Bin Cao, Gao Fei, Si-Cong Zhou, Yiwen Ding, Leilei Liu
This internally validated model integrates preoperative anatomical (McGoon ratio, RVOT diameter), intraoperative (CPB time), and postoperative physiological (inotropic score, oxygenation, PVOV) factors. It may serve as a potentially useful tool for early risk stratification and could support a dual-phase assessment strategy to guide targeted interventions in pediatric TOF repair. External validation in independent cohorts and prospective clinical implementation studies are warranted before routine clinical adoption.
BACKGROUND: Predicting occult central lymph node metastasis (CLNM) in papillary thyroid microcarcinoma (PTMC) is important for individualized surgical planning and risk-stratified decision-making, particularly in younger patients. This study aimed to develop and internally validate a simple four-variable nomogram integrating age, American College of Radiology Thyroid Imaging Reporting and Data System (ACR-TIRADS), BRAF V600E status, and Bethesda cytology for preoperative prediction of occult CLNM in PTMC.
METHODS: We retrospectively analyzed 130 PTMC patients (≤10 mm on pre-operative ultrasound) who underwent thyroidectomy with central lymph node dissection (CLND) at a single Chinese tertiary center (March 2020-December 2022). A four-variable logistic-regression nomogram (Model E) was developed integrating age, ACR-TIRADS category, BRAF V600E mutation status (from fine-needle aspiration), and Bethesda cytological classification. Internal validation used 5-fold stratified cross-validation. Calibration was assessed by the Hosmer-Lemeshow test and calibration curves; clinical utility by decision curve analysis (DCA).
RESULTS: CLNM was confirmed in 73/130 patients (56.2%). The 5-fold cross-validated area under the receiver operating characteristic curve (AUC) was 0.686, with acceptable calibration (Brier score 0.215; Hosmer-Lemeshow P=0.43). DCA demonstrated positive net clinical benefit across threshold probabilities of 0.31-0.95. Pre-specified but exploratory subgroup analyses suggested higher apparent discrimination in patients younger than 45 years (n=60; bootstrap median AUC 0.760, 95% CI: 0.574-0.888) and in BRAF wild-type cases (n=35; 0.864, 95% CI: 0.705-0.977), although CIs were wide and overlapping. In this PTMC cohort, BRAF V600E was not independently associated with CLNM; an inverse point estimate was observed but did not reach statistical significance and should be regarded as exploratory.
CONCLUSIONS: This simple, four-variable nomogram provides clinically useful pre-operative CLNM risk assessment in PTMC, particularly in younger and BRAF wild-type patients-subgroups in which preoperative risk stratification is most challenging. External multi-center validation is the necessary next step.