¹⁸F-Fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET) has been used for cancer screening in healthy subjects as well as for the diagnosis, staging, and assessment of treatment response in various cancers given that most malignant tumors are ¹⁸F-FDG-avid with increased glucose metabolism (1-4). Since ¹⁸F-FDG PET for oncologic purposes usually covers the entire body or torso (base of the skull to mid-thigh), clinically unexpected, incidental hypermetabolic lesions are frequently found. For example, previous studies have reported incidental uptake in secondary primary cancer in cancer patients at the time of initial tumor staging (5), primary parotid tumor (6), or thyroid gland (7-18).

Incidental thyroid uptake on PET is classified into two categories: focal or diffuse thyroid uptake (DTU). According to previous reports, focal thyroid uptake has been detected in 1-4% of cases and is reportedly associated with malignancy in 20-50% of cases (7-11). There were several studies regarding DTU on PET or PET/computed tomography (CT) that reported incidences ranging from 0.6% to 3.5% (12-17). Some studies have reported that DTU can be a normal variant, while several previous reports demonstrated that diffuse thyroid FDG uptake could be an indicator of benign thyroid disease and related to Hashimoto's thyroiditis. Chen et al. (15) reported that diffuse thyroid ¹⁸F-FDG uptake most likely represents chronic thyroiditis. Karantanis et al. (16) also reported that increased ¹⁸F-FDG uptake in the thyroid gland is associated with chronic lymphocytic thyroiditis. However, these studies involved small numbers of subjects, and there was a possibility of selection bias given that further diagnostic work-up, including in vitro thyroid function tests (TFT) and neck ultrasonography (USG), was only performed in patients with increased thyroid uptake. In other words, we cannot be certain that an absence of significant thyroid uptake excludes the possibility of abnormalities in the thyroid gland. Therefore, the clinical significance of DTU on PET remains unclear.

We retrospectively investigated the clinical significance of incidental DTU in subjects without prior history of cancer and thyroid disease by comparing the results of ¹⁸F-FDG PET and the results of in vitro TFT, serum anti-microsomal antibody (AMA) levels, and/or neck USG that were all acquired on the same day as part of a general health examination.

The characteristics of each group were described in Table 1. Among all 2062 scans, diffuse thyroid ¹⁸F-FDG uptake was found in 137 (group I; incidence, 6.6%). The proportion of female subjects was significantly higher in group I than in group II. The percentage of subjects with DTU in the female group was 16.0% and 5.2% in the male group, which was significantly different (p < 0.001).

Serum TSH and AMA levels were significantly higher in group I than in group II (p < 0.001). An elevated AMA level (55.5% vs. 3.3%, p < 0.001) and DPC on USG were more frequently observed in group I than in group II (48.7% vs. 3.9%, p < 0.001). The FT₄ level was slightly lower in group I than in group II (p < 0.05). The T₃ level was not statistically different between the two groups. The proportion of subjects with any abnormal results in serum FT₄, T₃, TSH, or AMA levels or with DPC on USG was significantly higher in group I than in group II (71.5% vs. 10.6%, p < 0.001). Among group I subjects who had any abnormal results in serum FT₄, T₃, TSH, or AMA levels or DPC on USG, 49.0% (48/98) had two or more abnormal findings. The remaining 50 subjects had only one abnormal finding. In contrast, among group II subjects with any abnormal results in serum FT₄, T₃, TSH, or AMA levels or DPC on USG, 89.7% (183/204) had only one abnormal finding, including one abnormal TFT result in 94 subjects, an increased AMA level in 50, and DPC on USG in 39. The remaining 21 subjects had two or more abnormal findings. Also, the AMA levels of group I subjects with elevated AMA were significantly higher than those of group II subjects with elevated AMA (2922.4 ± 6402.6 vs. 53.9 ± 459.2, p < 0.001). In addition, there were no significant differences in abnormal TFT, AMA levels, or DPC on USG between female and male subjects in group I.

