Surgical treatment for pelvic giant cell tumor: a multi-center study

Giant cell tumors of the bone rarely affect the pelvis, with an incidence of only 1.5 to 6.1 % of bone GCTs [4‐6]. Only 73 pelvic lesions had reported in literatures from 1949 to 1999 [7], and the largest number of cases with pelvis GCTs presented in one study was no more than 30. In this study, 29 patients with pelvis GCTs from five institutions had been collected for analysis. To our knowledge, it is the largest number of cases in the existing literatures as well as the only multi-center study on pelvis GCTs.

The local recurrence rate for giant cell tumors of the pelvis seems to be higher than that of any other location in the skeleton for western or eastern [23, 24], due to the complex anatomy and the large size these lesions can attain before diagnosis. The local recurrence rate can be as high as 43 % for therapeutic options other than wide resection [7, 13‐16, 25]. Sanjay et al. [16] reported that 3 of 15 patients with pelvic GCTs had local recurrence after intralesional surgery, whereas no one recurred in 2 patients who underwent wide resection. Balke et al. [13] reported that 1 of 16 patients with pelvic GCTs suffered local recurrence after intralesional surgery, whereas none of 3 patients with wide resection recurred. Guo et al. [8] reported that 4 of 13 patients with pelvic GCTs had local recurrence after intralesional surgery, whereas 0 of 14 patients suffered recurrence after wide resection. In our study, 2 of the 21 patients who underwent wide resection had local recurrences, whereas one of 8 patients who underwent intralesional surgery had local recurrence. No significant difference in recurrence was found between intralesional surgery and wide resection (p = 0.814). It is generally accepted that wide resection can reduce the risk of recurrence because the adjacent tissue can be resected with the tumor to achieve an adequate resection margin. For some pelvic GCTs, wide resection is hard to perform owing to the complex anatomic structures in the pelvis. Intralesional surgery is an adequate option for some patients based on careful preoperative planning. Initial surgical treatment is of great importance to patients on account of the fact that recurrence of the tumor may make it unresectable; as a consequence, it is vital for patients with pelvic GCTs to be operated by surgeons with sufficient knowledge of bone tumor. The tumor size and Campanacci classification should be taken into account in the preoperative plan. If the tumor is not limited to one region defined by the Enneking and Dunham classification system for pelvic tumors, we believe wide resection could be a reasonable choice. If the Campanacci classification is grade III that the bone walls are destroyed by the tumor and a large soft tissue mass forms, wide resection is recommended to ensure a safe margin.

Treatment options for pelvic GCTs include denosumab, serial embolization, interferon, radiation therapy, and surgical treatment. Denosumab is a fully humanized monoclonal antibody against the RANK ligand which has demonstrated significant activity in patients with unresectable or recurrent GCTs in the bone [9]. Denosumab was approved by the FDA for the treatment of adults and skeletally mature adolescents with GCT of the bone that is unresectable or where surgical resection is likely to result in severe morbidity in 2013. However, patients did not accept the denosumab treatment in China because the denosumab was not approved by CFDA and the patients’ agreement was hard to get. Although serial embolization [10] and interferon [11] have been accepted in the GCT treatment, it is still necessary to perform surgical treatment thereafter. The obvious advantage of radiation therapy is the avoidance of additional surgical morbidity, while its disadvantages include local effects of radiation therapy and the potential for radiation-induced sarcoma. Leggon et al. [7] reviewed literatures of pelvic and sacral GCTs, which revealed that patients treated with radiation therapy had an incidence of 44 % of local recurrence and 12 % of radiation-induced sarcoma.

