Early and late lesion-to-non-lesion ratio of thallium-201-chloride uptake in the evaluation of "cold" thyroid nodules.

Forty-nine consecutive patients with "cold" thyroid nodules were studied using early and late visual and semiquantitative measurements of 201Tl uptake in the nodule to differentiate benign from malignant nodules. The visual method compared 201Tl uptake in the nodule to the normal thyroid tissue. The semiquantitative method used a lesion-to-non-lesion (L/N) ratio of the same areas. Both measurements were carried out early (15 min) and late (3 hr) following 201Tl injection. The reproducibility of the method for the early and late measurements was tested for intraobserver and interobserver variability as well as for repeatability coefficients. The visual method resulted in 43% sensitivity and 79% specificity for the detection of malignant nodules. The L/N method showed that an early threshold of 1.55 chosen by receiver characteristic analysis had a sensitivity of 57% and a specificity of 86%, while the late ratio of 0.99 had a sensitivity of 100% and a specificity of 62%. It is concluded that a L/N 201Tl uptake method performed 3 hr following 201Tl injection is superior to a visual scoring method as well as to the early L/N 201Tl uptake in detecting malignant thyroid nodules.