Background: A technical goal of mandibular reconstruction is to optimize reconstruction plate contouring to maximize bony contact and narrow the distance between the reconstruction plate and osseous free flap. It is hypothesized that there is an association between plate-related complications and reconstruction plate contouring parameters. However, no study has evaluated this hypothesis quantitatively.
Objective: The purpose of this study is to retrospectively analyze the impact of plate contouring parameters on complication rates in oromandibular reconstruction using 3D modeling.
Methods: Postoperative computed tomography (CT) images were obtained from patients with oral squamous cell carcinoma with mandibular invasion who underwent segmental mandibulectomy and osseocutaneous/osseus free flap reconstruction at the Toronto General Hospital between 2003-2014. Patients were included if postoperative CT scans were performed within one year of surgery. Computer generated three-dimensional models of the reconstructed mandible and mandibular hardware were generated using Mimics v18.0 (Materialise, Leuven, Belgium). Quantitative analysis of mandibular reconstruction was performed to calculate (1) plate surface contact area, (2) overall percentage of plate contact and (3) mean plate to bone gap, defined as the distance between the reconstruction plate and bone. Demographic data and details of postoperative course were abstracted retrospectively including complications (surgical site infection, hardware exposure, intraoral dehiscence, neck dehiscence and orocutaneous fistula) and further treatment (i.e. postoperative radiotherapy). Univariate and multivariate analyses (multiple linear regression) were performed and included factors such as demographics, smoking history, diabetes and radiotherapy.
Results: Ninety-four patients were identified (M=58; F=36), and included in our study. Types of osseocutaneous/osseous free flaps used for reconstruction included the fibular (n=54; 57%) and scapular free flaps (n=40; 43%). Postoperative complications included surgical site infection (n=18), hardware exposure (n=14), intraoral dehiscence (n=12), neck dehiscence (n=15) and orocutaneous fistula (n=9), with some patients having more than one complication. On multivariate analyses, adjusting for smoking history, radiation and diabetes, a larger mean plate to bone gap was associated with both plate exposure (Exposed:1.67±0.5mm vs unexposed:1.11±0.4mm; p=0.02), and formation of orocutaneous fistula (Fistula: 1.49±0.7mm vs No fistula:1.14±0.4mm; p=0.02). The overall plate surface contact area and percentage of plate contact were not associated with postoperative complications.
Conclusion: This study provides quantitative confirmation that reconstruction plate contouring parameters are associated with the development of plate exposure and orocutaneous fistula. Improving plate conformance to bony constructs is a critical goal when fashioning a reconstruction plate. Overall plate surface contact area and percentage of plate contact did not impact complication rate.