Feasibility and Optimal Time Point of [68Ga]Gallium-labeled Prostate-specific Membrane Antigen Ligand Positron Emission Tomography Imaging in Patients Undergoing Cytoreductive Surgery After Systemic Therapy for Primary Oligometastatic Prostate Cancer: Implications for Patient Selection and Extent of Surgery

Take Home Message Prostate-specific membrane antigen positron emission tomography baseline imaging provides better information on residual disease after systemic therapy than preoperative imaging. Early application should be considered in high-risk localized or oligometastatic patients, as it has the potential to guide local therapy even at later stages.


Introduction
The introduction of prostate-specific membrane antigen (PSMA) targeted molecular imaging using positron emission tomography (PET) has significantly enriched the diagnostic and therapeutic landscape of prostate cancer (PCA). It is an accurate imaging modality for local, regional, and distant staging in primary, recurrent, metastatic, as well as castration-resistant PCA [1][2][3][4][5][6][7][8]. The most widely used and only guideline-recommended indication is at the time of biochemical recurrence (BCR) after local treatment with curative intent [4,[9][10][11], to differentiate between local and metastatic recurrence.
In patients with presumed high-risk localized disease, the use of PSMA-PET has led to a disease state shift, partly due to its improved sensitivity for oligometastases compared with standard imaging. Many oligometastatic PCA patients diagnosed by PSMA-PET show no evidence of systemic disease when assessed using standard imaging, which includes a bone scan (BS) and computer tomography (CT) [2,7,8,12,13]. This has led to an increased use of local radical therapy in patients with oligometastatic PCA, commonly after a response to systemic therapy [14,15].
In a previous prospective study, we have demonstrated that preoperative [ 68 Ga]gallium-PSMA HBED-CC conjugate 11-PET ([ 68 Ga]Ga-PSMA-11 PET) had high diagnostic accuracy for local staging in systemic therapy-naïve patients [2]. We hypothesized that the added information gained through PSMA-targeted molecular imaging might improve clinical decision-making for patients undergoing cytoreductive sur-gery after systemic therapy for oligometastatic PCA, potentially helping to fine-tune the tailoring of local therapy. However, as the effects of systemic therapy might reduce the sensitivity of preoperative [ 68 Ga]Ga-PSMA-11 PET, the ideal time point of its use remains unclear.
The aim of this study was to assess the feasibility and compare the diagnostic accuracy of [ 68 Ga]Ga-PSMA-11 PET images taken at baseline and before the initiation of systemic treatment, and preoperative images, using histopathology after cytoreductive surgery as reference.

Patient selection
This was an ethics-approved retrospective analysis of a prospectively maintained database (NCT02971358). We queried our database of patients treated with radical prostatectomy (RP) and superextended pel- was also assessed on a per-patient level, with and without the prostate region. Additionally, we performed an exploratory logistic regression analysis for the association of pathologic complete response with preoperative prostate-specific antigen (PSA), duration of ADT, docetaxel, enzalutamide, and radiographic complete response, as well as for the association of pathologic negative LNs and preoperative PSA, duration of ADT, docetaxel, enzalutamide, and radiographic complete response in LNs.
Owing to the exploratory nature of our study, statistical significance was considered for p < 0.05, but not in a confirmatory matter. Thus, no adjustments for multiplicity were performed. All tests were two sided and conducted using STATA-14 (StataCorp, College Station, TX, USA).

Results
The final analysis contained 20 patients. Baseline characteristics of patients are shown in Table 1 Fig. 1).
In our exploratory logistic regression analysis, there was a significant association between a pathologic complete response (in the prostate and LNs) and a complete radiographic response on preoperative [ 68 Ga]Ga-PSMA-11 PET/ MRI (odds ratio 32, 95% CI 1.39-737.56; p = 0.03). There was no significant association between pathologic complete response and preoperative PSA, duration of ADT, docetaxel, or enzalutamide. There was also no significant association between pathologic complete response in LNs and PSA, duration of ADT, docetaxel, enzalutamide, or radiographic complete response in LNs (Table 3). As there were no multiple associations, no multivariable analysis was performed.

