⁶⁸Ga-PSMA 11 ligand PET imaging in patients with biochemical recurrence after radical prostatectomy – diagnostic performance and impact on therapeutic decision-making

Prostate cancer (PC) is responsible for 11% of all cancer-related deaths [1]. Despite attempted local therapy with curative intent, up to 40% [2] of the patients experience disease recurrence. Most of these patients have prostate-specific antigen (PSA)-only recurrence without any clinical evidence of metastases [3]. Management of patients with PSA-only recurrence [i.e., biochemical recurrence (BCR)], is a clinical challenge as some may have localized disease only potentially benefiting from local therapy. Avoidance of overtreatment with its unnecessary side effects while ensuring effective durable disease control remains dependent on identification of the site of disease recurrence [4]. Indeed, early salvage treatment can confer durable local disease control, prolonging survival [5‐7]. Patients with BCR after radical prostatectomy (RP) with curative intent are often offered salvage radiotherapy (RT) alone, androgen deprivation therapy (ADT) alone or a combination of these two [3]. Salvage RT is most effective when PSA levels are below 0.5 ng/ml [8‐10]. At this threshold, the actual imaging techniques are limited in their sensitivity to differentiate local versus distant recurrence. Contrast-enhanced computerized tomography (CT) and bone scan (BS), the current standard diagnostic tests for staging of PC, have, unfortunately, inadequate sensitivity for the detection of PC in PSA ranges below 10 ng/ml [11].

The development of molecular imaging and the use of specific target probes, like choline-based positron emission tomography (PET), has already improved the diagnostic accuracy in patients with BCR after RP [12, 13]. Although the detection rates markedly increased when applying choline-based hybrid imaging [14, 15], the need for more accurate and earlier detection of metastases at lower PSA levels is necessary for the tailoring of individualized salvage therapies. In this context, [⁶⁸Ga]Ga-PSMA^HBED-CC conjugate 11 (PSMA) has emerged as a promising radiopharmaceutical [16‐18], albeit still not approved and, therefore, considered experimental [19]. Previous studies have shown that PSMA-PET imaging can detect lesions at lower median PSA levels. However, the verification of these lesions has not been assessed to our knowledge. Through verification, one can ensure that these lesions are not false positives and that all lesions have been detected. Moreover, we hypothesize that PSMA-PET demonstrates high sensitivities also at lower median PSA levels and contributes a major impact on the clinical decision-making process. Unfortunately, none of the previous reports, to our knowledge, addressed the clinical implications of this technology in the clinical decision-making regarding therapy. The aim of this study was to evaluate the diagnostic performance of PSMA-PET in the early detection of metastases in patients with BCR at lower PSA levels after RP for clinically non-metastatic PC and to assess its impact on therapeutic decision-making compared to standard imaging with CT/MRI.

Treating PC patients with BCR after RP, which is defined as a PSA value of >0.2 ng/ml, is a common but difficult situation. Differentiation of local versus systemic disease and its targeted therapy with the best long-term outcome remains an unmet need. While some patients may benefit from early systemic therapy, others may benefit from local therapy with curative intent. Metastasis-directed therapies are increasingly considered to improve oncologic success while lowering unnecessary side effects. To obtain better outcomes, such therapies need to be administered at lower PSA values while tumor burden is presumably low [5, 8]. Therefore, there is a rising demand for reliable and accurate diagnostic tools. Unfortunately, the detection of metastases at especially low PSA values is not reliable enough with the current guideline-recommended imaging modalities, although choline-based PET [11, 14, 15] has already markedly improved the diagnostic performance of imaging in this setting. In this context, PSMA-PET has emerged as a promising, more accurate method. It has, indeed, been investigated for the staging of recurrent PC after definitive local treatment in retrospective studies [16‐18, 22, 23]. However, these studies did not consider all the various issues in the management of these patients. To address this unmet need, we wanted: first to confirm/validate the diagnostic performance of PSMA-PET in patients with BCR after RP; second, to compare the diagnostic performance of PSMA-PET to the current standard imaging in this setting; third, to verify the lesions detected by PSMA-PET as true; and four, to assess how often the PSMA-PET changed the clinical decision-making regarding therapy.

We confirmed the high diagnostic performance of PSMA-PET in patients with BCR after RP. The overall detection rates in retrospective studies published before [17, 18, 23] ranged from 82.2% to 89.5%. Our data are congruent with these mentioned with an overall detection rate of 85.5% at a comparatively lower median PSA level at the time of imaging of 1.04 ng/ml, showing a higher detection efficacy than those published for all known choline-based PET tracers [13, 24, 25]. Our detection rates were from 65% for a PSA value of 0.2 to <0.5 ng/ml up to 100% for PSA values ≥2 ng/ml. As ADT might have an unknown influence on PSMA expression at such early stages of recurrence, we put emphasis on a highly selected and homogenous patient cohort, as we included patients with BCR after RP only, of which a majority (59%) already had postoperative RT. We excluded patients with ongoing ADT as a difference to published data [18]. Nevertheless, it has to be mentioned that an overall detection rate of 85.5% still implies a false negative rate of 14.5%, as it is assumed that any rising PSA level after a nadir of <0.2 ng/ml is representing cancer recurrence.