The analysis of thyroid functional status revealed that 1918 scans (93.0%) had euthyroid status, 9 (0.4%) had hyperthyroidism, 19 (0.9%) had hypothyroidism, 14 (0.7%) had subclinical hyperthyroidism, and 102 (5.0%) had subclinical hypothyroidism. Abnormal thyroid function was more frequently observed in group I (26.3% vs. 5.6%, p < 0.001). The most common thyroid disease in group I was subclinical hypothyroidism (83.3%). In group I, subclinical hyperthyroidism, hypothyroidism, and hyperthyroidism all exhibited the same prevalence of 5.6% (Table 2).

To evaluate the thyroid functional status according to the combination of DTU on PET and DPC on USG, subjects were divided into 4 groups; neither DTU nor DPC, presence of DPC without DTU, presence of DTU without DPC and presence of both DTU and DPC. In regard to the results, the thyroid functional status was significantly different between the groups (Table 3). In the group with both DTU and DPC, a significantly lower FT₄ level, higher TSH and AMA levels, and higher prevalence of subclinical hypothyroidism were observed than in the other groups. In the group without DTU or DPC, significantly lower TSH and AMA levels, and a higher prevalence of euthyroid status were observed than in the other groups.

On visual grading analysis, the mean uptake score was 0.16 ± 0.65. There were significant differences in the visual grading score according to the functional status of the thyroid (p < 0.001, Fig. 2). Post hoc LSD testing showed significantly higher uptake parameters in the subclinical hypothyroidism group than in the euthyroid, hypothyroidism, and subclinical hyperthyroidism groups. When subjects were divided into three groups according to the visual grading score (0 vs. 1-2 vs. 3-4), the proportion of abnormal TFT results (5.6% vs. 18.3% vs. 38.2%, respectively) or elevated AMA levels (3.3% vs. 43.9% vs. 72.7%, respectively) was significantly increased according to the visual grading score (p < 0.001). There were significant differences in FT₄, TSH, and AMA levels (p < 0.001) according to visual grading score group; however, there were no significant differences in T₃ levels (Fig. 3). More specifically, FT₄ was significantly decreased according to the visual grading score while TSH and AMA levels were significantly increased according to the visual grading score. Post hoc LSD tests showed that the subgroup with a visual grading score of 3-4 had significantly higher TSH and AMA levels and lower FT₄ levels compared to both of the other subgroups. The subgroup with a visual grading score of 0 had significantly lower TSH and AMA levels than the other two subgroups (Fig. 2). The proportion of subjects with any abnormal results in serum FT₄, T₃, TSH, or AMA levels or DPC on USG was significantly and gradually increased according to the visual grading score group (0 vs. 1-2 vs. 3-4 = 10.6% vs. 58.5% vs. 90.9%, p < 0.001).

Representative cases of subjects with DTU and without DTU are illustrated in Figure 1. A PET image of a 54-year-old male patient revealed no definite ¹⁸F-FDG uptake in the thyroid gland (Fig. 1A). The visual grading score was 0 and the AMA level was normal (1 U/mL). Thyroid function status was euthyroid (T₃, 157 ng/dL; FT₄, 1.11 ng/dL; TSH, 0.67 µIU/mL). The next PET image of a 54-year-old male patient demonstrated mildly increased FDG uptake in the thyroid gland (Fig. 1B). The visual grading score was 2 and the AMA level was 196 U/mL. Thyroid function status was euthyroid (T₃, 101 ng/dL; FT₄, 1.17 ng/dL; TSH, 1.82 µIU/mL). On the other hand, PET image of a 54-year-old male patient demonstrated diffuse ¹⁸F-FDG thyroid uptake (Fig. 1C). The visual grading score was 4 and the AMA level was 42300 U/mL. Thyroid functional status was subclinical hypothyroidism (T₃, 96 ng/dL; FT₄, 0.84 ng/dL; TSH, 13.97 µIU/mL).

