Three-dimensional computed tomographic volume rendering techniques in endoscopic thoracoplasty.

[BACKGROUND CONTEXT] Thoracoplasty is occasionally necessary to achieve an acceptable cosmetic result in the presence of a rib hump, especially in previously fused young adults with scoliosis. This usually requires the resection of four to five rib segments, and the morbidity associated with open posterior resection or of open anterior resection by means of thoracotomy is often considerable, apart from leaving an unsightly scar. We thought the use of an endoscopic internal rib resection technique would achieve the desired result with less morbidity. Our experience with using the technique of three-dimensional (3-D) computed tomographic (CT) volume rendering to plan our portals and releases for endoscopic scoliosis correction led us to believe the same techniques could be adapted to plan and endoscopically carry out the thoracoplasty.

[PURPOSE] To define the utility of 3-D CT volume rendering of the spine and thorax in the preoperative planning of endoscopic rib resection.

[STUDY DESIGN] A prospective evaluation of the utility of 3-D CT volume rendering for preoperative planning in patients with scoliosis undergoing endoscopic thoracoplasty for correction of rib humps.

[PATIENT SAMPLE] Four consecutive patients with previously fused scoliotic spines and pronounced right-sided rib humps requiring operative correction were selected.

[OUTCOME MEASURES] Outcome was assessed at a minimum follow-up of 6 months by clinical examination, patient satisfaction with the clinical result, and repeat helical CT scanning with 3-D reconstruction.

[METHODS] Four patients with previously fused scoliotic spines and pronounced rib humps underwent helical CT scanning with 3-D volume rendering, prior to endoscopic corrective surgery. All four patients had right-sided rib humps requiring corrective rib resection for cosmetic reasons. Using the technique of 3-D volume rendering, a vector plane was created to mirror the left scapula, and its intersections on the right chest wall were noted. The ribs to be resected were marked, and the length of rib resection was measured from the vector plane's intersection points with the ribs. In this way an estimate of the resection required to achieve the desired final position of the right elevated scapula could be determined. Entry portals were also estimated with vector lines to achieve optimal access to each rib. During surgery, the portal sites were assessed for access to the selected ribs. Also, the extent of rib resections was compared with the estimates. The final clinical outcome was assessed by clinical examination, patient satisfaction with the cosmetic result, and repeat helical CT scanning with 3-D reconstruction.

[RESULTS] The male to female ratio was 1:3, and the average age was 21 years. Our average estimated blood loss was 307 ml and average hospital stay was 4.75 days. The estimated portal sites were accurate and did provide for direct access to each selected rib involved in the deformity. We were able to resect the ribs at the points suggested by 3-D CT volume rendering, with the lengths of our resected segments matching our preoperative estimates. In all cases the elevated right scapula did descend into the rib resection bed, thus balancing the shoulder heights. An excellent cosmetic result was achieved in all cases as evaluated by clinical examination, patient's perception, and repeat helical CT scanning.

[CONCLUSION] The technique of 3-D CT volume rendering with vector plane estimates provides a reliable estimate of the rib resection required to achieve a cosmetically acceptable correction of the rib hump through minimally invasive techniques.