Management impact of 18F-DCFPyL PET/CT in hormone-sensitive prostate cancer patients with biochemical recurrence after definitive treatment: a multicenter retrospective study

Purpose The aim of this study was to investigate whether an early, accurate identification of disease using 18F-DCFPyL PET/CT imaging resulted in a change of decision on treatment management, for individual patients with biochemically recurrent (BCR), hormone-sensitive prostate cancer. Methods In this retrospective study, a total of 253 patients with BCR who underwent restaging 18F-DCFPyL PET/CT were assessed. Two urologists specialized in uro-oncology were asked to formulate a preferred treatment for each patient before and after knowing the results of the 18F-DCFPyL PET/CT. Results Out of 253 patients, 191 (75%) underwent robot-assisted radical prostatectomy (RARP) as primary therapy, and 62 (25%) external beam radiation therapy (EBRT). In 103/253 cases (40.7%), a preferred treatment change based on the 18F-DCFPyL PET/CT findings was reported. In patients post-RARP, a positive 18F-DCFPyL PET/CT (OR 6.21; 95%CI 2.78–13.8; p < 0.001) and positive pathological lymph node status (pN1) (OR 2.96; 95%CI 1.15–7.60; p = 0.024) were significant predictors for an intended change of management, whereas a positive surgical margin (OR 0.42; 95%CI 0.20–0.88; p = 0.022) was inversely associated with an intended change of management. Conclusion In this study, we found a significant impact of 18F-DCFPyL PET/CT on the intended management of patients with biochemically recurrent hormone-sensitive prostate cancer. A positive 18F-DCFPyL PET/CT scan, positive pathological lymph node status, and a negative surgical margin status were significantly associated with increased odds of having a change of management based on 18F-DCFPyL PET/CT findings. Supplementary Information The online version contains supplementary material available at 10.1007/s00259-021-05222-5.


Introduction
Prostate cancer (PCa) is associated with increasing age. In elderly men it is the second most frequent malignancy after lung cancer [1]. Robot-assisted laparoscopic radical prostatectomy (RARP) and external beam radiation therapy (EBRT) are two important curative treatment options for localized PCa. However, 20-50% of patients who have undergone RARP or EBRT will experience biochemical recurrence (BCR) of disease within 10 years [2][3][4][5][6].
Since the introduction of prostate-specific membrane antigen (PSMA) positron emission tomography/computed tomography (PET/CT), the detection of metastases in patients with BCR at low prostate-specific antigen (PSA) values has improved substantially [7,8]. PSMA is a transmembrane protein which is overexpressed in malignant prostatic tissue [9,10]. In addition to the widely used 68 Gallium-PSMA-11 [11], 18 F-DCFPyL is a promising novel, second-generation 18 Fradiolabeled tracer [12]. Both radiotracers demonstrated improved detection of metastases compared to conventional imaging techniques, such as computed tomography (CT), bone scan, or magnetic resonance imaging (MRI) [8,13,14]. Moreover, multiple studies showed a high diagnostic accuracy of PSMA PET/CT imaging in patients with BCR, with a variable positive predictive value of 78-99% [15,16].
When a modern imaging technique is associated with an enhanced diagnostic accuracy, it is expected that the treatment strategy will change due to an improved staging [17]. Indeed, Calais et al. [18] showed that a change in management was seen in 54 out of 101 (53%) patients who underwent 68 Ga-PSMA PET/CT for BCR after curative therapy for PCa. Comparably, Song et al. [19] showed a similar percentage of patients (60%) with a change of disease management for 18 F-DCFPyL PET/CT in a small cohort of 72 patients with BCR after RARP or EBRT. However, as both studies used actual implemented management as outcome instead of the intended treatment based on PSMA PET/CT, other factors such as patient preferences might have influenced the results. Consequently, if these management decisions were truly based on the PSMA PET/CT findings remains unclear. Therefore, the present study investigated the role of modern imaging for PCa in a large cohort of patients, in whom disease management was assessed both with and without the knowledge of the outcome of 18 F-DCFPyL PET/CT, in patients with BCR after RARP or EBRT.

Materials and methods
This retrospective study was conducted by the Prostate Cancer Network the Netherlands, i.e., Amsterdam UMC VU University (VUmc), the Netherlands Cancer Institute (NCI), and the Noordwest Ziekenhuisgroep (NWZ) Alkmaar, in the period December 2016 to December 2019. Approval of the institutional review board of VUmc (VUmc2020.048) and NCI (IRBd19-182) was obtained for this study, waiving the need to receive informed consent. All patients included from NWZ have given written informed consent.

