Giant cell tumor of bone: Is curettage the answer?

Background: Giant cell tumors (GCT) are neoplasms of mesenchymal stromal cells with varied manifestations. There is no uniform accepted treatment protocol for these tumors, Materials and Methods: 49 cases of proven giant cell tumors of appendicular skeleton, 27 prospective and 22 retrospective constituteed this study. The retrospective cases were collected by using computerized data base collection method. The patients were evaluated clinically, radiologically and by histology. Companacci grading and Enneking staging was used in the study. Two treatment modalities were used a) extended curettage (with/ without bone grafting/ cementation) or b) wide excision and reconstruction with a prosthesis or arthrodesis. Functional evaluation was done by Enneking’s system. Chi square tests, mannwhitney test and ANOVA were used for statistical analysis. Results: The average age was 26.82 years (16-50 years). 25 patients (51%) were recurrent GCT at presentation. The commonest site was lower end of femur (16 cases, 32.65%) and upper end of tibia (13 cases, 26.53%). 40 (81.63%) tumors had less than 5 mm of subchondral bone free of tumor. 35 (71.43%) tumors were Enneking’s surgical stage III and companacci grade III. Pathological fractures were seen in 12 (24.49%) cases. Intra-lesional currettage was used in 28 and enbloc excision in 19 patients and 2 (4.08%) underwent amputation. The average follow up period was 18.6 months (range 2-84). One recurrence was seen in a grade III recurrent distal radial lesion in the intralesional curettage group (3.57%) Enneking’s functional score with intralesional curettage (25.41) was better than enbloc excision (21.37). Enbloc excision had higher rates of infections (36.84 % Vs 25%) and soft tissue coverage problems (21.05% Vs 0). Conclusion: Intralesional therapy has a better functional outcome and less complications than enbloc excision, albeit with a high recurrence rate which can however be effectively treated with repeat extended curettage.


G INTROCUTION
The treatment of GCT has changed from amputation at the beginning of century to curettage and excision. The current iant cell tumors (GCT) of bones have been described suggestions have varied from curettage for all lesions 1,3 to histologically as neoplasms of undifferentiated wide excision for each tumor. 4 Increased recurrence rates mesenchymal stromal cells with the presence of seen with curettage alone led many to use adjuvants. This abundant giant cells; radiographically presenting as an eccentric study was conducted on all the patients with appendicular lytic lesion at the ends of long bones and clinically as a benign lesions treated since 2000, comparing intralesional curettage but often locally aggressive lesion. It has a tendency towards with wide excision, for recurrence, complications and local recurrence and occasionally malignant change.
residual limb function.
Bloodgood, 1 in 1912, coined the term giant cell tumor and emphasized the benign nature of this tumor. Modern view of GCT began in 1940 when Jaffe and associates proved these tumors as a benign aggressive. 2,3 This terminology is misleading, because 3% of giant cell tumors are primarily malignant or will undergo malignant transformation and metastasize. 2

MATERIALS AND METHODS
Ours was a combined retrospective-prospective study.
Patients treated between January 2000 to January 2006 were included in the study. Patients treated from January 2000 to January 2004 were studied retrospectively from previous hospital records and followed up regularly at regular intervals. Cases from January 2004 to January 2006 were studied prospectively and followed up regularly. A total of 49 patients were studied, with 27 in the prospective and 22 in the retrospective group. Average follow up period was18.6 months (range 2-84 months).  When necessary, standard internal fixation was used. Wide excusion was done when the tumor as assessed by recent Xrays and MR scane had extended to within 2 mm of articular cartilage or when it it had breached the cortex to extend to surrounding soft tissues. Post-operatively, patients were followed up at weekly intervals in first month, fortnightly for next 2 months and monthly thereafter. X-rays were taken at every visit after the eight weeks and then every six weeks. Functional evaluation was done by Enneking's system. This system is applicable in evaluating limb salvage surgeries. metaphyseal or diaphyseal) was carefully evaluated. Size of This evaluates pain, function and emotional acceptance, radiolucent area were recorded as occupying less than one besides dexterity as a measure of upper limb functions and half, one-half and more than one-half diameter of bone in walking ability, gait as a measure of lower limb functions. A-P view. Thickness of the subchondral bone at adjacent It was done 6 months post surgery or in the last follow up articular surfaces was measured radiologically and recorded in patients with less than 6 months followup (2 cases). Chi as more than 5 mm, 5 mm or less or zero. 1 Campanacci 5 square tests, Mann-Whitney test and ANOVA were used grading was used for cortical breach. Grade I tumor had a for statistical analysis. well marginated border of a thin rim of mature bone and the cortex was intact or slightly thinned but not deformed.

