124 I Positron Emission Tomography/Computed Tomography Versus Conventional Radioiodine Imaging in Differentiated Thyroid Cancer: A Review

In ¹³¹I therapy-naive patients

In 2004, Freudenberg et al. (11) demonstrated the value of ¹²⁴I PET and PET/CT compared with post-therapy ¹³¹I WBS. All 12 patients were positive (i.e., had at least 1 positive lesion) on the post-therapy ¹³¹I WBS, ¹²⁴I PET, and ¹²⁴I PET/CT. However, ¹²⁴I PET/CT identified more lesions than ¹³¹I WBS in 9/12 (75%) patients. Overall, ¹²⁴I PET/CT visualized more lesions (69 foci) compared with ¹²⁴I PET (60 foci), while post-therapy ¹³¹I WBS identified the least (50 foci). The lesion detection rate was 100%, 87%, and 83% for ¹²⁴I PET/CT, ¹²⁴I PET, and post-therapy ¹³¹I WBS, respectively. The ¹²⁴I PET/CT fusion display did not reveal additional lesions relative to separately reported ¹²⁴I PET and CT readings, but the CT provided identifiable anatomical structures in the ¹²⁴I PET imaging (11). A limitation of their study is that six patients had serum Tg <1 ng/mL in the absence of antibodies, inadequate TSH stimulation, urine iodine >150 μg/L, and/or ¹³¹I uptake >10% in the thyroid bed. If these six patients were excluded from the analysis, the lesion detection rate would be 100%, 79%, and 58% for ¹²⁴I PET/CT, ¹²⁴I PET, and post-therapy ¹³¹I WBS, respectively. They showed that ¹²⁴I PET/CT is more useful than ¹²⁴I PET and post-therapy ¹³¹I WBS in advanced DTC patients.

In another study, Capoccetti et al. (12) compared ¹²⁴I PET/CT and post-therapy ¹³¹I WBS in 67 postsurgical patients who underwent initial ¹³¹I therapy. Although the overall diagnostic accuracy of ¹²⁴I PET/CT could not be calculated from the information provided, the ¹²⁴I PET/CT and the ¹³¹I WBS were in total agreement (i.e., number, site, and extent of disease) in 58/67 (87%) patients, which included the same number of thyroid remnants (182 foci) in 58/67 (87%) patients and the same number of lymph nodes (63 foci) in 21/67 (31%) patients. In discordant cases, ¹²⁴I PET/CT detected more pathological foci, which were mainly than post-therapy ¹³¹I WBS in 5/67 (7.5%) patients. Conversely, the WBS detected more pathological foci than ¹²⁴I PET/CT in 4/67 (5.9%) patients, and these included additional thyroid remnants, lymph nodes, and disseminated lung metastases. Additional lesions found on ¹²⁴I PET/CT upstaged 11/67 (16%) of patients and modified management in 2/67 (3%) of all evaluated patients.

In mixed cohort of ¹³¹I therapy-naive and non-naive patients

In a retrospective study by de Pont et al. (3), the diagnostic accuracy of ¹²⁴I PET/CT was compared with planar ¹³¹I WBS and SPECT/CT. Twenty consecutive patients underwent planar ¹³¹I WBS, low-dose CT attenuated ¹³¹I SPECT, and high-dose CT attenuated ¹²⁴I PET/CT. Five of these patients had previous ¹³¹I therapies with a cumulative activity of 2.78–14.99 GBq (75–405 mCi). In the patient-based analysis, one additional patient had uptake on the ¹²⁴I PET/CT compared with ¹³¹I SPECT/CT; however, this paratracheal remnant was not seen by other imaging modalities. In the lesion-based analysis, ¹²⁴I PET/CT identified 57 (92%) of a total of 62 foci, ¹³¹I SPECT/CT identified 50 (81%), and ¹³¹I WBS identified 39 (63%). On further analysis, 13 patients did not have adequate TSH stimulation >30 mIU/L and/or serum Tg level >5 ng/mL (TSH stimulation unknown) at the time of ¹²⁴I administration. If these patients were removed from the analysis, then ¹²⁴I PET/CT detected one additional DTC lesion compared with ¹³¹I SPECT/CT; and if the Tg was >10 ng/mL, then 3/3 (100%) were positive on ¹²⁴I PET/CT when ¹³¹I SPECT/CT was positive. In two patients, the tumor stage was changed based on ¹²⁴I PET/CT findings. This study showed that ¹²⁴I PET/CT is predictive not only of post-therapy ¹³¹I planar WBS but also of ¹³¹I SPECT/CT for lesion detection and staging differentiated thyroid carcinoma.

