Clinical trials are mandatory for improving surgical cancer care.

Abstract

IN THIS ISSUE OF THE JOURNAL, WEEKS AND COLLEAGUES for the Clinical Outcomes of Surgical Therapy (COST) Study Group present the results of a multicenter randomized trial comparing quality-of-life (QOL) outcomes for patients with colon cancer who had open colectomy vs laparoscopic-assisted colectomy (LAC). While the findings that LAC provided only modest advantages compared with open colectomy are important for surgeons treating patients with colon cancer, the study also emphasizes the need for high-quality, rigorous randomized trials for improving surgical cancer care and underscores the importance of physicians to increase their efforts to enroll patients in such studies. The COST study was designed to compare the safety, efficacy, QOL outcomes, and cost of LAC vs standard open colectomy in patients with primary colon adenocarcinoma. The primary objective of the COST study is to determine whether disease-free survival and overall survival following LAC and open colectomy are equivalent. Mobilization of the bowel is conducted laparoscopically, and resection and anastomosis of the bowel are performed externally in the usual fashion. Several uncontrolled case series have suggested that patients undergoing LAC for a variety of conditions (such as diverticular disease) experience more rapid return of bowel function compared with conventional open colectomy. These findings have challenged the traditional approach of waiting until the passage of flatus, a bowel movement, or both, before feeding patients following open colectomy. However, a recent study demonstrated that 65 (73%) of 89 patients tolerated early oral feeding on day 1 or 2 following open colorectal surgery for various conditions. Such a single-institution series questions the theoretical benefits of the laparoscopic approach including reduced length of stay because of an earlier feeding pattern. Although the COST study is ongoing, the target accrual of patients needed to detect a predefined clinically meaningful QOL difference between treatment groups has been reached. For evaluation of the primary study end point, disease-free survival, the study design called for accrual of 1200 patients. However, the data monitoring committee recognized that the entire cohort would not be required to evaluate QOL end points. After a per-protocol planned interim analysis of the first 200 enrolled patients to estimate the variability in the pain distress item and the global rating scale, more patients were required to achieve the criteria for the global rating scale. The total target accrual of 416 patients for the QOL analysis was established. Between September 1994 and February 1999, 449 consecutive evaluable patients were enrolled in the QOL component of the COST study, and were surveyed with standardized, previously validated instruments for assessing symptoms and QOL among patients with cancer. The findings from the COST study demonstrate that compared with open colectomy, the laparoscopic-assisted procedure for colon cancer resulted in a statistically significant but clinically modest decrease in the duration of in-hospital postoperative analgesia and in length of stay (0.8 days). These differences did not translate into statistically significant improvements in patients’ QOL or symptoms in either the immediate postoperative period or during the 2-month follow-up. Previous studies with a QOL component have found some benefit for minimally invasive surgery, most commonly in surrogate end points, such as postoperative pain or analgesic requirements, and the magnitude of the differences observed in the COST study are consistent with these reports. However, as the authors point out, any QOL benefit must be weighed against the effectiveness of LAC in curing cancer. To their credit, the COST Study Group demanded more definitive evidence of a meaningful benefit than most prior studies in patients with nonmalignant conditions. Moreover, while advocates of the superiority of LAC for QOL outcomes might challenge the COST study on the basis of sample size, the target patient enrollment was calculated prospectively to provide adequate power to detect differences regarded as clinically meaningful.