Inclusion and exclusion criteria of patients
In this study, we included 253 consecutive patients with biochemically recurrent, hormone-sensitive PCa, after RARP or EBRT, who underwent an 18 F-DCFPyL PET/CT between December 2016 and December 2019. BCR was defined as a PSA level ≥ 0.2 ng/mL after RARP [20] and > 2.0 ng/mL after EBRT [21]. Patients with rising PSA values after RARP or EBRT, and who did not meet the criteria for BCR but still underwent 18 F-DCFPyL PET/CT for restaging purposes, were also included in this analysis.
Patients who underwent primary therapy, other than RARP or EBRT (e.g., brachytherapy, focal therapy of the prostate), were excluded. Also patients who priorly underwent PSMAbased imaging for BCR with a tracer other than 18 F-DCFPyL such as 68 Ga-PSMA-11 or 18 F-PSMA-1007 were excluded. Moreover, patients who had received adjuvant androgen deprivation therapy (ADT), any hormonal therapy (HT) at the time of performing 18 F-DCFPyL PET/CT, or any salvage treatment other than salvage radiation therapy (SRT) after RARP were not eligible for this analysis. Lastly, patients with incomplete clinical data were excluded from the study (Fig. 1).

Assessment on intended treatment of individual cases before and after 18 F-DCFPyL
Two urologists (AV, PvL), specialized in PCa surgery and care at the Amsterdam UMC and at the NCI, assessed all individual cases with BCR after RARP or EBRT. Anonymized patient charts, including patients' clinical, pathological, and biochemical characteristics, were presented to both urologists independently by two medical researchers (DM, PO). An intended treatment advice was given for all cases, firstly without the knowledge of the results of the 18 F-DCFPyL PET/CT and, secondly, with the findings of 18 F-DCFPyL imaging. Thereafter, a consensus meeting was organized in which all cases and patient histories were rediscussed, and a definitive treatment advice was given for every individual case Fig. 2.
After collecting the selected treatment advices, four different treatment categories were described, based on their clinical value (i.e., 1 to 4), further subdivided into a total of nine treatment allocations (i.e., a to i; Table 1): (1) local treatment (salvage radiotherapy (RT) to the prostatic fossa, with or without HT (a), salvage focal therapy (b)); (2) locoregional treatment (whole pelvis RT with (c) or without (d) salvage RT to the prostatic fossa, with or without HT, and salvage pelvic lymph node dissection (sPLND) (e); (3) metastasis-directed radiotherapy (MDT) of distant (oligo)metastatic lesions (f); and (4) systemic treatment ((delayed) step-up HT (g), chemotherapy (h), and a combination of both(i)).

Patient data and patient charts
All demographic data (age at the time of treatment, age at the time of performing the PSMA PET/CT), clinical parameters (clinical tumor stage, history of PCa treatment), biochemical parameters (initial PSA level, PSA level at time of scan), radiological data ( 18 F-DCFPyL PET/CT findings at the time of BCR), and pathological data (biopsy tumor features, pathological T-stage, radical prostatectomy Grade Group (GG), surgical margin status, pathological lymph node status) were collected for all patients. Furthermore, for patients who underwent EBRT, the number of fractions, dose, and number of months of hormonal therapy were reported. For each case, previous urological or radiation oncology treatments (such as salvage RT) were recorded, as was the biochemical follow-up of patients who underwent RARP or EBRT. 18

F-DCFPyL PET/CT imaging
At Amsterdam UMC, location VUmc, imaging was performed using a Philips Ingenuity TF (Philips Healthcare®, the Netherlands/USA) PET/CT system. 18 F-DCFPyL was synthesized via direct radiofluoration at the on-site cyclotron facility, compliant to good manufacturing practices (GMP [22,23]). The median tracer dose administered was 311 MBq (interquartile range (IQR) 301-322 MBq). PET images were acquired approximately 120 min after intravenous injection. The scan trajectory included mid-thigh to skull base, with 4 min per bed position.
At NCI, imaging was performed using a Philips Gemini TF-II or Vereos Digital PET/CT (Philips Healthcare®, the Netherlands/USA). 18 F-DCFPyL was administered as an intravenous bolus injection with a median dose of 197 MBq (IQR 189-207 MBq). Scanning commenced after an incubation period of approximately 60 min, with 2 min per bed position over the complete scan range.
At NWZ, imaging was performed using a Siemens Biograph TruePoint-16 (Siemens Healthineers, Germany) PET/CT scanner. Scanning was performed after approximately 120 min, with a median tracer dose of 290 MBq (IQR 280-323 MBq). PET images were made mid-thigh to skull base, with 5 min per bed position.
PET images were combined with either a low-dose CT scan (120-140 kV, 30-80 mAs with dose modulation) or a diagnostic CT scan (130 kV, 110mAs). All PET images were corrected for scatter, decay, and random coincidences; attenuation correction was performed using CT images.

