Table 3

 Desai et al. (2020) [99]

USA

Adolescent and young adult (AYA) cancer survivors (n = 1025), 18–40 years old; 64 identifying as sexual minorities.

Quantitative

A cross-sectional study analyses using multivariable logistic regression tested associations between sexual minority status and self-reported anxiety and depression.

Sexual minority AYA survivors had 1.88 higher odds of anxiety compared to heterosexual counterparts, but no statistically significant rates of depression. More social support was associated with less likelihood of depression.

Strengths: Use of validated scales (GAD-7 and PHQ-8); one of very few studies focused on AYA survivors.

 Russell et al. (2016) [100]

USA

Adolescent and young adult (AYA) cancer survivors (n = 56) including SGM (n = 22) and heterosexual (n = 34) survivors

Qualitative

AYA survivors were interviewed by telephone; asked about pre- and post-diagnosis thoughts regarding relationships, parenthood, fertility, and how/ if fertility risks were conveyed to them during treatment.

Both SGM and heterosexual survivors reported post-diagnosis dating challenges. Straight survivors had greater fertility concerns (p < .05). SGM survivors were more likely to be open to raising a non-biological child or never parenting. Straight survivors were more likely to be unsatisfied with information provided about fertility, but SGM survivors were just as likely to not be informed about potential infertility risks.

Strengths: Only AYA cancer survivorship study known to date that examines differences by sexual orientation; diversity of types of cancer and treatment modalities. Limitations: Small sample size limits subgroup analyses; mostly white sample.

 Bazzi et al. (2018) [32]

USA

BrC survivors: (n = 339 heterosexual women, n = 201 WSW)

Quantitative

Cross-sectional national survey recruited from Army of Women using multivariable regression with primary outcome as resilience.

Sexual orientation was not associated with resilience, but WSW who were unemployed had less resilience than employed counterparts whereas heterosexual women had no differences based on employment status.

Strengths: Large sample diverse in socioeconomic status, cancer stage, and type of treatment; use of validated scales (ISEL-6, Mini-MAC, RS-14). Limitations: Sample is partially one of convenience, mostly White, and highly educated; self-report data; cross-sectional design.

 Boehmer et al. (2011) [10]

USA (Massachusetts Cancer Registry)

Nonmetastatic BrC survivors (n = 257 heterosexual women, n = 69 WSW)

Quantitative

Multinomial regression with weighting of subpopulations; primary outcome was weight.*

While WSW in the general population were more likely to be overweight and obese, WSW cancer survivors were not statistically more likely to be overweight/ obese than heterosexual counterparts. This finding suggests that WSW may be motivated by cancer to reduce overweight.

Strengths: Recruitment from a population-based registry; diversity of education, socioeconomic status, cancer stage, and treatment modality. Limitations: Data reported from one state; self-report data; potential bias in reporting weight; cross-sectional design.

 Boehmer et al. (2012) [28]

USA

Heterosexual women (n = 257) and WSW (n = 181) diagnosed with nonmetastatic BrC

Quantitative

Using a telephone survey, clinical and demographic characteristics and HADS were assessed. Demographic and clinical factors were compared with t tests and chi-square tests and then these characteristics were compared to anxiety and depression assessed via with least squares regression.

The study hypothesis, that WSW who had been diagnosed with BrC had higher rates of anxiety and depression was not confirmed, but sexual orientation was associated with anxiety and depression through interactions with clinical and demographic factors, with younger age and decreased financial means associated with worse anxiety and depression.

Strengths: Use of validated scale (HADS).

Limitations: Sample is partially one of convenience, mostly White, and highly educated; self-report data; cross-sectional design.

 Boehmer et al. (2012) [10]

USA (Massachusetts Cancer Registry + national convenience sample)

Nonmetastatic BrC survivors (n = 257 heterosexual women, n = 181 WSW)

Quantitative

Least square regression separately run for physical component and mental component summary scales of the SF-12 on each demographic and clinical characteristic, controlling for sexual orientation.*

Overall, WSW and heterosexual women were comparable in QOL. WSW from the registry were more likely to be White, educated, and employed. Only WSW with low/ middle income had worse physical health than heterosexual counterparts. WSW who experienced more discrimination reported worse physical health.

Strengths: Use of validated scale (SF-12). Limitations: Sample is partially one of convenience, mostly White, and highly educated; self-report data; cross-sectional design.

 Boehmer et al. (2012) [11]

USA

Nonmetastatic BrC survivors (n = 257 heterosexual women, n = 181 WSW)

Quantitative

Least square regression was used for each demographic and clinical characteristic, controlling for sexual orientation.*

WSW appeared more resilient than heterosexual counterparts with some exceptions: unemployed WSW experienced greater anxiety than heterosexual women, and WSW who underwent radiation therapy were more depressed than heterosexual counterparts. WSW reported higher rates of discrimination, which was associated with more depression.

Strengths: Use of a validated measure (HADS); sample size. Limitations: Sample is partially one of convenience, mostly White, and highly educated; self-report data; cross-sectional design; low percentage of variance explained by models.

 Boehmer et al. (2012) [29]

USA

Nonmetastatic WSW BrC survivors without recurrence (n = 22)

Qualitative

Semi-structured telephone interviews ranging from 30 to 150 min; coding based on grounded theory.

Themes included: 1) BrC is a women’s, not a lesbian, issue; 2) I can manage my identity in the context of BrC; 3) I am better off than my heterosexual counterparts (e.g., less emphasis on body image, empathic female partners)

Strengths: Adaptations to interview guide to maximize neutrality. Limitations: Convenience sample, mostly White, and highly educated; self-report data.

 Boehmer et al. (2012) [12]

USA

Nonmetastatic WSW BrC cases and heterosexual controls (n = 85 cases, n = 85 controls)

Quantitative

Using a conceptual framework for heterosexual BrC survivors, generalized estimating equations identified explanatory factors of sexual function between cases and controls.*

Sexual function was predicted by self-perception of sexual attraction and urogenital symptoms for both WSW and heterosexual women; for partnered women, postmenopausal status and dyadic cohesion was predictive of sexual function; HRQOL was less explanatory for WSW’s sexual function compared to heterosexual women.

