1. A Descriptive and Comparative Analysis
Transgender Participants in the HIV Vaccine Trial Network’s HVTN 505 Trial:
A Descriptive and Comparative Analysis
Shelly Karunaa, Doug Grovea, Gail Brodera, Chuka Anudeb, Scott Hammerc, Magda Sobieszczykc, and Michele Andrasika,d
aVaccine & Infectious Disease Division, Fred Hutchinson Cancer Research Center, Seattle, WA; bDivision of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD; cDivision of Infectious Diseases, Department of Medicine, Columbia University College of Physicians and Surgeons, New York, NY; dDepartment of Global Health, University of Washington, Seattle, WA
BACKGROUND
RESULTS
Background
HIV prevalence in the US transgender (TG) population (2.64%) is five times the national average (0.45%), yet TG individuals constitute a small proportion of HIV vaccine trial participants (ppts). HVTN 505, a Phase 2b trial to assess the safety and efficacy of a DNA/rAd5 HIV vaccine regimen, was the HVTN’s first trial to specify the eligibility of TG women born male, distinct from MSM.
Methods
Data was analyzed for the HVTN 505 modified-intent-to-treat (MITT) cohort (n=2496), which included ppts who were HIV-uninfected at enrollment. To identify factors that may impact future engagement, enrollment, and retention, ppts reporting gender identity differing from birth sex (TG) were compared to ppts reporting concordant gender identity and birth sex (cisgender, CG).
Results
Forty-four (1.8%) TG ppts were compared to 2452 (98.2%) CG ppts in the MITT cohort. Of TG ppts, 14% identified Hispanic, 43% White, 41% African-American, 11% “other,” and 2% each as Asian or multi-racial. TG ppts were more likely to be non-white (57% TG vs. 24% CG; p<.001) and younger (mean 29 vs. 32 years old; p=0.04). TG and CG ppts did not differ in average number of pre-existing conditions (PECs) or adverse events (AEs), including total serious AEs. There were more deaths among TG ppts (n=2 or 5% vs. n=7 or 0.29%; p=.01) and more TG ppts missed at least one study visit (34% vs. 21%; p=.01). There were no differences in numbers of social harms or benefits, or in numbers of interim HIV tests. Differences in HIV incidence were not statistically significant (3 TG ppts or 4.3/100 person-yrs vs. 84 CG ppts or 2.1/100 p-y; p=0.37).
Conclusions
HVTN 505 TG ppts were more likely than CG to be characterized by demographic factors associated with HIV incidence, including non-white race and young age. Though overall HIV incidence was higher among TG ppts, the difference was not significant. Missed visits were more frequent among TG ppts; there was no difference in PECs and AEs.
Baseline Demographics in HVTN 505
In March 2011 the Institute of Medicine (IOM) published a report of lesbian, gay, bisexual, and transgender health in which it noted that transgender individuals (TG) in the US have increased risks of adverse health outcomes, generally.1 With respect to HIV infection, specifically, prevalence and incidence rates are much higher among TG than among cisgender (CG) individuals, with prevalence among black TG approaching 25% or more in some studies (Fig 1).2,3 This is largely attributed to complex relationships between known predictors of HIV incidence such as sero-discordant primary partners, unprotected anal intercourse, sex work and drug use, as well as the interaction of interpersonal, social, and structural variables with these predictors (Fig 2).2-7 
Social Impacts in HVTN 505
9% of participants reported social harms due to trial participation (14% of TG; 9% of CG ppts), most commonly a negative personal experience with family, friends, or significant others.
87% experienced social benefits to trial participation. The most common benefit was feeling good helping others, reported by 75% of TG and 70% of CG participants.
Other benefits included: receiving risk reduction counseling (50% TG, 42% CG); receiving medical care (46% TG, 29% CG); and positive impacts on personal relationships (36% TG, 31% CG).
