¹⁸F-FDG–Avid Thyroid Incidentalomas: The Importance of Contextual Interpretation

DISCUSSION

Our data demonstrated high mortality—median overall survival of 20 mo—in this large cohort of patients with ¹⁸F-FDG–avid TI undergoing ¹⁸F-FDG PET/CT at a comprehensive cancer center. However, this high mortality appears to be due almost exclusively to the poor prognosis associated with the underlying malignancy rather than to adverse outcomes related to incidentally detected ¹⁸F-FDG–avid thyroid cancer. In contrast, the ¹⁸F-FDG–avid TI had no discernible clinical impact—excluding that associated with medical investigation and management of this asymptomatic incidental finding—in 361 (99.7%) of 362 patients after a median of 24 mo of follow-up; it is uncertain if earlier intervention would have prevented the single death from incidental medullary thyroid carcinoma. This observation highlights the importance of considering the prognosis of the underlying malignancy when deciding whether further investigation and management are appropriate, as the potential benefits of intervention may not be realized because of the short life expectancy.

Efforts to identify the cohort of patients who will potentially benefit from further evaluation and treatment have not been successful. Most previous studies have focused on guiding the decision to biopsy by using the intensity of focal ¹⁸F-FDG uptake to identify the patients at highest risk of thyroid cancer. Although some studies have indicated an SUV_max threshold (7–9), many have found no such association (10–12), and a large metaanalysis by Bertagna et al. (3) confirmed that there is significant overlap of ¹⁸F-FDG avidity between benign and malignant lesions, without an SUV_max threshold to guide the decision to biopsy. This finding is consistent with a growing body of literature indicating that the ¹⁸F-FDG avidity of TI is often driven by mechanisms other than poorly differentiated malignancy. For example, it is increasingly recognized that oncocytic/Hürthle cell lesions (irrespective of benign or malignant etiology) demonstrate intense ¹⁸F-FDG avidity (13) due to an intrinsic mitochondrial defect driving inefficient glycolytic metabolism (14,15). Other causes of very intensely ¹⁸F-FDG–avid benign TI include degenerate nodules (16), follicular adenomas, and adenomatous hyperplasia (17). The limitation of this approach is evident from the low ¹⁸F-FDG uptake (SUV_max, 4.4) in the only clinically significant thyroid cancer in our study (Fig. 2) compared with the intensely avid benign follicular adenoma in the opposite lobe (SUV_max, 16).

In contrast, our study findings enable identification of those cases with a poor prognosis from underlying malignancy, in which further evaluation is unlikely to be of clinical benefit (Fig. 4), and emphasize the need to consider the broader clinical context. The presence of ¹⁸F-FDG avidity within the underlying malignancy conferred a significant crude hazard ratio of 8.5 for mortality (P < 0.001) beyond the overall poor prognosis seen in our entire patient cohort. This finding is consistent with the large body of literature confirming the superior prognostic value of ¹⁸F-FDG avidity across a spectrum of malignancies, including non–small cell lung carcinoma (18), gastroenteropancreatic neuroendocrine tumors (19), medullary thyroid carcinoma (20), and established metastatic differentiated thyroid carcinoma (21). This finding suggests that clinicians should take a cautious approach to investigating ¹⁸F-FDG–avid TI in the setting of ¹⁸F-FDG avidity in a known or suspected malignancy, especially if metastatic, because, unsurprisingly, an advanced tumor stage according to the criteria of the American Joint Committee on Cancer was also a statistically significant prognostic indicator, with a crude hazard ratio of 3.0 (95% CI, 2.3–3.9).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Intensely ¹⁸F-FDG–avid TI (SUV_max, 16) was identified on staging ¹⁸F-FDG PET/CT scan of 28-y-old woman with stage IIIB poorly differentiated cervical carcinoma and persistent ¹⁸F-FDG–avid pelvic nodal disease. Ultrasound-guided cytologic evaluation demonstrated 5-mm papillary thyroid carcinoma with no adverse features, which was treated with total thyroidectomy. (A and B) Baseline ¹⁸F-FDG–avid thyroid nodule (A) and subsequent cervical carcinoma metastases to supraclavicular nodes (B). (C and D) Baseline ¹⁸F-FDG–avid pelvic lymphadenopathy (C) with subsequent disease progression (D). Patient died 10 mo after baseline PET/CT scan from progression of metastatic poorly differentiated cervical carcinoma.

Although this high prognostic value of ¹⁸F-FDG avidity in oncology has contributed to a level of anxiety associated with ¹⁸F-FDG–avid TI, the lack of an SUV threshold for these lesions is supported by increasing evidence that ¹⁸F-FDG avidity associated with incidentally detected thyroid cancer does not add prognostic value beyond TNM staging (22). A potential explanation for this paradox is that a proportion of incidentally detected apparently ¹⁸F-FDG–avid thyroid malignancies actually represent incidental findings in pathology specimens, further supporting a more conservative approach to managing ¹⁸F-FDG–avid TI. A careful histopathologic and imaging correlation revealed that 20% (3/15) of presumed ¹⁸F-FDG–avid incidental thyroid carcinomas in fact represented incidental malignancies identified by the pathologist and did not correlate with the avid lesion at all (16). This finding is plausible given that autopsy studies have identified incidental thyroid malignancy in up to 36% of specimens (23). In another study, exclusion of such microcarcinomas based on their not being clearly associated with the imaged abnormality because of partial-volume artifacts reduced the incidence of true ¹⁸F-FDG–avid incidental thyroid malignancy from 32% to 14% (24).