Strengths: Case-control design; use of validated scale (SF-12); amount of variance explained by models (nearly half). Limitations: Convenience sample, mostly White, and highly educated; self-report data; use of a sexual measure designed for heterosexual women (FSFI); cross-sectional design.

 Boehmer et al. (2013) [13]

USA (Massachusetts Cancer Registry + national convenience sample)

Nonmetastatic BrC survivors (n = 257 heterosexual women, n = 181 WSW)

Quantitative

Multiple regression models with stepwise variable selection (p = .10); model fit reported with R² statistics.*

WSW had less cognitive avoidance coping than heterosexual peers. Social support and having a partner were more strongly associated with better mental and physically health, respectively, for WSW v. heterosexual counterparts.

Strengths: Use of validated scales (TPS, ISEL-6, Mini-MAC, BFS); large amount of variance explained in models. Limitations: Sample partially one of convenience, mostly White, and highly educated; cross-sectional design; self-report data.

 Boehmer et al. (2013) [14]

USA (Massachusetts Cancer Registry + national convenience sample)

Nonmetastatic WSW BrC survivors (n = 161 lesbians, n = 19 bisexual women)

Quantitative

Multiple regression models with stepwise variable selection (p = .10); fit reported with R² statistics.*

Lesbian and bisexual women did not differ in physical or mental health; however, women with female partners fared better than women who were with male partners or unpartnered.

Strengths: Use of validated scales (TPS, Mini-MAC, QLQ-BR23, SF-12); large amount of variance explained in models. Limitations: Small bisexual sample (n = 19); sample partially one of convenience, mostly White, and highly educated; cross-sectional design; self-report data.

 Boehmer et al. (2013) [15]

USA (Massachusetts Cancer Registry + national convenience sample)

Nonmetastatic BrC survivors (n = 257 heterosexual women, n = 181 WSW)

Quantitative

Multiple regression (for linear variables) and logistic regression (for dichotomous variables) models with stepwise variable selection (p = .10); fit reported with R² statistics or pseudo-R² statistics.*

WSW generally had lower blood pressure and fewer comorbidities than heterosexual counterparts. However, the impact of mastectomy and radiation in worsening arm symptoms was twice as strong for WSW compared to heterosexual peers. Having health insurance was associated with fewer side effects, an effect three times stronger for WSW v. heterosexual peers.

Strengths: Use of validated scale (QLQ-BR23). Limitations: Sample partially one of convenience, mostly White, and highly educated; cross-sectional design; self-report data.

 Boehmer et al. (2014) [16]

USA

Convenience sample of WSW (n = 85 with history of BrC, n = 85 never-diagnosed)

Quantitative

Case-control study examining sexual frequency, desire, ability to reach orgasm and pain using multiple general linear models or logistic regression for categorical variables.*

Groups did not differ in risk of sexual dysfunction or overall functioning, but cases had lower sexual frequency, less desire and ability to reach orgasm, and more pain during sex.

Strengths: Case-control design. Limitations: Use of a sexual measure designed for heterosexual women (FSFI); cross-sectional design.

 Boehmer et al. (2015) [18]

USA

Convenience sample of WSW (n = 85 with history of BrC cancer, n = 85 never-diagnosed)

Quantitative

Case-control study assessing self-reported physical activity, fruit and vegetable intake, weight, QOL, anxiety and depression using multiple general linear models or logistic regression for categorical variables.*

Groups did not differ in health behaviors, BMI, QOL, anxiety, and depression. Both groups were a majority overweight or obese, around 13-15% reporting depression and 37-45% reporting anxiety. More physical activity correlated with lower weight, less depression, and better mental health in both WSW groups.

Strengths: Case-control design; use of validated scales (HADS, SF-12). Limitations: Cross-sectional design.

 Boehmer et al. (2016) [17]

USA

Sample recruited from prior registry-based study plus a sample drawn from the Army of Women (n = 167 matched BrC survivor/ caregiver dyads)

Quantitative

Multiple logistic regression on fear of recurrence (FOR) using propensity score matching (p < .10). Simultaneous equation models were used to avoid endogeneity, since primary outcomes were patient and caregiver influence on each other’s FOR.

Survivor FOR was explained by years since diagnosis, co-residence with partner, caregiver receiving counseling, survivor ISEL scores, receipt of chemotherapy, and sexual orientation. Caregiver FOR was explained by years since survivor’s diagnosis, caregiver’s discrimination score, caregiver’s social support, survivor’s anti-estrogen therapy, survivor’s comorbidities, and sexual orientation. For both groups, caregiver FOR influenced survivor FOR, but not vice versa. Between groups, WSW survivors and caregivers had less FOR than heterosexual survivors and caregivers.

Strengths: Study design allowed for modeling of causal relationships for FOR. Limitations: Caregiver gender and sexual orientation were not considered; sample lacked racial diversity.

 Boehmer et al. (2018) [108]

USA

Sample recruited from prior registry-based study plus a sample drawn from the Army of Women (n = 167 matched BrC survivor/ caregiver dyads)

Quantitative

Multiple logistic regression on stress using propensity score matching (p < .10). Simultaneous equation models were used to avoid endogeneity, since primary outcomes were patient and caregiver influence on each other’s stress.

WSW survivor and caregiver stress were similar to heterosexual peers; however, WSW dyads showed interdependent stress associations where heterosexual dyads did not.

Strengths: Use of validated scales (ISEL, MSPS, DAS). Limitations: Cross-sectional design.

 Boehmer et al. (2020) [19]

USA

Nonmetastatic, non-recurrent BrC survivors of various sexual orientations (n = 167)

Quantitative

BrC survivors surveyed by telephone were assessed for QOL; propensity score weighting accounted for differences by sexual orientation in age and length of dyadic relationships; simultaneous equation models assessed dyads.

There were no differences in QOL by sexual orientation 6-7 years post-diagnosis; sexual minority dyads showed greater dependence on partner QOL scores than heterosexual dyads

Strengths: Propensity score weighting; use of simultaneous equation modeling; dyadic assessment; use of validated measures (SF-12, ISEL-SF, MSPSS). Limitations: Cross-sectional design; small comparative heterosexual group.

