Preoperative Identification of Medullary Thyroid Carcinoma (MTC): Clinical Validation of the Afirma MTC RNA-Sequencing Classifier.

Background: Cytopathological evaluation of thyroid fine-needle aspiration biopsy (FNAB) specimens can fail to raise preoperative suspicion of medullary thyroid carcinoma (MTC). The Afirma RNA-sequencing MTC classifier identifies MTC among FNA samples that are cytologically indeterminate, suspicious, or malignant (Bethesda categories III-VI). In this study we report the development and clinical performance of this MTC classifier. Methods: Algorithm training was performed with a set of 483 FNAB specimens (21 MTC and 462 non-MTC). A support vector machine classifier was developed using 108 differentially expressed genes, which includes the 5 genes in the prior Afirma microarray-based MTC cassette. Results: The final MTC classifier was blindly tested on 211 preoperative FNAB specimens with subsequent surgical pathology, including 21 MTC and 190 non-MTC specimens from benign and malignant thyroid nodules independent from those used in training. The classifier had 100% sensitivity (21/21 MTC FNAB specimens correctly called positive; 95% confidence interval [CI] = 83.9-100%) and 100% specificity (190/190 non-MTC FNAs correctly called negative; CI = 98.1-100%). All positive samples had pathological confirmation of MTC, while all negative samples were negative for MTC on surgical pathology. Conclusions: The RNA-sequencing MTC classifier accurately identified MTC from preoperative thyroid nodule FNAB specimens in an independent validation cohort. This identification may facilitate an MTC-specific preoperative evaluation and resulting treatment.

Identification of Hürthle cell cancers: solving a clinical challenge with genomic sequencing and a trio of machine learning algorithms.

BACKGROUND: Identification of Hürthle cell cancers by non-operative fine-needle aspiration biopsy (FNAB) of thyroid nodules is challenging. Resultingly, non-cancerous Hürthle lesions were conventionally distinguished from Hürthle cell cancers by histopathological examination of tissue following surgical resection. Reliance on histopathological evaluation requires patients to undergo surgery to obtain a diagnosis despite most being non-cancerous. It is highly desirable to avoid surgery and to provide accurate classification of benignity versus malignancy from FNAB preoperatively. In our first-generation algorithm, Gene Expression Classifier (GEC), we achieved this goal by using machine learning (ML) on gene expression features. The classifier is sensitive, but not specific due in part to the presence of non-neoplastic benign Hürthle cells in many FNAB. RESULTS: We sought to overcome this low-specificity limitation by expanding the feature set for ML using next-generation whole transcriptome RNA sequencing and called the improved algorithm the Genomic Sequencing Classifier (GSC). The Hürthle identification leverages mitochondrial expression and we developed novel feature extraction mechanisms to measure chromosomal and genomic level loss-of-heterozygosity (LOH) for the algorithm. Additionally, we developed a multi-layered system of cascading classifiers to sequentially triage Hürthle cell-containing FNAB, including: 1. presence of Hürthle cells, 2. presence of neoplastic Hürthle cells, and 3. presence of benign Hürthle cells. The final Hürthle cell Index utilizes 1048 nuclear and mitochondrial genes; and Hürthle cell Neoplasm Index leverages LOH features as well as 2041 genes. Both indices are Support Vector Machine (SVM) based. The third classifier, the GSC Benign/Suspicious classifier, utilizes 1115 core genes and is an ensemble classifier incorporating 12 individual models. CONCLUSIONS: The accurate algorithmic depiction of this complex biological system among Hürthle subtypes results in a dramatic improvement of classification performance; specificity among Hürthle cell neoplasms increases from 11.8% with the GEC to 58.8% with the GSC, while maintaining the same sensitivity of 89%.

Performance of a Genomic Sequencing Classifier for the Preoperative Diagnosis of Cytologically Indeterminate Thyroid Nodules.

