Individual prediction of lateral neck metastasis risk in patients with unifocal papillary thyroid carcinoma

Neural Network and Logistic Regression Models Based on Ultrasound Radiomics and Clinical-Pathological Features to Predict Occult Level II Lymph Node Metastasis in Papillary Thyroid Carcinoma

RATIONALE AND OBJECTIVES

Papillary thyroid carcinoma (PTC) often metastasizes to lateral cervical lymph nodes, especially in level II. This study aims to develop predictive models to identify level II lymph node metastasis (LNM), guiding selective neck dissection (SND) to minimize unnecessary surgery and morbidity in low-risk patients.

METHODS

A retrospective cohort of 313 PTC patients who underwent modified radical neck dissection (MRND) between October 2020 and January 2023 was analyzed. The patients were randomly assigned to a training cohort (70%) and a validation cohort (30%). Five predictive models were developed using neural networks (NNET) and logistic regression (LR) based on ultrasound radiomic features, clinical-pathological data, or a combination of both. Each model's performance was evaluated based on accuracy, area under the receiver operating characteristic curve (AUC), sensitivity, and specificity in predicting occult level II LNM. SHapley Additive exPlanations and nomogram were used to interpret the most important features in the models.

RESULTS

The occurrence rate of level II LNM was 28% in the cohort. Among the five predictive models developed, the LR-radiomics signature model demonstrated the highest performance, achieving an accuracy of 96.8% and an AUC of 0.989 in the validation set. In comparison, the NNET-radiomic + clinical feature model achieved an AUC of 0.935, while other models exhibited moderate to low accuracy and AUCs ranging from 0.699 to 0.785. The decision curve analysis demonstrated that the LR-radiomics signature model provided the greatest clinical utility, offering the highest net benefit across a range of decision thresholds for identifying occult level II LNM.

CONCLUSION

Our study developed predictive models using ultrasound-derived radiomic features and clinical-pathological data to assess the risk of occult level II LNM in PTC. The LR-radiomics signature model demonstrated high accuracy, making it a valuable tool for guiding personalized treatment decisions, by informing MRND for high-risk patients and supporting SND for low-risk patients to minimize unnecessary surgical interventions and optimize clinical outcomes.

Preoperative circulating tumor cells level is associated with lymph node metastasis in patients with unifocal papillary thyroid carcinoma

OBJECTIVE

Unifocal papillary thyroid carcinoma (PTC) refers to thyroid cancer that has only one isolated lesion, it has also the possibility of lymph node metastasis (LNM). Circulating tumor cell (CTC) has been used to assist in the assessment of tumor progression, but the relationship between CTCs levels and LNM in unifocal PTC patients is unclear.

METHODS

The clinical records (age, gender, Hashimoto's thyroiditis, thyroid function, tumor size, invaded capsule (thyroid cancer penetrating the capsule), clinical stage, and LNM) of unifocal PTC patients in Meizhou People's Hospital were analyzed retrospectively. Receiver operating characteristic (ROC) curve analysis was used to determine the cutoff value of CTCs levels to distinguish LNM. The relationship between CTCs level and clinical features was analyzed. Logistic regression analysis was used to evaluate the relationship between CTCs and LNM.

RESULTS

A total of 507 unifocal PTC patients were included, and 198(39.1%) patients with LNM. The critical value of CTCs was 9.25 FU/3mL by ROC analysis, and 288(56.8%) unifocal PTC patients with preoperative CTC-positive(≥ 9.25 FU/3mL). The patients with positive CTCs had higher proportions of normal thyroid function (91.3% vs. 84.5%, p = 0.018), and LNM (44.1% vs. 32.4%, p = 0.008) than patients with negative. High preoperative CTCs level (≥ 9.25/<9.25 FU/3mL, odds ratio(OR): 1.653, 95% confidence interval(CI): 1.115-2.451, p = 0.012), tumor size > 1 cm (OR: 3.189, 95% CI: 2.069-4.913, p < 0.001), and invaded capsule (OR: 1.521, 95% CI: 1.005-2.302, p = 0.047) were associated with LNM among unifocal PTC in multivariate logistic regression analysis.

CONCLUSIONS

High preoperative CTCs level (≥ 9.25 FU/3mL), tumor size > 1 cm, and invaded capsule were associated with LNM among unifocal PTC.

