Prognostic Factors for Refractory Outcome in Localizing TIO: Experience in a Tertiary Center.

Abstract

Context: Tumor-induced osteomalacia (TIO), a paraneoplastic disorder characterized by renal phosphate wasting, is cured by surgical removal of the culprit tumor. Despite correct localization, some remain refractory to intervention, resulting in substantial long-term medical complications.

Objective: We aim to identify risk factors associated with a refractory outcome.

Methods: This is a retrospective cohort of 44 patients with TIO diagnosed from 1998 to 2023 who underwent targeted intervention following successfully localization. Cure was defined as maintenance of normophosphatemia without supplementation for ≥1 month, maintained at last follow-up.

Results: Twenty-nine patients achieved cure and 15 had a refractory outcome. On univariate Cox regression, the HR for predicting cure was 3.43 (95% CI 1.45-8.11, P = .005) for patients diagnosed after 2013 (compared to before), and that for a negative surgical tumor margin was 2.56 (95% CI 1.20-5.45, P = .015) compared to positive/unspecified margins. After adjustment for year of diagnosis, tumors originating from soft tissue (HR 2.72 vs bone, 95% CI 1.22-6.09, P = .015) or located outside the spine (HR 0.22 for spine vs nonspine, 95% CI 0.05-0.96, P = .043) had higher chances of cure. Size of tumor, age, gender, or baseline biochemistry including levels of fibroblast growth factor (FGF)23, phosphorus, 1,25-dihydroxyvitamin D, or alkaline phosphatase were not predictive of cure. Postprocedural FGF23 was the best biochemical marker of cure (area under curve 0.899, 95% CI 0.764-1.00, P < .001).

Conclusion: Tumors diagnosed within the past decade with clear resection margins had more favorable prognoses. With regards to tumoral factors, baseline biochemistry was uninformative in predicting cure, while bone and/or spine localizations were associated with a refractory outcome.