The Malignancy Risk Assessment of Cytologically Indeterminate Thyroid Nodules Improves Markedly by Using a Predictive Model

Improving the TIR3B oncological stratification: try to bridge the gap through a comprehensive presurgical algorithm

PURPOSE

Indeterminate cytology still puzzles clinicians, due to its wide range of oncological risks. According to the Italian SIAPEC-IAP classification, TIR3B cytology holds up to 30% of thyroid cancer, which justifies the surgical indication, even if more than half of cases do not result in a positive histology. The study aim is to identify potential clinical, ultrasound or cytological features able to improve the surgical indication.

METHODS

Retrospective analysis. A consecutive series of TIR3B nodules referred to the Endocrine Unit of Careggi Hospital from 1st May 2014 to 31st December 2021 was considered for the exploratory analysis (Phase 1). Thereafter, a smaller confirmatory sample of consecutive TIR3B diagnosed and referred to surgery from 1st January 2022 to 31st June 2022 was considered to verify the algorithm (Phase 2). The main clinical, ultrasound and cytological features have been collected. A comprehensive stepwise logistic regression was applied to build a prediction algorithm. The histological results represented the final outcome.

RESULTS

Of 599 TIR3B nodules referred to surgery, 451 cases were included in the exploratory analysis. A final score > 14.5 corresponded to an OR = 4.98 (95% CI 3.24-7.65, p < 0.0001) and showed a PPV and NPV of 57% and 79%, respectively. The Phase 2 analysis on a confirmatory sample of 58 TIR3B cytology confirmed that a threshold of 14.5 points has a comparable PPV and NPV of 53% and 80%, respectively.

CONCLUSIONS

A predictive algorithm which considers the main clinical, US and cytological features can significantly improve the oncological stratification of TIR3B cytology.

The Impact of BRAF V600E Mutation Allele Frequency on the Histopathological Characteristics of Thyroid Cancer

BACKGROUND

A BRAF V600E mutation in papillary thyroid cancer (PTC) has been shown to be associated with aggressive behavior. Nevertheless, not all BRAF V600E PTCs behave aggressively. Allele frequency (AF) is the number of mutated molecules divided by the total number of wild-type molecules at a specific location in the genome. The relationship between BRAF V600E AF and the histopathological features of thyroid malignancies is not well understood. We hypothesized that the BRAF V600E AF will correlate directly with aggressive histopathological behavior. The aim of this study was to examine this relationship.

METHODS

A retrospective chart review was performed for patients treated for BRAF V600E thyroid malignancies from 2019 to 2022 at McGill University tertiary care hospitals (n = 317). Patients with BRAF V600E-positive malignancies that included information on AF were included (n = 44). The correlation between AF and tumor histopathological features was analyzed.

RESULTS

Out of the 44 nodules with a BRAF V600E mutation, those with aggressive features of PTC had a mean AF of 25.8%, which was significantly higher than the non-aggressive group with a mean AF of 10.25% (p = 0.020). Additionally, there was a statistically significant difference in mean AF between patients with a positive sentinel LN (29%) and those with a negative sentinel LN (17.8%) (p = 0.021). Classical PTC was present in 29.5% (13/44) of nodules, with a mean AF of 15.6%. The tall cell subtype was found in 64% (28/44) of nodules, with a mean AF of 23%. Solid and hobnail subtypes were less common in this study, and there was no statistically significant relationship between AF and histopathological subtypes (p = 0.107). Nodules smaller than 1cm had a mean AF of 13.3%, while nodules ranging from 1 2cm had a mean AF of 20.6%, and those larger than 2cm had a mean AF of 27.7%. However, no statistical difference was observed between AF and nodule size (p = 0.160).

CONCLUSION

In this study, BRAF V600E mutations in conjunction with AF help to determine whether thyroid malignancies will display aggressive behavior. This pre-operative finding can help thyroid specialists to determine the extent of thyroidectomy and whether lymph node dissection is required.

