Back to Annual Meeting Program
Predictors of Favorable Urinary Outcomes After Prostate Cancer Treatment
Peter Chang, MD1, Meredith M. Regan, Sc.D.2, John T. Wei, MD, MS3, Larry A. Hembroff, Ph.D.4, Jeff M. Michalski, MD, MBA5, Chris S. Saigal, MD6, Mark S. Litwin, MD, MPH6, Daniel A. Hamstra, MD, PhD3, Irving D. Kaplan, MD1, Joseph Aronovitz, MD, PhD1, Jay P. Ciezki, MD7, Eric A. Klein, MD7, Adam S. Kibel, MD8, Louis L. Pisters, MD9, Deborah A. Kuban, MD9, David P. Wood, MD3, Howard M. Sandler, MD10, Rodney L. Dunn, MS3, Martin G. Sanda, MD11.
1Beth Israel Deaconess Medical Center, Boston, MA, USA, 2Dana Farber Cancer Institute, Boston, MA, USA, 3University of Michigan School of Medicine, Ann Arbor, MI, USA, 4Institute for Public Policy and Research, Michigan State University, East Lansing, MI, USA, 5Washington University School of Medicine, St. Louis, MO, USA, 6UCLA Center for Health Sciences, Los Angeles, CA, USA, 7Glickman Urological and Kidney Institute, Cleveland Clinic, Cleveland, OH, USA, 8Brigham and Women's Hospital, Boston, MA, USA, 9MD Anderson Cancer Center, Houston, TX, USA, 10Cedars-Sinai Medical Center, Los Angeles, CA, USA, 11Emory University School of Medicine, Atlanta, GA, USA.
While prostate cancer (PCa) treatment is known to be associated with bothersome urinary side effects, previous studies have shown that a subset of patients may actually perceive an improvement in their urination after PCa treatment. However, information on what factors may predict favorable urinary outcome is sparse. Our objective is to identify factors that predict a favorable change in patient-reported overall bother with urination.
The previously described PROST-QA cohort (NEJM 2008) is a multicenter prospective cohort of 1201 men with early stage PCa who had elected radical prostatectomy (RP), external beam radiotherapy (XRT), or brachytherapy (BT) from 2003 to 2006 at 9 university-affiliated hospitals. We analyzed the 1021 men who completed 2-year quality-of-life follow-up.
The EPIC-26 overall urinary bother question is a global assessment of the degree to which urination is a problem for a subject, measured on a five-point likert scale (no problem to big problem). Our primary outcome is defined as the change in overall urinary bother between baseline and 2 years post-treatment, categorized into a five-item ordinal scale: major worsening, minor worsening, no change, minor improvement, and major improvement, where “minor” describes a one point change, and “major” describes a 2 or greater point change.
We used ordinal logistic regression to identify the factors that predict a more favorable urinary outcome across all levels of change in urinary bother (measuring the odds of having not only an “improved” outcome, but also a “less poor” outcome). We considered the following covariates: age, race, marital/living status, education, number of comorbidities, obesity, prostate size, D’Amico risk category, use of prescription medications for urinary problems, baseline lower urinary tract symptoms (LUTS), and baseline urinary incontinence. We measured LUTS using the AUA Symptom Index.
On multivariable ordinal logistic regression, after controlling for all other covariates, the presence of moderate to severe baseline LUTS was the most powerful predictor of good outcome; subjects with baseline AUA-SI score ≥ 8 had an approximately 3 times higher odds of favorable outcome after treatment (Table). The effect of baseline LUTS was especially strong in subjects who underwent RP compared to XRT and BT (odds ratio 4.29 vs 1.89 and 1.86, respectively).
Pretreatment urinary medications use also predicted favorable outcome (OR 1.41; p = 0.04). Baseline urinary incontinence was not a significant outcome predictor. RP and XRT were both significant positive predictors compared to BT, but there was no significant difference between RP and XRT. Cohabitation, obesity, and 2 or more comorbidities were negative predictors.
The presence of baseline lower urinary tract symptoms is the most powerful predictor of a favorable change in overall urinary bother after PCa treatment. Furthermore, patients taking urinary medications for baseline LUTS who undergo RP are the most likely to have an improved outcome. These results highlight the often overlooked importance of LUTS in PCa patients’ urinary health.
Univariate and Multivariable Predictors of Favorable Urinary Outcome after Prostate Cancer Treatment
|a) Univariate||b) Multivariable|
|Variable||OR||95% CI||p-value||OR||95% CI||p-value|
|RP||2.05||1.56 - 2.71||<0.0001||1.99||1.50 - 2.65||<0.0001|
|XRT||1.90||1.37 - 2.64||0.0001||1.81||1.30 - 2.51||0.0004|
|RP vs XRT||1.08||0.81 - 1.43||0.5984||1.10||0.83 - 1.47||0.5084|
|Baseline urinary meds||1.89||1.37 - 2.61||0.0001||1.41||1.01 - 1.96||0.0443|
|Mod-severe baseline LUTS||3.09||2.41 - 3.96||<0.0001||2.80||2.17 - 3.63||<0.0001|
|Baseline incontinence||1.67||1.16 - 2.40||0.0061||1.18||0.81 - 1.71||0.3899|
|Age > 65||1.06||0.84 - 1.33||0.6425|
|Black race||0.84||0.56 - 1.26||0.3904|
|Missing race||1.35||0.57 - 3.21||0.4916|
|Living together||0.78||0.58 - 1.06||0.1126||0.75||0.55 - 1.03||0.0723|
|College graduate||0.91||0.72 - 1.15||0.4139|
|2 or more comorbidities||0.85||0.67 - 1.09||0.2074||0.82||0.64 - 1.06||0.1335|
|BMI > 35||0.72||0.49 - 1.06||0.0957||0.71||0.48 - 1.04||0.0769|
|D’Amico risk group|
|Low risk (reference)|
|Intermediate risk||1.05||0.83 - 1.34||0.6897|
|High risk||1.47||0.98 - 2.20||0.0618|
|Intermediate vs High risk||1.40||0.92 - 2.12||0.1138|
|Prostate size > 50 cc||1.36||1.06 - 1.75||0.0159|
Back to Annual Meeting Program