RESULTS
Grade II tumor had relatively well defined margins but no radio-opaque rim. Grade III tumors had fuzzy borders.
A total of 49 cases, 22 (44.89%) cases from retrospective group and 27 (55.10%) patients from the prospective group staging was used preoperatively. Stage I is were studied and followed up. The age of the patient at defined as a latent (inactive) lesion that is asymptomatic, presentation varied from 16 to 50 years. 28 cases (57.14%) intracompartmental and histologically benign. Stage II has were in 3 rd decade of life. The male: female ratio was 1.5:1. been defined as active, symptomatic and intracompartmental.
The commonest site was distal femur in 16 cases (32.65%), Stage III is an aggressive lesion that is extra compartmental.
followed closely by upper end tibia in 13 cases (26.53%). Recurrent lesions were compared with primary lesions for Eight cases (16.33%) were seen in distal end radius, while any radiographic aggressiveness like Campanacci grade 4 cases (8.16%) each were seen in lower end tibia and and size on A-P radiograph. Recurrence was considered to proximal femur. Lesions were also seen in proximal humerus be present when there was progressive lysis of more than 5 (2 cases, 4.08%), lower end humerus (1 case, 2.04%) and mm at cement-bone or graft-host interface or if there was 2 nd metacarpal (1 case, 2.04%) Lower limbs accounted for an absence of a sclerotic rim at the above said interface. 7 37 cases (75.51%) while 12 cases (24.49%) involved upper Peripheral calcification around a soft tissue mass of uniform extremity. 25 (51%) of our patients had recurrence on density was the criteria for recurrence in soft tissues. 8 presentation, while the other 24 (49%) were primary. The initial treatment details of recurrent cases were not available Two treatment modalities were used: Extended curettage accurately in all cases as they were operated in other hospitals. (with/without bone grafting/cementation ) or wide excision Epiphyseo-metaphyseal was the commonest location in 44 and reconstruction with a prosthesis or arthrodesis. The type (89.8%) cases. In 39 cases (79.6%) tumors occupied more of reconstruction was decided by a affordability of the patient than half the width of bone on A-P radiographs, while the rest for prosthesis. Extended curettage was [ Figure 1] done when 10 (20.4%) had sizes less than half width of bone. 30 lesions atleast 2 mm of subarticular bone was free of the tumor (61.22%) had subchondral tumor free bone of less than 5 Enneking 6 with no soft tissue spillage as assessed on a recent MRI. Extended curettage was done using a high speed (70000 rpm, Midasrex © ) burr. Phenol (1 ml of melted phenol mixed with 10 ml of normal saline) was used as chemical adjuvant in all cases. The subchondral region of cavity was then packed with autogenous bone graft obtained from iliac crest. Either fibular cortical autograft or irradiated corticocancellous allograft bone obtained from bone bank was used. Polymethylmethacrylate (PMMA) was also used in some cases when the size of the intact articular surface after extended curettage needed additional support along with bone grafts. This was to prevent prolonged immobilization.   Out of 49 patients, 28 (57.14%) were treated with curettage. After curettage, bone grafting alone was used in 23 (82.14%) patients, while one patient, a fourth recurrence in proximal tibia was treated with only bone cement. Four patients were treated with both bone grafting and bone cement after curettage [ Figure 1], two of them receiving morcellized allograft in addition to autologous bone cement and bone graft. Bone cement was used where it was deemed necessary to provide additional structural support to articular cartilage, in addition to bone grafting. No structural allograft were used in our study.
Nineteen (38.77%) patients were treated with wide excision and 2 (4.08%) underwent amputation. Out of 19 patients 13 had GCT around the knee joint. Out of which 7 were treated with arthrodesis, six with prosthesis. Four lesions of distal radius were [ Figure 2] excised and non vascularized     Table 2].