Ultrasound Parameters Can Accurately Predict the Risk of Malignancy in Patients with "Indeterminate TIR3b" Cytology Nodules: A Prospective Study

The increase in the incidence of thyroid nodules with cytological findings of TIR3b requires the identification of predictive factors of malignancy. We prospectively evaluated 2160 patients from January 2018 to June 2022 and enrolled 103 patients with indeterminate cytology TIR3b nodules who underwent total (73 patients) and hemi-thyroidectomy (30 patients). Among them, 61 had a histological diagnosis of malignancy (30 classic papillary thyroid carcinoma, 19 had follicular papillary thyroid carcinoma variant, 3 had Hurtle cell carcinoma and 9 had follicular thyroid carcinoma), while 42 had a benign histology. Clinical, ultrasonographic and cytological characteristics were recorded. In addition, BRAF mutation was analysed. Patients with a histological diagnosis of malignancy had a higher frequency of nodule diameter ≤11 mm (p = 0.002), hypoechogenicity (p < 0.001), irregular borders (p < 0.001), peri- and intralesional vascular flows (p = 0.004) and microcalcifications (p = 0.001) compared to patients with benign histology. In contrast, patients with benign histology had more frequent nodules with a halo sign (p = 0.012) compared to patients with histological diagnosis of malignancy. No significant differences were found in BRAF mutation between the two groups. Our study suggests that the combination of ultrasonographic and cytological data could be more accurate and reliable than cytology alone in identifying those patients with TIR3b cytology and a histology of malignancy to be referred for thyroidectomy, thus reducing the number of patients undergoing thyroidectomy for benign thyroid disease.

Application of machine learning methods to guide patient management by predicting the risk of malignancy of Bethesda III-V thyroid nodules

OBJECTIVE

Indeterminate thyroid nodules (ITN) are common and often lead to (sometimes unnecessary) diagnostic surgery. We aimed to evaluate the performance of two machine learning methods (ML), based on routinely available features to predict the risk of malignancy (RM) of ITN.

DESIGN

Multi-centric diagnostic retrospective cohort study conducted between 2010 and 2020.

METHODS

Adult patients who underwent surgery for at least one Bethesda III-V thyroid nodule (TN) with fully available medical records were included. Of the 7917 records reviewed, eligibility criteria were met in 1288 patients with 1335 TN. Patients were divided into training (940 TN) and validation cohort (395 TN). The diagnostic performance of a multivariate logistic regression model (LR) and its nomogram, and a random forest model (RF) in predicting the nature and RM of a TN were evaluated. All available clinical, biological, ultrasound, and cytological data of the patients were collected and used to construct the two algorithms.

RESULTS

There were 253 (19%), 693 (52%), and 389 (29%) TN classified as Bethesda III, IV, and V, respectively, with an overall RM of 35%. Both cohorts were well-balanced for baseline characteristics. Both models were validated on the validation cohort, with performances in terms of specificity, sensitivity, positive predictive value, negative predictive value, and area under the receiver operating characteristic curve of 90%, 57.3%, 73.4%, 81.4%, 84% (CI95%: 78.5%-89.5%) for the LR model, and 87.6%, 54.7%, 68.1%, 80%, 82.6% (CI95%: 77.4%-87.9%) for the RF model, respectively.

CONCLUSIONS

Our ML models performed well in predicting the nature of Bethesda III-V TN. In addition, our freely available online nomogram helped to refine the RM, identifying low-risk TN that may benefit from surveillance in up to a third of ITN, and thus may reduce the number of unnecessary surgeries.