In this study, we investigated the clinical significance of incidental DTU on ¹⁸F-FDG PET in subjects without a history of cancer. Our data revealed that the presence of DTU was closely correlated with increased serum AMA and TSH levels and the presence of DPC on USG in non-cancer subjects. In 56% of patients with DTU, the level of serum AMA was increased. In 23% of patients with DTU, the level of serum TSH was increased. The group with a combined presence of DTU and DPC had the lowest FT₄ level, highest TSH and AMA levels, and the highest prevalence of subclinical hypothyroidism. These suggest that DPC may provide an additional contribution to thyroid functional abnormality and an elevated AMA level. In addition, the degree of increased thyroid uptake was significantly associated with an increase in serum AMA and TSH levels. However, T₃ was not associated with DTU. There was a very small difference in FT₄ between group I and II. These results suggest that further diagnostic work-ups such as a TFT and thyroid autoantibody check is necessary to rule out benign thyroid disorder such as autoimmune thyroid disease, when incidental DTU is observed in a routine interpretation of PET images, especially showing high ¹⁸F-FDG uptake.

There was a large discrepancy in the incidence of DTU between our study (6.6%) and previous PET reports (0.6-3.5%) (14, 19, 20). This is probably due to different detection rates derived from different consensuses regarding DTU. For several years, incidentally detected ¹⁸F-FDG uptake on whole-body PET/CT scans has been of interest to our group (21, 22). As such, we have made an effort to carefully review thyroid uptake encountered in routine clinical practice. The use of PET/CT might also contribute, in part, to the difference between our data and that of other studies due to the precise lesion localization and evaluation that can be achieved with PET/CT. As a consequence, we observed a relatively large number of cases with incidental thyroid uptake, and the overall intensity of thyroid uptake was lower than that of the previous reports.

There have been previous studies regarding DTU on PET or PET/CT (12-17, 23). With the exception of a few studies, the majority of previous studies evaluated non-cancer subjects in a manner similar to ours. The major aspects that differentiated our study from previous research were that we included all consecutive scans during a set period and only excluded subjects with known thyroid disease irrespective of the presence of DTU. Additionally, in our study the results of TFT and AMA were available in all subjects. In two previous reports, a control group without DTU was artificially selected (14, 16) and included subjects with known thyroid disease in the group, and comparative analyses of TFT and other tests were not performed for the group with DTU (16). Furthermore, the number of subjects was the largest in this study compared with previous studies.

In our study, 72% of subjects with DTU (group I) had an abnormal TFT, AMA level or DPC on USG, which was similar with the incidences of previous studies ranging from 59% to 100% (12-16). In contrast, 89% of subjects without DTU (group II) had no abnormal findings in serum FT₄, T₃, TSH, or AMA levels or DPC on USG. Most subjects in both groups had euthyroid status. In addition, the degree of the ¹⁸F-FDG visual uptake score was significantly associated with the risk of abnormal TFTs, an elevated AMA level, and serum levels of TSH and AMA. These findings suggest that the most plausible cause of DTU may be autoimmune thyroid disease, which corroborates the results of previous studies (12-17). Though subclinical hypothyroidism with autoimmune thyroiditis was the most frequently found in our study and other previous reports, other autoimmune thyroid disorders such as Graves' disease and chronic autoimmune thyroiditis may cause DTU (15, 17, 24). Additionally, the presence and degree of DTU seems to reflect the severity of thyroid cell damage caused by an autoimmune mechanism. Although there have been several studies regarding DTU on PET, the uptake mechanism of autoimmune thyroid disease still remains uncertain. In the process of chronic autoimmune thyroiditis, there is lymphocyte infiltration, inflammation, and tissue destruction (14). Activated lymphocytes within the thyroid gland affected by inflammation and autoimmune disease are reported to show intensive glycolysis (24). In a destructive step of autoimmune thyroiditis, T-helper type 1-mediated mechanisms lead to the infiltration of activated macrophages into the thyroid gland (25). Considering the results of our study and those of previous studies, we suggest that ¹⁸F-FDG uptake may be increased due to the lymphocytes and activated macrophages that have accumulated in the damaged thyroid gland as a result of an autoimmune mechanism.

