Incidence and Significance of Incidental Focal Thyroid Uptake on ¹⁸F-FDG PET Study in a Large Patient Cohort: Retrospective Single-Centre Experience in the United Kingdom

Abstract

Objective

To assess the prevalence and pathological nature of incidental focal thyroid uptake on ¹⁸F-FDG (2-[¹⁸F]-fluoro-2-deoxy-D-glucose) PET (positron emission tomography) and examine the role of the maximum standardised uptake value (SUV_max) to differentiate benign from malignant thyroid pathology.

Material and Methods

¹⁸F-FDG PET reports were retrospectively reviewed. Incidental focal tracer uptake in the thyroid was noted in 147 patients (0.5%). Patients with known primary thyroid malignancy were excluded. The final diagnosis was made following ultrasonography of the neck, fine-needle aspiration cytology (FNAC) or histopathology of the surgically resected specimen where surgery was indicated. A Mann-Whitney U test was used to compare the SUV_max of benign and malignant thyroid pathology. Receiver operating characteristic (ROC) analysis was performed to identify an SUV_max cutoff in differentiating benign from malignant pathology.

Results

A final diagnosis was achieved in 47/147 (32%) of the patients. The diagnoses included benign lesions in 36 patients and malignancy in 9 patients. In 2 patients, FNAC demonstrated indeterminate follicular lesions; however, surgical excision was not performed. There was a highly significant difference in the mean SUV_max of malignant focal thyroid uptake (15.7 ± 5.9) compared to that of benign lesions (7.1 ± 6.8) with a p value of 0.000123. An SUV_max of 9.1 achieved a sensitivity of 81.6%, specificity of 100% and area under the curve of 0.915 in the ROC analysis differentiating benign from malignant disease.

Conclusion

The malignancy potential of incidental focal thyroid uptake remains high and warrants prompt and appropriate follow-up by the clinician. The SUV_max may aid in further characterisation of the lesion and its management.

Introduction

The prevalence of thyroid nodules in the general population is high and is reported to be between 8 and 65% [1]. Due to the significant advances in imaging technology and the increased use of neck imaging, detection of unsuspected thyroid nodules, known as incidentalomas, is on the rise.

¹⁸F-FDG (2-[¹⁸F]-fluoro-2-deoxy-D-glucose) PET/CT (positron emission tomography/computed tomography) has been increasingly used for assessment of various malignancies and plays an integral role in cancer management. ¹⁸F-FDG is a glucose analogue and the mechanism of ¹⁸F-FDG uptake and detection of tumours is based on the higher glycolytic metabolism and the higher expression of membrane glucose transporter (GLUT) proteins in the malignant tissue [2]. Incidental diffuse and focal thyroid uptake is often seen on ¹⁸F-FDG PET/CT study. Diffuse uptake in the thyroid has been reported in approximately 0.6-3.3% of the ¹⁸F-FDG PET studies and is often due to a benign aetiology [3]. The prevalence of focal uptake within the thyroid (incidentaloma) on ¹⁸F-FDG PET has been noted to range from 0.2 to 10.1% in various studies. This is clinically more significant due to its high risk of malignancy in these lesions and the reported risk of malignancy is varied (8-64%) [4]. Malignancy identified within the thyroid incidentaloma on ¹⁸F-FDG PET has been noted to be of a higher grade/aggressive subtype [5], requiring prompt evaluation by the clinician. This may create a management dilemma for the referring clinicians [6].

The maximum standardised uptake value (SUV_max) assessed by ¹⁸F-FDG PET is a semi-quantitative measure of glucose metabolism, which is useful in the estimation of tumour grade or aggressiveness and as a marker in assessment of response to treatment. It is defined as the maximum uptake in the lesion scaled by the administered activity and patient weight or height [7]. Some studies claim a beneficial role of the SUV_max in differentiating benign from malignant thyroid pathology, but this has not been replicated in other studies and therefore remains controversial [8,9].

The aim of this study was to assess the pathological nature of the focal thyroid incidentalomas detected on ¹⁸F-FDG PET and the role of the SUV_max in differentiation of benign from malignant thyroid pathology in these patients.

Materials and Methods

This was a retrospective study reviewing ¹⁸F-FDG PET or PET/CT scan reports of 29,300 studies performed in the nuclear medicine department at our institution between January 1999 and December 2013 for various oncological and non-oncological indications. Institutional review board approval was obtained for the study.

The search criteria ‘uptake in the thyroid’ was applied to these scan reports, which provided 147 results as having incidental focal tracer uptake in the thyroid. Patients with an established diagnosis of a malignant primary thyroid neoplasm were excluded from the analysis. Data including age, sex, primary malignancy site, indication for the PET study and the SUV_max of the focal thyroid uptake were recorded.