Coexisting Molecular Alterations Increase the Risk of Malignancy in Thyroid Nodules with Copy Number Alterations

Molecular mutations and alterations play a role in thyroid tumorigenesis. Different alterations are associated with different clinical and pathological characteristics. Copy number alterations (CNAs) are known to be present in some thyroid tumors; however, their idiosyncratic clinicopathological implications are not yet well elucidated. A retrospective chart review was performed to identify patients with CNAs on pre-operative molecular testing results who subsequently underwent surgical treatment between January 2016 and April 2022 at McGill University teaching hospitals. Of the 316 patients with thyroid nodules who opted for molecular testing with ThyroSeqV3 followed by surgery, 67 (21.2%) nodules were positive for CNAs, including 23 Bethesda III, 31 Bethesda IV, 12 Bethesda V and 1 Bethesda VI nodules. On surgical pathology, 29.9% were benign and 70.1% were malignant or non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP). Among those that were malignant/NIFTP, 17.02% were considered to be aggressive cancers. The presence of other molecular alterations was found to be an independent predictor of malignancy in multivariate analysis (OR = 5.087, 95% C.I. = 1.12-23.04, p = 0.035). No unique factor was correlated with aggressiveness; however, CNA-positive thyroid nodules that were associated with high-risk mutations such as BRAF V600E, TP53, NTRK1/3 fusion, or PTEN mutation with high allele frequency (AF) ended up being aggressive cancers. Most of the CNA-positive thyroid nodules resulted in follicular patterned tumors in 41 (65.2%) cases and oncocytic tumors in 20 (29.9%) cases. This study demonstrates that 70.1% of surgically resected thyroid nodules with CNAs were malignant/NIFTP. Most CNA-positive thyroid nodules were either oncocytic patterned tumors or follicular patterned tumors. Furthermore, CNA-positive thyroid nodules were more likely to be malignant if they were associated with other molecular alterations or mutations.

Identification of potential diagnostic and prognostic biomarkers for papillary thyroid microcarcinoma (PTMC) based on TMT-labeled LC-MS/MS and machine learning

OBJECTIVES

To explore the molecular mechanisms underlying aggressive progression of papillary thyroid microcarcinoma and identify potential biomarkers.

METHODS

Samples were collected and sequenced using tandem mass tag-labeled liquid chromatography-tandem mass spectrometry. Differentially expressed proteins (DEPs) were identified and further analyzed using Mfuzz and protein-protein interaction analysis (PPI). Parallel reaction monitoring (PRM) and immunohistochemistry (IHC) were performed to validate the DEPs.

RESULTS

Five thousand, two hundred and three DEPs were identified and quantified from the tumor/normal comparison group or the N1/N0 comparison group. Mfuzz analysis showed that clusters of DEPs were enriched according to progressive status, followed by normal tissue, tumors without lymphatic metastases, and tumors with lymphatic metastases. Analysis of PPI revealed that DEPs interacted with and were enriched in the following metabolic pathways: apoptosis, tricarboxylic acid cycle, PI3K-Akt pathway, cholesterol metabolism, pyruvate metabolism, and thyroid hormone synthesis. In addition, 18 of the 20 target proteins were successfully validated with PRM and IHC in another 20 paired validation samples. Based on machine learning, the five proteins that showed the best performance in discriminating between tumor and normal nodules were PDLIM4, ANXA1, PKM, NPC2, and LMNA. FN1 performed well in discriminating between patients with lymph node metastases (N1) and N0 with an AUC of 0.690. Finally, five validated DEPs showed a potential prognostic role after examining The Cancer Genome Atlas database: FN1, IDH2, VDAC1, FABP4, and TG. Accordingly, a nomogram was constructed whose concordance index was 0.685 (confidence interval: 0.645-0.726).

CONCLUSIONS

PDLIM4, ANXA1, PKM, NPC2, LMNA, and FN1 are potential diagnostic biomarkers. The five-protein nomogram could be a prognostic biomarker.

