Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: Mr Maisonneuve and Dr Lowenfels call attention to the possibility that histological misclassification could, at least in part, account for the long-term survival observed in a considerable proportion of patients in our study. Actuarial 5-year survival was 22.5% in the group receiving adjuvant treatment with gemcitabine compared with 11.5% in the observation group. According to the original pathology reports, only 11 patients (4 in the gemcitabine group and 7 in the observation group) had tumors with histology other than adenocarcinoma, suggesting that the probability of a significant bias in favor of the adjuvant group was low. This assumption was confirmed by a modified analysis of our survival data that excluded all nonadenocarcinoma patients; this resulted in only minor changes in the qualified overall survival curves in either group (log-rank P = .01 compared with P = .02 in the original analysis). However, I agree about the potential problem of misclassification, both with regard to histological diagnosis and resection margin status. We have initiated a post-hoc central pathology review in a high-standard reference laboratory, and intend to report these findings when they are available.
Oettle H. Adjuvant Chemotherapy With Gemcitabine for Patients With Resectable Pancreatic Cancer—Reply. JAMA. 2007;297(23):2581-2582. doi:10.1001/jama.297.23.2582-a