PET/CT imaging was performed 60 min after injection of ¹⁸F-FDG (5 MBq/kg of body weight). Standard patient preparation prior to the study included a fasting period of at least 4-6 h and a serum glucose level <7 mmol/l (120 mg/dl) before ¹⁸F-FDG administration.

The PET scan report in patients with a focal uptake in the thyroid gland recommended further evaluation of the uptake with ultrasonography (USG) with and without fine-needle aspiration (FNA) and where appropriate referral to the head and neck, endocrinology or endocrine surgery teams for further management. The final diagnosis for the focal thyroid incidentalomas was made by USG of the neck, FNA cytology (FNAC) or histopathology of the surgically resected specimen, where available. The prevalence of thyroid incidentalomas on ¹⁸F-FDG PET or PET/CT and the rate of malignancy in focal uptake were assessed. The SUV_max of focal uptake was noted in patients with a focal thyroid incidentaloma and available final diagnosis.

Statistical Analysis

Statistical analysis was performed using commercially available software package SPSS version 16.0 (SPSS Inc., Chicago, Ill., USA). A Mann-Whitney U test was used to compare the SUV_max of benign and malignant thyroid incidentalomas. Statistical significance was defined as p < 0.05. Receiver operating characteristic (ROC) analysis was performed to identify an SUV_max cutoff in differentiating benign from malignant thyroid incidentalomas.

Results

Incidental focal uptake of tracer in the thyroid was observed in 147 patients. The final diagnosis was achieved in 47/147 (32%) by FNAC in 31 patients, histopathology of the surgical resection specimen in 10 and following neck USG in 6 patients (table 1). In the rest of the patients, final cytological or histopathological diagnosis was not available due to several factors, such as further patient management in other hospitals, advanced primary malignancy with widespread metastatic disease, poor prognosis, low clinical index of suspicion or death. The final diagnosis showed benign lesions in 36 patients, papillary thyroid cancer (PTC) in 5 patients (10.6%), secondary metastases in 3 patients and recurrence of lymphoma in 1 patient (table 1). In 2 patients, FNAC demonstrated indeterminate follicular lesions, but surgical excision was not performed. Thus, 9 out of 47 cases (19.1%) of thyroid focal uptake showed malignant involvement. In 5 patients with PTC, completion thyroidectomy histopathology results were available for 3 patients, which showed pT3N1 disease in 2 patients and pT1aN0 disease in 1 patient.

Table 1.

Patient characteristics with focal uptake in the thyroid where a final diagnosis was available

No.	Age, years	Sex	Primary	SUVmax	Final diagnosis	Modality for final diagnosis
1	67	F	Breast	7	Benign nodule	FNAC
2	68	F	Breast	3.4	Benign nodule	USG
3	52	F	Colorectal	1.7	Benign nodule	FNAC
4	75	M	Colorectal	3.8	Benign nodule	FNAC
5	72	F	CUPS	14.5	Benign nodule	FNAC
6	68	F	CUPS	2.4	Benign nodule	FNAC
7	46	F	Lymphoma (HL)	1.6	Benign nodule	FNAC
8	59	F	Lung	7.5	Benign nodule	FNAC
9	64	F	Lung	5.8	Benign nodule	FNAC
10	84	M	Lung	2.9	Benign nodule	USG
11	58	F	Lung	5.6	Benign nodule	FNAC
12	65	M	Lung	7.8	Benign nodule	FNAC
13	73	F	Lung	5.9	Benign nodule	Histopathology
14	68	F	Lung	5.4	Benign nodule	Histopathology
15	76	F	Melanoma	6.7	Benign nodule	USG
16	91	F	Melanoma	3.9	Benign nodule	USG
17	46	F	Lymphoma (NHL)	4.9	Benign nodule	FNAC
18	72	F	Lymphoma (NHL)	2.5	Benign nodule	FNAC
19	43	F	Lymphoma (NHL)	3.4	Benign nodule	FNAC
20	84	M	Lymphoma (NHL)	39.0	Benign nodule	FNAC
21	43	F	Lymphoma (NHL)	4.5	Benign nodule	FNAC
22	79	M	Lymphoma (NHL)	4.5	Benign nodule	FNAC
23	35	F	Lymphoma (NHL)	5.0	Benign nodule	FNAC
24	82	F	Lymphoma (NHL)	9.9	Benign nodule	USG
25	53	F	Lymphoma (NHL)	2.1	Benign nodule	Histopathology
26	51	F	Lymphoma (NHL)	10.3	Benign nodule	FNAC
27	71	F	Lymphoma (NHL)	2.5	Benign nodule	FNAC
28	66	M	Lymphoma (NHL)	5	Benign nodule	Histopathology
29	61	M	Oesophagus	9.1	Benign nodule	FNAC
30	48	F	Oesophagus	20.5	Benign nodule	FNAC
31	81	F	SqCC (anterior abdominal wall)	15	Benign nodule	FNAC
32	58	M	SqCC HN	15	Benign nodule	FNAC
33	77	F	Small fibre neuropathy	5.5	Benign nodule	FNAC
34	75	F	Stomach	4.5	Benign nodule	USG
35	47	M	Thymus	4.5	Benign nodule	FNAC
36	81	M	Colorectal	15	Metastases	Histopathology
37	65	M	SqCC HN	9.2	Metastases	FNAC
38	65	F	Lung	10.7	Metastases	FNAC
39	84	M	Lymphoma (NHL), lung	11.5	NHL	FNAC
40	66	F	Lung	15.2	PTC	Histopathology
41	58	M	SPN characterisation	17.4	PTC	Histopathology
42	46	M	Lymphoma – Sézary syndome	15.2	PTC	Histopathology
43	48	F	Lymphoma (NHL)	29.4	PTC	Histopathology
44	49	F	SqCC HN	17.8	PTC	Histopathology
45	65	F	Lung	3.8	THY3	FNAC
46	49	M	Thymoma	8.3	THY3	FNAC
47	37	F	Melanoma	2.1	Thyroiditis	FNAC

