Screening for medullary thyroid carcinoma: experience with different immunoassays for human calcitonin.

Calcitonin measurements in patients with nodular thyroid disease are helpful for early diagnosis and therapy of medullary thyroid carcinoma. We compared three commercial calcitonin assays routinely used: "CIS" (France), "Medgenix" (Belgium) and "Nichols-Advantage" (USA). In addition, we evaluated a two-step modification of the Medgenix-test and an enzyme immunoassay from Sangui (Japan). Method comparison studies revealed deviations from linear relationships between all routine assays. While histograms of CIS and Nichols results were similar at low concentrations with their highest frequency below 1 pg/ml, Medgenix showed a broad peak around 3 pg/ml. Correlation coefficients were 0.69 (CIS versus Medgenix) and 0.91 (CIS versus Nichols). In thyroidectomized patients, calcitonin was not detectable with CIS and Nichols, but the Medgenix test, which was more susceptible to interference, measured about 6 pg/ml. The immunoassay of Sangui showed insufficient analytical sensitivity. About 70-80% of calcitonin levels initially > 10 pg/ml were reproduced in the basal levels of the subsequent pentagastrin test. Average ratios (stimulated/basal level) were slightly higher for CIS and Nichols than for Medgenix. Prediction of a pathological stimulation from basal calcitonin was insufficient with CIS and Medgenix assays. If a calcitonin concentration of > 100 pg/ml (CIS) is considered as an indication for surgery, equivalent values are 71 pg/ml for Medgenix, and 96 pg/ml for Nichols. Using these criteria, about one third of patients who underwent pentagastrin stimulation showed pathological reactions. Reliable and sensitive calcitonin assays used for the screening of sporadic medullary thyroid carcinomas in patients pre-selected by nodular goiter are important for health and economic reasons, because they enable early therapy.