In a cohort of 106 postoperative ¹³¹I therapy-naive and 31 ¹³¹I therapy non-naive patients who received a cumulative activity of 1–47 GBq (37–1739 mCi) of ¹³¹I, Ruhlmann et al. (16) reported a high level of agreement between ¹²⁴I PET/CT and post-therapy ¹³¹I WBS and SPECT/CT. Seven patients had poorly differentiated carcinoma. Sixty-one of the 137 (45%) patients were positive both on the 25 MBq (0.68 mCi) ¹²⁴I PET/CT imaged at 24 and 120 hours and on the 1.0–10.0 GBq (27–270 mCi) ¹³¹I WBS and SPECT/CT imaged at 5–10 days or both. Of the 61 patients, 59 (97%) patients were positive on ¹²⁴I PET/CT and 60 (98%) patients were positive on post-therapy ¹³¹I imaging. The true positive was defined as positive on both ¹²⁴I and ¹³¹I, which was 95% (58/61), while 76 patients had no uptake on both scans. In the lesion-based analysis, a total of 227 metastatic lesions were found, of which ¹²⁴I PET/CT detected 223 (98%) and ¹³¹I imaging detected 225 (99%); the level of agreement between ¹²⁴I PET/CT and ¹³¹I imaging was 97% (221/227). They observed that a diagnostic activity of ¹²⁴I, ∼1% of the therapeutic activity of ¹³¹I, may be sufficient to achieve a high level of agreement between ¹²⁴I PET/CT and ¹³¹I WBS, including SPECT/CT. Ruhlmann et al. suggested that due to the high level of agreement, ¹²⁴I PET/CT is adequate in the detection of radioiodine-avid metastases and has much lower radiation exposure compared with ¹³¹I post-therapy scans.

Gulec et al. (13) reported a prospective study in patients who did or did not have prior ¹³¹I therapy and demonstrated that ¹²⁴I PET/CT imaging had high lesion detection sensitivity and offered the additional advantage of quantitation. In the lesion-based analysis of the 15 patients enrolled, a total of 46 lesions were identified by ¹²⁴I, ¹³¹I, and/or ¹⁸F FDG PET/CT, of which the individual scans identified 37, 28, and 16 of the lesions, respectively. ¹²⁴I PET/CT clearly demonstrated superior imaging by identifying 22.5% more foci of radioiodine-avid lesions than ¹³¹I imaging and by providing better anatomical detail of remnant thyroid tissue.

In ¹³¹I non-naive patients

Phan et al. (7) showed in patients who had been previously treated with ¹³¹I that of the 11 patients positive on the post-therapy ¹³¹I WBS, 9 (82%) patients showed uptake on the ¹²⁴I PET. Two patients had uptake only on the ¹²⁴I PET scan with no anatomical correlate confirmed by neck ultrasound, ¹⁸F-FDG PET, or magnetic resonance imaging (MRI), whereas two other patients showed only uptake on the post-therapy ¹³¹I WBS, of which one was confirmed by MRI but not ¹⁸F-FDG PET. Seven patients were negative on both scans. After excluding five patients with undetectable serum Tg (<0.3 ng/mL) after THW, then all seven (100%) patients who had positive post-therapy ¹³¹I WBS were also positive on ¹²⁴I PET scans. The ¹²⁴I PET scans in this group of patients adequately detected metastatic lesions also seen on post-therapy ¹³¹I scans in previously ¹³¹I-treated patients.