Importance: Use of next-generation sequencing of RNA and machine learning algorithms can classify the risk of malignancy in cytologically indeterminate thyroid nodules to limit unnecessary diagnostic surgery. Objective: To measure the performance of a genomic sequencing classifier for cytologically indeterminate thyroid nodules. Design, Setting, and Participants: A blinded validation study was conducted on a set of cytologically indeterminate thyroid nodules collected by fine-needle aspiration biopsy between June 2009 and December 2010 from 49 academic and community centers in the United States. All patients underwent surgery without genomic information and were assigned a histopathology diagnosis by an expert panel blinded to all genomic information. There were 210 potentially eligible thyroid biopsy samples with Bethesda III or IV indeterminate cytopathology that constituted a cohort previously used to validate the gene expression classifier. Of these, 191 samples (91.0%) had adequate residual RNA for validation of the genomic sequencing classifier. Algorithm development and independent validation occurred between August 2016 and May 2017. Exposures: Thyroid nodule surgical histopathology diagnosis by an expert panel blinded to all genomic data. Main Outcomes and Measures: The primary end point was measurement of genomic sequencing classifier sensitivity, specificity, and negative and positive predictive values in biopsies from Bethesda III and IV nodules. The secondary end point was measurement of classifier performance in biopsies from Bethesda II, V, and VI nodules. Results: Of the 183 included patients, 142 (77.6%) were women, and the mean (range) age was 51.7 (22.0-85.0) years. The genomic sequencing classifier had a sensitivity of 91% (95% CI, 79-98) and a specificity of 68% (95% CI, 60-76). At 24% cancer prevalence, the negative predictive value was 96% (95% CI, 90-99) and the positive predictive value was 47% (95% CI, 36-58). Conclusions and Relevance: The genomic sequencing classifier demonstrates high sensitivity and accuracy for identifying benign nodules. Its 36% increase in specificity compared with the gene expression classifier potentially increases the number of patients with benign nodules who can safely avoid unnecessary diagnostic surgery.

Pharyngoesophageal diverticula simulating thyroid nodules: An unusual occurrence with unique features.

Pharyngoesophageal diverticula (PED) of the Zenker's and Killian-Jamieson types arise in close proximity to the thyroid gland, and may rarely be confused with a thyroid nodule on ultrasonography. In this brief report, we detail the cytologic, clinical, and radiologic findings of three PED that were thought to be thyroid nodules, and were subjected to fine-needle aspiration (FNA). The patients were females with an age range of 51-64 years. All three patients had multiple thyroid nodules, and two patients reported symptoms attributable to the diverticulum. Nodule sizes ranged from 1.0 to 2.7 cm, and either the right or left thyroid lobe could be involved. Microcalcifications were present by ultrasonography in all three cases. FNA of these thyroid nodule mimics showed squamous cells with granular or amorphous debris, bacterial and/or fungal colonies, inflammation, and food particles. These cytologic features, particularly the presence of vegetable or meat fragments, are characteristic, and have also been reported in the few previous reports of PED. The presence of a diverticulum was confirmed with imaging studies in all our patients. Although a rare occurrence, the inadvertent FNA of a PED masquerading as a thyroid nodule is important to recognize, as a recommendation for appropriate radiologic studies could potentially avoid inappropriate therapy for thyroid disease.

Identification of parathyroid tissue in thyroid fine-needle aspiration: A combined approach using cytology, immunohistochemical, and molecular methods.

OBJECTIVES: Parathyroid (PT) lesions can be difficult to recognize in thyroid fine needle aspirations (FNAs), and when not identified correctly, PT cells may be mistaken for potentially abnormal thyroid cells. We therefore studied the utility of combining cytology, immunohistochemistry, and a molecular classifier to identify PT cells in thyroid FNAs. METHODS: Thyroid FNAs were received in CytoLyt, and were evaluated initially using The Bethesda System for Reporting Thyroid Cytology (TBSRTC). The PT molecular classifier was performed along with the Afirma Gene Expression Classifier (GEC) on samples with indeterminate cytology. Immunohistochemistry (IHC) for PT was performed on all samples using Cellient cell block sections. Clinical and ultrasound information was collected, when available. RESULTS: PT tissue was identified in 60 thyroid FNAs. Forty-seven (47) samples had cytologic features that were suggestive of PT cells, and were subsequently confirmed with IHC. Thirteen (13) samples were not recognized as PT, and were considered to be either Bethesda III or IV indeterminate thyroid nodules; a PT gene expression signature was subsequently detected by the GEC. These samples were also confirmed as PT by IHC. Clinical and ultrasound features were suggestive of a PT lesion in only a third of cases. CONCLUSIONS: Cytologic features, coupled with IHC, can identify intrathyroidal PT cells in the majority of CytoLyt samples. However, a significant minority (22%) of these FNAs may be misclassified as indeterminate by TBSRTC criteria, and molecular detection of the PT tissue can be helpful to potentially avoid an additional biopsy or diagnostic surgery. Diagn. Cytopathol. 2017;45:526-532. © 2017 Wiley Periodicals, Inc.

