Comparative analysis of PTH Washout and 99mTc-MIBI SPECT/CT in presurgical localization for patients with primary hyperparathyroidism.

PURPOSE: Our study aimed to evaluate the performance and concordance of parathyroid hormone wash-out (PTHw) and 99mTc-MIBI+SPECT/CT (MIBI) in preoperative localization for patients with primary hyperparathyroidism (PHPT), and assess potential factors influencing their diagnostic utility including patient characteristics, lesion size and comorbidities.

METHODS

This was a real-life, retrospective study involving 125 participants, who underwent ultrasound-guided fine needle aspiration biopsy with PTHw determined by electrochemiluminescence (Elecsys analyser) and subtraction scintigraphy together with SPECT/CT images.

RESULTS

PTHw was positive in 84.8% and MIBI in 84.0% of cases, with concordant findings in 73.6% of patients. In MIBI-negative cases, PTHw accurately localized PA in 70.0%, while MIBI identified lesions in 68.4% of PTHw-negative cases. Larger lesions were associated with improved performance for both techniques, although a significant size-dependent effect (≥10 mm) was observed only for MIBI (χ2 = 3.92, p = 0.048; φ = -0.18). The right inferior gland was the most frequent lesion location detected by PTHw (28.0%). PTHw correlated positively with serum PTH (r=0.29), PTHw-to-serum PTH ratio (r=0.84), total calcium (r=0.27), ionized calcium (r=0.31) and alkaline phosphatase (r=0.31), and inversely with phosphate (r = -0.22) and vitamin D (r= -0.20).

CONCLUSION

Our findings reinforce the complementary role of ultrasound-guided parathyroid hormone washout and 99mTc-MIBI with SPECT/CT, in the preoperative PHPT imaging diagnostics, particularly when one modality yields negative or inconclusive results. Additionally, our results highlight the need for individualized imaging strategies based on specific patient profiles. Larger prospective studies with histopathological verification are warranted to corroborate these observations and refine preoperative localization protocols.