The superiority of ¹²⁴I PET/CT scan compared with post-therapy ¹³¹I WBS at 72 hours was shown by Pettinato et al. (9) in 26 post-therapy patients treated with dosimetrically guided activities of 1.95–11.46 GBq (53–310 mCi) of ¹³¹I. In this study, ¹²⁴I PET/CT images correlated with every post-therapy ¹³¹I WBS and showed a superior image quality due to the better resolution and lesion detectability of PET/CT compared with WBS. Fifteen patients were positive, and 11 patients were negative on both scans. Of note, both patients with tall cell variant and one with Hurthle cell were negative on both scans, and more discussion of histology is summarized under section heading ¹²⁴I PET in various DTC histologies. Although a total of 34 lesions were measured for dosimetric analysis on 13 of the 15 ¹²⁴I PET/CT-positive patients, no individual lesion-based analysis was available for further evaluation.

However, in a retrospective review of seven ¹³¹I post-therapy patients with elevated serum Tg, Lammers et al. (5) showed poor sensitivity of ¹²⁴I PET/CT in detecting post-therapy ¹³¹I-positive metastases. The reasons for poor sensitivity may be due to patients not receiving a low-iodine diet, and five of the seven patients were prepared with rhTSH injections. Two patients had Hurthle cell cancer. Of the six patients with a positive post-therapy ¹³¹I WBS, only one patient (1/6 [17%]) was also positive on the ¹²⁴I PET/CT scan, and this patient was a true-positive based on CT.

In patients with positive serum thyroglobulin levels and negative imaging

From a prospective study, Khorjekar et al. (14) performed a retrospective study of 12 patients with elevated serum thyroglobulin levels, negative diagnostic ¹³¹I/¹²³I scan, negative ¹²⁴I PET/CT scan, subsequent “blind” ¹³¹I therapy, and subsequent post-therapy ¹³¹I scan. Of 12 patients, 10 patients had abnormal uptake on the post-therapy ¹³¹I scan, resulting in a false-negative rate of 83% in this select patient population. Hence, there may be a subgroup of patients with negative diagnostic radioiodine (¹²³I, ¹³¹I, or ¹²⁴I) imaging who may still benefit from ¹³¹I therapy; however, outcomes of the “blind” ¹³¹I therapy were not evaluated.

This was further supported by a prospective multicenter observational cohort study by Kist et al. (15). Seventeen post-therapy DTC patients with biochemical evidence of recurrence but negative neck ultrasound were enrolled. The patients underwent rhTSH-stimulated ¹²⁴I PET/CT imaging and a THW-stimulated “blind” ¹³¹I therapy to test the hypothesis that negative ¹²⁴I PET/CT can predict negative post-therapy ¹³¹I WBS. Their results showed that only five (56%) ¹²⁴I PET/CT scans were positive on the nine positive post-therapy ¹³¹I scans. In all of the eight patients with negative post-therapy ¹³¹I scans, the ¹²⁴I PET/CT scans were also negative. Of the 12 ¹²⁴I PET/CT scans with no pathologic uptake, the false-negative rate was 33% (4/12). Lesion-based analysis showed that the post-therapy ¹³¹I WBS revealed 14 lesions versus 8 on ¹²⁴I PET/CT. Of note, one patient with disseminated lung metastasis was positive on the post-therapy ¹³¹I scan but negative on ¹²⁴I. Kist et al. demonstrated that rhTSH-stimulated ¹²⁴I PET/CT was not adequate to avoid futile “blind” ¹³¹I therapy because of its high false-negativity rate (33%; 4/12) in the patient-based analysis. If ¹²⁴I PET/CT guided the decision whether or not to administer “blind” ¹³¹I therapy in this select group of patients, then 44% (4/9) of patients would have been denied potentially beneficial ¹³¹I therapy.

