Thyroid fine-needle aspiration (FNA) is an excellent test to evaluate the risk of malignancy of a thyroid nodule and is practiced in a wide variety of settings with excellent results. The addition of molecular testing has not only increased our knowledge of the underlying genetics of thyroid carcinoma but has improved the specificity of the test. Nevertheless, the percentage of indeterminate aspirates (Bethesda categories III and IV) within a laboratory can range widely (approximately 5%-30%). There are many possible causes for this variation, including differences in the underlying rate of malignancy within a laboratory and interobserver variation between cytologists. Consequently, quality metrics for thyroid FNA may be of particular value to ensure that clinicians are receiving consistent and reproducible results for this test.

In this issue of Cancer Cytopathology, Gokozan and colleagues1 combine both cytologic and molecular results to define quality metrics for thyroid FNA. In a manner similar to the atypical squamous cells of undetermined significance/squamous intraepithelial lesion ratio and human papillomavirus positivity rates in gynecologic cytology, these authors demonstrate that the Bethesda2 category III/VI ratio, along with neoplasia-associated genetic alterations, may be of use in determining whether they are overcalling or undercalling cases. There are many attractive aspects to this proposal. Although the atypical/malignancy ratio has been previously proposed3 as a way to compare the performance between laboratories that may have a difference in the underlying rate of malignancy, there are limitations to this test,1 and the measure has not been widely adopted. The addition of molecular results may allow cytologists to more accurately characterize the performance of an individual within their laboratory.

Nevertheless, this article raises several issues that still need more study. Although the current study proposes optimal standards for indeterminate aspirates that are in the middle of reported ranges, what represents the best performance is not yet entirely clear. One could make an argument that the best performance is actually at the highest rate of indeterminate cases (30%). If only 1 test is available for a process, one usually seeks a balance between sensitivity and specificity. In contrast, when one has 2 tests available for the same process, one should screen with the most sensitive test and then confirm with the more specific test. In this scenario, one would want to increase the indeterminate rate as much as possible to improve the sensitivity of the test and then use molecular testing to improve the specificity. Unfortunately, molecular testing currently is not entirely specific, and cytologists know this and instead try to seek a balance between sensitivity and specificity. In contrast, some clinicians insist that they want any case with any type of atypia sent for molecular testing, making it difficult to strike a balance between sensitivity and specificity. Although these 2 approaches are currently in conflict with each other, the specificity of molecular testing has consistently improved and likely will continue to improve in the future. If this happens, then increasing the rate of indeterminate cases may be best for our patients.

Second, Gokozan et al1 have chosen to focus on Bethesda category III rather than grouping categories III and IV together. Indeed, they state that some cytologists are undercalling cases when they are categorizing more cases as category III rather than category IV. Although this is a controversial topic, my bias is that any quality metric that focuses on the difference in rates of Bethesda category III and IV is not focusing on what is important. Before molecular testing, the small but possibly significant difference in risk of malignancy might have been important; however, now that both category III and IV cases are managed exactly the same way, the distinction between the 2 categories is something cytologists care a lot more about than clinicians and patients.

But these are minor points compared with the larger point these authors are convincingly making. Molecular testing in thyroid cytology is only going to become more and more important. Not only will the utility of specific molecular alterations be more clearly defined, but the settings in which molecular results may be able to provide value may also increase. For example, as noted in the second edition of Bethesda,2 some studies suggest that the most commonly used criteria for adequacy in thyroid FNA (6 groups of 10 cells) are too strict, but most cytologists are unwilling to use less stringent adequacy criteria. Could molecular testing be used to reassure cytologists and thus reduce the number of unnecessary repeat aspirations? This article provides an excellent starting point as our field defines quality metrics that incorporate both cytologic and molecular results to ensure the best care for our patients.

Funding Support

No specific funding was disclosed.

Conflict of Interest Disclosures

The author made no disclosures.