The primary endpoint of this study was to determine the detection efficiency of [89Zr]Zr-PSMA-617 PET/CT with respect to suspicious lesions in a larger cohort of patients with BCR and negative conventional PSMA PET/CT.

Secondary endpoints were the magnitude of lesion uptake of [89Zr]Zr-PSMA-617 and contrast to background over time, as well as a comparison of values of related PET variables with each other and with those of [68Ga]Ga-PSMA-11 in the conventional scan for sites of positive [89Zr]Zr-PSMA-617 PET/CT findings. The remaining secondary endpoints were near-term safety of [89Zr]Zr-PSMA-617 PET/CT, i.e. adverse events or vital sign abnormalities observed during or shortly after the procedure that we considered to be related to [89Zr]Zr-PSMA-617 PET/CT and the follow-up after [89Zr]Zr-PSMA-617 PET/CT-based treatment (e.g. radiotherapy or surgical resection of the metastasis) including serum PSA levels and imaging (e.g. follow up PSMA PET/CT).

The cohort comprised 38 consecutive men with BCR, defined as increasing prostate-specific antigen (PSA) following primary (curative-intent) treatment, who from 25 October 2021–8 May 2023, underwent [68Ga]Ga-PSMA-11 PET/CT with negative findings, and then, within a short period [89Zr]Zr-PSMA-617 PET/CT. All imaging took place at Saarland University Medical Center. To be included, patients also had to have unaltered prostate cancer treatment during the interval between the conventional scan and the experimental scan. None of the patients showed clinical progression between [68Ga]Ga-PSMA-11 or [89Zr]Zr-PSMA-617 PET/CT.

Table 1 summarizes patient and imaging characteristics of the study sample. This cohort was middle-aged to elderly, with Gleason stage 7 disease in approximately 2/3. The PSA levels obtained at the time of [89Zr]Zr-PSMA-617 PET/CT were low (median [minimum–maximum] 0.52 [0.11–2.50] ng/mL). Primary treatment included prostatectomy in all patients. The analysis conformed to the Declaration of Helsinki and received ethics approval from the Institutional Review Board of the Ärztekammer des Saarlandes/Saarbrücken (approval number: 170/22, 13 September 2022). All patients provided written informed consent for [89Zr]Zr-PSMA-617 PET/CT and allowed publication of de-identified patient data.

[68Ga]Ga-PSMA-11 scans were acquired using a median (minimum–maximum) 148.5 (111–184) MBq of radiotracer. PET/CT was performed ~ 1 h post-infusion, following standard procedures [18]. Findings were classified visually by consensus among three nuclear medicine physicians who were experienced PET/CT readers (SE, FK, FR); medical histories and prior images were available to aid interpretation. Negative scans were defined as those lacking pathological uptake.

[89Zr]Zr-PSMA-617 PET/CT took place a median (minimum–maximum) 38 (5–126) d after [68Ga]Ga-PSMA-11 PET/CT. Scans with the novel tracer were performed 1-h, 24-h, and 48-h post-administration. The 1-h scan was acquired to allow direct comparison with the prior conventional scan. The 24-h and 48-h scans evaluated imaging that would be expected to yield appropriate images with the long-lived radionuclide. After intravenous infusion of a median (minimum–maximum) 123 (84–166) MBq of [89Zr]Zr-PSMA-617, immediately followed by 500 mL of NaCl 0.9%, whole-body PET/CT images, extending from the vertex to the mid-femur, were obtained using a Biograph mCT 40 system (Siemens Medical Solutions, Knoxville, TN, USA). PET acquisition time was 3 min/bed position for the 1-h scan, 4 min/bed position for the 24-h scan, and 5 min/bed position for the 48-h scan. For attenuation correction and anatomical localization, low-dose CT was performed at a 120-keV x-ray tube voltage. Tube current modulation with CARE Dose4D software (Siemens Healthineers, Erlangen, Germany) was employed, with 30 mAs as the reference. A soft tissue kernel (B31f/Be32) and a slice thickness of 5 mm (increment: 2–4 mm) were used to reconstruct data, which also were corrected for decay, randoms, and scatter. To reconstruct PET images, an iterative 3-dimensional ordered-subset expectation maximization algorithm (3 iterations; 24 subsets) was applied, and Gaussian filtering was carried out to a transaxial resolution of 5 mm at full width at half maximum. Respective matrix and pixel sizes were 200 × 200 and 3.0 mm. [89Zr]Zr-PSMA-617 PET/CT was performed on a compassionate use basis per the German Pharmaceutical Act § 13 (2b). Attending nuclear medicine specialists oversaw the procedure, including requisitioning the radiopharmaceutical, which was manufactured in-house [12].

[89Zr]Zr-PSMA-617 PET/CT findings also were classified visually by consensus, by the same readers as with the conventional scans. Again, since interpretation was made within everyday practice, medical history and prior images were accessible. Lesions were deemed to be suspicious for prostate cancer if they were discernible on the 24-h scan and/or the 48-h scan in typical sites of pathologic uptake of PSMA-targeted tracers. Scans were deemed to be negative if no pathologic foci were seen.

For each scan (1-h, 24-h, 48-h post-injection), key PET variables reflecting radiotracer uptake and contrast were determined for each suspicious lesion on [89Zr]Zr-PSMA-617 PET/CT. To assess uptake, the maximum standardized uptake value (SUVmax) was measured using SyngoVia Enterprise VB 60 software (Siemens Healthineers, Erlangen, Germany). For comparison, uptake was measured at the corresponding site in the early imaging of [89Zr]Zr-PSMA-617 and [68Ga]Ga-PSMA-11 PET/CT, even though no lesion could be clearly identified at this site.

To characterize contrast, the tumor-to-muscle ratio (TMR), tumor-to-liver ratio (TLR), tumor-to-contralateral region ratio (TCR), and tumor-to-background ratio (TBR) were calculated. These variables were defined as the SUVmax of the lesion divided by the mean SUV (SUVmean) of the respective comparator. SUVmean was determined in spherical volumes-of-interest in the comparator location, i.e., gluteal muscle for TMR, central liver for TLR, the contralateral region for TCR, and the directly surrounding region for TBR.