 Brown & McElroy (2018) [21]

USA

WSW diagnosed with BrC

Mixed methods

Purposive and referral sampling were used to recruit WSW diagnosed with BrC to complete an online survey. Bivariate analyses were conducted using cross-tabulations, chi-square statistics, and difference of mean t tests to compare those in the sample who identified their sexual orientation as “queer,” “questioning,” or “other” (n = 9) and those who identified their gender as “transgender,” “genderqueer,” or “other” (n = 11) to the rest of the sample. NVIVO was used for thematic analysis of open-text questions.

Compared to the rest of the sample, those who identified their sexual orientation as “queer,” “questioning,” or “other” or their gender as “transgender,” “genderqueer,” or “other” were more likely to report having bilateral mastectomy without reconstruction, to think that disclosing SOGI to providers affected their care, to use LGBT-specific support groups, and to report that their current level of social support is below average. Thematic analysis revealed themes related to self-disclosure of SOGI to providers, need for recognition and support of partners, need for appropriate social supports for patients and partners, and impact of BrC treatment on intimate relationships

Strengths: Novel study comparing the experiences of lesbian and bisexual cisgender women with BrC to other SGM populations with BrC.

Limitations: Study participants who were not cisgender lesbian and bisexual women were identified as “queer” although many did not identify that way. Few people of color were included; predominantly White sample.

 Brown & McElroy (2018) [22]

USA

WSW BrC survivors (n = 68) ages 18–75

Mixed methods

Purposive and referral sampling were used to recruit WSW BrC survivors to complete an online survey. Bivariate analyses were conducted using cross-tabulations and chi-square tests to determine differences between those electing to choose bilateral mastectomy without reconstruction versus those who did not. NVIVO was used for thematic analysis of open-text questions.

25% of the sample elected to “go flat” or not receive breast reconstruction. “Flattopers” were more likely to identify as genderqueer, be out to their providers, and participate in SGM support groups compared to the rest of the sample. There were not significant between-group differences for the BITS. Qualitative themes from open-text responses included reasons for “going flat,” interactions with health care providers, gender policing/ heterosexism during treatment, and mixed physical and emotional outcomes of treatment choices.

Strengths: This is one of very few studies to report transgender/ genderqueer outcomes of BrC in their own words; use of a previously developed scale (BITS). Limitations: Cross-sectional design; predominantly white sample.

 Jabson, Donatelle, & Bowen (2011) [20]

USA

WSW BrC survivors (n = 68)

Quantitative

Purposive sampling via known WSW gathering places recruited WSW BrC survivors to participate in an online survey focused on perceived discrimination, social support, stress, and QOL; regression models examined predictive value of independent variables (perceived discrimination, support, stress) on QOL.

Most WSW (92%) reported being treated similar to heterosexual peers. Thirty-nine percent of participants indicated they were perceived as heterosexual by their health care team. Perceived social support and perceived discrimination were statistically significant predictors of better QOL, because perceived heterosexuality was a construct of the discrimination scale and associated with better QOL.

Strengths: use of validated scales (BSS, QOL-CSV, PSS) and adaptation of previous discrimination scale that showed strong reliability (α=.75). Limitations: Predominantly White, educated, insured, partnered, economically stable convenience sample; missing data may skew results toward the null.

 Jabson, Donatelle, & Bowen (2011) [33]

USA

BrC survivors (n = 143 heterosexual, n = 61 WSW women)

Quantitative

Convenience sample of 204 BrC survivors were recruited to an online survey. Means and standard deviations of global QOL and four subscales (physical, psychological, social, and spiritual wellbeing) were compared by sexual orientation (heterosexual v. WSW).

Overall QOL as well as subscales of QOL did not statistically differ between groups.

Strength: Use of validated scale (QOL-CSV). Limitation: Predominantly White, educated, insured, partnered, economically stable convenience sample; missing data may skew results toward the null.

 Jabson & Bowen (2014) [25]

USA

BrC survivors (n = 143 heterosexual women, n = 68 WSW)

Quantitative

Convenience sample of 211 BrC survivors were recruited to an online survey. Means and standard deviations of perceived stress were compared by sexual orientation.

WSW had higher perceived stress than heterosexual peers in regression modeling.

Strength: Use of validated PSS. Limitation: Predominantly White, educated, insured, partnered, economically stable convenience sample; missing data may skew results toward the null.

 Kamen et al. (2017) [26]

USA

WSW BrC survivors (n = 201) recruited through the Army of Women (n = 172 lesbian, n = 29 bisexual women).

Quantitative

WSW with stage 0-III BrC completed surveys capturing demographic and clinical factors, minority stress factors, psychosocial resources, and psychological distress factors; linear regression used to examine associations between demographic and clinical characteristics and distress; associations between minority stress, psychological resources, and psychological distress assessed using partial correlations and controlling for demographic and clinical factors associated with distress; structural equation modeling tested direct and indirect effects on distress; statistically significant indirect effects interpreted as mediation.

Discrimination, resilience, and social support were significantly associated with depression after controlling for age, education, income, employment and past chemotherapy. Discrimination, negative identity, resilience, and social support were significantly associated with anxiety. Depression and anxiety were correlated (r = .48). Outness and negative identity were significantly positively associated with distress. Resilience and social support were negatively associated with distress. Discrimination had an indirect association with distress mediated by resilience.

Strengths: First study to demonstrate resilience as a positive resource for WSW to buffer the effects of discrimination on distress; use of validated scales (LGB Identity Scale, RS-14, ISEL-SF, HADS). Limitations: Self-report, cross-sectional nature of study; lack of sociodemographic diversity in sample.

 Matthews et al. (2002) [23]

USA

WSW with BrC

Qualitative

Using a standardized methodology, focus groups were conducted with a convenience sample of WSW (n = 13) and heterosexual women (n = 28) diagnosed with BrC in the past 5 years. Thematic analysis and representative case study methods were used.

WSW reported higher stress associated with diagnosis, lower satisfaction with care received from physicians, and a trend toward lower satisfaction with available emotional supports. Overall QOL did not differ between groups.

Strengths: Multiple independent coders; standardized focus group methods with experienced moderator.

Limitations: Small sample size.

