A search for predictors of low-energy fractures in patients with primary hyperparathyroidism after parathyroidectomy

Introduction . Decreased bone mineral density (BMD) and low-energy fractures (LEF) are frequent and threatening complications of primary hyperparathyroidism (PHPT). After radical surgical treatment of PHPT, BMD recovery takes a long time, so the risk of LEF development is still elevated. Data on predictors of postoperative fractures in this cohort of patients remain very limited. Aim . To determine risk factors for the development of low-energy fractures in patients with PHPT within two years after radical surgical treatment. Materials and methods . The study group included 247 patients who underwent surgical treatment in Endocrinology Research Centre. In 2 years after surgery all patients were interviewed for the development of postoperative LEFs. The obtained data were used for comparative subgroup analysis, ROC-analysis and binary logistic regression construction. Results . Patients with postoperative LEFs were different from patients without LEFs in only one parameter – radius BMD (p = 0.0005, 0.455 [0.374; 0.501] g/cm2 vs 0.569 [0.452; 0.644] g/cm2, respectively). For this parameter, ROC analysis resulted in an AUC = 0.757 (95% CI: 0.674 to 0.841), indicating its average predictive ability. Quantitative parameters with significant difference and difference at the level of statistical trend were used in the construction of binary logistic regression. The constructed model allows predicting new LEFs with an accuracy of 94%. Conclusions . The occurrence of LEFs in patients with PHPT after radical PTE remains an actual problem; therefore, it is necessary to develop methods for their prediction and prevention. The presented mathematical model can be used as a medical decision support system for screening patients in order to exclude low-risk groups for the occurrence of LEFs in the postoperative period.