Discussion
In this study, we found preoperative [ 68 Ga]Ga-PSMA-11 PET/MRI after systemic therapy to have mediocre clinical accuracy on a region-based analysis. However, sensitivity and NPV were not sufficient for basing further treatment decisions, such as the template of resection, on these images. As baseline staging, [ 68 Ga]Ga-PSMA-11 PET-based hybrid imaging, however, showed more desirable parameters of diagnostic accuracy with high sensitivity and NPV, and a very low negative LR, for basing further localized treatment on it (example in Fig. 2A and B). On a perpatient level, the results were similar.
Several trials evaluating PSMA-targeted molecular imaging for primary staging have been reported. Even though there is one large prospective evaluation of [ 68 Ga]Ga-PSMA-11 PET as baseline staging before surgery reporting very high sensitivity of 85% and an AUC of 0.92 in the proPSMA study [8], most other trials have shown less favorable outcomes. Hope et al. [7] have reported sensitivity and NPV of 40% and 81%, respectively, in a phase 3 trial including 277 patients who underwent RP and extended pelvic lymph node dissection, along with Morris et al. [13] showing sensitivity and NPV of 40.3% and 83.2%, respectively, in their phase II/III trial evaluating [ 18 F]F-DCFPyL, another PSMA-targeted tracer. In our study, the PSMA-PET images taken as baseline staging, prior to systemic therapy, not only provided better sensitivity (95.65%), NPV (98.39%), and negative LR (0.06) for the presence of vital tumor tissue in a specific area, than preoperative imaging after systemic therapy, but interestingly were also higher than the values reported for primary staging of patients treated with immediate surgery in any of the previously mentioned studies [2,7,8,13].
One explanation for this could be that additional small tumor foci and micrometastasis, which are missed even by molecular imaging at initial staging, will usually be present on final pathology. Yet these lesions seem to be treated adequately with systemic therapy. This information has the potential to allow for immediate selection of the most suitable candidates for systemic therapy followed by cytoreductive surgery, right at the time of diagnosis. In this case, the extent of resection could be selected at baseline, as all areas with initially positive lesions should be resected; yet, no further extension seems to be necessary.
In the preoperative setting, [ 68 Ga]Ga-PSMA-11 PET showed very high specificity for residual disease, and a positive lesion after systemic therapy might be indicative of very aggressive disease; yet, the use of molecular imaging provided no additional information when compared with morphologic imaging using MRI alone. However, while the difference between MRI and CT is considered marginal for the staging of LNs [18,19], whole-body MRI is also not considered standard and might have reduced the benefit gained from molecular imaging.
Additionally, several studies have established the favorable diagnostic accuracy of PSMA-PET in PCA patients with BCR after local therapy showing superiority over standard imaging in some studies [4,[20][21][22]. Abufaraj et al. [11] analyzed PSMA-PET imaging in the setting of BCR in patients  undergoing salvage sePLND. The authors also performed a region-based analysis using the same regions as in our study, reporting diagnostic accuracy ranging from 95% to 98% and NPV ranging from 93% to100% depending on the location. Both in primary staging and at the time of BCR, the patients included in these trials had higher PSA levels at the time of imaging than that at preoperative imaging in our cohort. In this regard, in our study, the accuracy of preoperative [ 68 Ga]Ga-PSMA-11 PET/MRI before cytoreductive RP was very similar to patients exhibiting similar PSA values in the BCR setting [21] and to patients undergoing primary imaging with immediate surgery. Currently, it is unclear whether patients with oligometastatic PCA stand to benefit from localized therapy, with the optimal sequence of therapy being currently evaluated in clinical trials. Some ongoing studies include or allow PSMA-based molecular imaging within their protocols. A retrospective analysis of the prospective STAMPEDE (systemic therapy in advancing or metastatic prostate cancer: evaluation of drug efficacy) trial has shown a benefit of local therapy with radiation in patients with low metastatic burden PCA, detected by standard imaging in addition to ADT [23]. Other recently published STAMPEDE data have shown benefit for the combination of intensified systemic therapy with abiraterone, ADT, and local radiation of the prostate, in a very-high-risk localized disease cohort on conventional imaging, which might be similar to a cohort of oligometastatic patients using molecular imaging [24], further supporting the combination of systemic and local therapy for these patients. There are very few reported prospective randomized trials on oligometastatic PCA. Two trials have included patients with recurrent oligometastatic PCA after local treatment and were randomized between observation and stereotactic radiation [25,26]. In this setting, both Phillips et al. [25] and Ost et al. [26] could show an advantage to metastasis-directed therapy using radiation in terms of progression-free survival. Parikh et al. [27] have initiated a prospective trial evaluating the combination of upfront local treatment with RP, conducted with concomitant ADT for 6 mo and possible radiation therapy. Molecular imaging after 2 mo of ADT is included within their trial. In this and other trials [28], if patients are considered for inclusion and therefore local treatment, complete resection or radiation of all present metastatic lesions is pursued. Surgical resection in the form of sePLND or radiation of LN metastasis, however, comes with substantial morbidity [29]. Ideally, previous systemic therapy and accurate imaging would allow optimization of the extent of local therapy.
To our knowledge, this is the first study describing the diagnostic performance of [ 68 Ga]Ga-PSMA-11 PET before and after systemic treatment, with histologic verification. While histology as a reference standard allows valuable insights into this imaging modality, there are many limitations. First, while our database is maintained prospectively, this study was a retrospective analysis. To avoid some of the inherent problems with this design, we used very strict inclusion criteria to evaluate a homogeneous cohort of patients with complete data. However, these criteria and the generally low prevalence of patients presenting with this stage of disease have led to a small number of patients included within this study, which also presents a substantial limitation. There is a selection bias, as patients with immediate disease progression during systemic therapy would not be considered for surgery, thus not be included in our prospective database; anecdotally, this was a very rare event. In addition, patients with bone metastasis were included in our study and histologic verification of these lesions was not feasible. As the intention of this trial was only hypothesis generating, all images and also pathologic specimens were reviewed only by a single specialist in each area; thus, we cannot exclude the presence of interobserver variability.

Conclusions
As baseline imaging, [ 68 Ga]Ga-PSMA-11 PET has significantly better diagnostic accuracy, sensitivity, and NPV than images obtained preoperatively in systemically pretreated patients. If a patient is suitable for local treatment and complete resection of all residual tumor is intended, [ 68 Ga]Ga-PSMA-11 PET images taken prior to systemic therapy are significantly more accurate in selecting the relevant LNs for resection.
This imaging modality has the potential to improve our diagnosis as well as guide us for localized therapy at later stages of treatment. Further research on [ 68 Ga]Ga-PSMA-11 PET imaging, especially as baseline imaging, in localized high-risk or oligometastatic PCA and its incorporation into prospective trials are needed.  Other: None.
Financial disclosures: Bernhard Grubmüller certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
Funding/Support and role of the sponsor: None. Data sharing: The complete dataset used for this analysis is available in a pseudonymized fashion upon request, and after acquisition of IRB approval as well as data sharing agreements between institutions.