Early acquisition of [18F]FDOPA PET/CT imaging in patients with recurrent or residual medullary thyroid cancer is safe—and slightly better! – European Journal of Hybrid Imaging

Our study demonstrates that MTC lesions on the early acquisition (15 min p.i.) are significantly more [¹⁸F]FDOPA avid than lesions on the late acquisition (60 min p.i.) with a median WR of − 33%. This confirms the results of previous studies (4–6), which also show a higher intensity of MTC lesions on early imaging. However, although lesions were more radiotracer avid on early acquisitions and 3 more TP lesions were found, we found no statistically significant difference in the detection rate of lesions between early and late acquisition. TP lesions that were only seen on the early images were all localized to cervical lymph nodes, suggesting that the rapid uptake and washout of [¹⁸F]FDOPA in MTC lesions, among other things, could be dependent on anatomical location. However, this hypothesis is not consistent with the study by Soussan et al. (2012), and in the current study, the range of WR in the lesions of the cervical lymph nodes is quite wide (− 41% to + 25%), and without a clear tendency specific for this anatomical site (Additional file 2: Table S2). This fits well with our statistical analysis, which shows no significant difference in WR between different anatomical sites of lesions. Thus, it can be concluded that location of MTC lesions has no significant impact on WR in our study. In a single patient, an FP lesion was detected only on the early acquisition (Fig. 2). The reference standard in this case was a biopsy, which described the lesion as a lymph node with reactive changes in the form of severe follicular hyperplasia and without malignancy-suspected cells. As mentioned in the EANM practice guideline for PET/CT in MTC (Giovanella et al. 2020), [¹⁸F]FDOPA uptake may in rare cases be due to inflammation which we assume is the reason for the activity uptake in our FP finding.

Several studies have investigated the value of [¹⁸F]FDOPA PET/CT in the detection of recurrent MTC, and as previously mentioned, these vary from a few minutes up to approx. 90 min in acquisition times (Taralli et al. 2020; Treglia et al. 2013; Soussan et al. 2012; Golubić et al. 2017; Caobelli et al. 2018). However, only a few studies have made a comparison of at least two acquisition times (Taralli et al. 2020; Treglia et al. 2013; Soussan et al. 2012) to obtain a time estimate for the most optimal (maximal) [¹⁸F]FDOPA uptake p.i., and only one of these is a dynamic study (Taralli et al. 2020). Among those studies, only the study by Treglia et al. (2013) reports similar acquisition times (15 and 60 min p.i.) as in our study, although they do not provide data on the exact acquisition times actually used or the variation of these between subjects. However, the WR observed by Treglia et al. (2013) is comparable to our result (mean WR − 28% ± 13% versus median WR − 33% (range − 57 to + 50%)). Soussan et al. (2012) observed a mean WR of − 40% but with widely varying acquisition times, especially the late acquisition which was obtained with a median time of 94 (range 70–150) min, making it difficult to compare with our standardized protocol. Standardized acquisition times are not only essential in the comparison of different studies; in treatment monitoring, where any disease progression or regression must be assessed, the scans should be performed with the same acquisition time protocol in order to be directly comparable. The dynamic study by Taralli et al. (2020) observed a median WR of − 41% (range − 88% to + 10%), which is not far from our WR value. However, their study is not directly comparable to ours both because the acquisition times are given in time intervals (2–5 min, 5–10 min, and 40–45 min, respectively) and because the early and late times do not match our exact times of 15 and 60 min, respectively.

In a clinical context, it is desirable to perform only one scan and the question is whether to do this early or late. In our study, none of the PET scans visualized lesions on the late acquisition that were not already detected on the early acquisition. Therefore, it can be assumed that no MTC lesions would have been missed if only the early acquisition had been performed. In previous studies (Taralli et al. 2020; Treglia et al. 2013) as well as in the present, no statistically significant difference was observed between the two acquisitions in the detection of MTC lesions on a patient/scan basis, indicating that the outcome for patients is generally the same whether only the early or only the late acquisition is performed. However, although the choice of either early or late acquisition did not have a clinical impact on patient outcome in our study, the scan-based median WR of -33% indicates (1) that it is safe to omit the late acquisition, and (2) that image acquisition, reading and reporting may in fact be simplified by doing so. In the scanning example in Fig. 5, the early acquisition visualizes two lesions in cervical lymph nodes that are not visible in the late image. In addition, activity uptake in three other lesions (thoracic lymph nodes) is seen to persist, albeit reduced, from early to late acquisition. These three lesions might as well have had a higher WR and have been washed out on the late image, as well as the two cervical lymph node lesions. Therefore, if no early imaging had been performed in such a case, the patient would have been visually declared free of MTC lesions. If only a late acquisition is made, a situation may therefore potentially arise where FN results are obtained.

Other studies have indicated that the early acquisition may be performed very shortly after radiotracer injection. Thus, Taralli et al. (2020) observed a rapid maximal [¹⁸F]FDOPA uptake after only a few minutes. Such short interval between injection and scan may simplify patient logistics considerably.

Our study is the first to apply whole-body acquisition in both the early and late phases, providing us the opportunity to analyze WR for several different anatomical sites in addition to the neck and thorax, which is a strength of the study. As for the study limitations, some may see it as a drawback that multiple scans from the same patient are included as separate cases, which could potentially lead to data distortion. Patients may have different biological behaviors in their metastases and thus different WR of [¹⁸F]FDOPA. A patient with several scans could therefore affect the overall result more, than a patient who has only had one scan. In our study, however, we do not consider this as a limitation, as the WR variation in general is wide. When looking at the same type of anatomical lesion site, this variation applies both interindividually and intraindividually when comparing scans from the same patient. We therefore do not believe that results from one patient with multiple scans are the basis for bias in our study. A second limitation of the study is that local tissue perfusion and blood volume may have an impact on the individual lesion washout rate. Clinically, however, it is difficult to deal with this limitation, as the perfusion can not only vary from patient to patient, but also between the individual patient’s visits in case of multiple scans. Thirdly, another limitation could be sample size. As mentioned, MTC is a rare type of cancer, and the previously mentioned studies (Taralli et al. 2020; Treglia et al. 2013; Soussan et al. 2012) include a sample size of between 15 and 21 scans. We therefore consider our sample size of 27 dual-phase scans to be relatively large in a rare type of cancer.