3319 - Identification of Key Clinical and Dosimetric Predictors for Acute Treatment-Induced Toxicity following Postoperative Prostate Cancer Radiotherapy

Purpose/Objective(s): This study evaluated the incidence and predictors of acute toxicity in prostate cancer patients receiving post-operative radiotherapy. Identifying key clinical, pathological, and dosimetric factors provides a basis for treatment optimization and improved post-radiotherapy quality of life.

Materials/Methods: Prostate cancer patients receiving postoperative radiotherapy at Peking University Third Hospital between January 2019 and January 2023 were retrospectively reviewed. Patients with prior pelvic surgeries, brachytherapy, or incomplete follow-up were excluded. Acute genitourinary and gastrointestinal toxicity outcomes within 6 months were evaluated using CTCAE v5.0. Clinical and dosimetric parameters were summarized with descriptive statistics. For high-incidence endpoints, LASSO regression with 10-fold cross-validation was employed to mitigate multicollinearity. For low-incidence endpoints (n<20) or data separation, Firth’s penalized likelihood regression was utilized to minimize bias. Multivariable models incorporated variables with univariate P<0.1 and established clinical factors. Statistical analyses were performed in R (v4.5) using glmnet and logistf packages, with significance defined as P<0.05.

Results: A cohort of 108 patients was analyzed. No significant predictor was found for acute gastrointestinal toxicity. The incidence of G1+ urinary urgency, G2+ incontinence, G1+ hematuria, and G1+ dysuria was 45.4%, 13.0%, 9.3%, and 17.6%, respectively. LASSO-selected Bladder Dmax was the sole independent predictor for acute urinary urgency (OR=1.58, 95% CI: 1.07–2.42, P=0.026). Multivariable Firth regression identified positive surgical margin (Margin+) as a significant risk factor for incontinence (OR=4.51, 95% CI: 1.10–45.4, P=0.047), while dosimetric factors showed no independent correlation. Both Margin+ (OR=10.24, 95% CI: 1.35–143, P=0.026) and pre-operative PSA (OR=1.52, 95% CI: 1.01–5.73, P=0.049) were significantly associated with increased risk of hematuria. Pre-operative PSA demonstrated a significant univariate association with dysuria (OR=1.59, P=0.040), reflecting the impact of baseline urinary status.

Conclusion: Acute urinary toxicity following radiotherapy is driven by an interplay between dosimetric and surgical factors. Irritative symptoms like urgency are primarily dose-dependent (Bladder Dmax), while more severe toxicities such as incontinence and hematuria are predominantly influenced by surgical complexity and baseline biological factors. Radiotherapeutic planning should prioritize the limitation of Bladder Dmax. Personalized monitoring is warranted for Margin+ and high-PSA patients to mitigate severe urinary morbidity.