145 consecutive patients, evaluated at our institution with PSMA-PET/CT (n = 68) or -PET/MRI (n = 77) for BCR between May 2014 and January 2017 were retrospectively analyzed. Initially, all patients were supposed to undergo PET/MRI. Nevertheless, patients with metal implants in the pelvic region were shifted to PET/CT due to unfavorable image quality and unknown effects on attenuation correction. Also, patients with implants not suitable for a 3-Tesla (T) system, claustrophobia and/or pain were shifted to PET/CT. A total of 28 patients were excluded due to initial treatment with RT of the prostate, leaving 117 patients for final analysis.
All patients were treated with RP according to guidelines recommendation. All surgical specimens were processed according to standard pathologic procedures, staged with the AJCC TNM classification and graded with the WHO/ISUP 2005 grading system .
BCR was defined as two consecutive PSA rises above 0.2 ng/ml. Follow-up was generally every 3 months for the first 2 years, then semiannually until the fifth year, then annually.
All reported investigations were conducted in accordance with the Helsinki Declaration and national regulations. The study was approved by the local Ethics Committee (permit 1440/17). [68Ga]Ga-PSMAHBED-CC conjugate 11-PET was produced and injected according to a compassionate use during the conduct of a prospective clinical trial (EudraCT: 2014–004758-33; Clinicaltrials.gov Identifier: NCT02659527).
Imaging protocol and analyses
PET/MRI was performed on a Biograph mMR (Siemens, Germany), capable of simultaneous data acquisition, consisting of a MRI-compatible PET detector integrated in a 3.0-T whole-body MRI scanner. The PET component uses a 3-dimensional (3D) acquisition technique and offers an axial field of view (FOV) of approximately 23 cm and a transversal FOV of 45 cm with a sensitivity of 13.2 counts per second/kBq.
Local PET of the pelvis comprised a 10-min listmode acquisition, starting 60 min after injection. Partial body PET (skull base to thigh) was performed with 4 bed positions, with a 4-min sinogram mode each. Reconstruction parameters for PET were: 3 iterations/21 subsets; summation of the 10-min pelvic acquisition for visual and semiquantitative analysis. MRI-based attenuation correction was applied using DIXON-VIBE sequences comprising in- and opposed-phase as well as fat- and water-saturated images.
The integrated 3-T MRI is performed with the following sequences and parameters: pelvis: T2w turbo spin echo (tse) axial: matrix size: 512 × 512, in-plane resolution: 1.1 × 0.8 x 5 mm; FOV: 263 × 350 mm; TR: 3600 ms; TE: 103 ms. T1w turbo spin echo (tse) coronal: matrix size: 384 × 384, in-plane resolution: 1.0 × 0.9 × 5 mm; FOV: 221 × 350 mm; TR: 600 ms; TE: 12 ms. Diffusion-weighted imaging (DWI): matrix size: 192 × 192, in-plane resolution: 2.6 × 2.0 × 5 mm; FOV: 285 × 380 mm; b-values: 0.600 s/mm2; TR: 9200 ms; TE: 85 ms. Partial-body MRI simultaneous with PET: T2w HASTE: matrix size: 256 × 256, in-plane resolution: 1.56 × 1.5 × 6 mm; FOV: 380 × 380 mm; TR: 1400 ms; TE: 121 ms. T1 VIBE matrix size: 195 × 320, in-plane resolution: 1.6 × 1.2 × 3 mm; FOV: 309 × 380 mm; TR: 4.56 ms; TE: 2.03 ms. Sagittal spine sequences after PET: T1 tse: matrix size: 320 × 320, in-plane resolution: 1.4 × 1.1 × 3 mm; FOV: 263 × 350 mm; TR: 666 ms; TE: 9.6 ms. T2 STIR: matrix size: 320 × 320, in-plane resolution: 1.4 × 1.1 × 3 mm; FOV: 263 × 350 mm; TR: 3500 ms; TE: 43 ms.
Sixty minutes before PET/MRI acquisition start, patients received an intravenous injection of of 2 MBq/kg body weight [68Ga]Ga-PSMAHBED-CC conjugate 11 intravenously. For improved image quality, forced diuresis with 20 mg of furosemide was applied intravenously before the PSMA application and all patients in PET/MRI received a bladder catheter.
PET/CT was performed, from the vertex to the upper thigh, using a 64-row, multi-detector hybrid system (Biograph TruePoint 64; Siemens, Erlangen, Germany), with an axial FOV of 216 mm, a PET sensitivity of 7.6 cps/kBq and a transaxial PET resolution of 4–5 mm (full-width at half-maximum, FWHM). PET was performed 90 min after an intravenous administration of 2 MBq/kg body weight [68Ga]Ga-PSMAHBED-CC conjugate 11 with 4 min/bed position, four iterations per 21 subsets, a 5-mm slice thickness, and a 168 × 168 matrix, using the point-spread-function (PSF)-based reconstruction algorithm TrueX. CT maps were used for PET attenuation correction. Venous-phase CE-CT was obtained after the intravenous injection of 100 ml of a tri-iodinated, non-ionic contrast medium at a rate of 2 ml/s; a tube voltage of 120 mA, a tube current of 230 kV, a collimation of 64 × 0.6 mm, a 3-mm slice thickness at a 2-mm increment, and a 512 × 512 matrix.
PET/CT and PET/MRI reviews were separately assessed by an experienced reader for prostate hybrid imaging (MH), and CT/MRI alone was also separately assessed by an uro-radiologist (PB). For MRI and CT alone, AGFA IMPAXX EE software was used; for PET/MRI and PET/CT, Hermes Hybrid 3D (Hermes Medical Solutions Stockholm) was used.
Radiologic assessment was performed according to RECIST 1.1 criteria. For hybrid assessment, a focal uptake above the surrounding background in a morphologically visible structure (e.g. lymph node any size or bone) or soft tissue in the prostate bed as well as corresponding areas of restricted diffusion capacity in MRI DWI was assessed as a positive finding. Known “false” positive lesions such as slight focal uptake in the area of the paravertebral sympathetic ganglia were ignored . We considered as confirmation of the lesions either a histologic assessment such as biopsy or salvage lymphadenectomy as well as a PSA decline when radiation therapy was directed to the suspicious area. In case of PET/MRI, a positive corresponding diffusion restriction was also taken as a non-invasive confirmation of an area with potential higher cellular density and most likely tumoral tissue. Additional lower values for the apparent diffusion coefficient (ADC) strengthened the confirmation (Fig. 1).
A standard clinical interdisciplinary tumorboard with the unblinded hybrid imaging results in the mentioned observational period was available for each patient included in this study. Additionally, we retrospectively performed a condensed tumorboard, where only the clinical history, CT/MRI and bone scan results of the respective anonymized patients were presented. The study tumorboard consisted of experts in urology/uro-oncology, pathology and radiation oncology. The real therapy was blinded at the time of discussion. The majority decision of the retrospective tumorboard, blinded to hybrid imaging, was then compared to the real therapies, which were given to the patients according to the institutional tumorboard decision taken out of the clinical records.
Descriptive statistics of categorical variables focused on frequencies and proportions. Means, medians and interquartile ranges (IQR) were reported for continuously coded variables. The Mann-Whitney U test and chi-square test were used to compare the statistical significance of differences in medians and proportions, respectively. Univariable logistic regression was used to assess the predictive ability of PSA for PSMA-PET positivity. Statistical significance was considered at p < 0.05. All tests were two-sided. Statistical analyses were performed using STATA v.14.1 (StataCorp LP, College Station, TX, USA).