Obstructing Colonic Cancer: Failure and Survival Patterns over a Ten-Year Follow-Up After One-Stage Curative Surgery Paolo G. Setti Carraro, M.D.,* Marco Segala, M.D.,* Bruno M. Cesana, M.D.,t Giorgio Tiberio, M.D.* From the *Istituto di Chirurgia d'Urgenza, Universit& di Milano, Ospedale Maggiore Policlinico, and tLaboratorio di Epidemiologia, Ospedale Maggiore Policlinico, Instituto di Ricovero e Cura a Carattere Scientifico, Milan, Italy BACKGROUND: Large-bowel cancers that present as ob- structing lesions have a poor prognosis. However, little is known of the reasons for the dismal survival and of failure patterns after potentially curative treatment. METHOD: An observational study and multivariate analysis were con- ducted to identify determinants of survival and to compare recurrence patterns between obstructing and nonobstruct- ing tumors after primary resection and anastomosis as cur- ative treatment. RESULTS: Over a period of ten years (1980- 1989), 528 patients with colonic cancer were treated at one institution. The cancer was obstructing in 179 cases and nonobstructing ha 349. One-stage primary resection and anastomosis as curative treatment were performed in 107 obstructed and 256 nonobstructed patients. Three hundred thirty-six potentially cured survivors (94 in the former group and 242 in the latter) were followed for a median of 55 months. During follow-up, local recurrence occurred in 37 patients (12 obstructed (12.8 percent) and 25 nonob- structed (10.4 percent), P = 0.44) and metastatic disease in 68 (25 obstructed (27.6 percent) and 43 nonobstructed (17.8 percent), P = 0.029). Multivariate analysis of survival showed that age over 70 years, Dukes stage, histologic grade, and recurrence were the only prognostic factors. No statistically significant determinant turned out for local recurrence, whereas at multivariate analysis for met- astatic and overall relapse, Dukes stage, positive nodes, and obstruction remained independent prognostic fac- tors. CONCLUSIONS: After one-stage emergency curative treatment, patients presenting with obstructing tumors of the colon have a smaller survival probability than that of patients with nonobstructing lesions. Local recurrence pattern is similar between groups. Conversely, obstruc- tion, along with pathologic stage and positive nodes, carries a significantly higher risk of metastatic tumor recurrence and death. [Key words: Colon cancer; Ob- struction; Surgery; One-stage surgery; Curative surgery; Survival; Follow-up; Recurrence] Setti Carraro PG, Segala M, Cesana BM, Tiberio G. Obstruct- ing colonic cancer: failure and survival patterns over a ten-year follow-up after one-stage curative surgery. Dis Co- lon Rectum 2001;44:243-250. • Address reprint requests to Dr. Setti Canaro: Istituto di Chirurgia d'Urgenza, Ospedale Maggiore Policlinico, IRCCS,Via F. Sforza 35, 20122-Milan, Italy. T umors of the large bowel that present with ob- struction have a poor prognosis, the five-year survival probability ranging only from 0.12 to 0.31.1-6 Even in patients undergoing curative treatment, the long-term survival probability does not exceed 0.23 to 0.40,1, 2, 7-10 which compares poorly with the 0.6011' 12 five-year survival probability for nonobstructing le- sions. The features that determine the poor long-term outcome after potentially curative treatment are not clearly understood and appear to be independent of variables such as tumor stage and histologic differen- tiation. 6 In addition, little is known about whether obstruction influences recurrence patterns. To identify determinants of survival and to assess whether any difference exists in patterns of recur- rence between obstructing and nonobstructing tu- mors, an observational study was performed in a selected group of patients who presented with or without obstruction and underwent one-stage cura- tive treatment. Selection was aimed to remove factors of potential prognostic relevance such as resectability rate and operative mortality. PATIENTS AND METHODS 243 Between 1980 and 1989, 528 consecutive patients with colonic cancer were admitted to one unit. Per- forated patients had been excluded from analysis• Obstruction was observed at presentation in 179 (33.9 percen0 of the patients (OP group), and the lesion was nonobstructing in the other 349 (NOP group). The diagnosis of obstruction was based on clinical, radiologic, and operative criteria as suggested by Fielding et al. 13 One hundred sixty-five patients of the original cohort (72 OP and 93 NOP patients) were