Approximately one third of prostate cancer patients experience biochemical relapse (rising PSA) following initial treatment [1, 2]. Differentiating between loco-regional recurrence and distant metastasis at an early time point is crucial to provide proper treatment. There is growing evidence for improved biochemical control rates when salvage treatment is initiated at low PSA levels [3]. The currently recommended medical imaging methods (MRI and bone scintigraphy) lack sensitivity for detection of nodal and skeletal metastases at low PSA levels [4]. Simultaneous PET/MRI has the potential to improve the detection accuracy in recurrent prostate cancer, since it combines the excellent soft-tissue contrast of MRI with the high molecular sensitivity of PET in a single imaging session. The aim of this exploratory study is to evaluate detection rate of simultaneous ¹⁸F-Fluciclovine PET/MRI for recurrent prostate cancer. The performance of combined PET/MRI will be compared to that of MRI-only, which is the current clinical standard.

Patients that fulfilled the European Association of Urology criteria for biochemical relapse following radical treatment (two consecutive measurements with PSA > 0.2 ng/ml following radical prostatectomy or PSA > 2.0 ng/ml above the nadir following definitive radiotherapy [5]), were examined with simultaneous ¹⁸F-Fluciclovine PET/MRI (3 T Biograph mMR, Siemens, Erlangen, Germany). An overview of the imaging protocol is given in Figure 1.

MR images were reviewed by an uro-radiologist and the combined PET and MR images were reviewed by a nuclear medicine physician. Suspicious findings based on MR alone and combined PET/MR were registered. In this initial phase of the study we report detection rates for MR, PET (elevated tracer uptake) and combined PET/MR (either suspicious findings based on MR or PET tracer uptake). Detection rates are calculated as the number of patients with suspicious findings divided by the total number of patients. The median PSA levels in groups of patients with and without suspicious findings were compared with a non-parametric median test.

So far N=16 patients with biochemical relapse following radical treatment for prostate cancer have been examined with combined ¹⁸F-Fluciclovine PET/MR. Patient characteristics of this initial cohort are listed in Table 1. The detection rate was 31.3% (findings in 5 of 16 patients) both for MR alone and for PET. The detection rate for combined PET/MR was 43.8% (findings in 7 of 16 patients). In 9 patients, there were no suspicious findings on MR or PET. Of the 7 patients with suspicious findings, 3 had suspicious lesions on MR that also had elevated tracer uptake (one in seminal vesicle remanences and two in lymph nodes). Two patients had lesions suspicious only on MR (prostate bed and lymph node), while two patients had suspicious PET lesions only (lymph nodes and prostate) (Table 2).

Figure 2 shows the distribution of PSA levels in the groups with and without suspicious imaging findings. Median PSA value for patients without suspicious findings was 0.4 ng/ml (range 0.2-3.3 ng/ml), not significantly different from median PSA value of patients with suspicious findings on PET/MR (median 1.1 ng/ml, range 0.3-12 ng/ml). The detection rate in patients with PSA<1 ng/ml was 37.5% (3 of 8 patients). One patient with PSA values as low as 0.3 ng/ml had a lesion with elevated tracer uptake (Figure 3). The patient with the highest PSA value (PSA=12 ng/ml) had no suspicious MR findings but had a small area with increased tracer uptake towards the base of the prostate, however this was considered a uncertain finding.

We have demonstrated that combined PET/MR can detect suspicious lesions in patients with low PSA levels. The overall detection rate in the 16 patients included so far in this study was 43.8%, which is in the same range as previously reported for ¹⁸F-Flucoclovine PET/CT [6] and ¹¹C-Choline PET-CT [7]. In two patients, the suspicious lesions were only detected by MR while lesions in two other patients only were suspicious based on ¹⁸F-Fluciclovine uptake. This could indicate that a combined PET/MR examination may improve local relapse or metastasis detection in patients with early biochemical prostate cancer recurrence following radical therapy, by offering complementary information. A limitation in this preliminary report is that there is no reference standard that can be used to categorize the findings as true of false positive. We aim to include 80 patients and to establish a reference standard based on PSA-levels, follow-up imaging, and lesion biopsies to define presence and absence of disease. Combined ¹⁸F-Fluciclovine PET/MR can detect areas suspicious for prostate cancer recurrence at low PSA levels.