Analytical Performance of a Gene Expression Classifier for Medullary Thyroid Carcinoma.

BACKGROUND: The aim of this study was to demonstrate the analytical validity of an RNA classifier for medullary thyroid carcinoma (MTC). METHODS: Fresh-frozen tissue specimens were obtained from commercial sources, and MTC diagnoses were confirmed by histopathology review. De-identified patient fine-needle aspiration biopsies (FNABs) and whole blood from normal donors were obtained. Total RNA was extracted, amplified, and hybridized to custom microarrays for gene expression analysis. Gene expression data were normalized and classified via a machine learning algorithm. Positive control materials were produced from MTC tissues and tested across multiple experiments and laboratories. Twenty-seven MTC tissue specimens were used to evaluate the sensitivity of the MTC classifier. Gene expression data from tissues and FNABs were used to model classifier response to mixtures of MTC samples with normal thyroid tissue, a benign thyroid nodule, a Hürthle cell adenoma, and whole blood. Select mixture conditions were confirmed in vitro. Assay tolerance to RNA input variation (5-25 ng) and genomic DNA contamination (30% by mass) was evaluated. The intra- and inter-run reproducibility and inter-laboratory accuracy of MTC classifier results were characterized. RESULTS: The MTC classifier sensitivity of 96.3% [confidence interval 81.0-99.9%] was determined retrospectively using 27 MTC confirmed tissue specimens. One false-negative result in a necrotic tissue implicated sample necrosis in reduced classifier sensitivity. Dilution modeling of MTC samples with normal or benign tissues showed consistent detection of MTC down to 20% sample proportions, with in vitro confirmation of 20% analytical sensitivity. Classifier tolerance to RNA input variation (5-25 ng), genomic DNA contamination (30% by mass), and an interfering substance (blood) was demonstrated with 100% accurate classifier results under all tested conditions. The maximum observed run-to-run score difference for a single FNAB sample was ∼1 unit compared with the average score difference between 38 MTC and non-MTC FNABs of ∼32 units. MTC classifier results for 20 tissues processed from total RNA in two different laboratories showed 100% concordance. CONCLUSIONS: The MTC classifier, offered as part of the routine molecular testing of cytology-indeterminate thyroid nodules, demonstrates robust analytical sensitivity, specificity, accuracy, and reproducibility.

The utility of cellient cell blocks in low-cellularity thyroid fine needle aspiration biopsies.

BACKGROUND: Low cellularity can be problematic in thyroid fine needle aspiration (FNA) biopsies. The Cellient cell block (CB) system has been reported to improve cell recovery compared to traditional methods. Therefore, we studied the utility of Cellient CBs in the evaluation of thyroid FNAs, with an emphasis on low-cellularity specimens. METHODS: ThinPrep slides were prepared from thyroid FNAs submitted in Cytolyt. After assessment using TBSRTC criteria, Cellient CBs were requested on samples with residual FNA material and an initial cytologic impression of non-diagnostic, AUS/FLUS, and on apparently benign samples with marginally adequate cellularity. The contribution of the CB findings to the final diagnosis was assessed. RESULTS: 965 cases of paired ThinPrep and CB sections were examined. Overall, the cell block findings resulted in a change of the initial ThinPrep impression in 15% (n = 148) of cases. The vast majority of these changed cases were initially inadequate for interpretation, and specifically, 31% (n = 123) of the non-diagnostic ThinPrep samples became diagnostic with a CB. The cell block findings contributed to a change in diagnosis in 8% (n = 23) of AUS/FLUS cases, and in less than 1% of low-cellularity benign samples. CONCLUSION: The use of CBs in low-cellularity thyroid FNAs has not been well described. In this study, we found that the contribution of CBs in this setting varied by TBSRTC category. Specifically, the samples that benefited most were initially non-diagnostic specimens and select cases of AUS/FLUS, while low-cellularity benign samples gained very little additional information. Diagn. Cytopathol. 2016;44:737-741. © 2016 Wiley Periodicals, Inc.

