Abstract
In their article in this issue of Thyroid, Jasim et al. (1) call attention to the rising incidence of thyroid nodules, most of which are benign. As they observe, many nodules are discovered incidentally on imaging studies, particularly high-resolution ultrasound, which detects nodules in the majority of adults. While some studies have suggested a true increase in the incidence of differentiated thyroid cancer, the majority of new cases represent diagnosis of subclinical disease, leading to treatment of malignant nodules that are not likely to be clinically relevant during the patient's lifetime (2). This trend has led to concerns about the cost and morbidity associated with workup of nodules through biopsy, genomic testing, and, in indeterminate cases, diagnostic thyroid lobectomy.
Multiple professional organizations have attempted to counter this trend by developing risk stratification systems that assign nodules to malignancy risk categories, usually expressed as a range of cancer probabilities. In particular, these systems employ size and common sonographic descriptors, such as solid composition, microcalcifications, and taller-than-wide shape to identify nodules that warrant biopsy. Importantly, all these ultrasound features represent intrinsic characteristics of a given nodule; that is, they apply regardless of the nodule's position within the gland. This also holds true for nonconventional techniques such as elastography and contrast-enhanced sonography.
Other, nonintrinsic features play a role in determining whether a nodule requires sampling, whether because of their association with cancer or their adverse effect on prognosis if a nodule turns out to be malignant. For example, gross extrathyroidal invasion into the musculature and other soft tissues is widely recognized as an indicator of malignancy and increases the likelihood of regional and/or distant metastasis. Lesser degrees of potential extrathyroidal extension, suspected because of capsular bulging or discontinuity, are also suspicious features on ultrasound (3). Indeed, several risk stratification systems advise practitioners to take these findings into account when deciding on management, but they are not formally incorporated into the risk assignment process, whether it is done by pattern recognition, assignment of points, or an intermediate approach.
The significance of a nodule's position (upper, mid, lower lobe, or isthmus) has received relatively little attention in the literature, but a handful of single-institution studies have uncovered a correlation with metastatic disease. In a retrospective study of 1066 patients published in 2012, Zhang et al. demonstrated a lower risk of central nodal involvement but a higher probability of lateral lymph node metastases for tumors in the upper third of the thyroid (4). Subsequently, Lai et al. showed that solitary solid intrathyroidal papillary thyroid cancers had a higher risk of lateral nodal metastases if they were not located adjacent to the trachea, while Campenni found an association between location of thyroid cancer in the isthmus and locoregional and/or distant metastases (5,6). More recently, Ramundo et al. established a higher risk of malignancy for tumors located in the mid lobe, but this study only included 227 nodules (7).
Jasim et al.'s study expands on previous investigators' work by analyzing data for 3241 nodules from six referral centers to evaluate whether location and malignancy are related. Not surprisingly, they found that overall, nodules occurred less frequently in the isthmus, likely reflecting the lesser volume of parenchyma compared with the right and left lobes. Isthmus nodules were also smaller and significantly more likely to be cancerous than nodules elsewhere in the gland, particularly the lower lobes, which had the lowest malignancy rate. In addition to their greater association with lateral nodal metastases, nodules in the isthmus are more likely to exhibit adverse prognostic findings such as extrathyroidal extension (8). While it is logical that nodules in this location are more likely to breach the capsule because of proximity, an explanation for their relatively increased cancer risk remains to be elucidated.
In their conclusion, Jasim et al. advocate meticulous examination of the isthmus during thyroid sonography because of the lesser size and higher cancer risk of nodules in this location. Moreover, in my experience, nodules may be missed if they extend exophytically from the isthmus, particularly if they arise from the pyramidal lobe. The authors further suggest raising the suspicion level for isthmus nodules in revisions to existing risk stratification systems, which begs the question as to precisely how to apply this recommendation. As well, all published guidelines advise practitioners not to rely solely on ultrasound-derived risk levels to determine management, be it fine needle aspiration biopsy, imaging follow-up, or no further action. Other clinical considerations, such as the patient's age, history of neck irradiation, family history of thyroid cancer, and comorbidities, appropriately influence the course of action in shared decision-making with patients. However, these factors are typically not considered in risk stratification algorithms, leaving physicians to apply them based on general guidance and professional experience.
The challenge is compounded by the number of systems in active use worldwide. Indeed, more than one may be employed by different departments in a single facility, leading to inconsistent management recommendations and confusion for patients and caregivers. To address this problem, the International Thyroid Nodule Ultrasound Working Group (ITNUWG), a multidisciplinary alliance of physicians with expertise in thyroid nodule sonography, is attempting to devise a unified international guideline that incorporates clinical history and ultrasound while taking steps to reduce interobserver variability in feature assignment (9). Working group participants understand that regional and geographic differences in thyroid cancer incidence, tumor types, biological behavior, and treatment options will also have to be considered, as will other factors yet to be identified. To a large extent, the success of this endeavor will depend on continued research into novel ultrasound-detectable findings, such as nodule location and perfusion, as well as refinements of the intrinsic features in use today. Studies such as the research by Jasim et al. will help inform the process as it moves forward.