Effect of perioperative blood transfusion on recurrence and survival in 232 primary high-grade extremity sarcoma patients.

Overview

abstract

BACKGROUND: Allogeneic blood transfusion (BT) has been implicated as an unfavorable factor influencing cancer recurrence and overall survival. METHODS: To investigate this, 232 consecutive localized, high-grade extremity soft tissue sarcoma (STS) patients admitted between January 1, 1983, and December 31, 1989, were analyzed from our prospective database by univariable and Cox multivariable statistical methods. RESULTS: Twenty-eight patients developed a local recurrence (LR). Factors found significantly unfavorable for the rate of developing an LR by uni- and multivariable tests were age > 60 years and positive microscopic margin. Eighty-nine patients developed a distant metastasis (DM) and 72 patients died of their tumor. Median follow-up of survivors was 48 months. Unfavorable factors for DM and tumor mortality (TM) by univariable analysis included large size, deep tumor (that involved or was below the superficial fascia), positive microscopic margin, invasion of a vital structure, operative blood loss, duration of operation, and perioperative BT (whole blood or packed cells -24 to +48 h of curative operation). Multivariable analysis found large size, deep tumor, and positive margin significant independent unfavorable factors for DM and TM. The effect of BT was not a significant independent prognosticator for LR, DM, or TM by multivariable analysis (p = 0.26, 0.56, 0.08, respectively). The only factor that was found to be significant in a multivariable analysis of factors contributing to postmetastasis survival was time < 6 months until metastasis (p = 0.008). BT had no significant impact on postmetastasis survival (p = 0.42). There was a significant association between BT and deep, large tumors. As the size of deep tumors increased from < 5, > or = 5 < 10, > or = 10 < 15, or > or = 15 cm, the amount transfused was 15, 16, 49, and 68% (p < 0.00001). Also, BT was significantly (p < 0.005) associated with low hematocrit at initial diagnosis, blood loss during surgery, and the length of the surgical procedure. CONCLUSIONS: These data emphasize the importance of size, depth, and margin on distant recurrence and death for localized high-grade extremity STS. In the absence of a randomized trial, the impact of allogeneic blood transfusion would appear to be due to its strong association with large size and deep tumor invasion. This study also highlights the importance of a multivariable analysis and long-term follow-up to better define this controversial question.