Correlation of RTOG 9704 (adjuvant therapy (rx) of pancreatic adenocarcinoma (pan ca)) radiation therapy quality assurance scores (RTQASc) with survival (S)

Abstract

4523 Background: RTOG 9704 demonstrated a marginal S advantage (p=0.054) in multivariate analysis (MVA) of Gemcitabine (G) over 5FU before and after 5FU+RT for patients (pts) with pan ca resected for cure from the pan head but not from non-head sites (ASCO 2006, ASTRO 2006). This analysis was undertaken to assess the impact of RTQASc on S, S by treatment (rx) arm, and toxicity by rx arm. Methods: This is a secondary analysis of a prospective, randomized, phase III trial of the RTOG, ECOG, and SWOG. RTQASc was graded as per protocol (PP) or less than (<) PP. Using prospectively defined guidelines, <PP scores were variation acceptable (VA), variation unacceptable (VU), or incomplete/not evaluable (I/NE). I/NE pts were excluded from further analysis. Toxicities were scored by CTC, v 2.0. S is expressed as median S in yrs. Results: 416 pts had RTQASc of PP (216, 52%) or <PP (200, 48%; 42% VA, 6% VU). Frequency of PP and <PP did not differ by rx arm (PP = 55% on 5FU arm and 48% on G arm). Looking at PP vs <PP frequency of Grade 3+ Heme and Non- Heme toxicity did not vary significantly on the 5FU arm but did show a trend of < toxicity for PP pts on the G arm ( Table ). In contrast, S was increased for all (head, non-head) PP pts (median S 1.74 vs 1.47 yrs, p=0.019) and, in MVA, score of PP significantly impacted on S (p=0.02) but rx arm did not. PP and <PP S curves began to diverge at 14–15 months post surgery. For head pts, in MVA, RTQASc (PP superior to <PP) and rx arm (G superior to 5FU) both correlated with S (p=0.04, p=0.03, respectively). On the G arm PP pts had S of 1.89 yrs, significantly > than S of VA (1.41yrs) and VU (1.37yrs) pts. Conclusions: In this study prospectively defined RTQASc significantly correlated with S and effect of rx arm on S and showed a weaker effect on toxicity (G arm only). Timing of appearance of RTQASc effect on S implies effect on tumor control. In this context failure to consider RTQASc may confound observed outcomes and confuse correct understanding of the importance of RT. [Table: see text] No significant financial relationships to disclose.