LETTERS TO THE EDITOR

To the Editor:

The article by Farrell et al, “Pancreatic resection combined with intraoperative radiation therapy for pancreatic cancer,”¹ may be overly optimistic about its conclusions. They are to be congratulated on their minuscule complication rate. However, I believe they have mainly demonstrated the safety of intraoperative radiation therapy, not its efficacy.

Using the Rocky Mountain Cancer Center data for comparable survival statistics may be misleading. The patients in this reference group all underwent pancreaticoduodenectomies, whereas only 71.5% of the study group had Whipple resections. The two patients with distal pancreatectomies may have had more localized or contained disease. With such a small number overall, this difference may significantly influence median or 5-year survival. A comparison by stage would further clarify this question, and a comparison by cell type would answer the related issue raised by the authors in their comments on survival. In addition, the patients in the study group received both intraoperative radiation therapy (IORT) and external beam radiation. Postoperative adjuvant radiotherapy, with or without 5-fluorouracil, has already been demonstrated to enhance local control, and therefore might prolong median survival with no impact on long-term results. How can the authors be certain that the effects demonstrated in their series are due to the benefit of IORT alone?

Further amplification of the differences between the study group and the controls would answer several other questions. Did any of the controls with malignancy receive adjuvant therapy, especially external beam radiation? Was adjuvant therapy related to complications in the control group? What were the survival curves for pancreatic and nonpancreatic carcinomas? And does IORT really protect against anastomotic leak or stricture?

As the authors indicate, their survival curve approaches the known statistic for survival at 5 years. It is unclear that their study group is large enough, or free enough from confounding factors, to declare IORT beneficial. I agree that this modality may be used without adding significant morbidity based on their data, but they have not demonstrated that it can prolong median survival when used in comparable populations. Their conclusion that IORT is at least as good as any other therapy may also be interpreted to read that we have yet to discover a uniformly useful adjunct to surgery for pancreatic cancer.

August 26, 1997

Christopher M. Huiras MD, FACS