6042 Background: Both concurrent CDDP/RT and C/RT have been shown in randomized trials to yield superior disease control compared to RT alone in LAHNC, but no randomized trial has compared them. We evaluated our center's experience with these regimens. Methods: From 3/1/06 - 4/1/08, 175 patients were retrospectively identified who received definitive treatment for LAHNC with CDDP (planned total dose 100 mg/m2 Q3 weeks X 3) and RT (n = 125) or C (400 mg/m2 load; 250 mg/m2 weekly) and RT (n = 50). Patients who received prior RT, additional systemic therapy, and/or surgery to the primary site were excluded. C was given for the following reasons: auditory 30%, renal 4%, cardiac 2%, performance status 18%, patient choice 16%, neuropathy 4%, unknown 2%, and a combination of factors 24%. The median age: CDDP group 56, 6% >71; C group 66.5, 40% >71. Additional CDDP and C features: male sex, 86 v 78%; stage IV, 70 v 68%; and oropharynx, 78 v 70%. Median RT dose (70 Gy), RT length (46 days), and Karnofsky performance status (KPS) (90%) were the same; alcohol/tobacco use was similar. Results: At a median follow up of 18.7 months, with death without local failure (LF) as a competing risk, the 18 month LF incidence rate was 2.5% in the CDDP group and 43.3% in the C group (p < 0.0001), with the latest event occurring at 16.5 months. The 18 month disease-free survival (DFS) and overall survival (OS) rates were 85.7 v 40.9%, and 96.8 v 73.1%, in favor of CDDP (p < 0.0001 for both). Initially, 21 variables were assessed for significance, and when Cox proportional hazards model was used for multivariate analysis to address prognostic imbalances, treatment with CDDP still predicted for improved LF, DFS, and OS (p < 0.0001 for LF and DFS; p = 0.0017 for OS). For OS analysis, the concordance probability estimates were .67 for using drug choice alone and .80 for using drug choice, T stage, RT dose, and KPS. Conclusions: CDDP/RT and C/RT were used to treat somewhat different populations with LAHNC. The observed superiority of CDDP/RT compared to C/RT in LF, DFS, and OS may reflect patient selection issues. However, preliminary multivariate modeling suggests that CDDP/RT remains the preferred option for fit patients pending further analyses and prospective studies comparing these regimens. [Table: see text]
Read full abstract