LETTERS TO THE EDITOR

To the Editor:

We read with great interest the paper titled “Does Prophylactic Octreotide Decrease the Rates of Pancreatic Fistula and Other Complications after Pancreaticoduodenectomy?” by Charles Yeo et al, ¹ and presented at the annual meeting of the American Surgical Association. However, we have some reservations on various aspects of the study design, analysis of data, and potential areas of bias.

1) The primary endpoints of the study were “pancreatic fistula, other complications and death,” but it appears that a statistical power calculation was made only for pancreatic fistula. If, for example, “death” was also a primary study endpoint, and given the publications of the Baltimore group with a death rate of 1%, one would have needed more than 1,000 patients to be randomized to examine a difference in the death rate using octreotide.

2) In the discussion after the paper, the primary author admits to a bias on the premise that he expected a negative trial at the outset. In this context, it is pertinent to note that the study was terminated (by whom?) after recruitment of only 64% of the patients, which was probably insufficient to reach a statistically valid result.

3) Pancreatic fistula was judged on the tenth postoperative day or later. How was the primary endpoint recorded when more than 50% of patients were discharged on postoperative day 9 or earlier?

4) The authors state that the management of drains, including their removal, was left to the discretion of the primary attending surgeon. Did this study have any independent monitoring committee (excluding the primary author) that ensured a rigid implementation of the study protocol?

5) The trial was designed as a double-blind study. However, the “placebo” was saline and the first author states in the discussion that the octreotide injections were more painful than the saline injections. In the multicenter European trials, 2,3,4,5 taking into consideration the fact of pain at the injection site, the octreotide solvent that caused similar pain was used as a placebo. Thus, can the Yeo, et al. study be truly termed double-blind?

6) The withdrawal rate of this study was 45% (172 of 383) including 40 patients that were excluded because they did not receive “at least a 5-day course of octreotide study drug.” How do the authors reconcile the study quality with the exclusion of these 40 patients from the final analysis? Against this background, how did the authors authenticate that the remaining patients received the study drug?

7) The authors state that “the octreotide and control saline placebo were identical in appearance, volume and labeling, thereby masking the nursing staff...” We are aware that octreotide (250 μg in a volume of 250 μL) is only available as a 5 ml vial and once opened, the stability is maintained for 2 weeks. Therefore, two 5 ml vials would be necessary for 22 injections (i.e., requirement of the study protocol) since a maximum of 20 injections are possible from a single vial. From the statement of costs incurred ($61/injection × 22 injections × 104 octreotide group patients = $139,568), it appears no octreotide was wasted. Does this imply that some patients shared the same 5 ml vial? If so, the double blind nature of the study is compromised.

The observations, based on our own past experience, raise an additional doubt; namely, that the study drug was repacked by the Investigational Drug Pharmacy from the commercially available 5 ml vials. If so, how was the therapeutic value of octreotide ensured? If the drug was not being replaced and the vials were not shared, then the stated costs incurred are greatly disproportionate to those mentioned in the article.

8) Regarding costs, the authors carefully present the exact costs involved for the whole study. They state that the “total cost of this study was $162,383.” It has been mentioned that “the patients received the octreotide study drug subcutaneously before surgery.” Becaues 383 patients were enrolled in this study, it is presumed that the study drug was prepared and reserved for all these patients. In this situation, we suspect that either the drug reserved for those 118 patients who underwent a total pancreatectomy was used for other study patients or the calculated costs in this study are incorrect. Through knowing that the first author is very precise in his cost analysis, there is obviously an error (because only 104 of 383 patients have been considered in the cost analysis, and the total comes to $139,568).

9) In 1999, the same authors published another randomized controlled trial (also presented at the American Surgical Association) about radical versus conservative pancreaticoduodenectomy. ⁶ In this paper, they state the study was conducted “between April 1996 and December 1997.” They mentioned that “as part of an ongoing clinical trial evaluating pancreatic fistula and other complications, approximately 25% of the patients in this series received postoperative octreotide (250 μg subcutaneously every 8 hours) for 7 days.”

A similar article by the same authors ⁷ states that “between January 1994 and December 1997 inclusive, pancreaticoduodenectomy was performed in 597 consecutive patients,” and that “octreotide was not used prophylactically in these patients.” However, in the 1999 paper, ⁶ the authors refer to the period between April 1996 and December 1997, and state that “as part of an ongoing clinical trial” comparing pancreaticoduodenectomy with and without extended retroperitoneal lymphadenectomy “approximately 25% of the patients in this series received postoperative octreotide.”

In the present paper ¹ there is no mention of these studies. ^6,7 We agree with the authors conclusion that “in their setting” octreotide is not helpful and not cost-effective. However, we would appreciate clarification of the issues we have raised.

M. W. Buchler, MD
C. Bassi, MD
A. Fingerhut, MD
I. Klempa, MD