Prevalence of thyroid carcinoma in nodules with thy 3 cytology: the role of preoperative ultrasonography and strain elastography

BACKGROUND

Fine needle aspiration (FNA) cytology, the gold standard in assessing thyroid nodules, is limited by its inability to determine the true risk of malignancy in Thy 3 nodules. Most patients with Thy3 cytology undergo surgery to establish a histologic diagnosis. The aims of this study were to evaluate the prevalence of malignancy in Thy3 nodules, to examine the ultrasound (US) characteristics that are associated with a high cancer risk and to assess the role of real-time strain elastography.

METHODS

Retrospective cohort study of 99 nodules with Thy3 cytology in 99 patients who underwent thyroidectomy over a three-year period. Grayscale US, Doppler and real-time strain elastography data were evaluated.

RESULTS

Eighty-one nodules (81.82%) were benign, 18 (18.18%) were malignant, and almost all were papillary thyroid carcinoma (PTC). Univariable analysis revealed irregular margins (p = 0.02), ill-defined borders (p ≤ 0.001), a taller than wide shape (p ≤ 0.001) and the elasticity score (p = 0.02) as significant predictors of malignancy. Multivariable analysis showed that ill-defined borders and the elasticity score were significant and independent factors associated with malignancy. All soft nodules (elasticity scores 1-2) were benign (sensitivity 100%, specificity 33%, NPV 100%, and PPV 23%). There was a higher rate of malignancy in Thy3a nodules than in Thy3f nodules (42.86% versus 11.54%) (p ≤ 0.001).

CONCLUSIONS

Irregular margins, ill-defined borders, a taller than wide shape and low elasticity were associated with malignancy. Elastography should be performed when evaluating Thy3 nodules.

The gray zone of thyroid nodules: Using a nomogram to provide malignancy risk assessment and guide patient management

Background

Thyroid nodules have a low prevalence of malignancy and most proven cancers do not behave aggressively. Thus, risk‐stratification of nodules is a critical step to avoid surgical overtreatment. We hypothesized that a risk management system superior to those currently in use could be created to reduce the number of clinically indeterminate nodules (i.e., the “gray zone”) by concurrently considering the malignancy risks conferred by clinical, ultrasonographic, and cytologic variables.

Methods

Thyroidectomy cases were reviewed from three institutions. Their benign versus malignant outcome was used to evaluate the variables for correlation. A binary logistic regression model was trained and, using indeterminate nodules with Bethesda III and IV results, validated. A scoring nomogram was designed to demonstrate the application of the model in clinical practice.

Results

One hundred thirty thyroidectomies (28% malignant) met inclusion criteria. The final logistic regression model included difficulty in swallowing, hypothyroidism, echogenicity, hypervascularity, margins, calcification, and cytology diagnosis as input parameters. The model was highly successful in determining the outcome (p value: 0.001) with a R²(Nagelkerke) score of 0.93. The area under the curve as determined by receiver operating characteristics was 0.91. The accuracy of the model on the training dataset was 93% (sensitivity and specificity 92% and 96%, respectively) and, on the validation dataset, 80% (sensitivity and specificity 91% and 67%, respectively).

Conclusions

We report a model for risk assessment of thyroid nodules that has the potential to significantly reduce indeterminates and surgical overtreatment. We illustrate its application via a straightforward nomogram, which integrates clinical, ultrasonographic, and cytologic data, and can be used to create clear, evidence‐based management plans for patients.

Atypia of undetermined significance/follicular lesion of undetermined significance: Asian vs. non-Asian practice, and the Singapore experience

