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

Abstract

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.

Low-cost Thyroid Nodule Diagnosis Using a Supervised Autoencoder Neural network

“…Thus, it would be interesting to develop less expensive methods to improve the risk stratification of malignancy among ITNs. In this sense, it is well known that certain routine characteristics may refine the risk of malignancy among one cytological category [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [51]. Some studies have led to the development of classification methods based on clinical, biological, radiological, and cytological data [23].…”

“…Some studies have led to the development of classification methods based on clinical, biological, radiological, and cytological data [23]. Those methods yielded scores and nomograms with performance ranging from 76.2 % to 91 % and 72.1 %to 91 %in terms of accuracy and AUC, respectively [14], [15], [16], [18], [19], [20], [21], [22], [23]. However, they were based on reliable but outdated mathematical models.…”

“…We also evaluated the metrics of four classical machine learning methods on the same dataset: a partial least squares-discriminant Of the initial features, we picked 12 features of higher importance: the TI-RADS score, the cytological category according to TBSRTC, the cytological subclassification according to Valderrabano et al [5], the nodule's size in echography, the presence of colloid in the cytologic sample, the age, clinical hyperthyroidism, the Hashimoto's thyroiditis, the gender, clinical hypothyroidism, the existence of a thyroidopathy of any kind, and the cellularity of the cytologic sample, in order of decreasing importance. Features like the TI-RADS score, the Bethesda category are well known to be associated with the risk of malignancy of one TN, and remain the spearheads of the current diagnostic strategy [14], [15], [16], [18], [19], [20], [21], [22], [23]. The cytologic subclassification according to Valderrabano et al is an emerging classification, which has already proven its ability to refine the risk of malignancy among ITNs [5].…”

“…In fact, classical nodule sub-classifications, such as the TI-RADS classification, only use a few features of the nodule [12], [13]. Moreover, more advanced efforts using modern methods either were realised using only echography or cytology data, or tried to integrate all sources of nodule information, to the avail of an ungeneralizable model because of a lacking modus operandi [11], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23]; the only successful methods exhibit prohibitive costs, as mentioned previously. In this paper, we take an integrative approach and use three sources of nodule information: clinical, echography and cytology data.…”

The diagnostic value of a nomogram based on multimodal ultrasonography for thyroid-nodule differentiation: A multicenter study

“…Thyroid cancer is among the most common endocrine malignancies. In recent years, the detection rate of thyroid nodules has increased due to the popularity of conventional ultrasonography (US), but the incidence of thyroid cancer is only 7–15% ( 1 , 2 ). As a result, it is necessary to accurately distinguish between benign and malignant thyroid nodules.…”