Bisphosphonate treatment of aggressive primary, recurrent and metastatic Giant Cell Tumour of Bone

Giant cell tumour of bone (GCTB) accounts for around 5% of all primary bone tumours. GCTB is an expansile osteolytic tumour which most often arises at the end of a long bone in a skeletally mature patient [1, 2]. Clinically, most GCTB patients present with bone pain due to enlargement of the tumour. GCTB is characterised by the presence of numerous giant cells that exhibit the phenotypic features of mature osteoclasts [3]. The functional definition of an osteoclast is that it is capable of resorbing bone and, not surprisingly, GCTB is a tumour that causes extensive osteolysis. GCTBs are actively growing tumours which are often locally aggressive, eroding the bone cortex and extending into surrounding soft tissues [1, 2]. Metastatic lesions, most commonly in the lungs, can also develop in approximately 2% of GCTB cases. The treatment of primary GCTBs is essentially surgical and includes curettage (with or without adjuvants such as phenol or liquid nitrogen) and excision of the affected area; however, surgical excision and reconstruction is difficult as the tumour commonly involves the epiphysis and the tumour arises in relatively young patients, thus limiting surgical options. Reported recurrence rates after curettage vary, depending upon the thoroughness of curettage and the adjuvant treatment employed, but are generally in the range of 25 to 50% [4‐7].

Bisphosphonates are widely used to inhibit osteolysis in conditions such as osteoporosis, Paget's disease and metastatic cancer; these agents act by inhibiting osteoclastic bone resorption by a variety of means [8‐10]. Bisphosphonates have been used to treat osteolytic benign tumour-like lesions such as fibrous dysplasia and Langerhans cell histiocytosis that are not amenable to surgery [11‐14]. There are also a few reports of aminobisphosphonate treatment of primary GCTBs which have shown a variable but generally beneficial effect on tumour size and recurrence rate [15, 16]. There are no reports, however, on the effect of bisphosphonate treatment on recurrent or metastatic GCTBs.

Tumour osteolysis is mediated by osteoclasts [17, 18], and GCTB is a tumour that contains numerous osteoclasts. Bisphosphonates act primarily by inhibiting osteoclast-mediated resorption and it would seem logical that GCTBs would benefit from bisphosphonate therapy. A number of major European bone tumour centres have tried using aminobisphosphonates to control disease progression in problematic GCTB cases. Included amongst these cases were persistently recurrent GCTBs and large, pelvic or sacral GCTBs that were deemed inoperable on account of difficulty in obtaining surgical clearance, as well as a number of GCTB cases with evidence of metastatic disease. The aim of this study is to report the outcome of aminobisphosphonate treatment of these GCTBs. We have also analysed the effect of zoledronic acid (one of the most commonly employed bisphosphonates) on lacunar resorption by osteoclasts isolated directly from GCTBs in order to provide in vitro evidence for the inhibitory effect of aminobisphosphonates on GCTB osteolysis.

GCTB commonly recurs after curettage and complete excision is often not possible due to the size and location of the tumour. This study reports the experience of four European centres which have employed aminobisphosphonates as a component of adjuvant treatment for problematic primary (inoperable), recurrent and metastatic GCTBs. Although results are difficult to assess as the problematic nature of the cases studied meant that results could not be compared with a single control group, and aminobisphosphonates were administered on an individual ad hoc basis, often in combination with other adjuvant therapies, they are of interest as they show that most tumours did not increase in size after bisphosphonate treatment and that control of tumour recurrence and pain was achieved in some cases.

Actively growing GCTBs are associated with extensive bone resorption. Bisphosphonate treatment is well-established in controlling osteolysis associated with skeletal metastasis. Paget's disease, osteoporosis, multiple myeloma [8, 9], and bisphosphonates have also been used to control osteolysis occurring as a result of fibrous dysplasia and Langherans cell histiocytosis [12‐14]. Two recent studies have reported the use of bisphosphonates in GCTB. Tse et al [16] studied 24 patients treated with either pamidronate or zoledronic acid given preoperatively and found that only one of 24 patients (4.2%) developed local recurrence following this treatment; this contrasted to the control group who had a recurrence rate of 30%. There is also a report on the use of intravenous bisphosphonate to control the growth of a giant cell tumour of the sacrum [15]. As in our series of cases, the patients receiving aminobisphosphonates for GCTB in these two reports, did not complain of significant treatment side effects.