In this study, only 11% of subjects without DTU and previous thyroid disease had one or more abnormal findings in serum FT₄, T₃, TSH, or AMA levels or DPC on USG, which is similar to a result of 8% in a previous study of 36 subjects without DTU (25). Furthermore, those subgroup subjects without DTU had significantly less severe abnormal findings on TFT and AMA levels than subjects with DTU. Therefore, we suggest that further evaluation regarding functional thyroid disease is not always necessary if subjects have no DTU on PET. In contrast, 59% of subjects with mild to moderate DTU (uptake score 1-2) had one or more abnormal findings in serum FT₄, T₃, TSH, or AMA levels or DPC on USG. This suggests that even mild or low-grade uptake in the thyroid should not be ignored because this uptake may be associated with overt or subclinical hypothyroidism requiring hormone therapy, which supports findings from previous studies (16, 17). Therefore, any kind of incidental DTU may require further evaluation, such as TFT and anti-thyroid antibodies with or without USG. Nevertheless, approximately 10% of subjects with severe DTU (uptake score 3-4) had normal TFT and AMA levels without DPC on USG. Although the cause of DTU in these subjects is not clear, checking a single anti-thyroid antibody instead of multiple kinds of anti-thyroid antibodies might influence the result.

In this study, we used a semiquantitative thyroid uptake score compared with reference organs instead of conventional standardized uptake value (SUV)-based methods, which demonstrated significant positive associations with TFT and AMA levels. Conversely, in another study maximum SUV of the thyroid was not significantly correlated with serum TSH or thyroperoxidase antibody levels (16). This might have resulted from the small number of subjects (about 20) and the lack of a control group. Furthermore, another possible cause is that they measured the maximum SUV which cannot fully reflect the entire thyroid gland uptake. In practice, only a moderate correlation was found when we analyzed the correlation of AMA level and maximum SUV in group I (r = 0.44, p < 0.001, unpublished data).

We found that the percentage of DTU among females was higher than in males. This is possibly related to a well-known fact that autoimmune thyroid disease occurs more frequently in women than in men. According to previous reports, hormones, especially estrogen, are thought to play an important role in women's higher risk of autoimmune disorders (26-28). Gender differences in immunity may also explain the difference in risk for autoimmune disorders. For example, women naturally have stronger inflammatory responses than men when immune reactions are generated (26-28).

Our present study had several limitations. Only subjects undergoing PET or PET/CT among populations receiving a general health examination were enrolled in this study. This might result in a significant discrepancy of the sex ratio and high incidence of office workers as an occupation, which may potentially induce a selection bias. Another possible limitation was the use of two different kinds of scanners and acquisition protocols; therefore, the SUV of thyroids could not be compared in this study. Instead, the thyroid uptake score was acquired by comparison with the uptakes of reference organs, which is a commonly used objective semiquantitative method. Moreover, it was not possible to examine the correlation of the clinical exam, such as subjects' symptoms, with the results due to the lack of relevant medical records in our health promotion center.

In conclusion, the presence of DTU in subjects without a history of cancer is closely correlated with increased AMA and TSH levels and the presence of DPC on USG. Furthermore, the degree of DTU increased according to the elevation of serum AMA and TSH levels. Therefore, the most plausible pathological cause of DTU may be thyroid cell damage caused by an autoimmune mechanism. Our results suggest that incidental DTU may deserve further diagnostic evaluation such as TFT with a thyroid autoantibody check to rule out the possibility of benign thyroid disease. Nevertheless, approximately 10% of subjects with severe DTU had normal TFT and AMA levels without DPC on USG, which warrant further investigations to find a cause.