Larger free flap size is associated with increased complications in lower extremity trauma reconstruction

Background: Free flap reconstruction after lower extremity trauma remains challenging with various factors affecting overall success. Increasing defect and flap size have been demonstrated to be a surrogate for overall injury severity and correlated with complications. In addition, larger free flaps that encompass more tissue theoretically possess high metabolic demand, and may be more susceptible to ischemic insult. Therefore, the purpose of our study was to determine how flap size affects microsurgical outcomes in the setting of lower extremity trauma reconstruction. Methods: Retrospective review of 806 lower extremity free flap reconstructions performed from 1979 to 2016 among three affiliated hospitals: a private university hospital, Veterans Health Administration Hospital (VA), and a large, public hospital serving as a level 1 trauma center for the city. Soft tissue free flaps used for below the knee reconstructions of traumatic injuries were included. A receiver operating curve (ROC) was generated and Youden index was used to determine the optimal flap size for predicting flap success. Based on this, flaps were divided into those smaller than 250 cm² and larger than 250 cm². Partial flap failure, total flap failure, takebacks, and overall major complications (defined as events involving flap compromise) were compared between these two groups. Multivariate logistic regression was performed to determine whether flap size independently predicts complications and flap failures, controlling for injury-related and operative factors. Results: A total of 393 patients underwent lower extremity free tissue transfer. There were 229 flaps (58.2%) with size <250 cm² and 164 flaps (41.7%) ≥ 250 cm². ROC analysis and Youden index calculation demonstrated 250 cm² (AUC 0.651) to be the cutoff free flap for predicting increasing flap failure. Compared to flaps with less than 250 cm², larger flaps were associated with increased major complications (33.6% vs. 50.0%, p =.001), any flap failure (11.8% vs. 25.0%, p =.001) and partial flap failure (4.8% vs. 14.6%, p =.001). Logistic regression analysis controlling for age, flap type, era of reconstruction, number of venous anastomoses, presence of associated injuries, presence of a bone gap, vessel runoff, and flap size identified increasing flap size to be independently predictive of major complications (p =.05), any flap failure (p =.001), partial flap failure (p <.001), and takebacks (p =.03). Subset analysis by flap type demonstrated that when flap size exceeded 250 cm², use of muscle flaps was associated with significantly increased flap failure rates (p =.008) while for smaller flap size, there was no significant difference in complications between muscle and fasciocutaneous flaps. Conclusion: Increasing flap size is independently predictive of flap complications. In particular, a flap size cutoff value of 250 cm² was associated with significantly increased flap failure and complications particularly among muscle-based flaps. Therefore, we suggest that fasciocutaneous flaps be utilized for injuries requiring large surface area of soft tissue reconstruction.