The diagnosis and treatment of recurrent Campanacci III GCTB of the distal radius are challenging due to the special anatomical structure and histological characteristics [4,6–9,26,]. Recent studies have reported a range of 69–96% H3F3A mutations in GCTB [27–28], and Luo et al. demonstrated that H3F3A mutations may contribute to the differential diagnosis of GCTB in non-long bones [29]. H3F3A hotspot mutations were not highly expressed in other giant cell tumor-rich lesions including chondroblastoma, osteosarcoma, aneurysmal bone cysts and cartilaginous mucinous fibroma, which helped us to identify giant cell tumors of bone and other giant cell-rich lesions.
Intralesional curettage combined with local adjuvant therapy had been recommended as an optimal treatment method for primary or recurrent Campanacci I or II giant cell tumor of bone in the distal radius because the patients got more clinical benefits despite the slightly higher local recurrence rates compared to en bloc resection [30]. However, the high recurrence rate was the most significant risk in the treatment of distal radius Campanacci III GCTB and there was uncertainty regarding the best treatment strategy [31]. Aoude et al. reported curettage and distal radius location were independent risk factors for local recurrence (LR) of GCTB [32]. Furthermore, all patients were recurrent cases in this cohort and the special anatomical structures in the distal radius may contribute to further recurrences. Therefore, en bloc resection was deemed as a primary choice to receive satisfactory local control. The standard reconstructive approach after the resection of a tumor of the distal radius remained uncertain because of the extremely high functional demands of the wrist.
The commonly used reconstructive modalities included: arthrodesis, allograft reconstruction, ulnar transposition, non-vascularized or vascularized autologous fibula implantation, and endoprosthetic replacement [33]. Despite acceptable postoperative function, allograft reconstruction was associated with a high rate of complications, including fractures, bone nonunion, and infection [34–35]. The ulnar transposition technique reduced the incidence of allograft bone complications (infection and bone nonunion), but its most common complication remained proximal radio-ulnar junction nonunion. this might require autologous cancellous bone grafting and internal fixation revision [36–37].
Autogenous non-vascularized structural ICBG had been a favourable reconstruction technique that provided simple autograft harvest, ample cancellous bone, robust wrist stability, consistent graft union, and wrist fusion. 95.6% (22/23) patients had bone union in both osteosynthesis sites 12 months after surgery and the proximal osteosynthesis site had a prolonged union time (8.5 ± 1.9 months) than the distal osteosynthesis site (4.5 ± 1.4 months). However, this result may be caused by the use of reconstruction plates and was still acceptable compared with former studies using arthrodesis reconstruction [38–41]. Clarkson et al. compared the functional results between vascularized (80, median MSTS-93 score) versus non-vascularized (90, median MSTS-93 score) autografts for wrist arthrodesis in distal radius GCTB, and non-vascularized autografts were deemed as a more convenient reconstruction technique which had comparable functional scores [42]. In our study, the MSTS-93 scores (91.5% ± 5.0%) were slightly better than the former study without statistical difference (p > 0.05) and we hypothesized that may be a result of the ceiling effect in distal radius bone defect reconstruction. One patient in our study had a revision with free vascularized fibular grafting (FVFG) because of bone nonunion, and the FVFG technique still had non-negligible advantages in prompting bony union and larger bone defects.
The denosumab use brought revolutionary changes and has been proven to improve tumor response and reduce surgical morbidity in the treatment of giant cell tumor of bone, especially in recurrent or unresectable GCTB [43–44, 46]. However, Errani et al. reported a high local recurrence rate (60%, 15/25) in extremities GCTB treated with preoperative denosumab and intralesional curettage with long-term denosumab administration courses (> 6 months). All 25 patients with preoperative denosumab therapy were diagnosed with remaining GCTB after pathological specimen examinations [16]. Moreover, the perioperative administration doses and duration of denosumab have been controversial in different indications. In former studies, Hindiskere et al. reported that short-term PADT (≤ 3 doses) was associated with no differences in clinical outcomes compared to long-term PADT and Liang et al. Reported that short-term PADT could elicit tumor responses and reduce the perilesional fibrosis and ossification to facilitate nerve-sparing surgeries in sacral GCTB [45]. Moreover, Zhang et al. used short-term PADT (≤ 6 doses) to treat unresectable or recurrent GCTB in axial and appendicular bones and had a total recurrence rate of 27% (3/11) after reoperation [47]. Our studies indicated that the denosumab administration could decrease the CT enhancement rate of GTCB, which proved that the blood supply of tumor lesions was reduced by the denosumab therapy. Moreover, the effect of three-month courses of denosumab administration decreased gradually compared with one-month courses, which indicated that the short-term denosumab courses had the most evident effects on reducing tumor blood supply. Yang et al. reported longer denosumab administration (6 months) had no significant difference with 3 months of denosumab administration in axial and appendicular bones and PET-CT results indicated the average SUV max values of the tumors decreased significantly, which showed the biological activity of tumor declined [24]. In our cases, patients who accepted 3 months of denosumab treatment had significant CT enhancement rate and blood supply decrease and we recommended 3 months of denosumab treatment.