A Genomic Alternative to Identify Medullary Thyroid Cancer Preoperatively in Thyroid Nodules with Indeterminate Cytology.

BACKGROUND: The use of calcitonin screening for the rare medullary thyroid cancer (MTC) is controversial due to questions of efficacy, accuracy, and cost-effectiveness. This study reports the results of a large prospective validation using a machine-trained algorithm (MTC Classifier) to preoperatively identify MTC in fine-needle aspiration biopsies in lieu of calcitonin measurements. METHODS: Cytology analysis on a prospective consecutive series of 50,430 thyroid nodule biopsies yielded a total of 7815 indeterminate (Bethesda categories III/IV) cases, which were tested with the MTC classifier. A prospective, consecutively submitted series of 2673 Bethesda III-VI cases with cytology determined locally was also evaluated. RNA was isolated and tested for the MTC Classifier using microarrays. RESULTS: Forty-three cases were positive by the MTC Classifier among 10,488 tested nodules (0.4%), consistent with the low prevalence of MTC. Of these, all but one was histologically or biochemically confirmed as MTC, yielding a positive predictive value (PPV) of 98%. Of the positive cases, only 19 (44%) had been specifically suspected of MTC by cytology, highlighting the limitations of light microscopy to detect this disease. Three surgically confirmed MTC cases that were detected by the MTC Classifier had low basal serum calcitonin values, indicating these would have been missed by traditional calcitonin screening methods. A pooled analysis of three independent validation sets demonstrates high test sensitivity (97.9%), specificity (99.8%), PPV (97.9%), and negative predictive value (99.8%). CONCLUSIONS: A clinical paradigm is proposed, whereby cytologically indeterminate thyroid nodules being tested for common malignancies using gene expression can be simultaneously tested for MTC using the same genomic assay at no added cost.

Extramedullary hematopoiesis involving the thyroid: Fine-needle aspiration features and literature review.

INTRODUCTION: The thyroid has rarely been documented as a site of extramedullary hematopoiesis (EMH). We report the largest series to date, with nine cases of EMH, and compare our findings with previous reports of thyroid EMH. MATERIALS AND METHODS: Thyroid nodule fine-needle aspirations (FNAs) were collected over a 4-year period. Thin layer preparations were examined and correlated with clinical features and ultrasound characteristics. A comprehensive review of the English literature was done, and the results were compared with the current series. RESULTS: During the study period, 172,939 thyroid FNAs were examined. Nine samples (0.005%) contained trilineage bone marrow elements. Nodule calcifications were present in 7 patients. None of the patients had a history of a blood disorder, nor was there any evidence of a thyroid malignancy. Fifteen reports of 18 patients with thyroid EMH were identified in the English literature. Nodule calcifications were reported in 10 patients. Thyroid EMH was associated with primary myelofibrosis in 4 patients, and with chronic anemia in 1; calcifications were absent in 3 patients, and were not specified in the remaining 2. None had evidence of a thyroid malignancy. CONCLUSIONS: Thyroidal EMH is an extremely uncommon finding. Clinical and sonographic features are nonspecific. Thyroid EMH is usually an incidental finding, most likely related to mature osseous metaplasia. An occasional association with hematologic disease has been reported, however. To date, there does not appear to be any association with thyroid malignancy and surgery is typically not indicated. Therefore, recognition at the time of FNA is essential.