The purpose of our study was to identify possible risk factors of patients with GCT of the long bones after curettage and packing the bone cavity with bone cement or bone allografts.

We retrospectively reviewed the records of 249 patients with GCT of the limbs treated at Musculoskeletal Oncology Department of our institution between 1990 and 2013, confirmed histologically and recorded in the Bone Tumor Registry. We reviewed 219 cases located in the lower limb and 30 of the upper limb. This series includes 135 females and 114 males, with mean age 32 years (ranging 5 to 80 yrs). According to Campanacci's grading system, 190 cases were stage 2, 48 cases stage 3, and 11 cases stage 1. Treatment was curettage (intralesional surgery). Local adjuvants, such as phenol and cement, were used in 185 cases; whereas in the remaining 64 cases the residual cavity was filled with allografts or autografts only.

Oncological outcome shows 203 patients alive and continuously disease-free (CDF), 41 patients NED1 after treatment of local recurrence (LR), 2 patients NED1 after treatment of lung metastases, 2 AWD with lung metastases. One patient died of unrelated causes (DOD).

LR rate was 15.3% (38 pts). Lung metastases rate was 1.6% (4 pts). In patients treated by curettage and cement (185 cases) LR was 12% (22 pts). Conversely, in patients treated curettage and bone allografts it was higher (16/64 cases), with an incidence of 25% of cases (p=0.004). Oncological complications seemed to be related with site, more frequently occurring in the proximal femur (p=0.037). LR occurred only in stage 2 or 3 tumors without statistical significance (p>0.05). The mean interval between the first surgical treatment and LR was 22 months (range: 3–89 mos). However, in the multivariate analysis no significant statistical effect on local recurrence rate could be identified for gender, patient's age, Campanacci's grading, or cement vs allografts. The only independent risk factor related to the local recurrence was the site, with a statistical significance higher risk for patients with GCT of the proximal femur (p= 0.008).

Our observation on the correlation of tumor location and risk of local recurrence is new. Therefore, special attention must be given to GCTs in the proximal femur. In fact, primary benign bone tumors in the proximal femur are difficult to treat due to the risk of secondary osteonecrosis of the femoral head or pathologic fracture.

Numerous methods of reconstructions have been reported. Among these, total hip arthroplasty (THA) or bipolar hip arthroplasty (BHA) should be avoided when possible as more cases are observed in young patients.

Therefore, we do not suggest different approach for the proximal femur. GCT in the proximal femur is much more difficult to treat than in other sites, but if curettage is feasible, the best way is to save the joint with a higher risk of local recurrence, knowing that the sacrifice of the hip articulation in case of recurrence is always possible with THA or BHA.