TSH stimulation method for ¹²⁴I PET/CT

The method of stimulating the patient's TSH in preparation of scanning may be an important factor. For example, Van Nostrand et al. (17) showed that more patients were positive on ¹²⁴I PET/CT after THW (63% [10/16]) than after rhTSH (29% [7/24]). Also, a greater number of foci were detected on THW than rhTSH ¹²⁴I PET/CT (117 vs. 17 lesions). In addition, de Pont et al. (3) reported a smaller percentage of discordant results between ¹²⁴I PET/CT and post-therapy ¹³¹I scanning in patients who were prepared with THW compared with rhTSH. In their study, partial or complete discordance was seen in 43% (6/14) of patients who underwent THW and 67% (4/6) of patients who received rhTSH. Wu et al. (20) reported one patient with diffuse lung metastasis who underwent two sets of ¹²⁴I PET/CT scans performed two months apart. In this patient, no discrete nodular uptake was seen on the ¹²⁴I PET/CT under rhTSH stimulation, but two of the largest nodules were seen after THW. Therefore, it is also important to specify whether different TSH stimulation methods were used to compare ¹²⁴I PET/CT and ¹³¹I scans (15).

In kinetic studies of lesional absorbed dose to administered activity ratio (LDpA) with ¹²⁴I-PET/CT under rhTSH and THW stimulation, the mean LDpA under THW was almost twice that of rhTSH (51.8 Gy/GBq in 66 iodine-avid metastases vs. 30.6 Gy/GBq in 71 iodine-avid metastases) (21). Even though this difference (21.2 Gy/GBq [CI +51 to −9 Gy/GBq]) was not statistically significant, the trend is that rhTSH has a lower LDpA than THW in metastatic lesions. However, this study was a comparison between two different groups of patients. A study by Plyku et al. (22) compared LDpA within the same patients in which three patients underwent THW and rhTSH ¹²⁴I PET/CT at least one month apart. They found that the absorbed dose per unit administered activity was higher in the THW study than in the rhTSH study. The ratios of the average tumor absorbed dose after THW compared with rhTSH ranged from 1.4 to 27.

Given the above differences between rhTSH versus THW in ¹²⁴I PET/CT imaging, this could be a possible explanation for the poor detection rate of ¹²⁴I PET/CT when compared with post-therapy ¹³¹I scans in some studies. For example, in the study by Kist et al., all patients received rhTSH stimulation for ¹²⁴I PET/CT, but received THW for post-therapy ¹³¹I scan (15).

Administered activity of ¹²⁴I and ¹³¹I

Most studies administered 74 MBq (2 mCi) of ¹²⁴I for diagnostic imaging. One case report administered an activity as low as 15 MBq (0.4 mCi) of ¹²⁴I (23), while activities as high as 222–284.9 MBq (6–7.7 mCi) of ¹²⁴I were used in a case series (1). Due to the limited literature and wide heterogeneity between studies, no definitive conclusion can be made regarding administered activities of ¹²⁴I. Future studies are warranted to evaluate the effect of different ¹²⁴I activities on image quality. Furthermore, in this review, the diagnostic ¹³¹I scans were performed with 37–74 MBq (1–2 mCi) of ¹³¹I. However, the range of ¹³¹I activity administered before the post-therapy ¹³¹I scans ranged widely from 1 GBq (27 mCi) (16) to 20 GBq (541 mCi) (4). Therefore, we need to be cautious when directly comparing the lesion detection rate of post-therapy ¹³¹I scans between each publication.

The stark difference in the amount of prescribed activity to obtain ¹²⁴I PET/CT and post-therapy ¹³¹I scans may be one of the reasons for a lower lesion detection rate on ¹²⁴I PET/CT in selected subgroups of DTC patients, such as in patients previously treated with ¹³¹I and in patients with more aggressive pathologies. The major effect of the prescribed activity on the quality of radioiodine imaging has been shown by a Wells et al. (24) publication where the post-therapy ¹³¹I scan may be positive in as many as 25–80% of patients with a negative diagnostic radioiodine scan. Furthermore, Khorjekar et al. (14) have shown that up to 83% (10/12) of DTC patients with a negative diagnostic ¹²⁴I PET/CT may still have a positive post-therapy ¹³¹I scan. With all other imaging parameters constant, a higher prescribed activity usually yields a higher lesion detection rate (25). Therefore, further studies are needed to evaluate the lesion detection rate of ¹²⁴I PET/CT in selected subgroups of DTC patients utilizing higher ¹²⁴I activities than the presently used 74 MBq (2 mCi).

