Previous Pre-Exposure Prophylaxis Use Among Men Who Have Sex with Men Newly Diagnosed with HIV Infection in King County, Washington

Abstract

Background:

Pre-exposure prophylaxis (PrEP) discontinuations are common and are associated with subsequent HIV acquisition. The population-level impact of PrEP discontinuations is unknown.

Methods:

Public health staff routinely asked men who have sex with men (MSM) with newly diagnosed HIV infection about their history of PrEP use as part of partner notification interviews in King County, Washington from 2013–2021. We assessed trends in the proportion of MSM who ever took PrEP and described reasons for PrEP discontinuation.

Results:

A total of 1,098 MSM were newly diagnosed with HIV during the study period, of whom 797 (73%) were interviewed, and 722 responded to questions about their history of PrEP use. Ninety-four (13%) reported ever taking PrEP. The proportion of MSM who ever used PrEP before HIV diagnosis increased from 2.3% in 2014 to 26.6% in 2020–2021 (P <0.001 for trend). The median time from PrEP discontinuation to HIV diagnosis was 152 days and median duration on PrEP was 214 days. Common reasons for stopping PrEP included self-assessment as being at low risk for HIV, side effects, and insurance issues. Nineteen men were on PrEP at the time of HIV diagnosis; mutations conferring emtricitabine/tenofovir resistance were identified for 8 (53%) of 15 men with available genotype data.

Conclusion:

Over 25% of MSM with newly diagnosed HIV from 2020–2021 had ever used PrEP. Over 50% who discontinued PrEP were diagnosed <6 months after stopping. Strategies to preempt PrEP discontinuations, enhance retention, and facilitate resumption of PrEP are critical to decrease new HIV diagnoses.

INTRODUCTION

When used consistently, pre-exposure prophylaxis (PrEP) is a highly efficacious tool for HIV prevention.¹ In some areas of the U.S., PrEP use among men who have sex with men (MSM) is quite high,^2,3 and population-based studies suggest that the intervention is also highly effective in practice.⁴ However, many people who initiate PrEP discontinue it, with PrEP retention estimates from sites across the U.S. ranging from 30–60% at 12 months.^5–7 Published studies have described HIV seroconversions at rates from 0.05–3.9 per 100 person-years^8–10 after PrEP discontinuation, suggesting that in some instances the decision to stop or not resume PrEP leads to missed opportunities to prevent HIV acquisition.

The extent to which PrEP discontinuations contribute to new HIV infections at the population level is uncertain, yet this information is crucial to inform and guide public health HIV program prevention efforts. We used population-based HIV partner services data to evaluate trends in prior or ineffective PrEP use and reasons for stopping PrEP among MSM with newly diagnosed HIV in King County, Washington, from 2013–2021.

METHODS

Data Collection and Study Population

Laboratories and medical providers in Washington State are legally required to report all positive HIV test results and new HIV diagnoses to the health department. Public Health–Seattle & King County routinely attempts to interview all persons with newly diagnosed HIV infection for purposes of surveillance and partner notification.¹¹ When community PrEP use began to increase in King County in late 2013, partner services interview forms were updated to include questions about PrEP use and reasons why persons with newly diagnosed HIV may have discontinued PrEP. Partner services staff used these structured forms to guide interviews with each person shortly after HIV diagnosis and collect information on demographics, sexual behavior, substance use, and history of PrEP use prior to HIV diagnosis. These data were entered into an electronic database and analyzed for this study. Reported PrEP use strategy (i.e., intermittent vs daily use) was not consistently asked throughout the study period.

The current analysis includes data from all cases of newly diagnosed and reported HIV in adult MSM in King County from October 2013 to August 2021. We excluded newly reported cases that HIV surveillance defined as a prior HIV diagnosis (e.g., previously known positive in another state or county). Partner notification staff attempted to define PrEP start and end dates for each man reporting a history of PrEP use. If interviews were incomplete or if the month or day of PrEP discontinuation were otherwise unable to be determined, we defined these dates as June or the 15^th day of the month, respectively. Men who initiated a partner services interview and reported ever being on PrEP, irrespective of the duration of use, were categorized as previous PrEP users. Users who either reported stopping PrEP or denied currently being on PrEP at the time of partner services interview were categorized as PrEP discontinuations. Cases for which the last reported PrEP dose taken postdated the HIV diagnosis were determined to be on PrEP at the time of HIV diagnosis.

Data Analysis

We used chi-square tests to evaluate differences in select demographics between the persons with and without a history of PrEP use. Based on county data suggesting use of PrEP increased since 2014 among MSM at greatest risk for HIV acquisition,³ we hypothesized that the frequency of PrEP use ever in persons newly diagnosed with HIV would increase over time; thus, we examined temporal trends in the proportion of men reporting PrEP use using the Cochran-Armitage trend test. All analyses were conducted using R version 3.5.1 (R Foundation for Statistical Computing, 2018) and used an alpha level of 0.05 for statistical significance.