Accuracy of papillary thyroid cancer prognostic nomograms: a systematic review

Objective

Current staging and risk-stratification systems for predicting survival or recurrence of patients with differentiated thyroid carcinoma may be ineffective at predicting outcomes in individual patients. In recent years, nomograms have been proposed as an alternative to conventional systems for predicting personalized clinical outcomes. We conducted a systematic review to evaluate the predictive performance of available nomograms for thyroid cancer patients.

Design and methods

PROSPERO registration (CRD42022327028). A systematic search was conducted without time and language restrictions. PICOT questions: population, patients with papillary thyroid cancer; comparator prognostic factor, single-arm studies; outcomes, overall survival, disease-free survival, cancer-specific survival, recurrence, central lymph node metastases, or lateral lymph node metastases; timing, all periods; setting, hospital setting. Risk of bias was assessed through PROBAST tool.

Results

Eighteen studies with a total of 20 prognostic models were included in the systematic review (90,969 papillary thyroid carcinoma patients). Fourteen models were at high risk of bias and four were at unclear risk of bias. The greatest concerns arose in the analysis domain. The accuracy of nomograms for overall survival was assessed in only one study and appeared limited (0.77, 95% CI: 0.75-0.79). The accuracy of nomograms for disease-free survival ranged from 0.65 (95% CI: 0.55-0.75) to 0.92 (95% CI: 0.91-0.95). The C-index for predicting lateral lymph node metastasis ranged from 0.72 to 0.92 (95% CI: 0.86-0.97). For central lymph node metastasis, the C-index of externally validated studies ranged from 0.706 (95% CI: 0.685-0.727) to 0.923 (95% CI: 0.893-0.946).

Conclusions

Our work highlights the extremely high heterogeneity among nomograms and the critical lack of external validation studies that limit the applicability of nomograms in clinical practice. Further studies ideally using commonly adopted risk factors as the backbone to develop nomograms are required.

Significance statement

Nomograms may be appropriate tools to plan treatments and predict personalized clinical outcomes in patients with papillary thyroid cancer. However, the nomograms developed to date are very heterogeneous, and their results seem to be closely related to the specific samples studied to generate the same nomograms. The lack of rigorous external validation procedures and the use of risk factors that sometimes appear to be far from those commonly used in clinical practice, as well as the great heterogeneity of the risk factors considered, limit the ability of nomograms to predict patient outcomes and thus their current introduction in clinical practice.

Construction of prediction models for determining the risk of lateral lymph node metastasis in patients with thyroid papillary carcinoma based on gender stratification

BACKGROUND

Lateral lymph node metastasis (LLNM) is associated with poor prognosis in patients with papillary thyroid cancer (PTC). The purpose of this study was to determine the risk factors for LLNM and establish prediction models that could individually assessed the risk of LLNM.

METHODS

A total of 619 PTC patients were retrospectively analyzed in our study. Univariate and multivariate analysis were performed for male and female patients, respectively, to assess relationships between clinicopathological features and LLNM. By integrating independent predictors selected by binary logistic regression modeling, preoperative and postoperative nomograms were developed to estimate the risk of LLNM.

RESULTS

LLNM was detected in 80 of 216 male patients. Of 403 female patients, 114 had LLNM. The preoperative nomogram of male patients included three clinical variables: the number of foci, tuner size, and echogenic foci. In addition to the above three variables, the postoperative nomogram of male patients included extrathyroidal extension (ETE) detected in surgery, central lymph node metastasis (CLNM) and high-volume CLNM. The preoperative nomogram of female patients included the following variables: age, chronic lymphocytic thyroiditis (CLT), BRAF V600E, the number of foci, tumor size and echogenic foci. Variables such as CLT, BRAF V600E, the number of foci, tumor size, ETE detected in surgery, CLNM, high-volume CLNM and central lymph node ratio were included in the postoperative nomogram. Above Nomograms show good discrimination.

CONCLUSIONS

Considering the difference in the incidence rate of LLNM between men and women, a separate prediction system should be established for patients of different genders. These nomograms are helpful in promoting the risk stratification of PTC treatment decision-making and postoperative management.

