O’Connell, M.

4006 Background: Lymph node (LN) metastasis is associated with decreased survival in rectal cancer. It has been suggested that at least 14 LN be evaluated for adequate staging. However, a large percentage of patients have fewer than the recommended number of LN examined. We hypothesized that LN ratio would be predictive of overall survival in rectal cancer. Methods: Data was analyzed from Intergroup 0114, a mature trial of postoperative adjuvant chemotherapy and radiation in T3/4 and/or LN positive rectal cancer. Survival was the same for all arms allowing the entire group to be considered as one. The primary endpoint evaluated was overall survival. A proportional hazards model was used to determine the relative prognostic impact of LN ratio compared to number of LN examined, number of positive LN, number of negative LN and AJCC nodal stage. LN ratio was defined as the number of positive LN divided by the total number of LN examined. Four groups were analyzed based on proportion of positive LN: =0.25, >0.25–0.50, >0.50–0.75 and >0.75. Results: 1,648 patients were evaluable. There were 251 T1/2, 1,251 T3 and 146 T4 tumors. 513 patients were N0, 743 N1 and 392 N2. Median number of LN was 9. LN ratio was predictive of 5-year overall survival with rates of 0.71, 0.56, 0.50 and 0.43 respectively when analyzed by quartile (p<0.0001). LN ratio remained significant when overall survival was analyzed by number of LN examined and grouped into <10, <15 and >15 nodes evaluated (p<0.0001 for all). LN ratio also predicted overall survival in N1 (p=0.04) and N2 (p=0.0002) disease. When comparing LN ratio (χ2=79.5, p<0.0001) to number of LN examined (χ2=4.7, p=0.03), number of positive LN (χ2=38, p<0.0001), number of negative LN (χ2=32, p<0.0001) and AJCC nodal stage (χ2=55.5, p<0.0001), LN ratio appears to be the strongest predictor of overall survival. Conclusion: LN ratio predicts overall survival in patients with resected rectal cancer. Importantly, this is true in patients who have had a small number of LN evaluated, in addition to those with a large number of LN examined. LN ratio also appears to be a stronger predictor of overall survival than other described LN prognostic factors. LN ratio may be a useful variable to stratify outcome in patients with node-positive rectal cancer. No significant financial relationships to disclose.

Purpose: Intergroup Study 0114 was designed to study the effect of various chemotherapy regimens delivered after potentially curative surgical resection of T3, T4, and/or node-positive rectal cancer. A subset analysis was undertaken to investigate the prevalence and influence of salvage therapy among patients with recurrent disease. Patients and Methods: Adjuvant therapy consisted of two cycles of fluorouracil (FU)-based chemotherapy followed by pelvic irradiation with chemotherapy and two more cycles of chemotherapy after radiation therapy. A total of 1,792 patients were entered onto the study and 1,696 were assessable. After a median of 8.9 years of follow-up, 715 patients (42%) had disease recurrence, and an additional 10% died without evidence of disease. Five hundred patients with follow-up information available had a single organ or single site of first recurrence (73.5% of all recurrences). Results: A total of 171 patients (34% of those with a single organ or single site of recurrence) had a potentially curative resection of the metastatic or locally recurrent disease. Single-site first recurrences in the liver, lung, or pelvis occurred in 448 patients (90% of the single-site recurrences), with 159 (35%) of these undergoing surgical resection for attempted cure. Overall survival differed significantly between the resected and nonresected groups (P < .0001), with overall 5-year probabilities of .27 and .06, respectively. Controlling for worst performance status at the time of recurrence does not alter this relationship. Patients who underwent salvage surgery had significantly increased survival (P < .001) for each site. Conclusion: Attempted surgical salvage of rectal cancer recurrence is performed commonly in the United States. The chance of a long-term cure with such intervention is approximately 27%.