Foundation grant (#1017493) and an unrestricted investigator-driven Gilead Sciences grant. Dr. Loutfy has research grants and honoraria from ViiV Healthcare, Merck Inc., and Gilead Sciences.
