Value of carcinoembryonic antigen monitoring in curative surgery for recurrent colorectal carcinoma.

PURPOSE: This study is designed to review a carcinoembryonic antigen (CEA)-driven postoperative protocol designed to identify patients suitable for curative reresection when recurrent colorectal cancer is identified. METHODS: A total of 285 patients who were operated on for colon or rectal carcinoma between 1981 and 1985 were evaluated (with CEA levels) every two months for the first two years, every three months for the third year, every six months for years 4 and 5, and annually thereafter. CEA levels above 5 microg were considered abnormal and were evaluated with diagnostic imaging and/or endoscopy. RESULTS: Follow-up was available for 280 patients (98.2 percent). Distribution of patients by Astler-Coller was: A, 14 percent; B1, 20 percent; B2, 39 percent; C1, 5 percent; C2, 21 percent. There were 62 of 280 patients (22 percent) who developed elevated CEA levels, with 44 patients who demonstrated clinical or radiographic evidence of recurrence. Eleven patients were selected for surgery with curative intent (4 hepatic resections, 1 pulmonary wedge resection, 2 abdominoperineal resections, 2 segmental bowel resections, and 2 cranial metastasectomies). Three of 11 patients (27 percent) benefited and have disease-free survivals greater than 60 months. Of the 223 patients without elevated CEA, 22 (9.9 percent) had recurrent cancer without any survivors. Overall, 3 of 285 patients (1.1 percent) were cured as a result of CEA follow-up. CONCLUSION: CEA-driven surgery is useful in selected patients and can produce long-term survivors.

Variations in colon and rectal surgical mortality. Comparison of specialties with a state-legislated database.

PURPOSE: This study was designed to examine variations in operative mortality among surgical specialists who perform colorectal surgery. METHODS: Mortality rates were compared between six board-certified colorectal surgeons and 33 other institutional surgeons using comparable colorectal procedure codes and a validated database indicating patient severity of illness. Thirty-five ICD-9-CM procedure codes were used to identify 2,805 patients who underwent colorectal surgery as their principal procedure between July 1986 and April 1994. Atlas, a state-legislated outcome database, was used by the hospital's Quality Assurance Department to rank the Admission Severity Group (ASG) of 1,753 patients from January 1989 to April 1994 (higher ASG, 0 to 4, indicates increasing medical instability). RESULTS: Colorectal surgeons had an eight-year mean in-hospital mortality rate of 1.4 percent compared with 7.3 percent by other institutional surgeons (P = 0.0001). There was a significantly lower mortality rate for colorectal surgeons compared with other institutional surgeons in ASG 2 (0.8 and 3.8 percent, respectively; P = 0.026) and ASG 3 (5.7 and 16.4 percent, respectively; P = 0.001). CONCLUSIONS: Board-certified colorectal surgeons had a lower in-hospital mortality rate than other institutional surgeons as patients' severity of illness increased.