Surgical treatment of tumor-induced osteomalacia: a retrospective review of 40 cases with extremity tumors

80% (32/40) of our study cohort was treated successfully through complete tumor removal surgery during an average follow-up period of 43 months, including 29 patients after primary surgery and three after secondary surgery. To gain complete remission of clinical symptoms, the tumor must be completely removed. For the eight patients in this cohort whose serum phosphorus levels were only partially normalized, their tumors were incompletely resected.

Soft tissue tumors usually had clear boundaries and were well-encapsulated, so complete tumor resection was straightforward. However, some tumors grew close to bone, so sometimes it was difficult to distinguish whether bone involvement existed; in these cases, some bone tissue was removed. In fact, some cortical bone was resected in three patients with soft tumors in this series, and all of them completely recovered. The prognosis for TIO patients with soft tissue tumor was good, as 88% of patients achieved normal serum phosphorus in this study. Of the three patients whose serum phosphorus did not return to normal, one had a malignant tumor (patient 1), one had two previous surgeries at the same site (patient 19), and one had an undetectable tumor (patient 39).

Surgical treatment of bone tumors seemed to be much more difficult, because it was often impossible to distinguish tumors with normal tissues during surgery. Preoperative preparation to determine curettage range was important and tumor curettage with at least 5 mm from tumor margins was suggested. However, only four of nine patients were successfully treated after primary tumor curettage. Harbeck et al. [15] reported a 34-year-old male patient with TIO, whose tumor was located to the left greater trochanter. Intraoperatively, a hand-held gamma probe after administration of ¹¹¹Indium pentetreotide 1 day before surgery clearly identified the tumor. This report provided us another way to identify the tumor during surgery.

Bone tumor resection appeared to be more effective than curettage. In our study, two patients with femoral head tumors received total femur head resection and THA and both achieved successful treatment at the last follow-up. In addition, three patients with failed primary tumor curettage surgeries received secondary tumor resection and prosthesis reconstruction, two of which were successfully treated. Nevertheless, these kinds of surgeries are often associated with greater morbidity with loss of some limb function. Prosthesis related problems cannot be ignored either. Less invasive treatment has been reported in the literature. Hesse et al. [16] reported a 40-year-old woman with a tumor in her right femoral head. To preserve her hip joint, CT-guided radiofrequency ablation was performed for two rounds. Complete ablation was achieved, and no tumor recurrence was observed at one year follow-up. Tutton et al. [17] reported another patient with a tumor in his right iliac bone. CT-guided percutaneous ethanol ablation and percutaneous cryoablation was performed. The tumor was completely destroyed during 12 months follow-up. Although these results are promising, larger sample sizes are still needed. These are alternative methods and could be considered for patients who refuse surgical treatment.

Compared with widely tumor curettage surgery, tumor resection surgery seemed to be a better choice for TIO. However, considering anatomical and functional factors, it’s difficult to perform tumor resection in all cases as function could be severely compromised. Tumor curettage could be considered in cases where surgical morbidity was a concern and that attempting to curettage beyond the expected tumor margin was recommended.

The mesenchymal tumor was typically benign whereas malignant tumor was seen as well [2,12,14,18-21]. Four malignant tumors were identified in our series of cases. Of them, two received amputations, one received tumor resection and the last one received tumor curettage. For the last two patients, radiotherapy was arranged even though only a little data had suggested it to be strongly effective [19]. In fact, it was difficult to predict the behavior of these tumors only based on the morphologic findings [2,18]. In our cohort, two of three tumors involving both soft tissue and bone were proved to be malignant, which might be another characteristic to remind us of the behavior of these tumors.

There were some limitations for this study. Firstly, MRI or USG was not routinely performed during the follow-up time for patients with unnormalized serum phosphorus; instead, serum phosphorus and 1, 25-dihydroxyvitamin D as well as bone mineral density were regularly monitored. Only for those with decreased serum phosphorus and intending for further surgical treatment, residual tumors were examined. In addition, the minimum follow-up time was only 12 months, thus long term results could not be concluded. At last, surgeries were not always performed as surgeons planned. Some patients refused further or more invasive operations, which might influence the overall efficacy assessment of surgical treatment.