showed a trendp = 0.07
ROC AUCs showed that TRAb-KR had no ability to predict recurrence (0.57, 95% CI 0.41–0.73; p = 0.40), which is why no appropriate cut-off could be determined, whereas TRAb-IM showed a trend towards correlation with recurrence (0.65, 95% CI 0.47–0.82; p = 0.07), even though not reaching statistical significance.