Targeted 11C–choline PET-CT/TRUS software fusion-guided prostate biopsy in men with persistently elevated PSA and negative mpMRI after previous negative biopsy

mpMRI has become the preferred method for detecting prostate cancer (PCa) foci after initial negative random biopsy (Schoots et al., 2015). In order to improve the use of mpMRI within urological and radiological community, the Prostate Imaging Reporting and Data System (PI-RADS) v. 1 and, recently, v. 2 score were established (Hamoen et al., 2015; Weinreb et al., 2016). Currently mpMRI is incorporated into national and international guidelines (https://www.nice.org.uk/guidance; http://www.uroweb.org) for PCa detection in men with prior negative biopsy findings.

Although the negative predictive value (NPV) of mpMRI is around 95%, some tumours may not be detected. In a preliminary study, 122 men treated with radical prostatectomy, preoperative mpMRI failed to detect GS ≥ 7 tumours and tumours >1 cm diameter in 28% of cases (Le et al., 2015). Filson C. et al. confirmed such data. They found that systematic biopsies revealed clinically significant PCa in 16% of men with no suspicion of prostate cancer at mpMRI biopsy (Filson et al., 2016). If biopsy is indicated on clinical grounds, a negative mpMRI, which could miss a third of significant prostate cancer (Futterer et al., 2015), should not preclude it. Furthermore, MRI may have absolute or relative contraindications and therefore, there is an unmet need to find alternative imaging for target biopsy.

Molecular imaging with PET is one of the most promising tools for the investigation of PCa (Scher et al., 2007; Hernández-Argüello et al., 2016). PET-guided prostate biopsy implementation could be not only plausible, but also mandatory in some subjects with contraindications to MRI or high clinical probability of prostate cancer, associated to negative MRI. Although radio-labelled 11C–choline PET studies have provided equivocal results, because of some overlap with benign prostate hyperplasia and prostatitis (Scher et al., 2007), the sensitivity for tumour foci in prostate gland can be high enough to complement MRI findings in the final diagnosis (Hernández-Argüello et al., 2016).

The aim of the current study was to assess the potential clinical impact of 11C–choline PET-CT/TRUS software fusion-guided prostate biopsy in men with persistently elevated PSA after negative biopsy and negative mpMRI.

Clinical data were part of a prospective on-going observational clinical study: “Diagnostic accuracy of target mpMRI/US fusion biopsy in patients with suspected prostate cancer after initial negative biopsy” approved by our Institutional Review Board (IRB) and Ethical Committee. Written informed consent was obtained from all individual participants included in the study. Patients with persistently elevated PSA, with or without ASAP and/or HG-PIN and negative DRE, after at least one negative biopsy (at least 12 cores for each biopsy course) and a negative mpMRI (PI-RADS v.2 < 3) or contraindications to MRI were included.

11C–choline PET/CT imaging technique

The radiopharmaceutical 11C–choline was synthesized using a General Electric TracerLab FXc module as previously described by (Pascali et al., 2000), and administered in a total amount of 250–400 MBq. Ten minutes after tracer administration, body axial images from the mid-thigh to the skull base were obtained with an integrated PET/CT Discovery 690, GE Healthcare. Low-dose CT images were acquired using an automated dose modulation (maximal 140 mA, 140 kVp), 64 × 3,75 mm collimation, 3.75 mm slice thickness, 0.5 s rotation time, and pitch 0,984:1. Emission images were obtained with a 3.0 min acquisition for bed position in the pelvis and 1.5 min for the remaining total body acquisition (axial FOV 1150 mm, matrix size 256 × 256) and subsequently reconstructed using an iterative algorithm (VUE Point FX, 3 iterations, 24 subsets). Reconstructed images of the pelvis were obtained separately and displayed for reading as sagittal, axial and coronal views on OsiriX MD Imaging workstation (Fig. 1). A correlation between PET/CT prostate volume and ultrasound prostate volume using the BK Medical, Analogic Ultrasound Group, Pro Focus, Transducer 8818, 6/9 MHz, was also performed. The ellipsoid formula was used (PV = p/6 x [length (cm)] x [width (cm)] x [high (cm)]). PET was considered positive when focal uptake of 11C–choline was superior to the background activity.

