DETECTING MEDULLARY THYROID CANCER AMONG ADULT PATIENTS WITH THYROID NODULE: IS THERE A BEST STRATEGY? A COST-EFFECTIVENESS ANALYSIS

Purpose:                                                                                                              

Thyroid nodule is frequent while thyroid carcinoma especially sporadic medullary cancer (SMTC) is rare. To detect SMTC, the panel of American Thyroid Association (ATA) decided that they “cannot recommend either for or against the routine measurement of serum calcitonin (CT)”. On the contrary, European Thyroid Association (ETA) stated that “routine serum CT measurement was recommended in the initial diagnostic evaluation of thyroid nodules”. Our objective was to compare the cost and the effectiveness of recommendations (ETA) or lack of (ATA).

Method:

A Markov model was built to estimate life-time discounted costs and effectiveness (3%). A fictive patient cohort of 48.9 years with thyroid nodule greater than 1 cm detected either by medical examination or by ultrasonography was involved. In ETA strategy, CT measurement was performed in the initial evaluation of every patient. In ATA strategy, CT measurement was performed in patients meeting criteria for surgery. In both strategies, hormonal measurement, ultrasonography and FNA were performed. Patients with CT more than 10 ng/L for women and 20 ng/L for men were further explored. Patients underwent surgery, adapted or not to the type of carcinoma (i.e., differentiated thyroid carcinoma or SMTC), inappropriate, or follow-up. A systematic review of the literature explored the sensitivity and specificity of CT, FNA and frozen section, the efficacy of surgical procedures, and the probabilities of adverse events. Outcomes were expressed as life-expectancy (LE) and quality-adjusted life expectancy. Costs (€ 2013) were estimated from the health insurance perspective.

Result:

In the base-case analysis, both strategies provided approximately the same LE (i.e., 20.61 for ETA and 20.57 for ATA) making ETA strategy slightly more effective than ATA. ETA strategy (3203.27€) was slightly less expensive (13€ per life-year) than ATA (3216.27€), thus dominating ATA strategy. The robustness of the model was tested by sensitivity analyses. Only changes, included in the range of epidemiological data found, in the specificity of FNA for differentiated thyroid cancer diagnosis and the sensitivity of FNA for SMTC made ATA strategy more effective than ETA. Long term quality-adjusted analyses revealed similar results with increased differences in effectiveness between both strategies.

Conclusion:

Currently, choosing between both ATA and ETA strategies turn out difficult. Some important epidemiological data are lacking. However, ETA strategy dominates slightly ATA strategy.