Our results show that most inoperable primary sacral and pelvic GCTBs did not continue to increase in size after commencing oral or intravenous aminobisphosphonate treatment with some cases showing radiological evidence of bone formation and infilling of lytic lesions. Treatment also resulted in control of cases of persistently recurrent GCTB. Bisphosphonates are recognised to have an analgesic effect on bone tumours [20], and a decrease in pain was noted in some of the treated cases. This was most notable in case 13 where significant pain relief was achieved following zoledronic acid treatment of a large sacral tumour compressing the sciatic nerve; this tumour stabilized in size, developed a sclerotic border and showed infilling with bone; this patient is currently pain-free. The effect of bisphosphonates on metastatic lung lesions of GCTB was more difficult to evaluate as it is known that these lesions can remain stable in size for many years or even regress. However, it was noted that lung nodules of GCTB did not increase in size and no further nodules developed in the course of bisphosphonate treatment.

The tissue specific targeting of bisphosphonates to the mineral component in bone is extremely useful in controlling the growth of bone lesions. Bisphosphonates are non-toxic analogues of pyrophosphate which are known to inhibit osteoclast-mediated bone resorption. Binding of bisphosphonates to hydroxyapatite results in changes in the physico-chemical structure of the hydroxyapatite crystal [8‐10]. Aminobisphosphonates act by inhibiting farnesyl diphosphate synthase, an enzyme in the mevalonate pathway. This results in inhibition of osteoclast formation from mononuclear phagocyte precursors, decreased osteoclast resorption due to effects on the cytoskeleton, vesicular trafficking and membrane ruffling, and increased osteoclast apoptosis [21‐23]. It has been shown that bisphosphonate treatment induces apoptosis of GCTB stromal cells and osteoclasts [24‐26]. An inhibitory effect of zoledronic acid on GCTB-derived osteoclast resorption in vitro was noted in this study; in three of the four cases osteoclast resorption was abolished or profoundly inhibited whereas in one case the decrease in resorption (21%) was more modest. This is in keeping with the findings of Lau et al [3], who also noted a pronounced but variable inhibitory response to zoledronic acid by GCTB-derived osteoclasts. This variation in the inhibitory effect of zoledronic acid on osteoclasts derived from different GCTBs may reflect differences in the therapeutic effect of bisphosphonates in our study.

GCTB contains numerous mature, functional osteoclasts. The rationale for using bisphosphonates to control the osteolysis associated with the growth of GCTBs is that these compounds reduce osteoclast numbers in bone and inhibit osteoclastic resorption. A similar rationale underlies treatment of GCTB with denosumab, an anti-RANKL antibody; this has been reported to result in clinical and radiological improvement in some cases [27]. Aminobisphosphonates have an established role in the treatment of tumour osteolysis associated with metastatic cancer and our results show that these drugs can achieve stabilisation of disease in problematic GCTB cases. The use of different bisphosphonates and protocols of treatment, as well as the effect of other adjuvant therapies, needs to be taken into consideration when evaluating the results of this study. None of the patients in our study reported significant side effects following bisphosphonate treatment but it should be borne in mind that long-term therapy with bisphosphonates can be associated with significant complications, notably osteonecrosis of the jaw [28].

Further studies to analyse the effect of aminobisphosphonates on GCTB would seem appropriate, particularly in the context of controlling the progressive bone destruction and enlargement of inoperable large primary GCTBs and problematic bone and soft tissue GCTB recurrences. It may be worthwhile examining whether combined denosumab and bisphosphonate treatment is more effective in controlling osteolysis or in alleviating bone pain; the latter can be assessed using standardised scores eg functional assessment of cancer therapy-bone pain [29]. Our study highlights the fact that there is no recognised protocol for aminobisphosphonate treatment of GCTB tumours in Europe (or elsewhere in the world to our knowledge), and that there is a need to establish standardised protocols not only for the administration of these drugs but also for definition of outcome measures that will permit a reliable assessment of treatment efficacy.