Dual-modal radiomics for predicting cervical lymph node metastasis in papillary thyroid carcinoma

BACKGROUND

Preoperative prediction of cervical lymph node metastasis (CLNM) in patients with papillary thyroid carcinoma (PTC) is significant for surgical decision-making.

OBJECTIVE

This study aims to develop a dual-modal radiomics (DMR) model based on grayscale ultrasound (GSUS) and dual-energy computed tomography (DECT) for non-invasive CLNM in PTC.

METHODS

In this study, 348 patients with pathologically confirmed PTC at Jiangsu University Affiliated People's Hospital who completed preoperative ultrasound (US) and DECT examinations were enrolled and randomly assigned to training (n = 261) and test (n = 87) cohorts. The enrolled patients were divided into two groups based on pathology findings namely, CLNM (n = 179) and CLNM-Free (n = 169). Radiomics features were extracted from GSUS images (464 features) and DECT images (960 features), respectively. Pearson correlation coefficient (PCC) and the least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation were then used to select CLNM-related features. Based on the selected features, GSUS, DECT, and GSUS combined DECT radiomics models were constructed by using a Support Vector Machine (SVM) classifier.

RESULTS

Three predictive models based on GSUS, DECT, and a combination of GSUS and DECT, yielded performance of areas under the curve (AUC) = 0.700 [95% confidence interval (CI), 0.662-0.706], 0.721 [95% CI, 0.683-0.727], and 0.760 [95% CI, 0.728-0.762] in the training dataset, and AUC = 0.643 [95% CI, 0.582-0.734], 0.680 [95% CI, 0.623-0.772], and 0.744 [95% CI, 0.686-0.784] in the test dataset, respectively. It shows that the predictive model combined GSUS and DECT outperforms both models using GSUS and DECT only.

CONCLUSIONS

The newly developed combined radiomics model could more accurately predict CLNM in PTC patients and aid in better surgical planning.

Lymph node ratio in lateral neck is an independent risk factor for recurrence-free survival in papillary thyroid cancer patients with positive lymph nodes

BACKGROUND

The purpose of this study was to explore the impact of the lymph node ratio on prognosis in papillary thyroid cancer patients with lymph node metastasis.

METHODS

Data from papillary thyroid cancer patients with positive nodes who were initially treated at our institution during 2015-2016 were analysed. Univariate and multivariate Cox proportional hazard models were adopted to predict prognostic factors. A receiver operating characteristic (ROC) curve was used to find the best cut-off value of the lymph node ratio (LNR). Kaplan-Meier curves were used to show the relationship between the LNR in the lateral neck and recurrence-free survival.

RESULTS

The median follow-up time was 64.6 months, and recurrence occurred in 16 of 662 patients (2.27%). Univariate analysis showed that male sex, primary tumour size (&gt;17 mm), visible extrathyroidal extension, LNR in the central neck (&gt;0.5), LNR in the lateral neck (&gt;0.10), and visible extranodal extension were significantly correlated with recurrence-free survival (RFS) (p &lt; 0.05). Multivariate analysis using the Cox proportional hazards model showed that the LNR in the lateral neck was an independent risk factor for RFS (p = 0.039; HR 14.76).

CONCLUSION

The LNR in the lateral neck was an independent risk factor for recurrence-free survival. For patients with a high lymph node ratio in the lateral neck, more frequent follow-up might be needed.

Preoperative prediction of central neck metastasis in patients with clinical T1-2N0 papillary thyroid carcinoma

BACKGROUND

Central neck metastasis (CNM) is common in patients with papillary thyroid carcinoma (PTC). However, the prediction of CNM risk remains poorly defined, especially for patients with clinically negative lymph nodes. We developed a preoperative clinical nomogram to predict CNM risk in patients with clinical T1-2N0 (cT1-2N0) PTC.

METHODS

Data from 436 patients with unifocal cT1-2N0 PTC were available. We analyzed the association between preoperative variables and CNM using univariate and multivariate logistic regression and developed a clinical nomogram based on the multivariate regression model. The nomogram was validated externally using an independent dataset.

RESULTS

The CNM rate was 25.5%. Three clinical variables were associated with CNM, including age, gender, and tumor size. We built a CNM nomogram integrating these three variables. It had a poor index of internal discrimination (C-index, 0.655; 95% CI 0.596-0.715) and a poor index of external discrimination (C-index, 0.690; 95% CI 0.611-0.769).