To our knowledge, there have been few studies concerning PADT therapy combined with en bloc resection and arthrodesis in the treatment of distal radius recurrent Campanacci III GCTB. Because long-term denosumab had been proven to increase the local recurrence rate in curettage surgery and the distal radius adjoined the neurovascular structures, tendons, and distal radial ulnar joint (DRUJ), we believed that achieving a safe surgical margin was extremely important. Soft tissue extension was also common in distal radius recurrent GCTB, therefore a clear demarcation between tumor and healthy tissues was necessary. PADT could result in extensive perilesional new bone formation and sclerosis which was conducive to lesion resection and reduced the risk of tumor cell extravasation, and this layer became thicker when the denosumab administration period was prolonged. En bloc resection (EBR) could resect the eggshell-like mineralization layer with remaining tumor cells and help to achieve a clear surgical margin. Tsukamoto et al. reported a local recurrence rate in distal radial GCTB of 30.6% with preoperative denosumab therapy and no LR difference was found between the curettage group and the EBR group [48]. However, this study did not consider the denosumab usage in different Campanacci grade GCTB and the denosumab administration in Campanacci I or II GCTB may increase the local recurrence rate treated with preoperative denosumab and curettage therapy. Tsukamoto et al. also reported clinical outcomes of reoperation for recurrent GCTB in extremities following en bloc resection with a second recurrence rate of 41.4% (12/29) and a third recurrence rate of 17.2% (5/29) [49]. A meta-analysis showed that recurrent GCTB patients had a 20.5% secondary recurrence rate and a 23.4% third recurrence rate after reoperation [4]. In our study, the second recurrence rate was 17.4% (4/23) after reoperation and the third recurrence rate was 8.7% (2/23) after the third operation, which was fairly low in the treatment of distal radius recurrent Campanacci III GCTB.
Recent studies have also reported some complications associated with denosumab administration including jaw osteonecrosis, atypical femoral fracture, hypercalcemia, arthralgia, headache, nausea, fatigue, and anaemia [14, 43–44]. The long-term denosumab safety profile had not been determined in recent studies, however, former studies reported fewer complications, especially Grade 2 or higher complications, in the short-term compared with long-term denosumab administration. Palmerini et al. reported 54 cases that accepted long-term denosumab therapy (median duration of 54 months) had 37% complications in total and 13% Grade 2 or higher complications and 43 cases accepted short-term denosumab therapy (median duration of 12 months) had 2% complications (one case of osteonecrosis of the jaw) [44]. Moreover, Hindiskere et al. also reported no Grade 2 or higher complications in the short-term denosumab courses (≤ 3 doses) [45]. In our study, 8.7% (2/23) patients had denosumab-related Grade 1 complications (1 case of asymptomatic hypocalcemia and 1 case of limb pain) and no Grade 2 or higher complications were found in this cohort.
There were also some limitations in our study. Firstly, the retrospective cohort had selection bias and confounding bias due to the lack of randomization. Secondly, the sample size was still limited in this study, because the recurrent Campanacci III giant cell tumor of bone was a quite rare disease in a single institution, even in a referral cancer center. And to our knowledge, this is the largest case series regarding the adjuvant preoperative denosumab and en bloc resection therapy for recurrent Campanacci III GCTB in the distal radius. Because of the limited sample size and non-randomization nature of our study, we cannot conclude that adjuvant denosumab combined with en bloc resection will significantly improve the local control in distal radius recurrent Campanacci III GCTB. Therefore, a prospective, randomized study is necessary to determine the efficacy and safety of denosumab in the treatment of recurrent Campanacci III GCTB in the distal radius.
Nevertheless, this study introduced a possible method for the treatment of distal radius recurrent Campanacci III GCTB with preoperative short-course denosumab therapy, en bloc resection and arthrodesis, which has few mentions in the former literature. Although there are many options for reconstruction after resection of giant cell tumours of the distal radius, this technique avoids the need for microsurgical techniques and bone banking, and the complication rate is comparable to other described techniques [50–51]. Because the technique was not compared with other reconstruction techniques, we cannot claim that this reconstruction is superior to other methods. However, in our preliminary report, we obtained satisfactory oncological and functional results and we believe that this technique is a reasonable alternative to resection and reconstruction of recurrent Campanacci III GCT tumours of the distal radius. Further future studies comparing the advantages and disadvantages of arthrodesis with other reconstructive methods, such as vascularised fibula grafts and carpal prostheses, are necessary to assess the potential advantages of one technique over the other.