Prostate cancer is the second most common cancer among gay, bisexual, and other men who have sex with men (GBM), with sparse research documenting potential disparities in treatment and rehabilitation outcomes. Restore 2 is the first NIH funded randomized controlled trial of a rehabilitation program designed to improve urinary and sexual dysfunction in GBM prostate cancer survivors. This analysis investigates the impact of Restore 2 on urinary incontinence over one year of participation. Specifically it aims to 1) describe change in the severity of urinary incontinence over one year; 2) determine if participating in the treatment arm improved severity of urinary incontinence over one year; and 3) identify subgroups based on demographic, health, and prostate cancer characteristics with greater urinary incontinence severity and response to treatment over one year.

One year longitudinal analysis of 400 GBM enrolled in the Restore 2 trial. Participants were randomly assigned to the intervention group (N=204) or to the usual care control group (N=197). The intervention was an online, tailored, multi-component rehabilitation program that addressed both urinary and sexual dysfunction. Participants selected which components they wanted to address. Treatment options for urinary incontinence included pelvic floor muscles exercises, while those for sexual dysfunction included sildenafil, masturbation exercises, vacuum pump exercises, and anal dilators.    

Participants were self-identified GBM prostate cancer patients in the United States recruited mainly from online dating sites, prostate cancer survivor support groups, and social networking sites. Eligible participants had to have been diagnosed with prostate cancer with curative treatment (e.g., prostatectomy or radiation) completed, ongoing, or scheduled within two months of baseline. Participants were excluded if they did not speak English fluently or lived outside the United States. Data was collected online using Qualtrics software© prior to participants’ being randomly assigned to a treatment or control group. Urinary incontinence type and severity was measured with the International Consultation on Incontinence Questionnaire – short form (ICIQ) at baseline, 3, 6, and 12 months follow-up. Adherence to pelvic floor muscle exercises were self reported as never, less than monthly, 1-3 times a month, 1-4 times a week, and 5-7 times a week at 3, 6, and 12 months. The following self-reported baseline demographic, health, and prostate cancer status characteristics were used to determine their association with urinary incontinence severity over time: age, race, obesity, number of alcoholic drinks consumed in a typical day, health related quality of life measured with the Functional Assessment of Cancer Therapy: General subscale (FACT-G), number of comorbidities, use of anti-incontinence medications, time since diagnosis, prostate cancer stage, and type of prostate cancer treatment. Descriptive statistics were used to describe sample characteristics and urinary incontinence severity. Linear mixed effects regression models with unstructured covariance matrixes were used to describe change in ICIQ scores over time.  Baseline demographic, health, and prostate cancer status characteristics significantly associated (p<.10) with ICIQ growth parameters in univariate models were included in a multivariate model to identify independent predictors of change.

Participants had a mean age of 63.5 years, were 5.3 years on average past treatment, and were treated with surgery (59%), radiation (28%), surgery and radiation (11%) or treatment other than surgery and radiation (2%). The mean ICIQ score at baseline was 6.6(4.9) and only 13 (3%) participants reported no urinary incontinence at all four data collection time points. Univariate models indicated that ICIQ scores varied at baseline (β=6.45, p<.0001), but did not change over one year (β=.01, p=.458). Treatment group assignment did not predict severity (β=-.17, p=.685) or change in ICIQ scores (β=-.0001, p=.50). Age, race, and prostate cancer stage were not associated with severity or change in ICIQ scores. In univariate models, the following characteristics were associated with ICIQ severity, but not change in ICIQ scores over time: obesity (β=1.64, p=.003 ), alcholic beverage consumption (β=-.82, p=.094), health related quality of life (β=-.08, p<.0001), comorbidities (β=-.08, p<.0001), use of anti-incontinence medications (β=1.83, p=.011), time since prostate cancer diagnosis (β=.07, p=.097), and adherence to pelvic floor muscles exercises (p<.0001). The type of prostate cancer treatment predicted the severity of ICIQ scores in the following rank order: radiation and surgery (β=1.50, p=.027), surgery only (p<.0001), radiation only (β=-2.14, p<.0001), other treatments (β=-3.31, p<.0001). The type of urinary incontinence also predicted severity of ICIQ scores in the following rank order: mixed urinary incontinence (β=8.63, p<.0001), stress urinary incontinence (β=6.98, p<.0001), insensible or continuous urinary incontinence (β=5.57, p<.0001), urgency urinary incontinence (β=4.50, p<.0001), and post-void urinary incontinence (β=4.41, p<.0001). Table 1 presents the multivariate model which indicated that ICIQ scores were more severe in the control group and were independently associated with obesity, health related quality of life, adherence to pelvic floor muscles exercises, and the type of urinary incontinence. Figure 1 illustrates differences between the treatment and control groups in the ICIQ scores predicted by the multivariate model.

Participating in an online, tailored, multi-component rehabilitation program for GBM prostate cancer survivors did not change incontinence severity over one year and no subgroups of participants emerged as having benefited more. Participants were not required to have urinary incontinence and were able to choose pelvic floor muscle exercises as a treatment option. Even though participants with more severe incontinence did more pelvic floor muscle exercises, they did not benefit. It may be that the dose of the pelvic floor muscle exercises was too low to see improvement. Alternatively, pelvic floor muscle exercises was the only treatment offered for urinary symptoms and is most effective for stress incontinence. Participants had other types of incontinence that may not be amenable to treatment with only pelvic floor muscle exercises. Additionally, this trial occurred during the COVID pandemic which may have decreased program participation. On average these participants were diagnosed with prostate cancer over five years ago and represent long term survivors. This may be a group with persistent long term incontinence symptoms after prostate cancer treatment that may need more intense or different treatments.

Incontinence severity did not change over one year for GBM enrolled in an online, tailored, multi-component rehabilitation program for urinary and sexual dysfunction after prostate cancer treatment. Incontinence treatment options in multi-component rehabilitation programs should include more than pelvic floor muscle exercises and account for symptoms that persist in long term prostate cancer survivors.