 Wheldon, Roberts, & Boehmer (2019) [27]

USA

Female BrC survivors stage 0-III (n = 330 lesbian, n = 525 heterosexual)

Quantitative

Tested a theoretical framework to explain differences in coping between lesbian and heterosexual BrC survivors; five subscales from the Mini-MAC Scale used to measure coping with BrC among women post-treatment; mediation analysis used to examine the explanatory power of life course factors (e.g., parenting and education) in explicating the association between sexual identity and coping responses.

Lesbian women had less avoidant coping strategies and lower levels of anxious preoccupation than heterosexual counterparts.

Strengths: Replicates other literature indicating resilience among lesbian BrC survivors. Limitations: Cross-sectional study with a non-random sample that is mostly white.

 White & Boehmer (2012) [24]

USA

Partnered WSW (n = 15) diagnosed with nonmetastatic BrC from 2000 to 2005

Qualitative

One-on-one semi-structured telephone interviews were recorded, transcribed, and analyzed to identify emergent themes.

Themes describing WSW survivors' perceptions of support included female partners as the singular and most valuable source of support; sense of support from partner due to conversations about survivor distress; perceived partner distress; increased partner burden; and sense of support from partner sharing in a life beyond cancer.

Limitations: Mostly White sample; small sample size.

 Baughman et al. (2017) [109]

USA

Queer survivors with a diagnosis of stage III CRC (n = 8)

Qualitative

Semi-structured telephone interviews

Participants reported economic challenges associated with insurance coverage, employment, and housing as well as social isolation.

Strengths: This is the only known study focusing on queer CRC survivors; Sample was diverse in sex, sexual orientation, and socioeconomic status. Limitations: Lack of racial/ethnic diversity in sample; lack of staging in CRC respondents.

 Boehmer et al. (2020a) [110]

USA

Colorectal cancer survivors with stage I-III disease from four state cancer registries (California, Seattle-Puget Sound, Georgia, and Florida) (n = 480)

Quantitative

Survivors were mailed a questionnaire which asked about sexual orientation. Respondents participated in a telephone interview assessing quality of cancer care (e.g., physician communication, nursing care, and coordination of care). General linear models and logistic regression was used to obtain models that best fit each quality of care measure.

There were no statistically significant differences between sexual minority cancer survivors (n = 127) and heterosexual survivor (n = 353) ratings of physician communication, nursing care, and coordination of care; however sexual minorities rated quality of care as excellent more often than heterosexual survivors.

Strengths: Population-based state registry recruitment; rigorous quantitative methods. Limitations: Primarily White sample; reports did not differentiate between male and female sexual minority survivors.

 Allensworth-Davies et al. (2016) [111]

USA

Gay men age 50+ with a diagnosis of PrC (n = 111)

Quantitative

Cross-sectional national survey using multivariate generalized linear modeling with primary outcome as masculine self-esteem.

Men who were comfortable disclosing their sexual orientation to their doctor had higher masculine self-esteem scores. Mental health was positively correlated with masculine self-esteem. This study distinguished experiences of gay PrC survivors from heterosexual counterparts in terms of stigma and resilience.

Strengths: use of several validated scales (SF-12, EIPC, PDRQ-9); control of confounding variables; diversity of the study population in terms of age, insurance type, employment status, and treatment protocol. Limitations: convenience sample; lack of racial/ethnic diversity of participants.

 Jägervall et al. (2019) [34]

Sweden

Gay men with a diagnosis of PrC (n = 11)

Qualitative

Participants were recruited through purposive sampling through SGM-identified networks and participant referral of other eligible men. Semi-structure interviews of 30–90 min were conducted.

Participants reported loss of ejaculate, erectile dysfunction, weaker orgasms, and penetration difficulties. These challenges were sometimes associated with feelings of loss, unattractiveness, and disability.

Strength: First study in Sweden exploring gay men’s experiences of sexual changes. Limitations: Small and non-diverse sample.

 Capistrant et al. (2016) [41]

USA

Gay and bisexual men who had been diagnosed with PrC recruited from a national cancer support group network (n = 30)

Qualitative

One-on-one interviews probed for experiences with providers; health; sexual functioning; relationships; and informational, instrumental, and emotional support throughout PrC.

Single men in the study reported a need for independence; partnered men indicated varying levels of dependence on partners for support; many participants wished for more support options tailored for gay and bisexual men. In contrast to literature describing heterosexual PrC survivors, most support for gay and bisexual men came from family and friends rather than partners.

Strengths: One of few studies of gay and bisexual PrC survivors. Limitations: The sample was not very diverse: almost all participants were White, gay, and HIV-. There was not clear theoretical basis for the analysis.

 Crangle, Latini, & Hart (2017) [35]

USA and Canada

MSM who had been diagnosed with PrC within the last 4 years (n = 92)

Quantitative

Convenience sample of MSM recruited through a variety of methods; demographic, medical information, and measures of attachment and illness intrusiveness were collected; mediation models were tested using bootstrapping to examine each attachment dimension on subscales of IIRS, controlling for age and days since diagnosis.

Younger age and greater anxious attachment were associated with greater illness intrusiveness. Greater anxious attachment was associated with less comfort with outness. Less comfort with being out to one’s provider mediated the association between greater anxious attachment and more illness intrusiveness. This means that comfort with outness could reduce illness intrusiveness for MSM with anxious attachment styles.

Strengths: use of previously developed scales (RQ; IIRS) and a newly developed Outness Inventory that demonstrated strong reliability (subscales α≥.86). Limitations: Cross-sectional design; self-report data; predominantly White, highly educated sample; variable internal reliability of the RQ.

 Hart et al. (2014) [96]

USA and Canada

SM who had been diagnosed with PrC within the last 4 years (n = 92)

Quantitative

Convenience sample of MSM recruited through a variety of methods; demographic, medical information, and measures of QOL, HRWOL, change in sexual activity, sexual side effects, satisfaction with care, self-efficacy for symptom management, disease-specific anxiety, illness intrusiveness, and “outness level” collected; mean scores were calculated and compared to published population means in studies using the same scale, where possible; open-text responses reported descriptively.