Ethics

These activities were conducted as part of a public health PrEP program evaluation and therefore were not considered human subjects research.

RESULTS

A total of 1,098 MSM were newly diagnosed with HIV infection during the study period. The median age was 32 years (interquartile range [IQR]: 26–41); most were White (55.9%) and 29% were Latinx/Hispanic (Table 1). Overall, 797 (72.6%) men completed a partner services interview. The majority (150, 70.4%) of men who were not interviewed were unable to be located or had moved out of the jurisdiction. Compared to men who completed the partner services interview, the non-interviewed men were more often Latinx/Hispanic (90.1% vs 21.8%) and were slightly older (median age: 36 vs 33 years); demographics were otherwise similar between the two groups.

Table 1.

Demographics, Select HIV Risk Factors and Reported PrEP Use Status for Men who Have Sex with Men Newly Diagnosed with HIV in King County Washington, 2013–2021

	Total MSM (N=1098)
Age at HIV diagnosis, years (median [IQR])	32 [26–41]
Current gender
Cisgender man	1073 (97.7)
Other gender identity	25 (2.3)
Race/Ethnicity
Latinx/Hispanic	237 (29.0)
Asian	90 (8.2)
Black	141 (12.8)
Multiracial	51 (4.6)
Other §	39 (3.6)
White	614 (55.9)
Reported any duration of PrEP use before HIV diagnosis †
Ever used PrEP¶	94
No prior PrEP use	619
Housing status
Permanent/stable	621 (83.7)
Not permanent/unstable	104 (14.0)
Institutionalized	5 (0.7)
History of substance use ever by any method	315 (40.2)
Crack cocaine	14 (1.9)
Cocaine	75 (10.4)
Methamphetamines	137 (18.4)
Heroin	32 (4.4)
Erection stimulants (sildenafil, etc.)	65 (9.0)
Alkyl nitrites or poppers	186 (25.6)
History of injection drug use
Yes	80 (10.8)
No	656 (88.6)
Unknown/refused	4 (0.5)
Used post-exposure prophylaxis ever in the past
Yes	26 (3.8)
No	642 (94.8)
Unknown/refused	9 (1.3)

All cells are n (%) unless otherwise stated.

Abbreviations: PrEP (pre-exposure prophylaxis); MSM (men who have sex with men); IQR (interquartile range)

^§Other includes Native Hawaiian/other Pacific Islander or American Indian/Alaska Native

^†Excludes 75 interviewees with missing data and 9 who were unsure or declined to respond about prior PrEP use

^¶Includes 19 men on PrEP at time of HIV diagnosis

Public health staff collected data on PrEP use from 722 (90.6%) interviewed men, of whom 94 (13%) reported ever using PrEP (Figure 1). Compared to men who had never used PrEP, PrEP users were somewhat more likely to be Latinx/Hispanic (29.6% vs 21.8%, P=0.186) and less likely to be Black (6.6% vs 14.9%, P=0.047). The majority of men with newly diagnosed HIV were White, and this percentage did not differ significantly between men with and without a history of PrEP use (59.2% vs 57.8%, P=0.899). History of any prior methamphetamine, erection stimulant, poppers or post-exposure prophylaxis use were more commonly reported among the prior PrEP use group (data not shown).

Figure 1:

Number of New HIV Diagnoses Among MSM in King County, Washington by PrEP Use Status, 2013–2021

Abbreviations: MSM (men who have sex with men), PrEP (pre-exposure prophylaxis)

Note: Data span from October 2013 to August 2021. Clustered bar graphs use leftward Y-axis scale and line graph uses rightward Y-axis scale.

The proportion of men who ever used PrEP prior to HIV diagnosis increased from 2.3% (3 of 133) in 2014 to 26.6% (33 of 124) in 2020 through August of 2021 (P<.0001 for trend). Median time from PrEP discontinuation to HIV diagnosis was 152 days [IQR: 18–366, including 18 (19%) imputed dates] and the median duration of reported PrEP use before discontinuation was 214 days [IQR: 53–616]. Nineteen men (2.6% of all interviewed men) were on PrEP at the time of HIV diagnosis (Figure 1), among whom reason for HIV testing when diagnosed was collected for 18; only 6 (33%) tested as part of routine PrEP care. Information about patterns of PrEP use (e.g., intermittent or event-driven) was not routinely collected.

