Use of a 3D-Printed Patient-Specific Surgical Jig and Ready-Made Total Sacral Endoprosthesis for Total Sacrectomy and Reconstruction
Objective. In the present study, the authors aimed to optimize the workflow of utilizing a 3D printing technique during surgical treatment for malignant sacral tumors, mainly on preparation of patient-specific surgical jigs and ready-made 3D-printed total sacral endoprosthesis. Methods. Three patients with a malignant sacral tumor received total sacrectomy with preoperative design of a patient-specific 3D-printed cutting jig and endoprosthetic reconstruction. Size of ready-made 3D-printed endoprosthesis was determined based on preoperative images, planned surgical margin, and size of the endoprosthesis. A patient-specific cutting jig was designed with a bilateral cutting slot matching the bilateral planes of the implant precisely. The tumor was removed en bloc through a single posterior approach only, being followed by reconstruction with ready-made total sacral endoprosthesis. Results. The mean time for preoperative design and manufacture of the surgical jig was 6.3 days. Surgical jigs were successfully used during surgery and facilitated the osteotomy. The mean operation time was 177 minutes (range 150-190 minutes). The mean blood loss was 3733 ml (range 3600-4000 ml). R0 resections were achieved in all the three cases proven by pathology. Evaluation of osteotomy accuracy was conducted by comparing preoperative plans and postoperative CT scans. The mean osteotomy deviation was 2.1 mm (range 0-4 mm), and mean angle deviation of osteotomy was 3.2° (range 0-10°). At a mean follow-up of 18.7 months, no local recurrence was observed. One patient had lung metastasis 15 months after surgery. Two patients were alive with no evidence of the disease. Conclusions. The patient-specific surgical jig and ready-made 3D-printed total sacral endoprosthesis can shorten the surgical preparation time preoperatively, facilitating accurate osteotomy and efficient reconstruction intraoperatively. The workflow seems to be feasible and practical.