CONCLUSIONS

We developed a preoperative nomogram to quantify the risk of CNM in unifocal cT1-2N0 PTC patients. However, our data showed that preoperative clinical parameters were not able to accurately predict the likelihood of CNM. Other variables need to be investigated to improve the prediction capability of this nomogram.

The incidence and features of Delphian lymph node involvement in patients with papillary thyroid carcinoma

Introduction

In papillary thyroid cancer patients, the extent of dissection is still a matter of debate. Evaluating Delphian lymph nodes (DLNs) during the surgery has been speculated as a valuable tool to determine the extent of dissection. Herein, we aimed to evaluate the incidence and features of DLNs involvement in patients with papillary thyroid carcinoma.

Method

We conducted this cross-sectional study among surgical cases of papillary thyroid cancer. Patients were divided based on their DLNs involvement status. Their age, gender, location of the mass, lymphatic involvement, tumor size, tumor characteristics, pathology report, and operation note features were compared between the two groups. Definitive pathology slides of the patients were evaluated regarding DLN features.

Results

Of the 61 patients (mean age: 38.2 ± 12.0), 45 (73.8%) were females. In 13 (21.3%) patients, DLNs involvement was reported. A statistically significant relationship was noted between DLNs involvement and other lymph nodes' involvement on the same side of the mass (P < 0.001), the opposite side (P = 0.041), and also central lymph nodes (P < 0.001). Vascular invasion was also significantly higher among patients with DLNs involvement (P = 0.012).

Conclusion

Since DLNs involvement is significantly associated with extensive nodal involvement, intraoperative evaluation of DLNs is recommended to establish the extent to which dissection should be performed.

Extranodal extension is an independent predictor of extensive nodal metastasis in T1 papillary thyroid cancer

PURPOSE

Extensive lymph node metastasis (ELM) can occur in some patients with T1 papillary thyroid cancer (PTC). However, the risk factors for ELM in patients with T1 PTC have not been fully explored. In this study, we aimed to examine the association between extranodal extension (ENE) and ELM in patients with T1 PTC.

PATIENTS AND METHODS

We identified 645 consecutive patients who had T1 PTC initially resected at our centre. Clinical and pathological data were reviewed. Univariate and multivariate logistic regression analyses were performed to identify the risk factors for ELM.

RESULTS

ELM was identified in 3.9% of T1 PTC patients, and ENE was identified in 8.1% of patients. ENE was associated with male sex, large tumour size, more positive nodes, and comprehensive surgical treatment. In multivariate analysis, three variables were independently associated with ELM, including ENE (odds ratio [OR], 11.15; 95% confidence interval [CI], 4.54 to 27.30; P < 0.001), age (OR, 0.96; 95% CI, 0.93 to 0.99; P = 0.022), and tumour size (OR, 1.18; 95% CI, 1.06 to 1.31; P = 0.002). Bilateral multifocality and sex were not independently associated with ELM.

CONCLUSION

ENE is a strong independent predictor of ELM in patients with T1 PTC. Patients with ENE-positive nodes might need extensive neck dissection.

Clinical implications of Delphian lymph node metastasis in papillary thyroid carcinoma

Background

As is known, identifying risk factors precisely for lymph node metastasis (LNM) plays a vital role in initial treatment for papillary thyroid carcinoma (PTC). Nonetheless, whether Delphian lymph node (DLN) metastasis has value in predicting LNM remains an open question. This study covered a sample of 1,575 patients, which is the largest sample group so far, aiming to assess the predictive validity of DLN metastasis in PTC.

Methods

This retrospective cohort study was conducted with 1,575 eligible PTC patients who underwent thyroid operation between July 2013 and December 2018 and clinicopathologic parameters of patients with DLN metastasis were compared with those without DLN metastasis.

