The present study describes our approach to diagnosis and treatment of complex prolactinoma cases using [18F]FET-PET/MRICR, demonstrating its added value for identification of remnants and thereby aiding in shared decision-making and surgical planning in selected patients with prolactinoma. [18F]FET-PET/MRICR was able to identify lesions as potential targets for TSS in patients in whom conventional MRI failed to localize (the extension of) the lesion, and aided in assessment of the likelihood of total resection.

Recently, the landscape of prolactinoma treatment has changed, from DAs being the primary treatment to TSS being considered a viable first-line alternative in non-invasive tumors [3, 5,6,7, 29]. Optimal visualization of tumor (remnants) is key for selecting surgical candidates. In the setting of an experienced team, functional imaging has added value for complex patients, in whom the lesion is not readily detectable using conventional MRI [9, 12, 13, 17, 20, 30]. [11C]MET-PET/MRICR is the technique most frequently used, whereas [18F]FET-PET/MRICR is an attractive alternative since no on-site cyclotron is required. Yet, limited data on [18F]FET-PET/MRICR for this indication are reported in literature [24, 25].

The purpose of the present study was therefore to determine if [18F]FET-PET/MRICR is useful in the management of complex prolactinoma cases. [18F]FET-PET/MRICR was performed for two indications: (1) for additional information on a potential target lesion, or (2) for radiological diagnosis when conventional MRI failed to identify a lesion. [18F]FET-PET/MRICR identified a positive lesion in 82% of patients, among which four out of six patients with negative conventional MRIs. Moreover, [18F]FET-PET/MRICR results influenced clinical decision-making in 88% of cases. Eight patients underwent TSS, confirming [18F]FET-PET/MRICR findings, and achieving the surgical goal in 86%. [18F]FET-PET/MRICR was inconclusive due to postoperative inflammation in one patient (#10), and failed to identify a lesion in two patients (#7, 9). Thus, [18F]FET-PET/MRICR was useful in most cases, aiding in clinical decision-making.

To create optimal circumstances for adenoma localization, DAs were withdrawn in advance. Based on the half-life of cabergoline, a withdrawal period of > 4 weeks was previously recommended [13]. No previous recommendations were made for quinagolide or bromocriptine. However, a period of > 1 week is advisable – based on their half-life and biological properties – provided the prolactin level has increased sufficiently (≥ 2xULN) after withdrawal. The required withdrawal period may, however vary based on the duration and dose of DA treatment, with highly responsive tumors that have shrunk considerably requiring a longer period. Furthermore, timing of postoperative functional imaging is complex. In the absence of evidence, we would advise waiting for ≥ 3 months if clinically feasible, as diffusely increased uptake in the sphenoid region was observed ten weeks postoperatively, impacting interpretability. Generally, clinical circumstances should be optimized to improve chances of identifying an active lesion.

Several remarkable observations need addressing. Firstly, even patients with only subtle prolactin level elevations (i.e. 1.0-2.5xULN) showed [18F]FET uptake, and this technique, therefore, seems to be highly sensitive for intrasellar lesions. Secondly, one patient with a negative [18F]FET-PET/MRICR showed signs of a chronic hypophysitis upon histopathological evaluation rather than prolactinoma tissue. Therefore, the preoperative diagnosis can be questioned in this patient. Thirdly, even though [18F]FET-PET/MRICR was performed in only those patients in whom TSS was seriously considered, 3 patients decided not to proceed with TSS after an operable lesion was identified. Hence, an algorithm needs to be developed to prevent unnecessary functional imaging, which is time-consuming and increases costs. We believe an experienced MDT overseeing the full trajectory is highly important for careful repeated counseling, and optimal use of this functional imaging modality.

In the present study, [18F]FET-PET/MRICR and [11C]MET-PET/MRICR showed good correspondence in uptake pattern in the two patients undergoing both examinations. Comparing the [18F]FET and [11C]MET tracers, a clear benefit of [18F]FET is its longer half-life, enabling its use in centers lacking a cyclotron. Moreover, [18F]FET-PET/MRICR may be more sensitive for small intrasellar lesions with low metabolic activity, as the normal pituitary tissue takes up less [18F]FET compared to [11C]MET [25]. However tracer wash out in the CS, as was observed in acromegaly [25], may hamper the interpretation of [18F]FET-PET/MRICR in cases with parasellar extension – particularly when the uptake is moderate (notably not worsening outcomes in the present study). While awaiting larger cohorts assessing the sensitivity of [18F]FET-PET/MRICR in the parasellar region, [18F]FET-PET/MRICR should be used cautiously in cases with possible CS involvement, especially in centers with limited experience with functional imaging.

[18F]FET-PET/MRICR uptake ratios – indicative of metabolic activity of the lesions – were not correlated to serum prolactin levels. For comparison, [11C]MET-PET/MRICR uptake ratios were analyzed in our recently published cohort of patients with a prolactinoma, showing no correlation to serum prolactin levels either (p = 0.443, r = 0.193) [13]. These findings opposed earlier studies [17, 20], which reported a linear relationship between methionine uptake and prolactin levels, although the statistical methods used were not reported. The present findings may be explained by the fact that our cohort consisted of complex cases with pretreated remnants. Additionally, in case of [18F]FET-PET/MRICR, wash out of tyrosine may play a role. Dynamic studies in larger populations with treatment-naïve adenomas should be performed for more adequate analysis of the relationship between tyrosine uptake and prolactin levels.

In this small study in which we share the results of complex cases involving [18F]FET-PET/MRICR, descriptives rather than predictive statistics were used. Not all [18F]FET-PET/MRICR results could be verified, as not all patients underwent TSS. Furthermore, it should be stressed that not all conventional MRIs preceding functional imaging were completely negative, yet were inconclusive due to uncertainty about the localization or extension of the lesion in patients who were pretreated. The discovery rates of [18F]FET-PET/MRICR in patients with a negative conventional MRI may therefore be lower.

In conclusion, [18F]FET-PET/MRICR may be of added value for assessment of prolactinoma remnants and their extensions, or to localize prolactinomas unidentifiable with conventional MRI, particularly when no on-site cyclotron is available.