CUPS = Carcinoma of unknown primary site; NHL = non-Hodgkins lymphoma; HL = Hodgkins lymphoma; SqCC = squamous cell carcinoma; HN = head and neck; SPN = solitary pulmonary nodule; THY3 = follicular lesion/follicular neoplasm.

The mean SUV_max of malignant focal thyroid uptake was 15.7 ± 5.9 and that of benign lesions was 7.1 ± 6.8. There was a highly significant difference between the SUV_max in two groups (p = 0.000123; fig. 1). All patients with malignant thyroid pathology had a high SUV_max of the focal uptake. Although most of the benign lesions showed low-grade tracer uptake, a high SUV_max up to 39.0 was observed.

Fig. 1.

Box plots of the SUV_max of benign and malignant focal incidentalomas.

ROC analysis was performed to identify the cutoff SUV_max for differentiating benign from malignant thyroid incidentalomas (fig. 2). The cutoff SUV_max identified was 9.1 (sensitivity: 81.6%, specificity: 100%, area under the curve: 0.915). Further, serial ¹⁸F-FDG PET studies were performed in 9 out of 47 patients, which showed a stable SUV_max over 2-12 months in 5 patients with benign pathology. Two patients had new focal uptake in the thyroid when compared to the previous study, which was confirmed as lymphoma recurrence in one and a benign lesion in the other patient. Other two benign incidentalomas showed a variable trend in the SUV_max. Figures 3, 4, 5 demonstrate focal thyroid uptake and its significance in a few cases from the study.

Fig. 2.

ROC analysis to identify the cutoff SUV_max to differentiate benign from malignant incidentalomas.

Fig. 3.

A 91-year-old-female with previous history of melanoma on ¹⁸F-FDG PET/CT study [axial CT (a), axial PET (b) and fused PET/CT (c)] showed incidental focal uptake in the right thyroid lobe with an SUV_max of 3.9. On USG evaluation the features of right thyroid lobe nodule were suggestive of benign pathology.

Fig. 4.

A 35-year-old-female with high-grade non-Hodgkin's lymphoma on ¹⁸F-FDG PET/CT study [axial PET, CT and fused PET/CT (a-c); coronal PET, CT and fused PET/CT (d–f); sagittal PET, CT and fused PET/CT(g-i)] showed incidental focal uptake in a hypodense nodule in the left thyroid lobe (red arrows; colours refer to the online version only) with an SUV_max of 5.0. In addition, intense uptake was noted in the right axillary lymph nodes (purple arrows; colours refer to the online version only). On histopathology, the left thyroid nodule was confirmed as a benign follicular adenoma.

Fig. 5.

A 58-year-old-male underwent ¹⁸F-FDG PET/CT study for assessment of a solitary pulmonary nodule (SPN). ¹⁸F-FDG PET/CT showed no uptake in the SPN, but there was intense focal uptake of tracer in a right thyroid lobe nodule [arrow in maximum intensity projection (a), axial CT (b), axial PET (c) and axial fused PET/CT images (d)] with an SUV_max of 17.4. On histopathology it was confirmed as a PTC.