We found that PSMA-PET detected lesions exclusively with no morphologically suspicious correlate in a majority of the patients. Similarly, Eiber et al. [18] reported that 33% of all the lesions were observed in PSMA-PET only without a suspicious correlate in CT, and that in 25% of the cases, PSMA-PET revealed more lesions than CT. Although this is in line with our data, it seems surprising, as we had significantly more (67%) lesions positive in PSMA-PET only without a suspicious RECIST 1.1 correlate in CT or MRI. Considering the limited role of standard imaging at low PSA levels [11], a reason might be the comparatively lower median PSA in our study, which is nearly half as high as in the study of Eiber et al. This is also true for PSMA-negative findings being assessed as RECIST-positive on CT or MRI that did not occur in our population.

Within our cohort, we found that exact histopathology as a gold standard of proof for PSMA-PET-positive lesions was only available in 16 (16%) patients undergoing salvage lymphadenectomy. However, the observed substantial decrease of PSA in the 45 (45%) patients undergoing PSMA-PET/CT(MRI)-guided RT also provided a regional confirmation. Altogether, 41 of the patients with metastases-directed therapies due to PSMA-PET demonstrated only a slight increase of the median PSA after 6 and 12 month from 0.12 ng/ml to 0.37 ng/ml. This implies a suitable approach for extending therapeutic windows. In 18 (18%) of the remaining 39 patients, a simultaneous PET/MRI was performed demonstrating a correlate with significant diffusion restriction in the PSMA-positive lesion. Although ADC maps are prone to artifacts and patients movement, in 12 (12%) of these patients, a significantly lower ADC signal, showing higher cellularity in the respective area, confirmed the results to a certain extend [26]. Although this is not a histological confirmation, combined PSMA-PET/MRI might have the ability to add diagnostic certainty. In local recurrence, [¹¹C]choline PET/MRI demonstrates superior detection rates compared to PET/CT due to the multiparametric MRI protocol [27]. Another study by Freitag et al. using PSMA-PET/CT vs. PSMA-PET/MRI in the evaluation of lymph node metastases demonstrated a significantly higher visibility of PSMA-positive lymph node correlates in MRI-DWI [28]. Taken into account, that studies already reported on the effective use of ADC cut-off values for assessing lymph node malignancy in PC patients [29, 30], it is most likely that PSMA-positive lymph nodes with MRI diffusion restriction have to be considered as malignant, although a dedicated PET/MRI study facing that issue has to be performed separately in a sufficient number of e.g. salvage lymphadenectomy patients. Despite these limitations, a reputable standard of reference was available in 79% (79/100) of PSMA-PET-positive lesions.

We found that PSMA-PET changed the clinical decision-making regarding therapy in 50 patients. In this context, to the best of our knowledge, this is the first study applying a retrospective, interdisciplinary tumorboard, given only conventional imaging results and being blinded to the results of the PSMA-PET and the original therapeutic decision. In this regard, it could be shown, that a positive PSMA-PET had changed the (conventional) treatment management in about three fourths of the patients with negative standard imaging (74.6%), of whom 86% were referred toward metastasis-directed therapies. In other words, 58% (29/50) of the patients with positive PSMA-PET and negative standard imaging did not undergo ADT, delaying its potential risks and side effects, and 36% (18/50) of the patients were not managed with a wait-and-see strategy, which would potentially lead to a progress in the disease and the further growth of metastases, which were not seen in the standard imaging. Instead, 13 of these patients (26%) underwent salvage lymphadenectomy and 30 patients (60%) underwent PSMA-PET/CT(MRI)-guided RT to pelvic or retroperitoneal lymph nodes or bone metastases.

Limitations of this study are its retrospective design which prevents a randomized assessment of the clinical impact of PSMA-PET including clinical follow-up and also provides limited information concerning the details of previous therapies such as the extend of the primary lymph node dissection in this consecutive clinical patient cohort. Due to the applied homogenization of the patient cohort, the final number of patients available for evaluation is reduced. Nevertheless, the results still remain significant. Although the majority of PSMA-positive sites could be acceptably confirmed, a true histopathological gold standard was only available in 16% of the PSMA-positive patients.

Our data confirm the superior detection rate of disease recurrence sites using PSMA-PET/CT–PET/MRI in patients with BCR after RP for PC compared to the guideline-recommended standard imaging modalities. Additionally, we could demonstrate a clinical impact toward metastases-directed therapies in a cohort of patients with low PSA level recurrences. PSMA-PET may improve early detection, treatment planning and survival in patients with BCR after RP with curative intent and therefore opens windows for targeted therapies before the onset of a systemic approach. Further confirmatory studies with a prospective randomized design are necessary to confirm these preliminary findings.