Abstract
A number of oncology societies have recommended either templates or synoptic reporting of important tumor features on pathology reports. The College of American Pathologists (1) and the Association of Directors of Anatomic and Surgical Pathologists (2) have published such recommendations for most human carcinomas; in these synoptics, some features are deemed “required,” while others are “optional.” The differences were predicated on literature data indicating which features were clinically useful and those that had not yet been studied sufficiently to be considered necessary to include. For thyroid carcinomas, especially papillary carcinoma and to a lesser extent follicular carcinoma, the features required in pathology reports are similar to those of the common cancers, but some are different and unique to the thyroid. In this issue of Thyroid, Kahn et al. (3) describe a quality assurance analysis of pathology reports of thyroid carcinoma resections, list the features that should be included, and indicate how often important parameters were included in the reports.
What features of thyroid cancer resections should be evaluated and listed on pathology reports for resections of these lesions? Table 1 lists the required features and includes tumor characteristics that are needed for estimating the tumor extent (staging) and possible predictive factors.
Tumor size and histologic type are primarily important. Whether the tumor has extended beyond the thyroid gland (extrathyroidal extension) and whether regional lymph nodes are involved also help to stage the malignancy. Tumor grade is not usually included in thyroid tumor pathology reports, although many synoptic schemes do include a list of known histologic subtypes.
It has been our experience that certain cytomorphologic and histologic factors should be noted even if these compose only a fraction of the tumor. Hence, we have concluded from anecdotal experience (and there is limited literature support for this) that in a well-differentiated tumor, the presence of foci of solid or trabecular growth pattern or numerous mitotic figures and/or necrosis should be mentioned since in recurrences, the percentage of these areas of tumors may increase and indeed may become the predominant morphologic subtype. In addition, in a usual or classical papillary carcinoma, the finding of areas of tall cell cytology should be noted since in nodal metastases or other recurrences, the tall cell histology may predominate. Some studies have indicated that 10% or more of these more aggressive areas should place the tumor in a poorly differentiated category (4).
In our view, extrathyroidal extension of tumor has been a controversial area to assess. To explore this, one must revert to the basic question, “Does the thyroid gland possess a true capsule?” It is well-known that the thyroid gland is incompletely enveloped by a connective tissue layer (i.e., external capsule), which is derived from pretracheal fascia (5,6). This “capsule” is often incomplete in the anterior midline, isthmus, and especially in the posterior aspects of the gland. These external limits of the thyroid gland are further muddied by the presence of skeletal muscle, fibroadipose tissue, and even detached islands of thyroid tissue (commonly encountered in nodular goiter and chronic, lymphocytic thyroiditis) within this so-called thyroid capsule. Therefore, the mere presence of a tumor nodule or island in close proximity to, or one that appears to be embedded within, the normal elements in the thyroid capsule, such as adipose tissue or skeletal muscle, does not always imply extrathyroidal extension. Based on this, one can understand the difficulty and the operator bias in determining microscopically whether or not a carcinoma is still confined within the thyroid or has extended beyond it. It has been suggested that a reaction in the form of inflammation and/or desmoplasia may be confirmatory for tumor invasion within the perithyroidal soft tissue. However, some authors have questioned this criterion because it may not be applicable in patients with severe thyroiditis or those who have undergone preoperative fine-needle aspiration because reactive changes can occur in both situations. Furthermore, reactive changes may not be present in all examples of extrathyroidal invasion.
We believe that in order to comment on extrathyroidal invasion the resected specimen should be well sampled, and a good number of sections should include the tumor and the inked surgical plane. In our practice, we use the discovery of tumor within perithyroidal adipose tissue and/or tumor whose outline is beyond the last normal thyroid follicles to define extrathyroidal invasion. We consider this to be minimal extrathyroidal extension, and although the AJCC staging system would place such a tumor in the stage III group, such cases are rarely associated with aggressive clinical behavior. On the other hand, extension into skeletal muscle or into neighboring organs is obviously predictive of extensive locoregional disease and is a cause for concern.
