BACKGROUND: Multiple methods of reconstruction after orbital exenteration for cancer have been described including skin grafts, regional flaps, and, most recently, microvascular free flaps. A unifying algorithm that incorporates all current methods has not been described.

OBJECTIVE: To evaluate the various methods of reconstruction and propose a comprehensive reconstructive algorithm based on the surgical defect, anticipated need for postoperative radiation therapy, and plans for postoperative prosthetic rehabilitation.

METHOD: A retrospective review of 75 cases (41 males, 34 females) involving orbital exenteration for malignancy at the University of Texas M. D. Anderson Cancer Center between 1999 and 2006 was performed. Defects were classified into three groups: 1) orbital exenteration only (25 patients), 2) extended orbital exenteration, involving removal of one or more orbital walls (43 patients), and 3) orbital exenteration with maxillectomy (7 patients). RESULTS: Reconstructive methods included skin grafts (7 split- and 11 full-thickness skin grafts), regional flaps (2 temporoparietal fascia and 4 temporalis muscle flaps), and microvascular free flaps (3 radial forearm, 25 anterolateral thigh, 23 and rectus abdominis myocutaneous free flaps). Reconstructions were classified as resulting in an “open” cavity (33 patients) in which a concave orbital socket that facilitates accommodation of an orbital prosthesis is produced or a “closed” cavity (42 patients) in which the orbital socket is filled with soft tissue to the level of the orbital rim. Twenty-one patients (28%) experienced complications (24% open vs. 31% closed cavity, p > 0.05; 22% skin graft reconstruction vs. 33% regional tissue reconstruction vs. 29% free flap reconstruction, p>0.05%). Complications included: infection (11%), sinonasal fistula (1%), partial graft loss (1%), flap loss (1%), donor site complication (5%), CSF leak (1%), wound dehiscence (1%), hematoma (3%), and osteoradionecrosis (3%). Thirty-eight patients (51%) received postoperative radiation therapy. Nine of these patients experienced complications at the surgical site, including 3 of 8 (38%) patients with skin graft reconstruction and 6 of 30 (20%) patients with regional or free flap reconstruction. Sixteen patients (21%) received an orbital prosthesis. Of this group, 7 patients (44%), wear their prosthesis regularly. Of the patients who received an orbital prosthesis, 7 had skin graft reconstruction (39% of patients reconstructed with a skin graft), 1 had temporoparietal fascia flap reconstruction (17% of patient reconstructed with a regional flap), 1 had radial forearm flap reconstruction, 5 had rectus abdominis myocutaneous flap reconstruction, and 2 had anterolateral thigh flap reconstruction (16% of patients reconstructed with a free flap). Three patients required a flap revision to accommodate their prosthesis, 2 with anterolateral thigh flaps and 1 with a rectus abdominis myocutaneous flap.

CONCLUSIONS: The method of reconstruction should be tailored to the defect as well as the postoperative needs of the patient. Skin grafts are often adequate when the defect is limited to the orbital contents alone and no postoperative radiation therapy is planned. Otherwise, regional or free flaps should be considered. When a prosthetic is planned, the goal should be to create an open cavity with a skin graft, regional flap, or thin free flap. Bulky fasciocutaneous or myocutaneous flaps are indicated when a closed cavity is preferred and no prosthetic is planned.