Our finding that the clinical decision not to investigate an ¹⁸F-FDG–avid TI was associated with a worse overall prognosis confirms that the underlying malignancy likely did influence the decision to investigate incidentalomas in this cohort. The exceptionally low incidence of persistent clinically significant thyroid cancer in this series suggests that this decision was generally made appropriately. The proportion of investigated ¹⁸F-FDG–avid TI in a large metaanalysis (3)—mean, 35% (range, 11%–100%)—is similar to our cohort (36%) and suggests that this principle is also applied variably in other centers.

Our study findings support a more cautious contextual approach to the investigation and management of ¹⁸F-FDG–avid TI. In the first instance, we advocate an individualized approach to further investigation, recognizing that further evaluation of ¹⁸F-FDG–avid incidentalomas may be associated with patient anxiety, additional costs, and potential morbidity associated with aggressive investigation (2). Notably, our findings suggest that the management of incidental thyroid cancer in this setting is likely to have negligible impact on the patient’s outcome despite additional risks associated with potentially unnecessary surgery and delay of definitive treatment of the underlying malignancy. We recommend limiting further investigation to patients with an otherwise favorable prognosis for whom diagnosis of incidental thyroid carcinoma may be clinically relevant. This decision involves careful consideration of factors associated with the patient (age, nononcologic morbidity), the underlying malignancy (persistent ¹⁸F-FDG avidity, stage), and local features such as adjacent ¹⁸F-FDG–avid neck nodes on the PET scan. In many patients (particularly those with a condition, such as lymphoma, for which ¹⁸F-FDG PET is routinely used in restaging), investigation of the ¹⁸F-FDG–avid TI can be deferred until after successful treatment of the underlying malignancy. As previously highlighted by Ginsberg, the reporting nuclear medicine specialist may practically convey this message by stating that “incidental FDG-avid thyroid nodule is of doubtful clinical relevance in the context of FDG-avid metastatic malignancy” (25).

If a diagnosis of papillary thyroid cancer is confirmed, we strongly recommend consideration of active surveillance (Fig. 5) (26), consistent with the recommendations of the American Thyroid Association guidelines for patients with a relatively short life expectancy in whom the benefits of intervention may not be realized (4). A period of observation provides a clearer understanding of the prognosis associated with the underlying malignancy in addition to the biology of the incidental thyroid carcinoma. Moreover, some of these patients will have serial ¹⁸F-FDG PET studies for the underlying malignancy, providing a further opportunity for surveillance of the incidental thyroid lesion without performing additional investigations. It is important that all clinicians, from the PET reporting nuclear medicine specialist to the referring oncologist, endocrinologist, and thyroid surgeon, be informed of this approach from the outset to provide a consistent message to the patient.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Mildly ¹⁸F-FDG–avid TI (SUV_max, 3.1) was initially identified on staging ¹⁸F-FDG PET/CT of 59-y-old man with metastatic melanoma in 2007. FNA (performed to exclude metastatic melanoma) confirmed incidental papillary thyroid malignancy, which remains stable under surveillance imaging and of limited clinical significance in setting of metastatic melanoma under active treatment.

The retrospective nature of the study is a potential limitation, and 28% of identified cases of ¹⁸F-FDG–avid TI were excluded from analysis. However, most of these are explained by the fact that the patient was attending our referral center only for PET imaging, not for clinical follow-up, and only a modest 8.4% of the total identified cases received less than 12 mo of clinical follow-up. Furthermore, a prospective study would be challenging to implement for a relatively rare incidental pathology in this patient cohort, and the overall proportion of investigated patients was similar to that seen in other studies. Importantly, the follow-up for meaningful clinical outcomes in all patients (irrespective of the extent of thyroid investigation) is an advantage of our study design that eliminates the potential for verification bias seen in most prior studies of ¹⁸F-FDG–avid incidentalomas. Referral bias to a comprehensive cancer center is a possible limitation of this study. However, we believe the results remain generalizable because malignancy remains the most common PET indication and most PET scans were performed for standard approved indications.

Our study identified only cases of ¹⁸F-FDG–avid thyroid cancer, and incidental thyroid cancer that was not ¹⁸F-FDG–avid was also likely to have been present in this patient population. Furthermore, metastatic thyroid cancer exists on a spectrum from well-differentiated (typically iodine-avid) to poorly differentiated (typically ¹⁸F-FDG–avid) disease described as the flip-flop phenomenon (27). However, both types of disease may be present in the same patient because of metastatic disease heterogeneity, and in one study up to 33% of metastatic lesions demonstrated both iodine and ¹⁸F-FDG uptake (28). This has important management implications in determining suitability for radioactive iodine therapy, and consequently, a combination of both iodine and ¹⁸F-FDG PET/CT imaging has been advocated in the treatment of advanced thyroid cancer (29).

A strength of this study is the focus on the presence of the clinically meaningful thyroid cancer outcomes of structural incomplete response or death in this patient population with a limited life span. Furthermore, the large population and long duration of follow-up—362 patients for a median of 2 y—adds to the validity of the study findings.