MSM reported significantly worse urinary and bowel function, greater bother of lack of ejaculation than heterosexual peers from other published studies, lower satisfaction with PrC care—but overall health status was similar. MSM reported significantly worse mental but not worse physical health functioning than heterosexual peers. Nearly half (49%) of MSM reported changes to erectile function and 40.2% indicated less frequency of sexual activity. MSM reported painful erections, climacturia, low libido, changes in self-image, partner struggling with relationship changes, and significant changes in sexual experiences due to lack of ejaculation.

Strengths: Use of validated scales (EPIC; SF-36; MSHQ; CapSURE; ILLS) and a newly developed Outness Inventory that demonstrated strong reliability. Limitations: Predominantly White, educated, and “out” self-selected sample; cross-sectional design; variation in study design of comparison groups.

 Hartman et al. (2013) [45]

Canada

Homosexual couples following one partner’s radical prostatectomy due to PrC (n = 6; i.e., three couples)

Qualitative

Interpretative phenomenological analysis using inductive coding.

Major themes included acknowledging, accommodating, and accepting sexual changes. Unlike research on heterosexuals, the role of open relationships was helpful in 2 of the 3 partners studied. These couples also benefited from communication (similar to heterosexual couples). For the third couple, sexual dysfunction was so significant that communication did not feel beneficial in helping with sexual health.

Strengths: This study provides a counternarrative to the dominant heterosexual assumptions about sexual health following radical prostatectomy. Limitations: The study was exploratory with a small sample.

 Lee, Breau, & Eapen (2013) [43]

Canada

MSM with PrC (n = 15)

Quantitative

Pilot study comparing post-treatment QOL in MSM who had surgery to MSM who had radiation for treatment of PrC.

While the sample size precluded statistical comparisons, the radiation group appeared to have fewer sexual side effects post-treatment in terms of retained ability for penetrative and receptive intercourse.

Strengths: Use of validated scales (EPIC, MSHQ). Limitations: Pilot study with small sample prevented statistical analysis; researcher-created sexual function survey not validated.

 Lee et al. (2015) [112]

Canada

MSM with PrC (n = 16)

Qualitative

MSM were interviewed face-to-face or via video conferencing and asked about sexual QOL after PrC. Interviews were recorded, transcribed, and analyzed.

Themes from semi-structured interviews included sexual dysfunction (e.g., erectile, urinary, ejaculation, and orgasmic), intimacy challenges, and lack of support for cancer and psychosocial needs. Sexual QOL and relationship confidence were lower for those with greater sexual dysfunction. Coping was challenged by lack of support.

Strengths: The first qualitative study exploring the impact of PrC on MSM survivors’ sexual experiences; rich data to develop a new QOL instrument specific to MSM PrC survivors. Limitations: Sociodemographic diversity not discussed.

 Hoyt et al. (2020) [113]

USA

Gay men who had been diagnosed with PrC (n = 11)

Qualitative

Focus groups (n = 3) with gay PrC survivors (n = 11) using conventional content analysis.

Major challenges for participants included minority stress, intimacy/sexuality concerns, impact on life outlook, healthcare experiences, social support and the gay community, and intersectional identities.

Strengths: 2-3-hour time for focus groups allowed for participant directed discussion; racial diversity in sample.

Limitations: Small sample size.

 McConkey & Holborn (2018) [44]

Ireland

Gay men with PrC (n = 8)

Qualitative

In-depth interviews based on phenomenology were conducted with gay PrC survivors; interviews were recorded and transcribed; data was divided into “meaning units”; credibility and trustworthiness were bolstered by reflexivity, memoing, field notes of interviewee behaviors, and peer review of thematic descriptions from the data.

Three major themes that emerged included: (1) the experience of diagnosis and treatment, marked by shock at diagnosis, overwhelm during decision-making, sexual impacts of treatment; and degree of access to a nurse specialist; (2) experiences of health care service, including disclosure and communication with the care team; and (3) sources of support (e.g., family, friends), heteronormativity of support groups, and lack of gay community resources.

Strengths: First-known study to explore gay PrC survivor experiences in Ireland. Limitations: Lack of racial, national, and educational diversity in sample (important since 14-23% of the gay population in Ireland is foreign born).

 Motofei et al. (2011) [39]

Romania

Romanian PrC survivors (n = 17 heterosexual men, n = 12 gay men)

Quantitative

Gay and heterosexual PrC survivors were asked about sexual functioning prior to and after starting bicalutamide monotherapy. A 2 × 2 factorial ANOVA compared heterosexual v. gay and pre- v. post-exposure to bicalutamide.

Mean IIEF scores were lower after bicalutamide exposure for the full group (p < .001) with greater reductions in scores for gay v. heterosexual survivors after exposure.

Strength: First-known study of gay PrC cancer survivors in Romania; only study found to examine sexual impact of a drug by sexual orientation; use of validated scale (IIEF). Limitations: Small sample size; potential for recall bias; binary design does not account for bisexuality.

 Polter et al. (2019) [38]

USA

PrC survivors who participated in the RESTORE study (n = 191) including HIV+ (n = 24) and HIV− (n = 167) MSM

Quantitative

Cross-sectional, online survey of MSM treated for PrC examined sexual function, bother, and HRQOL using MANOVA and multivariate linear regression to evaluate association of HIV status and HRQOL after controlling for demographic and sexual characteristics.

HIV+ status was associated with lower mean urinary, sexual, and bowel scores on the EPIC after controlling for demographic and sexual characteristics. HRQOL did not differ by HIV status.

Strengths: Use of validated scales (EPIC, SF-12). Limitations: Small number of HIV+ men in the sample; cross-sectional design; evidence of fraudulent responses (procedure used to omit 200 responses was not described).

 Rosser et al. (2016) [42]

USA

Gay and bisexual men who had been diagnosed with PrC (n = 19)

Qualitative

In-depth telephone interviews with gay and bisexual men who had radical prostatectomies.

Themes included shock at diagnosis; depression; anxiety, grief, loss of sexual confidence; changes in sense of “maleness,” gay/bisexual identity, sex-role identity; sex interest and partners; disclosure of cancer survivorship status; and changes to relationships including renegotiation of exclusivity with partners.

Strengths: One of few studies focused on gay and bisexual PrC survivors.

Limitations: Small sample size.

