Objective  Periacetabular metastasis treatment remains diverse without a standard protocol. This study aimed to propose a novel surgical strategy for periacetabular metastases based on a modified Harrington classification.

  Methods  We reviewed the data of 283 patients with periacetabular metastases who underwent surgeries at Peking University People’s Hospital from June 2003 to September 2021. There were 146 men and 137 women, and the mean age of the patients was (56.2±12.4) years. The most common histological diagnoses were metastatic lung cancer (68 cases), renal cancer (43 cases), and breast cancer (38 cases). A modified surgical classification for periacetabular metastases based on the Harrington classification was proposed. Surgeries were performed according to the original Harrington classification for patients with class I, Ⅱ, and Ⅳ lesions. For Harrington class Ⅲ lesions, three subtypes were further categorized based on the involvement of bone destruction and soft tissue mass volume; class Ⅲa lesions had bone destruction distal to the inferior border of the sacroiliac joint without giant soft tissue mass, class Ⅲb lesions had bone destruction extended proximally to the inferior border of the sacroiliac joint without a giant soft tissue mass, and class IIIc lesions had a giant soft tissue mass. Intralesional excision followed by the placement of Steinmann pins/screws with cemented total hip arthroplasty (THA) was performed for class Ⅲa and a few Ⅲb lesions, and en bloc resection followed by modular hemipelvic endoprosthesis replacement was performed for class Ⅲc lesions and the remaining Ⅲb lesions.

  Results   Among 283 patients, 279 underwent limb salvage surgeries. All surgeries were completed uneventfully according to the proposed modified surgical classification. The mean surgical time was (218.6±82.4) min, and the mean intraoperative hemorrhage volume was (1593.0±1162.5) mL. Postoperative complications occurred in 14.0% of 279 patients. The mean follow-up period was (19.6±13.1) months. The mean Musculoskeletal Tumor Society (MSTS)-93 functional score was 18.3±5.2. Among patients with class Ⅲ lesions, the MSTS-93 scores of patients who underwent intralesional excision followed by placement of Steinmann pins/screws with cemented THA and en bloc resection followed by modular hemipelvic endoprosthesis replacement were 18.6±5.8 and 19.3±4.9, respectively. Recurrences were noted in 15 patients (5.3%), including four who underwent en bloc resection and 11 who underwent intralesional excision.

  Conclusions  The indications for surgical interventions for pelvic metastasis are severe pain and difficulty in ambulation caused by metastatic lesions. According to our proposed modified Harrington classification, the outcomes of surgical treatment for patients with periacetabular metastases can be improved with no increased surgical risk.