Mandibular reconstruction outcomes for in-house patient-specific solutions

Segmental mandibulectomy and mandibular reconstruction are often performed for various benign and malignant head and neck conditions. Standard of care reconstruction involves titanium plate fixation with tissue transfer. The advent of computer-aided design and manufacturing (CAD/CAM) has enhanced aesthetic and functional outcomes in mandibular reconstruction by enabling patient-specific solutions like 3D-printed anatomic models. At an increasing number of institutions, these solutions can be produced in-house via point-of-care manufacturing. Since little has been published on the accuracy and outcomes of this approach, this study sought to evaluate the reconstructive accuracy and clinical outcomes of patients who received in-house patient-specific mandible models.

A retrospective chart review was conducted of 44 patients from a large midwestern academic medical center who received point-of-care patient-specific 3D printed models to assist in segmental mandibulectomy and reconstruction from December 2020 to June 2022. CAD/CAM models were produced from pre- and post-operative CT scans. Pre- and post-operative scans were aligned using a novel reference landmark-the maxilla. Measurements were taken by two different researchers at the mandibular condyles, coronoids, angles as well as a plane from the maxilla to the mandibular pogonion to determine reconstructive accuracy. Inter-rater reliability was assessed via intraclass correlation coefficient. Demographic, clinical, surgical, and radiographic variables were also collected to profile cohort characteristics and outcomes.

After exclusions due to poor or no post-operative imaging, 25 patients were included in the final analysis. Squamous cell carcinoma (n = 19) was the most common pathology, and males (n = 18) were represented more than females (n = 7). 96% (24/25) of patients had good plate adaptation and 96% (24/25) had good osteotomy adaptation. Reconstruction accuracy measured by comparing preoperative to postoperative anatomic alignment was very good, with an average absolute difference across all patients of only 3.10 mm. Inter-rater reliability between measurements was high with an average 0.98 intraclass correlation coefficient.

We present a novel method for measuring mandibular reconstruction accuracy through the use of the maxilla as the anatomic landmark. Furthermore, our profile of patients who underwent segmental mandibulectomy and reconstruction with the assistance of in-house produced 3D printed patient-specific models appears to result in suitable anatomic alignment of the reconstructed mandible and produce good clinical outcomes.