CORR Insights®: Is a Short-course of Preoperative Denosumab as Effective as Prolonged Therapy for Giant Cell Tumor of Bone?

Where Are We Now?

Giant cell tumor of bone (GCTB) is a locally aggressive, rarely metastasizing lesion that is usually found in the epiphyseal-metaphyseal region of the long bones. Surgical management most often consists of extended intralesional curettage to remove all disease. Because of the infiltrative nature of GCTB as well as the desire to preserve articular cartilage and subchondral bone to facilitate joint salvage, microscopic disease may persist after intralesional treatment. In the past, this has resulted in a substantial risk of local recurrence, as high as 40% in a recent study [5]. To reduce this risk, intralesional, cytotoxic chemical adjuvants have sometimes been used at the time of curettage. These include ethanol, phenol, hydrogen peroxide, and liquid nitrogen cryotherapy. The local recurrence rates in case series using these adjuvants are usually comparable to those in which no adjuvants are used [7, 10, 14].

Denosumab is a human monoclonal antibody that inhibits the receptor activator of nuclear factor-kappa β ligand. This receptor blockade inactivates osteoclastic giant cells in GCTB, thereby promoting bone formation rather than aggressive lytic bone destruction. It has shown promise in decreasing skeletal events in patients with metastatic breast cancer [12] and those with multiple myeloma [15]. As a result, there was substantial initial enthusiasm to use denosumab in patients with GCTB to facilitate joint salvage in large periarticular tumors that would otherwise be treated with en bloc resection and prosthetic reconstruction, decrease the local recurrence rate, and provide stability for inoperable tumors, such as those in the axial skeleton.

The initial treatment recommendation was 120 mg of denosumab on Days 1, 8, and 15 of Month 1, followed by monthly dosing of 120 mg for a further 5 months [8]. Despite initial promising results using this regimen [13], other authors have since noted a substantially higher local recurrence rate after intralesional curettage than in previous series without denosumab [3]. Furthermore, toxicities and malignant degeneration of GCTB have been reported [1, 2]. These findings have prompted some centers to consider a shorter course of treatment before surgical management in the hope that the purported benefits may outweigh the risks outlined above [11].

Hindiskere et al. [6] undertook a single-center, retrospective study to determine whether a shorter course of denosumab of three or fewer doses before surgical management of GCTB is associated with a higher risk of local recurrence or a difference in the clinical and radiographic response compared with the conventional longer regimen. The authors concluded that a short course of denosumab was no different from a long course in terms of Musculoskeletal Tumor Society scores, radiologic and histologic response, and recurrence-free survival. Based on this, they concluded that short-course denosumab treatment therefore may be considered in patients with GCTB when the benefit of joint salvage outweighs the risk of local recurrence and toxicities. The findings from this study help to more-clearly define the specific clinical scenarios in which denosumab, albeit at a reduced dose, may still play an important role.

Where Do We Need To Go?

Denosumab is clearly not the panacea that many thought it might be. Longer-term follow-up has revealed an unacceptably high risk of local recurrence in patients who underwent intralesional curettage after treatment with denosumab, which is reflected in the current study [6]. The 5-year local recurrence-free survival reported by Hindiskere et al. [6] is still substantially higher than that reported in the best studies without denosumab [9]. Therefore, in GCTBs for which extended intralesional curettage can be done safely without sacrificing the adjacent joint surface, there is likely no role for neoadjuvant denosumab. In these situations, good clinical results can be obtained with a reasonably low risk of local recurrence.

There are, however, three clinical scenarios in the presentation of GCTB that are still troublesome and in which denosumab may still play a role.

In the first, extensive subchondral and cortical bone destruction may preclude joint salvage, but with some restoration of bone stock, joint salvage may be undertaken. It is unclear when adequate restoration of bone stock occurs during denosumab therapy. The radiologic criteria that were used (RECIST, modified Choi criteria) by Hindiskere et al. [6] to assess response consider tumor size and were originally described for solid organ tumors. The Campanacci system, which was also referred to by the authors and was developed for benign aggressive tumors, is highly subjective. Studies that address the specific radiographic features of GCTB and its response to denosumab in an objective fashion are lacking.

