11023 Background: DOX remains critical in STS treatment. Controversy exists regarding its optimal administration route (BOL vs CIV). BOL vs CIV could affect toxicity and/or efficacy. A randomized trial to assess this is unlikely. We conducted a post hoc analysis to explore differences in these routes of DOX administration. Methods: Data from a prospective randomized phase III study of doxorubicin with or without evofosfamide (TH-302) were used. At the discretion of treating physician, BOL or CIV DOX could be used. Grade 3-5 hematologic, non-hematologic and cardiac toxicities and treatment response were explored using multivariable logistic regression. OS and PFS were analyzed using Kaplan-Meier and Cox proportional hazards. Results: 640 subjects were enrolled (556 BOL, 84 CIV). Baseline differences in age, extent of disease and prior radiotherapy were controlled for in regression models. Hematologic toxicity was associated with age, performance status (PS) and cumulative (CUM) DOX dose. Non-hematologic toxicity was associated with age, PS, receipt of prior radiotherapy and CUM TH-302 dose. Cardiac toxicity was only associated with CUM DOX dose. Odds of response were strongly associated with CUM DOX dose (mg/m2, OR = 1.011, p < 0.0001), and, to a lesser extent, with CUM TH-302 dose (g/m2, OR = 1.081, p = 0.0008), STS subtype and prior radiotherapy. Comparing CIV to BOL DOX, neither OS (median 21.7m vs 18.3m, HR = 0.85, p = 0.29) nor PFS (median 6.1m vs. 6.1m, HR = 0.89, p = 0.43) was affected by manner of DOX administration (CIV vs BOL). Cox analyses indicated that factors reflecting tumoral biology and host status, rather than treatment received, were associated with OS (PS, histologic STS subtype, histologic grade, receipt of prior radiotherapy) and PFS (PS, treatment-related toxicity). Conclusions: Our analyses provide no evidence for superiority of either BOL or CIV administration of DOX as regards toxicity or efficacy in STS treatment. Thus, the logistically simpler BOL administration of DOX should be favored over CIV administration.
Robotic-Assisted Extraperitoneal Paraaortic Lymphadenectomy is Associated with Fewer Surgical Complications: A Post-hoc Analysis of the STELLA-2 randomized trial.