Records included information on reason for PrEP discontinuation for 31 (33%) of 94 men who stopped PrEP. The most common reasons for stopping PrEP were self-determination of being low risk for HIV acquisition (12.9%), side effects (12.9%), and insurance issues (12.9%). Being homeless/unstably housed or moving/traveling away from the area were each reported by 9.7% of men who discontinued PrEP. Other less common reasons for PrEP discontinuation included pill burden, affordability, anxiety about seeing a provider, and having difficulty keeping appointments.

HIV genotype data were available for 15 (79%) of the 19 men who reported being on PrEP at the time of diagnosis. Wild-type virus was identified in 7 (47%) cases. The remaining 8 (53%) cases had one or more nucleoside reverse transcriptase inhibitor resistance mutations, including four cases of an isolated M184V and one case of M41L/M184V/T215E. High-level tenofovir resistance was detected for three cases: one with K65R/M184V and two with A62V/K65R/M184V mutations. Partner services staff were able to confirm that 18 (95%) of these men, including those with detectable resistance mutations, were successfully linked to HIV care and initiated on active antiretroviral therapy.

DISCUSSION

We found that a history of PrEP use among MSM newly diagnosed with HIV increased dramatically between 2014 and 2021, and that by 2020–2021 more than one in four MSM with newly diagnosed HIV infection in King County, WA had either previously taken PrEP and discontinued it or were inconsistently or ineffectively using PrEP. Over half of the former PrEP users were diagnosed with HIV by approximately 5 months after stopping PrEP.

The most common reasons that men discontinued PrEP were medication side effects, insurance problems and believing themselves to be at low risk for HIV. Anticipatory counseling about potential side effects, particularly when initiating PrEP, should be routine practice and may prevent patients from misjudging unrelated symptoms that could lead to stopping PrEP. While PrEP cost is often a considerable barrier for patients across the U.S., Washington State operates an expanded PrEP drug assistance program (PrEP DAP)¹² that significantly reduces out-of-pocket costs for medications, laboratory services and provider visits for both insured and uninsured patients. This program makes PrEP available to all MSM; however, many men may not be aware of this. Our findings highlight how success in overcoming barriers presented by cost requires not only funding to make PrEP universally available – which largely exists in Washington State – but also widespread knowledge that this funding is available, which may not yet be universal. Encouragingly, none of the four men providing a reason for PrEP discontinuation in 2020–2021 reported lack of insurance as a reason for stopping PrEP.

Though some PrEP discontinuations may be appropriate (e.g., during prolonged periods of abstinence from sex or while in mutually monogamous, seroconcordant partnerships), patients may misperceive when their risk for HIV changes and thus may not think to restart PrEP. Targeted interventions to improve identification and outreach for persons who have dropped off the PrEP care continuum should be a top priority. Strategies such as anticipatory guidance about how to readily resume PrEP if stopping it, periodic outreach to persons who have discontinued PrEP to counsel about perceived side effects and determine if they might benefit from resuming the intervention, and routinizing discussions about transitioning from daily to event-driven PrEP as a harm reduction approach may reduce the number of discontinuations and avert unnecessary HIV seroconversions. Additionally, offering alternative methods for PrEP follow-up such as telehealth could improve PrEP retention and ensure that persons can continue to undergo appropriate and timely HIV screening while taking PrEP.

Our study has both strengths and limitations. The study was population-based, not drawn from a specific clinic or group of clinics, enhancing its representativeness of the entire Seattle population of MSM diagnosed with HIV during the study period. However, in some instances PrEP use dates and reasons for discontinuation were either incomplete or missing. Only 66% of MSM with newly diagnosed HIV in King County were successfully interviewed and answered questions related to PrEP. Latinx MSM were less likely to be interviewed than non-Latinx MSM, and although we did not identify other factors significantly associated with receipt of partner services, it is possible that men included in the study were different in other ways – including in their history of PrEP use – from men not included in the study, limiting the representativeness of the study population. The generalizability of our findings to non-MSM populations or other parts of the U.S. is unknown.

Further, our study population only included persons diagnosed with HIV infection and details about consistency of PrEP use and adherence were limited. As a result, we were not able to determine if unsuccessful PrEP use might have been more common in one group of men than another. However, prior studies, including data from our county, demonstrate that PrEP discontinuations are more common among Black MSM and potentially MSM who use methamphetamines^6,13 – populations that also experience an elevated risk for HIV in the absence of PrEP – highlighting the need for more effective approaches to retaining these men on PrEP.

While PrEP remains an important tool for HIV prevention, the intervention is only effective when used consistently during periods of risk. PrEP discontinuations are common, and patients may be particularly at risk for HIV infection in the weeks to months after stopping. As we strive to maximize PrEP access and coverage in the U.S., it remains imperative to consider how to most effectively preempt inappropriate drop-offs along the PrEP care continuum and make resumption of PrEP easier as people enter into periods of higher sexual risk.