In the second scenario, the tumor may be far too extensive to undertake joint salvage in any circumstance, but restoration of some bone stock may prevent inadvertent intralesional spillage of the friable tumor at the time of en bloc resection. This friability, which is a common feature of GCTBs, is reduced with denosumab as the cellular tumor is replaced by a more solid, fibrous extracellular matrix. The histologic assessment used by Hindiskere et al. [6] is fairly rudimentary and based on a reduction in the number of giant cells and stromal cells without any regard for the bone architecture. A more rigorous pathologic assessment of the response to denosumab is warranted.

In the third scenario, typically in tumors with an axial location, long-term disease stabilization using denosumab indefinitely may be preferable to extensive surgical resection, with a high risk of complications and local recurrence. Hindiskere at al. [6] referred to this group of patients; however, these patients’ outcomes were combined with those of patients who underwent surgery. An assessment of a larger cohort of patients who are treated with denosumab alone would be warranted to ascertain the long-term effects of the drug.

These three troublesome clinical scenarios require further investigation regarding the best management. Ideally, any patient who presents with a GCTB in a major joint should be considered for joint salvage, given its superior functional outcome. Any intervention to facilitate this must be considered. In situations in which joint salvage is not possible under any circumstance, the surgeon should be confident that en bloc resection could be done with a minimal risk of local recurrence. Finally, in unresectable tumors such as those in axial locations, we should strive for lifetime disease control without the risk of progression or neurologic sequelae.

How Do We Get There?

Most clinical studies to date on the use of denosumab in GCTB have featured a “grab bag” of clinical scenarios. To determine the future role of denosumab, the three clinical scenarios mentioned above need to be considered separately.

In the first scenario, a cohort of patients with GCTB with subchondral bone loss should be treated with denosumab and followed with serial CT scans to assess the extent of restoration of bone stock, most notably in the subchondral region. The mean time to restoration of this bone stock might guide future treatment regimens; only one or two doses of denosumab might be necessary. Dual-energy x-ray absorptiometry scanning to assess bone density in the defect could also be used. Patient-derived pain scores, local recurrence, and patient function should be used as outcomes. In this way, a more GCTB-specific radiographic outcome score could be developed.

In the second scenario, a more-rigorous histologic assessment of the response of GCTB to denosumab should be developed. This may be challenging in clinical practice and would be more appropriate for animal studies. Animal models of giant cell tumor indeed exist, and it would be feasible to histologically examine animals that have been treated with denosumab in incremental doses to determine at what timepoint a hypercellular GCTB transitions to a more-fibrous stoma. This might guide any subsequent treatment regimen in humans.

The third scenario will require a longer-term follow-up study of the long-term effects of denosumab. Fortunately, such a cohort of patients has already been under surveillance [2]. The rarity of GCTB makes it essential for any prospective study to be multicenter. Fortunately, the worldwide oncology community has already demonstrated its ability to undertake such a collaborative effort [4].

Finally, ongoing translational research must continue in an effort to identify other therapeutic agents that may be used to treat this aggressive tumor. Denosumab essentially disables osteoclast-like giant cells, while neoplastic mononuclear stromal cells persist and are ostensibly responsible for local recurrences when they occur. Stromal cells are characterized by a mutation in the H3 histone, family 3A gene, specifically the point mutation G34W, which is the driver mutation of GCTB. Stromal cell lines expressing this mutation have been developed, and the addition of gemcitabine or the novel agent MK-1775 inhibits cell proliferation in vitro [8]. This and further studies with other agents targeting molecular alterations, specifically in stromal cells, are likely to have a more substantial and long-lasting inhibitory effect on GCTB in clinical practice.