Reply to Improved Survival With Adjuvant External-Beam Radiation Therapy in Lymph Node-Negative Pancreatic Cancer

Abstract

We appreciate the opportunity to reply to Dr. Chu's comments. We acknowledge that selection bias is a concern for retrospective studies. However, bias does not necessarily detract from their value; such studies can address and answer important questions. We attempted to limit selection bias in our report1 by removing early deaths. This adjustment does not entirely eliminate selection bias and is not equivalent to the randomization necessary to definitively prove the benefits of adjuvant radiation. To this end, we indicated in our report, rather, that adjuvant radiation therapy clearly has no survival disadvantage, an outcome less prone to selection or statistical bias.1 This conclusion is noteworthy in light of the findings of the European Study Group for Pancreatic Cancer (ESPAC-1) trial.2 We read with great interest the study conducted by Hazard et al.3 However, there were notable differences in their methodology compared with ours. First, Hazard et al investigated patients who received neoadjuvant and/or adjuvant radiation therapy, whereas we limited our study to patients who received adjuvant external–beam radiation. Second, Hazard et al described the extent of disease using the American Joint Committee on Cancer (AJCC) staging system by incorporating both T and N classifications. Therefore, early–stage disease was categorized as T1-T2N0M0, whereas the T3-T4N0M0 group was comprised of patients with direct extension of disease. Additional groups included patients with lymph node–positive disease (regardless of T classification) and patients with disease with both direct extension and lymph node positivity. Using this stratification, Hazard et al concluded that radiation therapy was advantageous in patients with direct extension and/or regional lymph node involvement with respect to overall survival, and advantageous in patients with lymph node–positive disease and disease with both direct extension and lymph node positivity with respect to cause–specific survival. Similarly, our study demonstrated that the primary benefit of radiation therapy with respect to overall survival was noted in the group of patients with T3N0 disease. As such, the findings of our study1 and those of Hazard et al3 are not necessarily contradictory. Furthermore, we agree that the adequate retrieval and identification of lymph nodes may be relevant in the accurate staging and classification of patients with pancreatic cancer.4 In summary, our study and that of Hazard et al1, 3 suggest that adjuvant radiation therapy may improve survival in all patients who have undergone curative surgery for pancreatic cancer. We advocate the implementation of methodologically sound prospective trials to definitively prove our conclusions. Avo Artinyan MD, MS*, Joseph Kim MD*, * Department of General Oncology Surgery, City of Hope, National Medical Center, Duarte, California.