Participant Characteristic
Transgender
n=44 n (%)
Cisgender
n=2452 n (%)
All Subjects
n=2496 n (%)
Age
Median (SD)
25.5yo (8.93)
29yo (9.02)
Race
White
19 (43)
1859 (76)
1878 (75)
Black/African American
18 (41)
406 (17)
424 (17)
Multiracial
1 (2)
89 (4)
90 (4)
Asian
1(2)
32 (1)
33 (1)
Native American/ Alaskan Native/ Native Hawaiian/ Pacific Islander
23 (1)
Other
5 (11)
40 (2)
45 (2)
Ethnicity
Hispanic
6 (14)
213 (9)
219 (9)
Figure 1. HIV prevalence. Overall sample refers to the general US adult population. Survey total and remaining circles refer to transgender respondents to the National Transgender Discrimination Survey.3
Self-Identified Gender Identity: Many Names
Comparisons of HVTN 505 Transgender & Cisgender Participants
Self-Identified Gender Identity
Frequency
Percent
Male
2,460
98.24%
Female 8
0.32%
Transgender Male 2
0.08%
Transgender Female 28
1.12%
Another identity [Queer (2); Gender Queer (1); Born male with female hormones (1); Not claiming male or female, no sex (1); Decline to state (1)]
Variable
Transgender
(n = 44)
Cisgender
(n = 2452)
p-value
Frequency
Mean (SD)
Missed Visits, # ppts with ≥ 1
15 (34.1%)
N/A
420 (17.1%)
< 0.01†*
Discontinuation of Vaccinations
3 (6.8%)
82 (3.3%)
0.19†
Pre-Infection, Pre-Unblinding Dropout
5 (11.4%)
380 (15.5%)
0.67†
Ppts Reporting Outside HIV Testing
6 (15.8%)
(n=38)
239 (10.5%)
(n=2286)
0.28†
Pre-existing Conditions
249ᶲ in 41 ppts (93.2%)
5.66 (4.6)
14,023 in 2,319 ppts (94.6%)
5.72 (4.6)
0.73† (Freq.) (Mean)
Adverse Events
Any AE
88 in 31 ppts (70.5%)
2.00 (2.1)
4,573 in 1,657 ppts (67.6%)
1.87 (2.3)
0.75† (Freq.) (Mean)
Severe AEs only
8 in 5 ppts (11.4%)
0.18 (0.5)
196 in 145 ppts (5.9%)
0.08 (0.4)
0.18† (Freq.) (Mean)
Death
2 (4.6%)
7 (0.3%)
0.01†*
TG are, thus, among those most in need of an HIV vaccine. Yet, TG experience high rates of stigma, discrimination, and prejudice, which create significant barriers to HIV vaccine trial participation.2,3,8 The phase 2b HVTN 505 trial assessed the safety and efficacy of a preventive DNA/rAd5 HIV vaccine regimen in at-risk MSM and transgender women (TGW) who have sex with men.  It was HVTN’s first trial to specifically include TGW, distinct from MSM.
Reasons for Enrolling in HVTN 505
† Fisher’s Exact Test *p-value significant at 0.05 level
ᶲSome of these conditions are referable to participants’ TG status (eg, gender dysphoria)
CONCLUSIONS & IMPLICATIONS FOR FUTURE TRIALS
HVTN 505 TG ppts were more likely than CG to be characterized by demographic factors associated with HIV incidence, including non-white race and young age.
Though overall HIV incidence was higher among TG ppts, the difference was not statistically significant, albeit with limited power to detect any difference.
Missed visits were more frequent among TG ppts; there was no statistically significant difference in other retention parameters (eg, DOV, dropout), or in PECs and AEs.
TG & CG ppts who meet high-risk criteria can be successfully recruited & retained, and contribute meaningfully to HIV vaccine clinical trials.
TG>CG ppts cited “a friend joined” as a reason to enroll, which may prove informative for future TG engagement & recruitment; both TG & CG cited participating primarily for altruistic reasons.
Figure 2. Social & structural factors contributing to HIV risk. Lines indicate a relationship as documented in the literature, but not directionality or causality.9
HIV Incidence in HVTN 505
Group
Number Infected
Total person-years follow-up
Rate
(infections/ person-yEArs)
95 % CI
for the rate
Transgender (n=44) 3
70.18
(4.27/100 p-y)
(0.0088,   )
Cisgender (n=2452) 84
(2.08/100 p-y)
(0.0166,   0.0258)
METHODS
Data was analyzed for the HVTN 505 MITT cohort (n=2496), which included ppts who were HIV-uninfected at enrollment.
Participants reporting birth sex differing from gender identity (TG) were identified by case report form (CRF) and were compared to ppts reporting concordant birth sex and gender identity (CG).
There is no statistically significant difference in HIV incidence between TG & CG ppts in HVTN However, there is little power to compare HIV incidence among these TG & CG groups.
If TG& CG group sizes are held constant, and HIV incidence rate in the CG ppts is held constant, 2 additional infections in the TG group would have led to an observed rate of 7.12/100p-y, p<0.05.