Hürthle Cell Hurdles: Why Do Tumors Derived From This Cell Type Refuse to Reveal Their Secrets and Weaknesses?

We are at an extraordinary time in the evolution of diagnostic medicine, integrating classical histologic diagnostics with theranostics, targeted molecular-based medical therapeutics focused on treatment of the specific molecular biology of individual disease subtypes, instead of gestalt therapies for entity type. Rather than diagnosing papillary thyroid carcinomas as variants with solely morphologic variability, we are moving toward understanding these thyrocyte-based lesions as morphologic and molecular entities that harbor both unique behavioral patterns and molecular targets for therapy. As an example, we have recently reported data on kinase fusion-related thyroid carcinomas,¹ including NTRK-related thyroid carcinomas,² that bear morphologic overlap and durable clinical responses to therapy.

Hürthle cell (oncocytic) lesions are ill-defined. They can be metaplastic events or neoplastic events. Neoplastic entities include Hürthle cell adenomas and carcinomas of the thyroid. Most Hürthle cell neoplasms of the thyroid are of limited biological potential, like oncocytomas of the kidney or salivary glands. However, morphologically similar Hürthle cell lesions of the thyroid can become highly biologically aggressive and arguably are the most dangerous neoplasms of the thyroid save for undifferentiated (anaplastic) thyroid carcinoma.

In the thyroid, Hürthle cell lesions, especially when large, are sources of great concern for pathologists, and their mere presence, particularly when thickly encapsulated, herald malignancy until otherwise excluded. Papillary thyroid carcinomas have distinct morphology and molecular genetics and have somewhat predictable behavior: most aggressive tumors are locally problematic, with extrathyroidal extension and/or local lymph node metastasis. Follicular thyroid carcinomas are more challenging to diagnose, with diagnosis based on the invasion of a tumor capsule and/or intracapsular/extracapsular vascular invasion. When widely invasive, these tumors are easy to diagnose; however, when minimally invasive, they may rely on subjective determinants for the appreciation of capsular and/or vascular invasion. Hürthle cell lesions share similarities of both tumor behaviors, a diagnosis requiring invasion, and behavior involving both vascular spread and lymphatic spread through the lymph nodes. Hürthle cell carcinomas (HCC) are known to be very aggressive and, when truly malignant, they have widespread distant metastases and poor outcomes. It is the difficulty in defining truly in truly malignant that is a problem for pathologists that transfers to cytologists. Most Hürthle cell lesions are indolent except for those that are not, and this can be very difficult to predict. Deciphering this sphinx’s riddle is one of the true prizes within the grasp of cytology.

Recent data from our group³ and colleagues at Memorial Sloan Kettering Cancer Center⁴ interrogated primary Hürthle cell carcinomas^3,4 and Hürthle cell carcinomas with matched metastases¹ for molecular changes. Our group noted an array of pathogenic mutations in HCC, including mutations in DAXX, TP53, NRAS, DKN1A, ARHGAP35, and TERT promoter. Like what is being discovered for oncocytic, mitochondria-rich lesions, we also found mutations in mitochondrial DNA-encoding complex I of the electron transport chain couple of widespread chromosomal loss to a near-haploid state. The Memorial Sloan Kettering Cancer Center group found similar anomalies in the mitochondrial genome TERT promoter and genes affecting the RTK/RAS/AKT/mTOR pathway and DNA damage repair. These findings are not typically explored with cytologic assay, and many of them are not investigated after surgical excision, especially because of their costly nature and requirement for fresh tissue. Furthermore, no large studies comparing HCC with Hürthle cell adenoma or Hürthle cell metaplastic changes associated with thyroiditis have been performed for comparative genetics to establish which markers are predictive for Hürthle cell lesion-related biologic behavior.

In this issue of Cancer Cytopathology,⁵ Gilani and colleagues run into a similar molecular wall that is best summarized in their condensed abstract. “Cases with a cytologic diagnosis of Hürthle cell neoplasm exhibit various molecular alterations, but no consistent association is observed in histologically benign or malignant cases.” These authors were able to study 279 patients diagnosed with Hürthle cell nodules, of which 85 had molecular testing performed. Molecular anomalies were detected in 51 nodules, and 34 nodules had no anomaly detected. Of the nodules that had surgical follow up (n = 137), 28 were classified as malignant, and 109 were classified as benign or nonneoplastic. Molecular changes were as likely to be found in histologically benign/nonneoplastic lesions as in neoplasms. This is a problem, and it is a problem that has not been resolved biologically to gain diagnostic insight for significant clinical use in cytopathology.

The current study is particularly important, frustratingly provocative, in fact, because we increasingly seek more minimally invasive procedures to gain as much information as possible. We are looking for a combination of cytopathologic features plus molecular features (a whole commercial industry) to provide insight into biological behaviors of thyroid tumors or into the responsiveness of tumors to targeted therapeutics.⁶ What this study points out to us, especially for Hürthle cell tumors, which present a particularly bipolar biologic treachery, is that we are still not able to discern, by cytologic criteria or with molecular adjuvant testing, which lesions will be most problematic and how we can best predict their behavior for purposes of preoperative planning, preoperative treatment, or postoperative management.