Results

The incidence of DLN metastasis, according to our research samples, is 24.4% (384/1,575 patients). And results show that DLN positivity was closely associated with adverse prognostic factors including younger age, larger tumor size, extrathyroid extension, tumor location in the isthmus or upper lobe of the thyroid, number of LNM >5, higher recurrence. After carefully adjusting important confounding factors, we find that in multivariate logistic regression analyses, DLN metastasis is an independent predictor for both central LNM (CLNM, adjusted OR =7.81, P<0.001) and lateral LNM (LLNM, adjusted OR =3.40, P<0.001). Moreover, the stratified analyses also show convincing evidence of a positive correlation between DLN metastasis and LNM in levels II-IV in the vast majority of subgroups.

Conclusions

The present study suggests that DLN metastasis is an independent risk factor for CLNM and LLNM of levels II-IV. The cervical lymph nodes should be meticulously evaluated to guide tailored treatment during operation in patients with DLN involvement.

Prediction of Cervical Lymph Node Metastasis Using MRI Radiomics Approach in Papillary Thyroid Carcinoma: A Feasibility Study

BACKGROUND

Cervical lymph node (LN) metastasis of papillary thyroid carcinoma (PTC) is critical for treatment and prognosis. To examine the feasibility of MRI radiomics to preoperatively predict cervical LN metastasis in patients with PTC.

METHODS

Between January 2015 and March 2018, a total of 61 patients with pathologically confirmed PTC were analyzed retrospectively. The patients were divided into cervical LN metastasis group (n = 37) and no cervical LN metastasis (n = 24). T2WI and T2WI-fat-suppression (T2WI-FS) images were collected. A number of radiomic features were automatically extracted from the largest section of tumor. Three types of classifier (the random forests, the support vector machine classifier and the generalized linear model) based on T2WI and T2WI-FS images of cervical LN metastasis and no cervical LN metastasis were constructed and evaluated with a nested cross-validation scheme.

RESULTS

Radiomic features extracted from T2WI images were more discriminative than T2WI-FS images. The random forests model showed the best discriminate performance with the highest area under the curve (0.85, CI:0.76 -1), accuracy (0.87), sensitivity (0.83), specificity (1.00), positive predictive value (PPV = 1.00) and negative predictive value (NPV = 0.88).

CONCLUSION

MRI radiomics analysis based on conventional T2WI and T2WI-FS can predict cervical LN metastasis in patients with PTC, and the radiomics is shown to be an assistant diagnosis tool for radiologists.

Prophylactic lymph node dissection in clinically N0 differentiated thyroid carcinoma: example of personalized treatment

Considering the 'differentiated thyroid carcinoma (DTC) epidemic', the indolent nature of DTC imposes a treatment paradigm shift toward elimination of recurrence. Lymph node metastases in cervical compartments, encountered in 20-90% of DTC, are the main culprit of recurrent disease, affecting 5-30% of patients. Personalized risk-stratified cervical prophylactic lymph node dissection (PLND) at initial thyroidectomy in DTC with no clinical, sonographic or intraoperative evidence of lymph node metastases (clinically N0) has been advocated, though not unanimously. The present review dissects the controversy over PLND. Weighing the benefit yielded from PLND up against the PLND-related morbidity is so far hampered by the inconsistent profit yielded by PLND and the challenging patient selection. Advances in tailoring PLND are anticipated to empower optimal patient care.

Development and validation of a nomogram for preoperative prediction of cervical lymph node involvement in thyroid microcarcinoma

Cervical regional lymph node involvement (CRLNI) is common in papillary thyroid microcarcinoma (PTMC), but the way to deal with cervical lymph node involvement of clinically negative PTMC is controversial. We studied data of patients histologically confirmed PTMC in the Surveillance, Epidemiology, and End Results (SEER) Program and Department of Surgical Oncology in Hangzhou First People's Hospital (China). We screened 6 variables of demographic and clinicopathological characteristics as potential predictors and further constructed a lymph node involvement model based on the independent predictors including age, race, sex, extension, multifocality and tumor size. The model was validated by both the internal and the external testing sets, and the visual expression of the model was displayed by a nomogram. As a result, the C-index of this predictive model in the training set was 0.766, and the internal and external testing sets through cross-validation were 0.753 and 0.668, respectively. The area under the receiver operating characteristic curve (AUC) was 0.766 for the training set. We also performed a Decision Curve Analysis (DCA), which showed that predicting the cervical lymph node involvement risk applying this nomogram would be better than having all patients or none patients use this nomogram.