Of 298 consecutive patients enrolled from April 2015 to September 2016, 15 patients (median age 71 yrs. ± 8.89; tPSA 13.5 ng/ml ± 4.3) fulfilled the inclusion criteria: negative mpMRI (PI-RADS v.2 < 3). Twelve mpMRI were performed by 3-Tesla device and 3 by 1.5 Tesla machine with endo-rectal coil.

Biopsy

No significant complications due to biopsy were recorded. PCa was detected in 7/15 (46.7%). Over 58 cores, 25 (43.1%) were positive. In one patient PCa was located in the peripheral zone (Figs. 1 and 2), while 6 patients had PCa in transition zones. Four patients harboured a significant PCa: Gleason score ≥ 7. Three patients had a Gleason score 6. One patient showed a granulomatous prostatitis and the others a benign prostatic hyperplasia. A spatial correlation between PET-CT and topography of index lesions was found in a patient who underwent radical prostatectomy (Fig. 2).

MpMRI has become the preferred method for detecting cancer-suspicious regions in the prostate, especially in men with a previous negative biopsy, and with a few caveats, it showed to be capable of identifying many serious cancers (Cash et al. 2016). Currently there is robust evidence that MRI is substantially more accurate than US guidance, as the latter usually fails to reveal a valid target within the prostate. mpMRI information was used according three different schemes: cognitive-, in bore- and software assisted TRUS fusion-biopsy; the latter is currently adopted in many urological departments. However, the standardization and validation of mpMRI, including quality assurance in acquisition, reading, and reporting, operator skill and experience in drawing ROIs under a grey scale and performing the biopsy along with the non-negligible rate of false negative, leave room to identify other imaging strategies.

PET imaging has been recently considered one of the most promising approaches for PCa detection. The prerequisite for the diagnosis of primary PCa in PET is the presence of a high tumour-to-background ratio and the correlation of imaging findings with whole gland pathological samples. Although controversies exist on the accuracy of 11-Choline PET for primary detection of PCa (Souvatzoglou et al., 2011), 11C–choline PET scanning could be considered an alternative in all those men with persistent risk of harbouring PCa after a negative mpMRI. The idea of PET guidance for biopsies in men with elevated PSA-levels and a negative mpMRI is not new. In 2008 a study was published in which patients with persistent elevated PSA and previous negative prostate biopsy, were investigated with 18F–choline PET/CT to delineate PCa and guide a repeat biopsy (Igerc et al., 2008). However, they used a cognitive technique and not a software fusion biopsy. In a recent randomized prospective trial, Taverna et al. confirmed the poor accuracy of 3-T mpMRI associated with systematic cognitive biopsies (Taverna et al., 2016). One of the strengths of our approach is that all the patients underwent software assisted fusion biopsy.

Patients with an aggressive PCa (GS > 6) presented with a high tracer uptake, median SUVmax 5.45 and SUVratio to background 1.57, although no statistically significant difference was found. Piert et al. reported that uptake of 11C–choline was significantly increased in high GS lesions (≥4 + 3), with a MIB-1/Ki-67 labeling index ≥5% (p = 0.01) in comparison to lesions with GS 3 + 4 or lower (Piert et al., 2009). In this context, we found a high sensitivity, but a low specificity of 11C–choline PET/TAC at software assisted target biopsy.

We must acknowledge some limitations of our study. First of all, the sample size remains small. We did not perform a consensus re-reading of the mpMRI by different prostate imaging experts. The difference of SUVratio between positive and negative patients was not significant and no speculation about a potential SUVratio cut-off between clinically significant vs. indolent PCa can be done. The combination of mpMRI and PET would be superior to both modalities alone, but although it could improve the specificity, it would significantly increase costs.

In conclusion, notwithstanding we showed a “proof of concept” for software PET-CT/TRUS fusion-guided target biopsy, we found a low specificity for 11C–choline PET/CT although a good correlation with pathological outcome. Further studies with a larger population or new tracers such as PSMA could be investigated in order to improve the accuracy of the PET/CT fusion software assisted biopsy in a specific sub set of patients with a suspected primary PCa.