Propensity-Weighted Comparison of Long-Term Risk of Urinary Adverse Events in Elderly Women Treated For Cervical Cancer

Int J Radiat Oncol Biol Phys. 2015 Jul 1;92(3):586-93. doi: 10.1016/j.ijrobp.2015.02.025. Epub 2015 Apr 16.


Purpose: Cervical cancer treatment is associated with a risk of urinary adverse events (UAEs) such as ureteral stricture and vesicovaginal fistula. We sought to measure the long-term UAE risk after surgery and radiation therapy (RT), with confounding controlled through propensity-weighted models.

Methods and materials: From the Surveillance, Epidemiology, and End Results (SEER)-Medicare database, we identified women ≥66 years old with nonmetastatic cervical cancer treated with simple surgery (SS), radical hysterectomy (RH), external beam RT plus brachytherapy (EBRT+BT), or RT+surgery. We matched them to noncancer controls 1:3. Differences in demographic and cancer characteristics were balanced by propensity weighting. Grade 3 to 4 UAEs were identified by diagnosis codes plus treatment codes. Cumulative incidence was measured using Kaplan-Meier methods. The hazard associated with different cancer treatments was compared using Cox models.

Results: UAEs occurred in 272 of 1808 cases (17%) and 222 of 5424 (4%) controls; most (62%) were ureteral strictures. The raw cumulative incidence of UAEs was highest in advanced cancers. UAEs occurred in 31% of patients after EBRT+BT, 25% of patients after RT+surgery, and 15% of patients after RH; however, after propensity weighting, the incidence was similar. In adjusted Cox models (reference = controls), the UAE risk was highest after RT+surgery (hazard ratio [HR], 5.07; 95% confidence interval [CI], 2.32-11.07), followed by EBRT+BT (HR, 3.33; 95% CI, 1.45-7.65), RH (HR, 3.65; 95% CI, 1.41-9.46) and SS (HR, 0.99; 95% CI, 0.32-3.01). The higher risk after RT+surgery versus EBRT+BT was statistically significant, whereas, EBRT+BT and RH were not significantly different from each other.

Conclusions: UAEs are common after cervical cancer treatment, particularly in patients with advanced cancers. UAEs are more common after RT, but these women tend to have the advanced cancers. After propensity weighting, the risk after RT was similar to that after surgery.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Aged, 80 and over
  • Brachytherapy / adverse effects
  • Case-Control Studies
  • Cystitis / epidemiology
  • Cystitis / etiology
  • Female
  • Hematuria / epidemiology
  • Hematuria / etiology
  • Humans
  • Incidence
  • Medicare / statistics & numerical data
  • Postoperative Complications* / epidemiology
  • Propensity Score
  • Proportional Hazards Models
  • Radiation Injuries / complications*
  • SEER Program
  • Socioeconomic Factors
  • Spasm / epidemiology
  • Spasm / etiology
  • United States / epidemiology
  • Ureteral Obstruction / epidemiology
  • Ureteral Obstruction / etiology*
  • Urinary Bladder Diseases / epidemiology
  • Urinary Bladder Diseases / etiology*
  • Urinary Fistula / epidemiology
  • Urinary Fistula / etiology
  • Urination Disorders / epidemiology
  • Urination Disorders / etiology
  • Uterine Cervical Neoplasms / mortality
  • Uterine Cervical Neoplasms / pathology
  • Uterine Cervical Neoplasms / radiotherapy*
  • Uterine Cervical Neoplasms / surgery*