The Bethesda System for Reporting Thyroid Cytopathology has paved the way for comparisons of the practice of thyroid cytology in many different regions. However, there have been comparatively few studies documenting differences between Asian and non-Asian practice. Here, we aim to compare a few key parameters between the two regions, focusing on the indeterminate category of atypia of undetermined significance (AUS)/follicular lesion of undetermined significance (FLUS). We compared its incidence, resection rates (RRs), risk of malignancy (ROM), rate of repeat fine needle aspiration (rFNA), ROMs of cytomorphologic subcategories of nuclear atypia (AUS-N) vs. architectural atypia (AUS-A), and, finally, the incidence of papillary thyroid carcinoma (PTC) vs. follicular neoplasms (FNs) in resected AUS/FLUS cases in Asian and non-Asian regions. Where possible, these metrics were compared with the Singapore experience from a tertiary referral institution. While the incidence of AUS/FLUS was similar in both regions, we found geographical differences in the RRs and ROMs, which may reflect a higher collective threshold for surgery in Asian countries. However, both cohorts showed higher ROMs in the AUS-N subcategory as compared to the AUS-A subcategory, supporting the subclassification of the AUS/FLUS based on the presence of nuclear atypia. We also observed a higher incidence of AUS-N coupled with a higher incidence of PTC in resected AUS/FLUS nodules in Asian cohorts, while AUS-A and follicular-patterned neoplasms featured more prominently in the non-Asian cohorts. These incidences may account for the starkly different molecular approaches that we noted-in Asian (chiefly Korean and Chinese) centers, BRAF mutational analysis was favored, while gene panels and gene expression classifiers were more frequently applied in non-Asian centers (chiefly in the United States of America). Overall, the data from Singapore appears more closely aligned to non-Asian trends, despite its geographical location in Southeast Asia and its predominantly Asian population. We conclude that there is significant heterogeneity in the outcomes of the AUS/FLUS categories between and within regions, which is only partially explained by regional variations, and may also reflect different regional diagnostic and management practices. This highlights the importance of understanding the local context in the interpretation of indeterminate Bethesda categories, rather than adopting a "one-size fits all" approach.

BRAFV600E mutation combined with American College of Radiology thyroid imaging report and data system significantly changes surgical resection rate and risk of malignancy in thyroid cytopathology practice

Background

Ultrasonography patterns and molecular testing may assist in stratifying the malignancy risk of indeterminate cytology diagnosis. The purpose of this study was to assess the value of fine needle aspiration (FNA) cytology in combination with American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS) and BRAF^V600E mutation in differentiating high-risk thyroid nodules.

Methods

From February 2010 to February 2014, 2,643 consecutive thyroid nodules from 2,399 patients (688 men and 1,711 women; mean age, 44.3±12.5 years) who underwent preoperative FNA biopsies were enrolled. The Jiangsu Institute of Nuclear Medicine has adopted TI-RADS stratification and BRAF^V600E mutation analysis as a routine procedure to assist in evaluating FNA cytopathology since January 2016. From February 2017 to July 2018, 1,905 thyroid nodules of 1,837 patients (501 men and 1,336 women, 49.5±12.8 years) who underwent preoperative ultrasound-guided FNA biopsies with BRAF^V600E mutation analysis and ACR TI-RADS grading data available were enrolled for comparison in this study.

Results

The cancer risk in nodules with BRAF^V600E mutation was 99.7% (905/908) according to the histological findings. The risk of malignancy (ROM) was found to increase with advancing ACR TI-RADS category. High-risk ultrasound features (TI-RADS 5) did show a good performance in predicting malignancy (98.1%). The combination of TI-RADS 5 and BRAF^V600E mutation reached the best diagnostic efficiency [sensitivity 97.7%, specificity 94.8%, positive predictive value (PPV) 99.6%]. It is apparent that after the implementation of ACR TI-RADS and the BRAF^V600E mutation analysis, the resection rates (RRs) of The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) I and III categories showed significant decreases (39.1% vs. 21.6% and 36.1% vs. 16.7%, respectively). In contrast, the risks of malignancy of TBSRTC I and III categories indicated substantial increases (41.5% vs. 80.0% and 34.6% vs. 50.0%, respectively). The ROM of thyroid nodules with nondiagnostic (ND, I) category showed the most significant increase of 41.5% to 80.0%.