 Rosser et al. (2017) [40]

USA and Canada

Gay and bisexual men who had been diagnosed with PrC (n = 193)

Quantitative

Online survey regarding sexual functioning measured using the Expanded Prostate Cancer Index Composite (EPIC) and a tailored Gay Sexual Functioning Inventory (GSFI).

MWM had worse urinary and hormonal function, but better sexual function than published norms. Most participants described their sexual functioning as fair to poor following treatment with less than a quarter of men reporting sufficient erections for insertive anal sex. Anal receptive men reported pain during sex after treatment. Over half of respondents reported urination problems during sex. Sexual functioning significantly predicted long-term positive health outcomes.

Strengths: Large sample.

Limitations: Convenience sample; predominantly White.

 Rosser et al. (2019) [114]

USA and Canada

MSM who had been diagnosed with PrC (n = 193)

Quantitative

Survey regarding rehabilitation treatments for sexual and urinary effects of PrC treatment—what they were offered, what they tried, and what their satisfaction was with the outcomes.

The most common problems reported were loss of ejaculate (93.8%), erectile difficulties (89.6%), change in sense of orgasm (87.0%), loss of sexual confidence (76.7%), changes to the penis (65.8%), increased pain in receptive anal sex (64.8%), urinary incontinence not related to sex (64.2%) and urinary incontinence during sex (49.2%). Of these factors only loss of ejaculate, erectile difficulties and nonsexual urinary problems were commonly discussed by clinicians during PrC treatment. Satisfaction with specific rehabilitation options varied widely.

Strengths: First-known study to investigate rehabilitation options to address sexual and urinary effects of PrC treatment. Large sample size.

Limitations: Mostly White MSM. Cross-sectional.

 Thomas et al. (2013) [47]

Australia

Australian MSM with a PrC diagnosis within the last 7 years (n = 10)

Qualitative

An asynchronous, online focus group was hosted over 4 weeks with MSM PrC survivors discussing impact of PrC on their lives.

Respondents mentioned accessing support, the challenges of incontinence and sexual changes, changes to sexual relationships, and divergent emotional responses (resilience v. negative outcomes). Respondents also indicated that general practitioners were more empathic than their urologists, and felt their emotional needs were not adequately addressed and that interactions with urologists were often distressing.

Strengths: Leveraging online technology to conduct qualitative work is innovative.

Limitations: Recall and self-selection bias; all-White sample prevents exploration of diverse MSM outcomes.

 Thomas et al. (2018) [115]

Australia

Australian PrC survivors (n = 813)

Quantitative

An online survey asked respondents about demographics, treatment modality for PrC, body image, self-esteem, sexual function and urinary function; a 2 × 2 ANCOVA was conducted to examine the main effect of two factors: sexual orientation and PrC diagnosis over six outcomes: self-esteem, urinary function, sexual function, appearance evaluation, health evaluation, and health orientation; differences in age and Gleason score were also examined.

Never-diagnosed respondents were statistically significantly younger than cancer survivors. Overall, gay respondents had statistically significantly higher age-adjusted self-esteem scores compared to heterosexual peers. PrC survivors had statistically significantly worse urinary and sexual function and health orientation than never- diagnosed peers. No statistically significant differences in outcomes were found between gay and heterosexual PrC survivors, although urinary function differences only narrowly failed to meet statistical significance (p = .054).

Strengths: Use of validated measures (EPIC, MBSRQ). Limitations: Cross-sectional design; small sample size of gay men with PrC; potential for self-selection (via social media recruitment) and self-report bias.

 Torbit et al. (2015) [93]

USA and Canada

MSM who received a PrC diagnosis within the prior 4 years (n = 92)

Quantitative

A multiple mediation design was used to test both self-efficacy and satisfaction with care on the relationship between physical symptom severity and FOR for PrC survivors.

Worse physical symptoms were associated with greater FOR. Self-efficacy and satisfaction of care mediated the statistically significant relationship between worse bowl function, worse hormone function, and worse sexual function with FOR, respectively. Self-efficacy and satisfaction did not mediate worse urinary function and FOR, but did explain 61% of the variance in the sample for that outcome.

Strengths: Use of a validated tool (EPIC) and tools from prior studies to measure self-efficacy and satisfaction with care. Limitations: Mostly White, educated, partnered sample; cross-sectional design; self-report data.

 Ussher et al. (2016) [94]

Australia

Australian PrC survivors (n = 124 MSM, n = 225 heterosexual men)

Quantitative

Participants were recruited through urology and primary care practices, support groups, SGM community groups, social media, and cancer research volunteer databases; multiple regression and independent samples t-tests assessed group differences; Pearson’s correlations assessed associations between MSM and heterosexual samples; multiple linear regression was used to identify meaningful predictor variables for HRQOL.

MSM were younger, less likely to be partnered, and more likely to have casual sex than heterosexual peers in the sample. MSM reported worse HRQOL, worse masculine self-esteem, lower satisfaction with care, higher psychological and cancer-related distress, greater ejaculation concerns, higher sexual functioning, and more sexual confidence at statistically significant levels compared to heterosexual peers.

Strengths: Use of validated tools (FACT-P, BSI-18, CSFQ-M, DSC, EPIC, MAX-PC, PrCQOL). Limitations: Differences between MSM and heterosexual samples (e.g., age, ethnicity, employment status, relationship status, and treatments received).

 Ussher et al. (2017) [36]

Australia, New Zealand, UK, USA

Australian MSM PrC survivors (n = 124) and their partners (n = 21); subset interviewed (n = 46 survivors, n = 7 partners)

Mixed Methods

An online survey of MSM PrC survivors (n = 124) and their male partners (n = 21) explored sexual experiences, relationships, and psychological wellbeing after treatment; a subset of this sample opted to also be interviewed (n = 46 survivors and n = 7 partners); descriptive statistics from the survey and themes from the interviews were reported.

Survivors reported erectile dysfunction, emotional distress, feelings of sexual disqualification, both negative and adaptive impacts on gay identity, loss of libido, climacturia, pain during anal sex, lack of ejaculation, and penile shortening.

Strengths: Use of validated measures (EPIC-Sexual Domain, CSFQ-M, FACT-P); mixed methods design.