¹²⁴I PET/CT scan initiation time and acquisition time

The rate of lesion detection varies by the day of ¹³¹I imaging after administration of ¹³¹I (26), so likewise, the ¹²⁴I PET/CT scan will also have varying rates of detection when imaged at different time points after administration of ¹²⁴I. Wu et al. (20) evaluated 14 sets of 5 time points (2, 24, 48, 72, and 96 hours) for 13 DTC patients highly suspected of metastatic disease. They found that a total of 37 lesions suspicious for metastasis were identified on ¹²⁴I PET/CT, of which 8 were seen on the 2-hour scan, 22 at 24 hours, 30 at 48 hours, 28 at 72 hours, and 27 at 96 hours. The 48-hour scan yielded a significantly greater number of distant metastases. As further support that 48 hours may be the optimal imaging time, Pettinato et al. (9) and Gulec et al. (13) showed that the 48-hour ¹²⁴I PET/CT images correlated well with post-therapy ¹³¹I WBS.

In regard to duration of acquisition, the majority of studies in this review had similar PET acquisition times of four to five minutes per bed position. Only Pettinato et al. imaged three to four bed positions with a duration of 15 minutes each. Lee et al. had an acquisition time of three minutes per bed frame. Kist et al. did not specify their acquisition time. Although comparisons between acquisition times should be made, there are too many confounding factors between these studies to make a fair conclusion. Future studies are warranted to compare administered ¹²⁴I activity, time of imaging from administration of ¹²⁴I, and image acquisition times for ¹²⁴I, ¹²³I, and ¹³¹I.

¹²⁴I PET/CT in ¹³¹I therapy-naive versus non-naive patients

A factor that may affect the detection rate of ¹²⁴I PET/CT in the literature is whether the study included only ¹³¹I therapy-naive patients, non-naive patients, or a mixed group of patients. In the reviewed studies, ¹²⁴I PET was either performed in post-surgical patients as initial evaluation, or performed in ¹³¹I-treated patients suspected of recurrent or metastatic disease.

For comparison of the lesion detection rate of ¹²⁴I PET/CT with diagnostic radioiodine scanning between naive versus non-naive patients, there was no study that just looked at naive patients. All five diagnostic ¹³¹I scan studies were performed in ¹³¹I therapy non-naive cohorts (6 –8,10,17). These five studies uniformly showed that ¹²⁴I PET/CT is better than diagnostic radioiodine scanning in non-naive cohorts.

We observe that compared with the ¹³¹I therapy-naive cohorts, the lesion detection rate of ¹²⁴I PET/CT in cohorts of non-naive patients with suspected or known advanced disease is lower (5,14,15). Studies in cohorts of only ¹³¹I therapy-naive patients that compared ¹²⁴I PET with post-therapy ¹³¹I scans included the work by Freudenberg et al. (11) and Capoccetti et al. (12); studies that analyzed a heterogeneous cohort included the one by de Pont et al. (3), Ruhlmann et al. (16), and Gulec et al. (13). This observation may be due to a relative increase in the proportion of dedifferentiated thyroid cancer cells in metastatic lesions as previous ¹³¹I therapies eradicated radioiodine-sensitive differentiated cells or as disease progression toward poorly differentiated variants or anaplastic transformation. However, the clinical utility of ¹²⁴I PET/CT in ¹³¹I therapy-naive patients may be low due to the lower incidence of distant metastasis in ¹³¹I-naive patients. Until the higher cost of ¹²⁴I per MBq or mCi decreases (13), it may be more economical to perform ¹²³I/¹³¹I diagnostic scans in the initial patient assessment, and to perform ¹²⁴I PET in patients with suspected recurrent and metastatic disease.

