Urinary Obstructive Problems Exposed But Hormonal Health-Related Quality-of-Life Concerns Eschewed in Prostate Cancer Quality-of-Life Study

Abstract

Pardo et al report the health-related quality-of-life (HRQOL) outcomes of a prospective, multicenter cohort of patients with early-stage prostate cancer who were treated through prostatectomy, external-beam radiotherapy, or brachytherapy; patients who received neoadjuvant hormonal therapy were excluded from the cohort. Compliance with follow-up was robust, with 84% of participants completing follow-up 3 years after primary treatment. The most common adverse HRQOL consequence of treatment was decline in sexual function; however, half of the Spanish patient cohort reported severe problems with sexual function at the pretreatment baseline, suggesting that this cohort had significantly worse HRQOL at baseline than did community-based or referral center–based, non-Hispanic American cohorts that have undergone similar studies (in which 25% reported severe sexual problems at baseline). Therefore, changes in HRQOL observed by Pardo et al (Table 2 and Fig 1) may be more generalizable than the absolute reported prevalence of HRQOL concerns at a given time in follow-up (Fig 2). Because patients who received neoadjuvant radiotherapy were excluded and the brachytherapy cohort was composed almost entirely of men with low-risk cancer, the radiotherapy and brachytherapy findings are likely not applicable to men with an intermediate risk or a worse risk of prostate cancer who undergo contemporary radiotherapy. Nevertheless, the study provides important information regarding HRQOL outcomes, including validation of the importance of urinary-irritative concerns and insight regarding the relationship of baseline function and eventual outcomes of primary prostate cancer treatment. Among men who reported problems with urinary obstructive problems at baseline (81 of 364 respondents who completed the study), Pardo et al report that 64% of those who underwent prostatectomy reported improvement in urinary function. This observation validates, in a prospective multicenter setting, that obstructive urinary symptoms warrant emphasis in medical decision making and patient counseling. The underappreciated impact of obstructive urinary symptoms was first suggested by a single-institution, cross-sectional HRQOL study of outcomes after prostatectomy, radiotherapy, or brachytherapy that was met with substantial skepticism when first presented at the annual meeting of the American Society of Clinical Oncology in 1999. Subsequently, Steineck et al queried HRQOL outcomes in a cross-sectional survey of survivors of prostate cancer who had been managed by prostatectomy as compared with those who had been managed by watchful waiting (with the latter having eventually undergone radiotherapy or hormonal therapy in the majority of cases) and suggested possible HRQOL benefit in obstructive urinary symptoms after prostatectomy. More recently, the American Prostate Cancer Outcomes and Satisfaction With Treatment Quality Assessment (PROSTQA) multicenter consortium study of outcomes after prostatectomy, radiotherapy, or brachytherapy showed that urinary obstructive HRQOL had a similar impact on patient satisfaction with overall prostate cancer outcome as did urinary incontinence or sexual problems. The study of outcomes also prospectively showed the benefit of prostatectomy in relieving obstructive urinary problems. The question of how HRQOL data might be best presented for clinical usefulness remains elusive and without a consensus standard. One approach is to report the proportion of patients who report a return to baseline; however, this approach can be confounded by floor effects among men with poor baseline function or by lack of sensitivity to possible HRQOL improvements, such as are evident in the obstructive urinary domain. Another approach is to present responses to specific items from the HRQOL instrument. For this purpose, items (such as the University of California, Los Angeles Prostate Cancer Index bother items) that have response options indicating extent of bother from a particular symptom (eg, big problem, moderate problem, small problem, very small problem, no problem) provide a convenient focal point. Pardo et al opted for the latter approach and—to facilitate presentation—collapsed small and moderate problems into a single category, as previously suggested. However, it is arguable whether small problems are appropriately bundled with moderate problems (rather than being bundled with very small problems or no problems). That Pardo et al observed improvement in erection functioning from being a small-tomoderate baseline problem to being no problem (in 22% of patients with brachytherapy) suggests that small problems may be more appropriately bundled with very small problems and no problems to reflect insignificant drift between the latter two categories (unless there is a rationale whereby brachytherapy might be expected to improve erection function). An important caveat for interpreting the comparison of HRQOL consequences among primary treatments was suggested by Pardo et al and warrants emphasis: the major, post hoc exclusion of nearly one third of the original cohort for this analysis. Men who received neoadjuvant hormonal therapy before external radiation or brachytherapy were enrolled onto the initial cohort but were excluded from the HRQOL analyses, resulting in a cancer severity bias wherein patients treated with prostatectomy had more severe cancer before treatment (58% intermediate to high risk) as compared with patients who underwent external radiotherapy without hormonal treatment (43% intermediate to high risk) or brachytherapy (11% intermediate to high risk). By excluding those patients treated with radiotherapy who received adjuvant hormonal therapy, the analysis excluded the only setting in which level I evidence indicates survival benefit from prostate cancer radiation. Three clinical scenarios have shown that primary treatment results in survival benefit for men with early-stage prostate Editorials