"THE PREDICTIVE POWER OF CITOLOGY IN PRE-OPERATIVE DIAGNOSIS RELATED TO THE HISTOLOGICAL OUTCOMES OF THYROID SURGERY"

ABSTRACT BACKGROUND: Thyroid fine needle aspiration (FNA) is the cornerstone of the diagnostic workup for thyroid nodules. According to the classification by the “Società Italiana di Anatomia Patologica e Citologia Diagnostica” (SIAPEC), TIR3A and TIR3B categories represent indeterminate findings with low and intermediate risks of malignancy, respectively, while TIR4 and TIR5 indicate a strong suspicion or cytological certainty of carcinoma. Clarifying the predictive power of TIR3A/TIR3B cytology on histological outcomes and the relationship between cytological class and histological features (angioinvasion, extrathyroidal extension, lymph node and distant metastasis, staging), may optimize the extent of surgery and follow-up. METHODS: Retrospective analysis of patients who underwent FNA between January 2017 and December 2020 at the Endocrinology Unit of the AOU of Modena. Among 4420 subjects who underwent FNA, we selected those with cytological outcome between TIR3 and TIR5 (n=1088). Then, we considered only patients with a histological report available after surgery (n=195) and divided them into two groups based on their preoperative cytology: TIR3-TIR3A-TIR3B (n=149) and TIR4-TIR5 (n=46). For each cytological category and group, we collected histotypes, age, type of surgery (lobectomy or total thyroidectomy (TT)), type of neck dissection (central compartment (CC) or lateral compartment (LC)), tumor size, TNM and AJCC 8th edition staging, extra-thyroidal extension (ETE), lymph node/distant metastases, vascular invasion, positive margins to tumor invasion, injury of the recurrent laryngeal nerve, and post-surgical hypoparathyroidism. RESULTS: Histology was available for 195 out of 1088 patients (17.9%); 149 TIR3A/B (76.4%) and 46 TIR4/5 (23.6%). Across groups, histotype distribution differed significantly (p<0.001): TIR3 resulted mainly in benign lesions (adenomatous hyperplasia 30.9%, follicular adenoma 13.4%), and were more rarely malignant, predominantly classic papillary thyroid carcinoma (PTC)(23.5%), whereas TIR4/5 clustered in classic PTC (60.9%). Within TIR3, TIR3B had a higher prevalence of malignant histological outcomes than TIR3A (overall PTC variants 56.4% vs 31.9%), supporting the predictive role of the cytological subclassification of indeterminate outcomes. Surgical strategy scaled with cytology risk (p<0.001): TIR4/5 patients were selected for major surgery, 60.9% undergoing TT + CC neck dissection and 10.9% selected for TT + LC dissection, whereas TIR3 more frequently underwent TT without lymphadenectomy (12.1% for TT+CC, 0.7% for TT+LC). More aggressive histopathology concentrated in TIR4/5: nodal metastases were more frequent (N1a 21.7%, N1b 15.2% vs 4.0% and 0.7% in TIR3; p<0.001) and positive margins occurred more often (R1 28.3%, R2 4.4% vs R1 9.4%, R2 0%; p<0.001). Complications increased with surgical extent: hypoparathyroidism was higher in TIR4/5 (p=0.001), whereas recurrent laryngeal nerve injury did not differ (p=0.305). Overall, cancers arising in TIR3A/B more frequently had a follicular-patterned features and limited locoregional involvement, while TIR4/5 were associated with adverse pathology and the need for more extensive surgery. CONCLUSIONS: A risk gradient in histological aggressiveness is confirmed, from TIR3A to TIR3B, up to TIR4–TIR5. TIR3B tumors are more malignant and aggressive than TIR3A tumors, justifying an extensive surgical approach and more stringent surveillance; in TIR4–TIR5, the greater surgical complexity reflects more advanced diseases, with a possible increase in expected complications. Overall, the TIR3A/TIR3B subclassification improves preoperative stratification of indeterminate nodules and supports individualized surgery and follow-up. Overall, these data suggest that the indeterminate category is not a pathological "doubt" but that the carcinomas that fall into these categories actually have biological differences even at histological level.