Does Denosumab Change the Giant Cell Tumor Treatment Strategy? Lessons Learned From Early Experience
Manish G Agarwal,Manit K Gundavda,Rajat Gupta,Rajeev Reddy
DOI: https://doi.org/10.1007/s11999.0000000000000243
Abstract:Background: Although giant cell tumors (GCTs) are benign, their aggressiveness and tendency to recur locally challenge the orthopaedic surgeon's ability to perform joint-preserving intralesional surgery with an acceptably low risk of local recurrence. Denosumab has emerged as a possible medical treatment of GCT because it seems to halt the progression of GCT, alleviate pain, and increase perilesional bone formation, but its exact role has been questioned, and specifically its efficacy and associated complications are not well characterized. Questions/purposes: (1) Does denosumab reduce the risk of recurrence after resection or intralesional surgery? (2) What are the complications associated with the use of denosumab? Methods: Fifty-four patients with 30 primary and 25 recurrent tumors between November 2013 and July 2016 were treated with denosumab after a confirmed histopathologic diagnosis of GCT. Another 17 patients in the same period were treated without denosumab. During the study period, we encouraged the use of denosumab in all patients except those who refused, could not afford it, or where it was contraindicated (eg, in pregnancy). In all patients undergoing intralesional surgery, we arbitrarily planned six doses before surgery. Variations in total doses before surgery were dependent on patient-related factors; in some, we gave less doses because patients expressed the inability to afford any more doses, whereas in some patients, extra doses were added when the patient wished to delay surgery as well as the because of surgeon judgment wherein in some patients, we stopped before six doses when we thought adequate bone had formed for intralesional curettage. The mean number of doses was 6.8 per patient (median, 6; range, 3-17) preoperatively. The minimum followup was 12 months (median, 27 months; range, 12-42 months). Every patient showed improvement clinically in terms of pain and halting of tumor progression within three to four doses. This was seen radiologically as a sharply defined soft tissue mass as well as hazy ossification within the tumor. For a case-matched comparison study, we identified controls as 34 patients undergoing curettage from the retrospective analysis of 68 patients curetted without denosumab between February 2010 and July 2016 matched to 25 denosumab-treated patients in terms of site, size, Campanacci grade, and recurrent versus primary status, and with a minimum 2 years followup for the control group. Fisher's exact test was used for statistical study. Patients undergoing resection were planned for surgery after three doses of denosumab to allow the tumor to solidify and potentially decrease tumor spillage at the time of surgery. The resections could not be case-matched for comparison owing to the smaller numbers. Results: We observed 14 recurrences out of the 37 curetted tumors (38%). In the case-matched analysis, 11 of 25 patients in the denosumab-treated curettage group had recurrences (44%) compared with seven of 34 (21%) in the nondenosumab-treated control group. The risk of denosumab-treated patients experiencing local recurrence as compared with the nondenosumab-treated patients was nonsignificant with a two-tailed p value of 0.085 (significance at p < 0.05) as derived from Fisher's exact test (odds ratio, 3.03; 95% confidence interval, 0.96-9.54). There was no recurrence in the resection group. Because we do not have a control group for resection, we are unable to comment on the importance of this finding. One major complication that we observed was a recurrence with malignant transformation in a patient with a proximal humeral GCT. We did not observe any other complications related to the denosumab therapy. Conclusions: Although we could not demonstrate a higher risk of local recurrence with preoperative denosumab for intralesional surgery in the dose and frequency we administered, we advise caution in its routine use for intralesional procedures because it may be important to curette up to margins on pretreatment imaging owing to the potential residual tumor within the denosumab-mediated thick bony shell, which may result in local recurrence. We believe that denosumab treatment before resection of a large tumor aids resection without tumor spillage, particularly where important structures like the neurovascular bundle are dissected away from the tumor margin, although we cannot confirm that it helps lower the incidence of recurrence. We are concerned regarding the malignancy-causing potential from our observation in one patient as well as reports of this by others and recommend judicious use of this drug in patients with GCT. Level of evidence: Level III, therapeutic study.