Some physicians request an assessment of resection margins. Very little data exist that define distance of tumor edge to margins as a determinate of risk for local tumor recurrence. However, if the tumor is transected (i.e., true positive margin), we believe this should be stated in the diagnosis or tumor template. This factor may influence decisions on postoperative therapy.
The presence of lymphatic invasion (in papillary carcinoma), often manifested as isolated psammoma bodies in lymphatic spaces in the gland (occasionally as microfoci of viable tumor), appears to be associated with nodal metastases more often than is found in lesions without these features. Vascular invasion (the sine qua non of follicular thyroid carcinoma diagnosis) can also be identified in papillary carcinomas; in our experience, it is often seen away from the main tumor mass and in the extraglandular component of the tumor. The number of veins involved by tumor thrombi should be estimated, and extensive venous invasion (considered by most observers to be greater than four foci) should be mentioned since such tumors appear to have a greater risk of metastatic disease (1,2).
Even though most current synoptic reporting schemes (5) lump lymphatic and vascular invasion into one category of “lymphovascular invasion,” we strongly believe that lymphatic and vascular invasion should be evaluated separately and listed in the template (7). Our reasoning involves the fact that sole lymphatic involvement by the tumor leads to spread to the regional lymph nodes, but vascular invasion predicts risk for hematogenous distant metastases to lung, bone, brain, and liver.
Many templates include the feature of perineural invasion, but this is a rare event; certainly if it is seen, it should be mentioned (1). The presence of lesions in the nontumoral thyroid should also be mentioned and listed in the report. Chronic, lymphocytic thyroiditis, hyperplasia, and benign follicular nodules are common benign findings (1,2).
The presence of a focus or foci of papillary microcarcinoma should be noted and the size(s) and extent of tumor given. If there appears to be increased numbers of C cells, this should be confirmed by appropriate immunostaining and mentioned in the report.
If lymph nodes are present in the thyroid lobectomy or thyroidectomy, or are received as a separate specimen, the number of involved nodes and the total number of nodes recovered should be included in the report. The total number of nodes found is important since it reflects both the completeness of the node dissection or sampling and the care with which the pathologic examination was carried out. Though optional, it is also important to document the extranodal extension of the tumor and presence of vascular invasion in the extranodal soft tissues. The finding of parathyroid tissue and its location, if known, should be noted as well.
In sum, the pathology report needs to include the size and histotype of the tumor, the presence of less-differentiated or higher grade foci, the presence of lymphatic or vascular invasion, and whether extrathyroidal tumor is present (including node metastases).
Is this all we need? In the 21st century, at the dawn of the era of personalized medicine, does the pathology report need to describe the results of ancillary tests? In this regard it is the standard of care to include results of steroid receptor analysis and her2neu (human epidermal growth factor receptor 2) status and other molecular markers in reports of breast cancer specimens (8). Similarly, it is becoming necessary to report the presence of human papillomavirus (by in situ hybridization, polymerase chain reaction, or p16 immunostaining as a surrogate marker) in oropharyngeal squamous carcinoma, epidermal growth factor receptor, and K-RAS mutation status in lung carcinoma and microsatellite instability in colonic adenocarcinoma (9,11). So what about the thyroid specimens?
At present, many institutions and private laboratories are using adjunct molecular analysis in fine-needle aspiration specimens diagnosed as AUS/FLUS, follicular neoplasm or suspicious for papillary carcinoma, according to the Bethesda classification scheme (12). The molecular assay either only detects the presence of the BRAF V600E mutation or multiple gene mutations and translocations. It has been shown that all fine-needle aspiration samples demonstrating BRAF-V600E mutations are papillary carcinoma. Many agree that it is appropriate to include the results of the molecular analysis in the original cytology report so that all the information for that specimen is collated in one report (13).
The value of BRAF mutational testing on the tissue of a resected papillary carcinoma remains unclear. Although initial reports (14,15) indicated a positive test could predict aggressive clinical behavior; more recent series cast doubt on this assessment in certain subgroups (16). Other molecular analyses for assessing prognosis do not appear as promising; in the future, additional marker mutations or translocations may prove to be helpful. Thus, if one evaluates the pathologic features discussed above and listed in the Table 1, one can likely predict possible adverse clinical outcomes, while expending fewer resources in this era of economic constraints.