(P 284) WOUND VAC TEMPORIZATION AFTER TUMOR RESECTION OF SOFT TISSUE SARCOMAS - A COST ANALYSIS IN A PROPENSITY-SCORE MATCHED POPULATION

Objective: Myxofibrosarcoma is a rare and locally aggressive soft tissue sarcoma with a high local recurrence rate. Although final histologic evaluation is the gold standard for margin assessment, results often take more than a week to come back and surgeons commonly rely on intraoperative frozen pathology, known for its low diagnostic accuracy. This often leads to positive (R1) margins after primary soft-tissue coverage while increasing the risk of local recurrence. At our center, we have modified our surgical approach and instead of performing immediate closure or soft-tissue reconstruction after tumor excision (single-stage excision/reconstruction [SS]), we now temporize the wound bed with vacuum-assisted closure (VAC) until final margins are negative. VAC temporization has shown superior local control rates compared to standard closure, with similar patient-reported outcomes. However, VAC use remains limited due to perceived higher costs of negative pressure wound therapy. Therefore, our study sought to compare the short- and mid-term costs between the SS and VAC groups.

Methods: Patients with a histologic diagnosis of myxofibrosarcoma who underwent either primary tumor resection or tumor bed excision and who had a minimum follow-up of one year were eligible for inclusion. Data on treatment costs were obtained from our institutional electronic data warehouse using the date of admission as the starting point. All billable items (including services) provided by our institution during the specified time period were included. We assessed both direct and indirect costs per episode of care. The main study outcomes were inpatient costs, 90-day costs, and one-year costs. We additionally calculated daily inpatient costs in order to adjust for the length of stay of each patient. Propensity score matching (PSM) based on age, sex, age-adjusted Charlson Comorbidity Index, and tumor size was conducted using a two-to-one nearest-neighbor matching without replacement and a caliper fixed at 0.1.

Results: A total of 139 patients underwent surgical resection of a myxofibrosarcoma at our institution. PSM was performed on the 50 patients with available cost data and 36 patients (13 in the SS group and 23 in the VAC group) were finally included. No significant differences in demographic and tumor characteristics were found between groups. We found no difference in median total inpatient costs, direct costs, and indirect costs between patients treated with SS excision/reconstruction and patients treated with VAC temporization. After accounting for length of stay, both median and mean daily inpatient costs were similar between groups (p=0.16 and p=0.26, respectively). We found higher median 90-day total costs in the VAC group (503.5 CU) compared to the SS group (265.5 CU, p=0.03). However, there were no differences in mean 90-day total costs (p=0.053), direct costs (p=0.052), and indirect costs (p=0.059) between groups. While median one-year total costs were also higher in the VAC group (503.5 CU) than in the SS group (272.8 CU, p=0.04), this was attributable to indirect costs. Notably, we found no difference in median one-year direct costs between groups (p=0.052). Median total number of tumor excisions was higher in the VAC group compared to the SS group (p=0.03). However, negative margins were achieved in all patients in the VAC group compared to 69% in the SS group (p=0.005). While oncologic outcomes were similar in the matched cohort, an additional analysis of the entire cohort revealed lower local recurrence rate (5.6% vs. 22.5%, p=0.02) and overall death (11.1% vs. 32%, p=0.02) in the VAC group.

Conclusion: VAC temporization had similar inpatient costs and post-operative outcomes to SS excision/reconstruction. While median 90-day and one-year costs were higher in the VAC group, mean costs did not differ. Future studies should use multi-institutional data to increase the statistical power and externally validate our findings.