Objective  To explore the methods for the reconstruction of lateral skull base defects after radical tumor resection.

  Methods  Patients who underwent lateral skull base radical tumor resection and reconstruction from January 1, 2012 to December 31, 2019 in the Department of Stomatology and Otorhinolaryngology in Peking Union Medical College Hospital, were enrolled in this retrospective study. These patients were divided into three subgroups depending on the type of defects. Group I, patients without skin defects; Group Ⅱ, patients with skin defects; Group Ⅲ, patients with larger jaw defects. Different reconstructive methods were applied for the three groups.

  Results  Sixty-two patients (63 procedures of defect repair) were enrolled in this study. The survival rate of tissue flap was 98.41%(62/63), and the primary wound healing rate was 90.48% (57/63). There were 34 patients in Group I. Among them, 6 cases received free fat transplantation; 14 cases were repaired with fat transplantation combined with local flaps; 13 cases were repaired with sternocleidomastoid muscle flap or combined temporal muscle flap; 1 case was repaired with temporal muscle flap combined with submandibular gland flap. Among them, two patients had postoperative cerebrospinal fluid leakage and delayed healing after dressing change. All patients had no postoperative infection and fat liquefaction, and the wounds healed by primary intention. The survival rate of tissue flap was 100%, and 94.12% (32/34) of the wounds healed by primary intention. There were 23 patients in Group Ⅱ. Five patients whose skin defects were less than 3 cm were reconstructed with local tissue flaps combined with advancement flaps; all flaps survived completely and the wounds healed by primary intention. The survival rate of tissue flap was 100%, and the primary wound healing rate was 100%. For those skin defects greater than 3 cm, free anterolateral thigh flaps were applied in 8 patients, pedicle flaps such as pectoralis flaps in 6 patients, 2 cases used forearm flap combined with pedicle skin flap or local tissue flap, 2 cases had pectoralis major myocutaneous flap or combined cervical and facial propulsion skin flap and submandibular gland flap, 1 case had rectus abdominis skin flap. One of the patients who had undergone high-dose radiotherapy developed venous embolism 3 days after repair with the anterolateral thigh myocutaneous flap, which was later repaired with pectoralis major myocutaneous flap rescue, and the remaining flaps survived completely; 1 patient with anterolateral thigh flap transplantation developed recipient hematoma on the second postoperative day, and the wound recovered after hematoma evacuation. One patient with rectus abdominis muscle repair developed cerebrospinal fluid leakage after surgery, and the skin lesions had delayed healing after dressing change. Two cases were repaired with pectoralis major myocutaneous flap and other pedicle skin flaps, and there was a small area of necrosis at the distal end. The wounds had delayed healing after dressing change. The wounds of the remaining patients healed at first stage; the first-stage wound-healing rate was 84.21% (16/19), and the flap survival rate was 94.74%(18/19). There were 5 patients in Group Ⅲ; 2 cases were reconstructed with fibular flap; 2 cases were reconstructed with free rib graft combined with temporal flap; 1 patient underwent femur-anterolateral thigh osteomyocutaneous free flap reconstruction. There were 10 patients whose tumors invaded the bones of the skull base and caused intracranial and external communication after surgery. Seven of them had no postoperative complications, and the remaining three had post-operative cerebrospinal fluid leakage, which delayed healing after dressing changes. There were 24 patients received simultaneous facial nerve reconstruction, including 12 cases reconstructed with greater auricular nerve transplantation, 8 cases with hypoglossal nerve, 3 cases with masseter nerve (+/-) greater auricular nerve, and 1 case with lateral femoral cutaneous nerve. Thirteen patients were followed up, and all had varying degrees of recovery of facial nerve function.

  Conclusions  Different reconstruction regimens should be taken according to different types of defects after radical tumor resection. The patients without skin defects can be reconstructed with the temporal muscle flap combined with autologous fat grafting or local tissue flaps. The patients with skin defects smaller than 3 cm can be reconstructed with the cervicofacial advancement flap combined with local tissue flaps. The most reliable free flap for lateral skull base reconstruction is the anterolateral thigh flap for the patients with skin defects larger than 3 cm. For those patients who were not suitable for microsurgical reconstructive operation, the pectoralis major muscle flap repair should be selected. If the defect is above the zygomatic arch, it is better to choose the pectoralis major muscle flap combined with cervicofacial advancement flap or local tissue flaps. The bone defect should be reconstructed appropriately if the patients presented with a large jaw defect. If conditions permit, the facial nerve defect reconstruction should be applied simultaneously.