Combined Forced Diuresis and Late Acquisition on [68GA]GA-PSMA-11 Pet/CT for Biochemical Recurrent Prostate Cancer: A Clinical Practice-Oriented Study

In the last decade, the clinical management of prostate cancer (PCa) patients with biochemical recurrence (BCR) after treatment with curative intent has been revolutionized by the introduction of PSMA PET/CT, which became the first-choice imaging method for restaging. Among the numerous ligands developed for PSMA-targeted imaging, [⁶⁸Ga]Ga-PSMA-11 has shown great promise in the setting of BCR, having high diagnostic accuracy and impact on clinical management.

However, identifying the site of recurrence with [⁶⁸Ga]Ga-PSMA-11 is partly limited by the tracer's urinary excretion, which may reduce the detection rate in patients having a residual disease in the prostate bed or in peri-ureteral nodes. A further potential shortcoming of [⁶⁸Ga]Ga-PSMA-11 urinary excretion is the inconsistent PET/CT scan interpretation, particularly for readers with lower levels of expertise. Indeed, in a multicentric study evaluating the [⁶⁸Ga]Ga-PSMA-11 PET/CT interobserver variability, readers with low experience showed only moderate agreement for the local restaging compared to nodal and distant metastases.¹ As a consequence, detecting local tumours can be challenging due to signal overlay by the excreted tracer.

In the latest version of the joint EANM/SNMMI procedural guidelines for [⁶⁸Ga]Ga-PSMA-11 PET/CT, the administration of Furosemide (20 mg) is eventually recommended shortly before or after tracer injection to improve the visualization of peri-ureteral and peri-bladder tissues.² Similarly, increased lesion detection has been reported with abdominopelvic delayed imaging up to 180 min post-injection (p.i.).² This increased accuracy has been related to an improvement in identifying lesions close to the ureter or the bladder or lesions with slower tracer accumulation due to low PSMA expression.² However, protocols for furosemide-induced forced diuresis and delayed abdominopelvic imaging and, more importantly, their combination is currently not standardised in the clinical setting. In particular, the added value in terms of diagnostic accuracy, reader's confidence, and interobserver agreement of a [⁶⁸Ga]Ga-PSMA-11 PET/CT protocol combining these procedures are still unknown. Given the practical implications associated with implementing this demanding protocol in clinical practice, it would be highly relevant to disclose specific clinical scenarios that might benefit from it.

We thus prospectively tested the systematic addition of forced diuresis late abdominopelvic acquisition after the standard [⁶⁸Ga]Ga-PSMA-11 PET/CT acquisition protocol. The study aims to verify the added value of this imaging protocol in terms of i) diagnostic accuracy, ii) readers’ confidence and iii) interobserver agreement among six PET readers with various levels of expertise.

One hundred prospectively recruited BCR patients were restaged with dual-phase [⁶⁸Ga]Ga-PSMA-11 PET/CT. Figure 1 summarizes our imaging protocol. Patients received 2 MBq/kg of [⁶⁸Ga]Ga-PSMA-11 followed by standard 60 min p.i. imaging. At 140 min p.i., all patients received 20 mg of Furosemide intravenously, followed by an acquisition of the pelvis and lower abdomen at 180 min p.i.

T phases.jpg

Figure 1: Study Protocol

Six PET-readers were classified as having low (<30 prior studies, n=2), intermediate (30-300 studies, n=2), or high level of experience (>300 studies, n=2) for [⁶⁸Ga]Ga-PSMA-11 PET/CT reporting.¹ All readers, blinded to reference standards and other diagnostic imaging procedures, rated standard acquisitions alone and standard plus forced diuresis late acquisitions in a stepwise fashion according to E-PSMA guidelines.³ Readers were asked to indicate their personal confidence to quantify the likelihood of the presence of PCa using an unstructured five-point confidence scale (from 1 = certain benign to 5 = certain pathological) as well as the PSMA reporting and data systems (PSMA-RADS) according to current guidelines.³ Readers had full access to patient history, PSA levels, PSA kinetics, and the pre-test probability assessment.

Additional forced diuresis late-phase abdominopelvic imaging slightly (though not significantly) increased the [⁶⁸Ga]Ga-PSMA-11 PET/CT diagnostic accuracy compared to the standard protocol in the restaging of PCa. However, it significantly improves the diagnostic accuracy for local uptakes rated by low-experienced readers (Table 1) and for nodal uptakes, particularly when rated as uncertainly pathological at standard imaging. Indeed, at the per-lesion analysis, forced diuresis late-phase imaging significantly increased nodal accuracy from 78.3% to 84.7% (p < 0.01), while its effect did not reach significance for local accuracy, which increased from 78% to 81.3% (p = 0.15).

Table 1: Changes in Diagnostic Accuracy According to the Reader’s Previous Experience

Obtained data showed that SUVmax kinetics may potentially guide the dual-phase [⁶⁸Ga]Ga-PSMA-11 PET/CT interpretation. Indeed, when SUVmax variation from standard to late forced acquisition imaging (ΔSUVmax) was included in uni- and multivariable logistic regression model containing standard imaging semiquantitative parameters indicated by the E-PSMA reporting guidelines (SUVmax, PSMA score, PSMA-RADS) in the prediction of PCa recurrences, it significantly and independently predicted the presence of local and nodal PCa recurrences in addition to the remaining semiquantitative imaging parameters.³

The dual-phase [⁶⁸Ga]Ga-PSMA-11 PET/CT imaging protocol increased the reader's confidence regardless of the prior experience. The mean overall confidence increased after reading standard + forced diuresis late-phase scans compared to standard acquisition regarding both the local and nodal status (both p<0.0001). The increase in the reader’s confidence level was confirmed for both local and nodal lesions when readers were grouped according to their previous experience. In both cases, readers with lower levels of expertise took the higher advantage of forced diuresis late-phase imaging. Interestingly, the analysis of PSMA-RADS changes did not reproduce the same findings observed considering the unstructured five-point confidence scale. Indeed, PSMA-RADS system showed less sensitivity in tracking the increase in reader’s confidence compared to confidence categories. Perhaps, the unstructured confidence assessment proposed by the present study had the advantage of more flexibility and allowed the PET reader to consider uptake kinetics from early to late-phase imaging, which is not included in PSMA RADS classes. Given the endorsement of dual-phase [⁶⁸Ga]Ga-PSMA-11 PET/CT imaging by the joint EANM/SNMMI guidelines, it might be helpful to add specific criteria to include PSMA uptake kinetics interpretation in the future PSMA-RADS v1.0 updates to overcome this limitation.²

Finally, the tested imaging protocol increased the interobserver agreement in identifying nodal recurrences. Agreement among observers for local restaging was coherent with previous literature and remained unaltered when forced diuresis late-phase imaging was added to standard imaging. By contrast, the addition of forced diuresis late-phase imaging increased the agreement among readers in identifying nodal recurrences.

In conclusion, the present data do not support the systematic use of a protocol based on standard acquisition followed by combined forced diuresis late-phase imaging in the clinical setting, considering that it is demanding for a nuclear medicine department. However, it allowed the identification of clinical scenarios that might benefit from it (i.e., when PSMA PET/CT is reported by low-experienced readers or when uncertain nodal uptakes are observed at the standard scan).

Written by: Matteo Bauckneht, MD, PhD, Nuclear Medicine, IRCCS Ospedale Policlinico San Martino, University of Genova, Genova, Italy

References:

Read the Abstract

Login