 Wassersug et al. (2013) [37]

International: Primarily USA, Australia, Canada, and UK

Men (n = 556) from 17 countries with a diagnosis of PrC (n = 460 heterosexual men and n = 96 MSM)

Quantitative

Logistic regression and Wald tests assessed outcomes including sexual health, urinary incontinence, and depression.

No between-group differences were found for urinary incontinence or erectile dysfunction; however, MSM were more bothered by sexual impacts of PrC than heterosexual peers.

Strengths: International reach, adaptation of validated scale (EPIC). Limitations: Sample is largely affluent with access to the internet.

 Wright et al. (2019) [116]

USA and Canada

MSM (n = 189) with a diagnosis of PrC recruited from Malecare, an online cancer support organization

Quantitative

Linear regression was used to compare participants with cats only, dogs only, both cats and dogs, or no pets on SF-12 mental and physical component scores.

Participants with pets had lower mental health scores than non-pet owners. Cat owners had better physical health than other groups.

Strengths: First study to look at companion animal ownership association with mental and physical wellbeing; use of validated scale (SF-12). Limitations: Convenience sample; cross-sectional design; inability to determine directionality of association; no heterosexual control group.

 Anderson et al. (2020) [46]

USA

Gay, older long-term AIDS survivor with moderate-to-severe demoralization and serious medical illness (e.g., cancer) (n = 18)

Quantitative

Participants were recruited for a single-arm open-label study of psilocybin-supported group therapy with 8–10 therapy visits and one psilocybin administration visit. Two-way repeated ANOVA was conducted at baseline, end-of-treatment and three-months post-treatment to assess mean demoralization using the Demoralization Scale II.

Clinically meaningful change in demoralization from baseline to 3-month follow up and no serious adverse events suggested that psilocybin-assisted group therapy could be an effective intervention for older long-term survivors of AIDS living with cancer or other serious medical illness to reduce demoralization.

Strengths: Use of validated scale (Demoralization Scale (II); focus on high-need, elderly population with serious medical illness; one of few interventions in the literature. Limitations: Feasibility study; larger study is needed to confirm results.

 Boehmer et al. (2011) [31]

USA (California)

CHIS respondents ages 18-70 (n = 122,345 CHIS respondents, n = 10,942 survivors)

Quantitative

Pooled data from CHIS 2001, 2003, 2005 using logistic regression; primary outcomes were prevalence of cancer and self-reported health.*

WSW had ≥2.0 odds of fair/poor health compared to heterosexual counterparts with greater risk for racial minorities and older women; greater prevalence and younger diagnosis of cancer were reported by MSM compared to heterosexual counterparts but self-reported health was not different for MSM.

Strengths: Large, population-based sample (CHIS); first-known study to report prevalence of cancer and self-reported health of cancer survivors by sexual orientation. Limitations: Data collected only from one state; self-reported nature of the data.

 Boehmer et al. (2019) [30]

USA

BRFSS respondents who had a past diagnosis of cancer (n = 68,593 heterosexual women, n = 1,931 WSW)

Quantitative

Secondary data analysis of 2014-2017 years of BRFSS data. Survivors were categorized with an access deficit if any one of the following were true: no health insurance, delaying care, avoiding care due to cost, and lacking a trusted physician. Weighted analysis computed odds ratios and 95% confidence intervals using cumulative logit models and logistic regression, taking into account confounders.

WSW reported more access to care deficits—including lack of health care coverage, having no personal physician, avoiding care due to cost, and being without an annual visit—compared to heterosexual peers (p < .0001). WSW with deficits had poorer physical and mental QOL and trouble concentrating compared to heterosexual peers.

Strengths: Use of a large, population-based sample. Limitations: Small WSW sample sizes prevented subanalyses.

 Boehmer et al. (2020b) [48]

USA

Cancer survivors from the Behavioral Risk Factor Surveillance System (2014–2018)

Quantitative

Multiple logistic regression was used to estimate cancer prevalence and odds ratios for a variety of health outcomes and behaviors for 954,800 cisgender and transgender individuals with a history of cancer. The sample included 1877 transgender women, 1344 transgender men, 876 gender nonconforming individuals, 410,422 cisgender men and 540,389 cisgender women.

Transgender men had a higher rate of cancer compared to cisgender men, but not cisgender women. No other prevalence differences were found. Gender nonconforming survivors reported greater physical inactivity, alcohol use and depression compared with cisgender peers. Transgender men reported poorer physical health and more comorbidities than cisgender men or women. Transgender women reported more diabetes than cisgender men or women and greater cardiovascular disease compared to cisgender women.

Strengths: Population-based study; one of very few studies focused on transgender and gender nonconforming population. Limitations: No notable limitations

 Bryson et al. (2018) [49]

Canada

SGM breast and gynecological cancer survivors (n = 81)

Qualitative

Purposive sampling used to recruit diverse sample of SGM BrC survivors across Canada; semi-structured interviews conducted to explore patient experiences of care, health outcomes and decision-making.

This study reported on perceptions of how intersectional identity influenced feelings of safety and interactions with health care providers. It provides evidence that cisnormative systems negatively shaped care experiences for genderqueer people. Relevant to the present review outcomes reported were: physical impacts of cancer treatment that resulted in altered experiences of gender in society; lack of preparation or hormonal treatment for surgery-induced menopause; and mental health effects associated with lack of hormonal treatment.

Strengths: Diversity of sample; rich exploration of narratives; large sample; one of two known studies of genderqueer cancer survivors sharing their experiences in their own words. Limitations: No notable limitations.

 Hutchcraft et al. (2020) [97]

USA

Lesbian and bisexual women who participated in the National Health Interview Survey 2013–2018

Quantitative

Weighted, multivariable logistic regression was used to estimate odds ratios of heterosexual, lesbian, and bisexual women with a history of cancer on health-related QOL and health behaviors.

With heterosexual women as the reference group, lesbian women were 58% more likely to self-report fair/poor health, almost twice as likely to report chronic obstructive pulmonary disease or heart conditions, and twice as likely to be current smokers; bisexual women reported three times the rate of psychological distress, twice the rate of heart conditions, nearly three times the rate of food insecurity, and were less likely to have a recent mammogram.

Strengths: National population-based sample, rigorous methods. Limitations: No notable limitations.