The average follow up was 18.6 months with a median of 19 months. Six patients in curettage group and 5 patients in wide excision group had a follow up of less than 12 months. The average functional score of 47 patients was 23.49. In 28 patients with curettage and bone grafting it was 25.41 while with wide excision it was 21.37, this difference being significant (P=0.017). Three patients following wide excision, with complications like exposed plate due to soft tissue coverage problems and recurrent infections had poor  Figure 3]. Two underwent amputation for whom functional score cannot be applied. All complications were more frequent with wide excision. Wound infection was seen in 7 cases (25%) treated with curettage, which was low compared to 7 cases (36.84%) with wide excision. Soft  utility of the limb. Extended curettage with bone grafting was done in 57.14% of our patients, this being the commonest modality of treatment in the series. This treatment modality has been shown to be effective by other authors. 1,9,11,13 42.86% underwent wide excision. Curettage with bone grafting was the commonest modality in primary cases (70.83%) while wide resection was the commonest treatment for recurrent lesions (48%). Pathological fracture was not a contraindication to curettage and bone grafting in this study as was opined by Dreinhofer. 14 was considered to be present when there was progressive complication rates than curettage. lysis of more than 5 mm at cement-bone or graft-host (P=0.013). Other complications like carpal impingement etc. were seen more frequently in patients of distal end Curettage in GCT is usually followed by adjuvant therapy radius treated with wide excision (31.58%).
either to achieve a more thorough tumor kill or to provide enhanced structural support around the joint. We used DISCUSSION phenol with curettage. Phenol is known to cause protein coagulation, damages deoxyribonucleic acid and causes Lack of a definite treatment protocol in treating GCT necrosis. 15 Good results using phenol has been shown by has resulted from a highly variable presentation of this other authors. Polymethylmethacrylate, cortico-cancellous ubiquitous bone lesion. Lack of studies from South East autograft and allograft, Pamidronate, etc. have all been Asia has further compounded the problem with a deficient effectively used following curettage in GCT. 15,16 Our data on outcome of GCTs in Asian population.
experience is based on use of PMMA, phenol and bone grafts, with good functional outcome and low recurrence The mean age of presentation was 26.8 yrs .
rates. We had only one recurrence among 28 patients with Presentation was commonest in third decade. This is in this procedure. Literature shows a variable recurrence rate accordance with previous studies. 1,2,5,10 Male-female ratio of 4.5% to 52%. 1,3,5,10,11 Meticulous pre-operative evaluation in our study was 1.5:1. Campanacci reported an equal sex including staging and surgical procedure (previous scar ratio for GCT. 5 We found 29 (59.18%) of our lesions around incorporation in the incision, clean dissection, large cortical the knee joint with 16 (32.65%) cases in distal end femur window, the use of high speed burr and use of adjuvants and 13 (26.53%) in upper end of tibia. Prognosis of GCT like phenol, cement 17 can reduce recurrence rates effectively. around knee joint is vital from a functional point of view.
The use of mechanical burr and process of exteriorization This has been shown by other authors as well. 2,10,11 is known to reduce recurrence rate. 18 On the contrary, several studies have ruled out any advantage of using an Pathological fracture was seen in 12 (24.49%) of our adjuvant. 19,20 Extended follow up may well be required patients, commonly in lower end of femur (5 cases or before evaluating success of the strategies used. 41.67%). Similar observations were made by Campanacci 5 and Turcotte. 11 It's presence, however did not affect the The mean follow up score at the last follow up was 23.5. final functional outcome in our study (P=0.564). Fifty one Curettage had better functional outcome when compared percent of our patients presented to us with first recurrence, to wide excision (P=0.017). Joint salvage is known to following a primary procedure done outside. Recurrence improve function. Wide excision was associated with more interface or if there was an absence of a sclerotic rim at the above said interface. 7 Lysis on host-graft interface can be analysed as graft has a different radiodensity than host bone until it is fully incorporated, when its density merges with that of host bone. Recurrent lesions did not have any greater tendency for fractures compared to primary tumors (P= 0.935).
According to Schajowicz, 12 curettage alone is an inadequate oncologic procedure for GCT but associated with better functional outcome compared to enbloc excision. Treatment is a balance between oncological adequacy and functional

CONCLUSION
The extended curettage is associated with a better functional outcome than wide excision. This is particularly important in the Indian context as most of our patients if given a choice, opt for curettage over an endoprosthetic replacement in view of financial constraints, functional limitations (particularly squatting) and more loss of productive working hours. Thus for Indian patients extended curettage with bone grafting in giant cell tumors is a cost-effective, easy technology option.