Therefore, it is quite clear that surgery plays an unparalleled role in the pelvis GCT treatments. It is widely accepted that intralesional surgery can spare nerve roots, pelvic support, and the hip and visceral structures, whereas it has a high risk of local recurrence, even for a recurrent GCT. Various adjuvant modalities have been used to supplement intralesional surgery in order for a low recurrence rate, which include the use of cytotoxic agents such as phenol [16], zinc chloride [26], ethanol [14], and physical adjuvants such as polymethylmethacrylate [13, 14, 27], cryosurgery [13], microwave inactivation [28], and a high-speed burr drill [13, 14]. Besides, a study shows that curettage based on CT classification can also lower the recurrence rate [29]. In this series, microwave inactivation had been employed for 3 patients before intralesional surgery and no recurrence occurred. It is obvious that microwave inactivation is performed before surgery other than other adjuvant treatments. The inactived tumor tissues do not have viable cells during curettage in theory. The difficulties for this method include protection of the surrounding normal tissues and complete inactivation for the whole tumor tissue. Furthermore, GCTs did not destroy the subchondral bone in the acetabular area for these 3 patients. We used polymethylmethacrylate for 2 patients in order for filling and inactivation, one of which suffered recurrence. The high-speed burr drill was employed in 2 patients to deal with residual GCT cells that may remain on the surface of the curettage cavity, and no recurrence occurred. We did not use nitrogen or other chemicals for the sake of avoiding chemical injury to the surrounding tissues from leaking adjuvants. Wide resection has been thought to minimize the chance of local recurrence, which is true for tumors in regions I and III, rather than the ones in region II. There is no doubt that wide resection is difficult to perform for large-size pelvic GCTs, especially those involving region II. An increased surgical morbidity is usual even for an experienced bone tumor surgeon.

The MSTS 93 score had been applied for patients’ functional evaluation in this study. Compared with intralesional surgery, the function of the patients who accepted wide resection appeared worse. However, it was not absolute for all patients that most patients who accepted wide resection for GCTs in regions I and III had good functions. For the GCTs involving region II or more than one region, favorable functional results were hard to achieve although some specialized prostheses had been developed to reconstruct the defect after acetabular bone resection. In this series, we preferred to use screws, titanium rods, and bone cement to reconstruct the integrity of the pelvis based on our experience. The advantages of this method include simple instruments, flexible fixed way, and relatively stable mechanical strength. The disadvantages include fixation loosening and breakup. The fixation-related complications were not observed in our patients, and most patients who accepted pelvis reconstruction could resume their routine activities within 3 months after surgery.

As we know, complications are common for pelvis tumor because of wide wound exposure, extensive soft tissue stripping, implant existence, local hematoma formation, and poor skin flap blood supply [30]. In this study, wound healing disturbance was the major complication, which occurred in 4 patients who underwent wide resection and pelvis reconstruction. Limb shortening, delayed infection, and dislocation happened in 3 patients, respectively. It is obvious that patients with GCTs in region II had a higher incidence of complication compared with those in regions I and III. For reducing complications, it is necessary to protect the skin flap and reduce soft tissue tension. Otherwise, drainage is very important for postoperative care.

It is necessary to alert readers to be aware of the limitations of this study. Firstly, the number of patients is still small although it is the largest series among the reports owing to the fact that pelvic GCTs are rare. It is hard to make any definitive statements regarding the differences of recurrences, complications, and postoperative function among different treatments without large sample statistical analysis. Secondly, this is a multi-centric retrospective study and the patients’ treatments were decided by five experienced bone tumor surgeons, respectively; consequently, the differences among surgical technologies cannot be avoided. Nevertheless, consensuses of treatments had been made by these surgeons, and postoperative situations of different treatment methods may provide valuable information for surgeons in the decision-making process. Thirdly, the minimum follow-up is short and additional local recurrences might occur with longer follow-ups. Nevertheless, 70 % of local recurrences occur within 2 years [31].

Treatment of pelvic GCTs remains a challenge for surgeons, especially for the ones involving region II. In this study, patients with GCTs that did not destroy the subchondral bone in region II achieved good functions and suffered no recurrence by undergoing microwave inactivation before intralesional surgery. Therefore, it could be an alternative treatment for some patients, although it should be determined on the basis of sufficient clinical evidence; as a consequence, a prospective study about the clinical efficacy of microwave inactivation for GCTs has already been launched by our team. It has great difficulties to achieve balance among recurrence, complications, and functions for the GCTs with benign histology and aggressive biological behavior; thus, detailed communication with patients is essential during the decision-making process.