¹²⁴I PET in various DTC histologies

Few studies reported on aggressive variants of thyroid cancer (5,9,12,16). However, of those that did report the histology, the trend was that ¹²⁴I PET/CT imaging may be less useful in aggressive histologies with increased false negatives on ¹²⁴I PET images. Of these studies, Pettinato et al. (9) reported that in three tall cell variant papillary thyroid cancers (PTCs) and one Hurthle cell cancer, 4/4 were negative on ¹²⁴I PET/CT and 3/3 were negative on the post-therapy ¹³¹I WBS. Lammers et al. (5) reported one Hurthle cell cancer that was false negative on ¹²⁴I PET/CT. Capoccetti et al. (12) reported four Hurthle cell carcinomas, as well as one clear cell, one tall cell, and three sclerosing variant PTCs, and Ruhlmann et al. (16) reported seven poorly differentiated thyroid carcinomas, but no individual data were available. As ¹²⁴I PET/CT is not widely available and the number of patient studies is small, future studies should also separate the results of patients by histology.

¹²⁴I PET in patients with disseminated lung metastases

¹²⁴I PET may have poor performance in the detection of disseminated miliary lung metastases. In four such patients who had a positive diagnostic or post-therapy ¹³¹I WBS, Kist et al. (15), Capoccetti et al. (12), and Gulec et al. (13) reported that all four cases were negative on ¹²⁴I PET/CT. Freudenberg et al. (4) retrospectively analyzed 70 consecutive patients with advanced DTC and found that only 1/7 patients with disseminated lung disease had visible uptake on ¹²⁴I PET. However, this may be due to confounding factors such as low ¹²⁴I activity or time of imaging as discussed previously. In this study, disseminated iodine-avid lung metastases were defined as lung metastases positive on post-therapy ¹³¹I WBS but negative on thoracic CT. In a quantitative analysis of the lung-to-background (L/B) ¹²⁴I uptake ratios, there was a significant difference in the L/B ratio between disseminated lung metastases and the control group (0.92 ± 0.31 vs. 0.41 ± 0.13; p < 0.001). Freudenberg et al. (27) suggested that the L/B ratio in ¹²⁴I PET is a useful tool for early detection of disseminated lung metastases and can be done as part of ¹²⁴I PET dosimetry. However, the L/B ratio should not be used alone due to overlapping scores. Despite the nondiagnostic quality of the attenuation CT, Freudenberg et al. state that an independent interpretation of the attenuation CT is important to improve the detection of disseminated lung metastasis on ¹²⁴I PET/CT. Additional methods to improve the detection of disseminated lung metastasis on ¹²⁴I PET/CT are needed.

¹²⁴I PET in ¹⁸F FDG PET-positive patients

A higher percentage of ¹⁸F-FDG PET-positive patients in the study cohort can decrease the ¹²⁴I PET positivity in patients; however, a positive ¹⁸F-FDG PET scan does not exclude positive ¹²⁴I PET uptake. The presence of FDG PET uptake is not mutually exclusive with the presence of ¹²⁴I PET uptake: 23/31 lesions in Freudenberg et al. (11), 23/76 lesions in Freudenberg et al. (28), 7/46 lesions in Gulec et al. (13), and 1/1 lymph node lesion in the study by de Pont et al. (3) were ¹²⁴I(+)FDG(+) of the true-positive lesions based on the study reference standard. In these 4 articles, a total of 54/155 lesions had both positive ¹²⁴I PET and ¹⁸F FDG PET uptake.

Based on the articles in this review, some patients with a negative ¹²⁴I PET scan and positive ¹⁸F FDG scan [¹²⁴I(−)/FDG(+)] may still be amenable to ¹³¹I therapy. Lee et al. (10) and Freudenberg et al. (28) showed that 4/4 patients with ¹²⁴I(−)/FDG(+) lesions treated with ¹³¹I alone achieved regression. Interestingly, the regression achieved was as long as 9–22 months. Therefore, ¹²⁴I(+)/FDG(+) or ¹²⁴I(−)/FDG(+) scans should not automatically preclude a further attempt at ¹³¹I treatment.

Future studies should be considered to compare the ¹⁸F-FDG SUVs in this subgroup of negative diagnostic radioiodine scans to help distinguish patients who may still benefit from ¹³¹I treatment with those who are truly radioiodine refractory. Moreover, future studies can evaluate the utility of both ¹²⁴I PET and ¹⁸F-FDG PET in providing additional diagnostic localization of metastasis in patients who have negative ¹²³I/¹³¹I diagnostic scans, and in helping to make ¹³¹I treatment decisions.