Can postoperative surveillance with serial CEA immunoscintigraphy detect resectable rectal cancer recurrence and potentially improve tumor-free survival?

Background: This study was performed to determine if postoperative serial m onitoring of rectal cancer patients can be performed with an immunoscintigr aphic imaging test for carcinoembryonic antigen (CEA). It was also of inter est to assess whether this test, in combination with standard monitoring pr ocedures used in an intensive surveillance plan, can result in the identifi cation of surgically salvageable patients. Study Design: Forty consecutive resected Dukes' B and C rectal cancer patie nts underwent a prospective, single-institution, surveillance trial of phys ical examination (including digital rectal examination), endoscopy, CT of t he abdomen and pelvis, liver ultrasound, chest x-ray, blood CEA, and CEA im munoscintigraphy with arcitumomab (CEA-Scan, Immunomedics, Morris Plains, N J) every 6 months for the first 2 years and every 12 months for the next 3 years after initial operation. Outcomes were compared with those from a sim ilar group of 69 patients treated previously at the same institution but wi thout CEA imaging. Results: A total of 219 CEA imaging studies were performed without any sign ificant adverse effects or immune responses, and resulted in lesion sensiti vity, specificity, accuracy, and positive and negative predictive values of 94.1%, 97.5%, 97.3%, 76.2%, and 99.5%, respectively. Of the 40 patients, 1 6 developed 22 surgically confirmed local or distant recurrences, and CEA i maging correctly disclosed 82% of these lesions pre-operatively. All of the patients found to have recurrences had at least one tumor site by CEA imag ing; only 6 of 16 had elevated blood CEA titers. On a patient-basis, there was a sensitivity of 100%, a specificity of 79.2%, an accuracy of 87.5%, an d positive and negative predictive values of 76.2% and 100%, respectively. The potential therapeutic benefit of serial arcitumomab imaging is suggeste d by the fact that 6 of 16 patients (37.5%) with recurrence underwent poten tially curative second-look operations, compared with 6 of 69 (8.7%) of a c omparable population studied at this institution during an earlier 6-year p eriod, using all of the same tests except CEA imaging. None of the patients in this historic control group survived more than 21 months, although the mean survival of the six patients resected for cure in the study population was 35 months (range 11 to 69 months). During 6 years of followup, three o f the six re-resected patients eventually died of cancer recurrence, two di ed from other causes (and were confirmed by necropsy to be tumor-free), and one patient is still free of disease in the sixth year. CEA scanning appea red to be more predictive of recurrence than blood CEA testing or other dia gnostic modalities. Conclusions: Arcitumomab inclusion in intensive surveillance of patients wi th resected rectal cancer can disclose tumor recurrence at a stage that all owed surgical salvage therapy in 37.5% of the 16 patients with recurrence w ho had second-look surgery, and in 13% the patients were free of disease du ring longterm follow-up. This pilot study suggests that a randomized prospe ctive trial comparing standard surveillance procedures to the use of CEA im aging added thereto should be undertaken. (J Am Coll Surg 2000;191:511-518, (C) 2000 by the American College of Surgeons).