Conclusions

ACR TI-RADS, together with BRAF^V600E mutation and cytological diagnoses, can assist in improving the prediction of malignancy of thyroid nodules, especially in the TBSRTC I and III categories.

Strategy to reduce unnecessary surgeries in thyroid nodules with cytology of Bethesda category III (AUS/FLUS): a retrospective analysis of 667 patients diagnosed by surgery

PURPOSE

Fine-needle aspiration (FNA) is widely used for the diagnosis of thyroid nodules detected by ultrasonography. However, the cytology of atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS) often leads to unnecessary thyroid surgery. This study aimed to identify a strategy to reduce unnecessary surgeries in patients with AUS/FLUS nodules.

METHODS

Medical records of 667 patients with the cytology of AUS/FLUS who underwent surgery from January 2007 to December 2017 were retrospectively reviewed. Clinicopathological data were analyzed to identify malignant factors in thyroid nodules with AUS/FLUS. Factors were compared between patients with thyroid cancer and those with benign thyroid nodules, using stepwise multivariate logistic regression and decision tree model.

RESULTS

Pathological thyroid cancer was identified in 193 (43.3%) patients. There was a significant difference in malignancy incidence with respect to family history, number of nodules, number of FNAs, ultrasonographic finding, lymphocytic thyroiditis, and BRAFV600E mutation. Multivariate analysis showed that ultrasonography (K-TIRADS 5) was the most influential independent predictor of malignancy in AUS/FLUS (odds ratio = 11.02, p < 0.001), followed by possessing BRAFV600E mutation (odds ratio = 4.54, p < 0.001). This strategy enabled 226 (89.3%) patients to avoid unnecessary surgeries based on the decision tree model. There was no node of repeated FNA in the decision tree model, which reduced the risk of malignancy (odds ratio = 0.35, p = 0.029).

CONCLUSION

K-TIRADS 5 and BRAFV600E mutation were predictive of malignancy in nodules of AUS/FLUS. These factors should be considered in strategies to reduce unnecessary surgeries for AUS/FLUS.

Comparison of 18F-FDG-PET and 18F-FDG-PET/CT for the diagnostic performance in thyroid nodules with indeterminate cytology: A meta-analysis

BACKGROUNDS

We performed a meta-analysis to compare F-FDG-PET and F-FDG-PET/CT for the diagnostic performance in thyroid nodules with indeterminate cytology by Bethesda classification.

METHODS

PubMed and Embase databases were searched for eligible studies from the earliest available date of indexing through September 2019. Only studies with clearly stated histopathology confirmation were included. Publication bias was assessed by Deeks funnel plot. The pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, positive predictive value, negative predictive value, and diagnostic odds ratio was determined by random-effect analysis, respectively. All diagnostic estimate indexes were determined separately for PET alone and PET/CT and were compared pair-wisely using Z-test.

RESULTS

We included 214 patients from five PET alone studies and 420 patients from 8 PET/CT studies in this meta-analysis. The range of the prevalence of malignancy was 11% to 27% for PET alone studies (Median, 20%) and 4% to 50% for PET/CT studies (Median, 24%). The sensitivity (0.95 vs 0.73, P < .01), negative likelihood ratio (0.20 vs 0.53, P = .04) and negative predictive value (0.99 vs 0.91, P < .01) of PET alone are significantly better than those of PET/CT. For PET/CT, Fagan nomogram indicated that when the pre-test probability was set at 24%, the negative post-test probability could decrease to 12%.

CONCLUSION

This meta-analysis reveals that in evaluating thyroid nodules with indeterminate cytology by Bethesda classification, the rule-out performance of F-FDG-PET is significantly better than F-FDG-PET/CT, although the latter represents a more objective and accurate technique. We hypothesize that the lack of precise localization of suspicious FDG uptake in the neck region may have contributed to this overvaluation for PET alone studies, and advocate that future studies be performed with PET/CT rather than PET alone to avoid misinterpretation and overvaluation in this scenario.