Discussion

¹⁸F-FDG is a glucose analogue and its mechanism of uptake is based on the higher glycolytic metabolism and the high expression of GLUT proteins in the malignant tissue [2]. The significance of incidental focal thyroid uptake on ¹⁸F-FDG PET study was first described in 2001 [10]. Many research studies are currently available on this topic, with different results on focal uptake of tracer on ¹⁸F-FDG PET [3,4,5,6,8,9,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38]; however, a large UK series has yet to be reported. We report one of the largest studies on the evaluation of incidental focal ¹⁸F-FDG uptake in the thyroid. The largest study by Bertagna et al. [4] included 49,519 patients, and focal thyroid uptake was identified in 1.5% of the patients. The authors reported that 34.1% of incidentalomas were malignant. Two studies with a large patient population by King et al. [26] and Kwak et al. [30] showed very low prevalence of focal incidentalomas in 0.2 and 0.6% of the patients, respectively, similar to our study. However, the prevalence of thyroid incidentalomas on ¹⁸F-FDG PET varies from 0.2 to 10.1% in different studies [4]. This could well be related to variation in geographic area, number of patients studied and patient characteristics.

The incidence of malignant neoplasm in our cohort was19.1% where a final diagnosis was available. Systematic review of previous studies on this topic by Soelberg et al. [39] showed malignancy in 34.8% of patients with focal uptake in the thyroid. However, our study showed a slightly lower incidence of malignant involvement of the thyroid in this population, similar to the King et al. [26] publication.

Previous studies have indicated that PTC and the follicular variant of PTC are the most prevalent thyroid cancer types, accounting for 81.1%, whereas lymphoma and secondary metastatic disease have been seen in only 4.1% of the patients [39]. Our study data is in concordance with previous findings showing PTC as the most frequent histopathological type of primary thyroid cancer in these patients. Further, we did not find any other histopathological subtypes of the thyroid cancer. It has been mentioned previously that malignancy identified within the thyroid on ¹⁸F-FDG PET may be more aggressive, which could be due to the fact that ¹⁸F-FDG PET has less sensitivity in identifying differentiated cancers [5]. Interestingly, our results demonstrated that all incidentalomas with primary thyroid malignancies had differentiated PTC, with 1 patient having pT1a disease. FDG avid lesions are likely to be those that express GLUT1 intensely. There is some evidence that phosphatase and tensin homologue (PTEN)-negative PTC have considerable GLUT1 expression, and the relationship between the increasingly understood genetic alterations (BRAF, PTEN, etc.) and FDG avidity merits further study [40]. Furthermore, contradicting the previous findings, we found that almost half (44.4%) of the patients with malignant incidentaloma were due to secondary metastases or lymphoma.

The SUV assessed by ¹⁸F-FDG PET/CT is a semi-quantitative measure of glucose metabolism, which often reflects metabolic activity frequently correlated with biologic aggressiveness and clinical behaviour of malignant lesions, though not specific for malignancy. Literature evidenceon the SUV_max in benign and malignant lesions vary with several reports showing a statistically significant difference [9,13,19,21,22,23,34,41] whilst many others have shown no significant difference [8,14,15,27,30,32,33,35,36]. Previous studies have also reported cutoff SUV values ranging from 3.5 to 5 in differentiated benign to malignant thyroid lesions [24,34,41]. In our study, a highly statistical difference was present between the SUV_max of benign and malignant lesions. We found a cutoff value of 9.1 had 81.6% sensitivity and 100% specificity in differentiating benign from malignant lesions.

Considering the literature evidence and following the findings from our study, it may be suggested that in patients with an incidental focal thyroid uptake, an SUV_max <3.3 and low clinical risk, one may be able to reassure the patients [42]. However, if there is a high clinical risk (such as previous history of radiation exposure and family history of thyroid cancer) or if the SUV_max is higher, the thyroid uptake should be regarded as suspicious for underlying malignancy and investigated further as appropriate. However, in patients with widespread metastases with poor prognosis, further investigation of incidental thyroid uptake may not be appropriate. The decision regarding investigating the incidental thyroid uptake needs to be made on an individual patient basis.

There are some limitations to our study which need to be mentioned. This being a retrospective study, a definitive diagnosis could not be available in many patients due to reasons such as further patient management in other hospitals, advanced primary malignancy with widespread metastatic disease, poor prognosis, low clinical index of suspicion or death. As data was collected over nearly 14 years, the studies were performed on different PET or PET/CT scanners. To our knowledge, however, this is the largest UK series on thyroid incidentalomas over the past 14 years. The results of this study are important as a lot of these findings were determined in a cohort of patients with poor prognosis or who may be terminally ill. It is also essential that patients with incidental focal thyroid uptake are properly evaluated and a collaborative protocol is established between referring clinicians, nuclear medicine physicians, endocrinologists and thyroid surgeons for the appropriate management of these patients.

Conclusion

Incidence of focal thyroid uptake on ¹⁸F-FDG PET study remains rare in our study cohort. The malignancy potential of these lesions, however, remains high and warrants prompt follow-up by the clinician. The SUV_max may aid in further characterisation of the lesion and its management. Incidence of malignant primary and secondary pathology within these lesions remains equally possible.

Disclosure Statement

The authors declare that they have no conflicts of interest.