Journal Club - December 2004

The effect of centralization of primary surgery on survival in ovarian cancer patients.
Tingulstad S, Skjeldestad FE, Hagen B.
Obstet Gynecol. 2003 Sep;102(3):499-505. [Abstract]

Summary
Tingulstad and coworkers conducted a historical prospective case-control study in Norway to examine the effect on survival of centralized referral of ovarian cancer patients for primary surgery. Beginning in 1995, a separate prospective study was initiated using the Risk of Malignancy Index algorithm as a triage tool for referral of women with suspected ovarian cancer to the regional teaching hospital for surgery. Prior to 1995, referral of patients with suspected ovarian cancer was dependent solely on the judgment of clinicians in the surrounding community hospitals. From 1995 through 1997, 38 patients underwent primary surgery for ovarian cancer at the regional teaching hospital after referral from community hospitals and comprised the population of study cases. For comparison, 76 historical control patients with ovarian cancer, matched for age, stage of disease, and type of chemotherapy, were identified as having undergone primary surgery at the community hospitals during the period from 1992 through 1994. An analysis of primary surgical outcome for patients with advanced-stage disease revealed that control patients were significantly less likely to under optimal cytoreduction (24%) compared to cases referred to the regional teaching hospital (48%, p=0.04). There were no significant differences between the two groups with respect to the type and duration of chemotherapy received. Univariate analysis of overall survival time revealed that patients operated on at the regional teaching hospital had a significantly longer median survival time (21 months) compared to control patients (12 months, p=0.01). Using a stepwise Cox regression model of overall survival, the authors identified residual disease following primary surgery and the completeness of chemotherapy as independent predictors of survival. Although treating hospital only approached statistical significance in the multivariate model of overall survival, there was a statistically significant interaction between residual disease and treating hospital (cases/controls) (p=0.01). Compared to patients undergoing optimal cytoreduction at the regional teaching hospital (reference group), patients operated on at community hospitals had statistically significantly worse survival outcomes whether they were left with optimal (HR = 6.5, 95%CI = 2.0 to 21.3, p=0.002) or suboptimal residual disease (HR = 8.9, 95%CI = 3.2 to 24.3, p=0.001).

Discussion
This is an important study in several respects. First, it indirectly indicates that a coordinated referral system for women with suspected ovarian cancer is feasible using existing diagnostic algorithms (in this case the Risk of Malignancy Index). Secondly, this study reaffirms the previously documented observation that specialized training in radical surgery for ovarian cancer predicts a higher likelihood of an optimal surgical outcome for patients with advanced stage disease.

Conclusion
These data suggest that while an aggressive surgical approach resulting in optimal residual disease is an important determinant of survival, the overall management undertaken at the regional teaching hospital was associated with a positive effect on survival over an above the outcome of the initial surgical procedure. Increased efforts to centralize ovarian cancer care in centers with particular expertise should be undertaken universally.

Back to Journal Club Archives