 Jabson, Farmer, & Bowen (2015) [91]

USA

Cancer survivors participating in the NHANES from 2001 to 2010 (n = 576 heterosexual women, n = 26 WSW).

Quantitative

NHANES data from 2001 to 2010 were pooled and 602 cancer survivors were identified. Between-group (WSW v. heterosexual) characteristics, health behaviors, and self-reported health were compared using chi-square and t-tests; logistic regression was used to compare WSW v. heterosexual aORs; propensity score adjustment used for sociodemographic variables.

4.3% of the sample self-identified as WSW. WSW were 2.5 times more likely to report past illicit drug use and 60% less likely to report current health as good compared to heterosexual peers.

Strengths: Population-based sample. Limitations: Small sample of sexual minority cancer survivors in NHANES data due to lack of data collection of sexual orientation from 2001 to 2006 limited the power of the study.

 Kamen et al. (2014) [117]

USA

Men who reported sexual orientation in the BRFSS in 2009 from Arizona, California, Massachusetts, Ohio, and Wisconsin (n = 14,354)

Quantitative

The complex sampling procedure in SPSS (v. 20.0) weighted the sample based on demographic variables and state of residence; statistically significant between-group differences were used as co-variates for a logistic regression and t-tests examining outcomes.*

Gay men were 82% more likely to report a cancer diagnosis (p < .05) and were more likely to report less exercise, more distress, and greater alcohol and/or tobacco use. These health behaviors were shown to continue after a cancer diagnosis for gay men.

Strengths: Population-based sample from five states; first study to examine cancer disparities among gay men. Limitations: Cross-sectional design of BRFSS; potential lack of disclosure of sexual orientation among respondents.

 Kamen et al. (2015) [92]

USA

LiveStrong survey respondents: n = 207 SGM, n = 4899 heterosexual cancer survivors in 2010

Quantitative

Propensity matched cancer survivors (n = 621 heterosexual v. 207 LGBT survivors) assessed for distressed, difficulties with social relationships, fatigue and energy; symptoms assessed through dichotomous yes/no items and analyzed using Poisson regression; subgroup analyses by sex conducted.

SGM men reported greater depression and more relationship difficulties compared to heterosexual counterparts. SGM women did not have differences compared to heterosexual peers.

Strengths: First-known study to examine psychological distress of sexual minority cancer survivors. Limitations: Cross-sectional design of the study.

 Kamen et al. (2015) [50]

USA

291 SGM cancer survivors (n = 159 MSM, n = 123 WSW, n = 7 transgender men, n = 2 transgender women)

Quantitative

Participant demographics, cancer diagnosis, experiences of care, support-related factors, and self-rated health were assessed through a researcher-developed survey; descriptive data reported; logistic regression used to compare outcomes.

Parental support was the strongest single factor associated with good health followed by having a partner present during cancer diagnosis.

Strength: One of the largest studies of SGM cancer survivors at the time of publication. Limitations: Researcher-created survey that has not been validated; self-report data; cross-sectional design; self-selection bias; recall bias of support and comparison of support at diagnosis with present self-reported health.

 Kamen et al. (2016) [95]

USA

Queer (n = 10) and heterosexual (n = 12) cancer survivors

Quantitative

Randomized controlled trial of a 6-week exercise intervention comparing survivor-only v. survivor-caregiver dyad using independent samples t-tests.

At baseline, queer survivors reported greater depression (p = .01) and fewer steps walked (p = .03) compared to heterosexual counterparts. Post-intervention, there were no differences between queer v. heterosexual survivors, but survivors with partner support had a significantly greater reduction in depressive symptoms compared to the survivor-only group.

Strengths: One of very few interventional studies to improve QOL of queer cancer survivors; use of validated scales (CES-D, STAI, DSQ). Limitations: Small sample size.

 Lisy et al. (2019) [51]

Australia

Australian cancer survivors (n = 2115)

Quantitative

Cancer survivors diagnosed between 2009 and 2013 were identified through the Victorian Cancer Registry and asked to complete a survey about demographics, QOL, social difficulties, and information needs; descriptive data reported as well as between-group differences (SGM v. heterosexual).

Of the 2115 Australian cancer survivors who responded to the survey, 33 (1.6%) disclosed SGM status. SGM survivors had significantly fewer financial, support, and communication challenges post-treatment but greater challenges with diet and lifestyle than heterosexual peers. SGM survivors were more likely to report anxiety/ depression and body image challenges, but not at a statistically significant level.

Strengths: First population-based survey of SGM cancer survivors in Australia; use of some (unspecified) validated measures. Limitations: Questionnaires were not validated; small SGM sample; sexual orientation was not decoupled from gender identity.

 Matthews et al. (2016) [118]

USA

SGM cancer survivors (n = 175)

Quantitative

Cancer survivors were recruited through SGM-serving organizations to take an 82-item online survey asking about demographics, cancer type, comorbid conditions, health behaviors, and QOL; descriptive statistics summarized demographics; multivariable models were created to explore associations with physical and mental subscales of the SF-12.

Lower physical QOL scores were associated with older age at diagnosis, breast or gynecological cancer, medical comorbidities, overweight or obesity, and cancer recurrence (p < .05). Lower mental QOL scores were associated with younger age at diagnosis, lack of physical activity, FOR, lower levels of social and emotional support, and participation in therapy/support groups (p < .05).

Strengths: Use of a validated QOL measure (SF-12); diversity of type and stage of cancer as well as geography distribution across USA Limitations: Cross-sectional, convenience sample; limited racial/ethnic diversity in sample; no comparison group.

 Seay et al. (2018) [119]

USA

SGM cancer survivors (n = 114)

Quantitative

LGBT cancer survivors were recruited through National LGBT Cancer Project for a survey capturing demographics, social support, posttraumatic stress, and cancer survivorship care needs.

Most respondents reported unmet needs (73%) and over half reported unmet psychological and sexual care needs. Those who reported that providers were not culturally competent had greater unmet needs (p = .0.01 and greater posttraumatic stress (p = 0.035).

Strengths: Community-based participatory approach to study design; use of validated surveys (SCNS-SF34, IES-6, MOS-SS). Limitations: Cross-sectional, limited participation by cisgender women and gender diverse persons, limited racial/ethnic diversity in sample.