Integrating histology and biomarkers to predict clinical outcomes in parathyroid tumors: a multicenter study.

AIM: This study aims to evaluate clinical, biochemical, and histopathological determinants of disease severity and postoperative outcomes in parathyroid tumors and to assess whether a composite histology-biomarker risk score provides an integrated framework to better characterize tumor behavior beyond conventional histological classification.

METHODS

This retrospective multicenter cohort study included patients with primary hyperparathyroidism caused by typical adenoma (TA), atypical parathyroid tumor (APT), or nonmetastatic parathyroid carcinoma (PC) diagnosed between 2000 and 2021 in three tertiary hospitals. Outcomes included preoperative parathyroid hormone (PTH) and calcium levels, postoperative hypocalcemia, and persistent hyperparathyroidism at 6 months. A composite unweighted risk score (range 0-5) was developed, assigning one point for each adverse feature: atypical/malignant histology, loss of parafibromin, Ki-67 ≥ 3%, and positive galectin-3 or PGP9.5 expression.

RESULTS

A total of 139 patients were analyzed (64.7% women; mean age 58.3 ± 13.6 years), including 90 TA, 20 APT, and 29 PC. Preoperative PTH and calcium levels increased progressively with the risk score (P < .001). Postoperative hypocalcemia rising from 0% in score 0% to 53% in scores ≥ 4, while persistent hyperparathyroidism increased from 6% to 45% (P < .01). Each one-point increase in the score was associated with postoperative hypocalcemia (odds ratio [OR] = 2.00) and persistent hyperparathyroidism (OR = 1.87), independent of age, sex, chronic kidney disease, and surgical extent. Model discrimination was good (area under the curve [AUC] of an analysis using receiver operating characteristic [ROC] curves 0.78-0.81) with stable calibration.

CONCLUSION

The composite histology-biomarker risk score was linearly associated with biochemical severity and postoperative outcomes, offering a practical tool to refine risk stratification beyond histological classification.