Image interpretation of 18 F-DCFPyL PET/CT imaging
Interpretation of the scans was performed by nuclear medicine physicians in the three hospitals, all of whom have an ample experience with PSMA PET reading (>400 scans). PSMA reporting was performed in accordance with the PROMISE criteria [24]. If at least one metastatic lesion or sign of local recurrence in the prostate/prostatic fossa (miTr) was found, the scan was considered to be positive (i.e., focal and higher uptake of the PSMA tracer compared to the surrounding tissue, not compatible with physiological uptake). Loco-regional lymph node metastases in the true pelvis were classified as miN1. miM1a was defined as lymph node metastatic disease outside the surgical template, whereas lesions that showed increased PSMA expression in the bones or the visceral organs were classified as miM1b and miM1c, respectively. This classification was conform the EAU guidelines [20].

Statistical analysis
Descriptive statistics, median, and interquartile range (IQR) were used to summarize numerical variables, whereas percentages (%) were used for categorical variables. The differences in change of management rates for PSMA-positive and PSMA-negative patients, as well as the difference between different therapies (RARP, RARP + SRT, and EBRT), were compared using the Chi-square test. Statistical significance was set at p < 0.05 [25].
In order to investigate the potential association between selected variables and the likelihood that would lead to a change of management of 18 F-DCFPyL PET/CT, a multivariable logistic regression was performed. Multiple parameters that could potentially predict change of management were studied: PSA value at time of the scan, 18 F-DCFPyL PET/ CT findings, RARP GG, surgical margin status, pathological lymph node status, pathological T-stage, and the administration of salvage therapies prior to scan. The dichotomous outcome variable was change of management based on 18 F-DCFPyL PET/CT imaging. As multiple pathological variables were included, the multivariable analysis was solely performed on the RARP group.

Impact on patient management
The treatment advice for all cases, both in absence and presence of the findings of 18 F-DCFPyL imaging, is presented in Table 4. Treatment categories and treatment allocations were given for the different clinical BCR indications. In 103 out of 253 cases (40.7%), restaging 18 F-DCFPyL PET/CT scan findings were reason to change treatment advice. For patients who  The proportion of patients in whom management of disease was changed was not statistically different between men undergoing RARP (44.0%), men undergoing RARP + SRT (31.7%), and those patients who underwent EBRT (38.7%), as initial treatment (χ2; p = 0.34).

Predictors of management changes
In order to analyze a homogenous group, only RARP patients have been taken into account. On multivariable logistic regression analysis, a positive 18 F-DCFPyL PET/CT scan (odds ratio (OR) 6.21; 95%CI 2.78-13.8; p < 0.001) or a positive pathological lymph node status (OR 2.96; 95%CI 1.15-7.60; p = 0.024) was significant predictors for an intended change of management, whereas a positive surgical margin (OR 0.42; 95%CI 0.20-0.88; p = 0.022) was inversely associated with an intended change of management. The PSA level at the time of the scan, pathological T-stage, RARP GG, and the administration of SRT were not associated with an intended management change (Supplementary Table 1).

Discussion
In the present study, we systematically assessed the impact of 18 F-DCFPyL PET/CT imaging on the change of management of patients with biochemically recurrent hormone-sensitive PCa after RARP or EBRT. The medical charts of 253 patients were presented to two separate urologists specialized in urooncology, who chose from various treatment options, initially without knowledge of PSMA-based imaging and subsequently with the results of restaging 18  Previous studies that assessed the impact of 68 Ga-PSMA-11 PET/CT on PCa management found promising results for the proportion of changes in disease management. Albisinni  (24) RARP robot-assisted laparoscopic radical prostatectomy, SRT salvage radiation therapy, EBRT external beam radiation therapy, PET positron emission tomography, CT computed tomography et al. [26] found a change of management in 76% (99/131) of patients who underwent 68 Ga-PSMA-11 PET/CT for BCR after radical prostatectomy or multimodality treatment. One of the main differences to our study is that a total of 14 different treatment options for relapse were used compared to only 4 treatment categories in our study. Due to this higher number of treatment options, a change of management was more likely to occur. Moreover, continuing surveillance (withholding hormonal therapy) was often chosen as treatment option in that study. Therefore, patients in whom systemic treatment was proposed pre-PSMA, and active surveillance post-PSMA, were considered as having a change of management. In our study, surveillance was pooled with (delayed) stepup hormonal therapy to assess actual change of management instead of merely delaying treatment options until further notice. Lastly, while the percentage of patients who underwent RARP as primary curative therapy was almost comparable to that of our series (i.e., 81% versus 75% in our study), the median PSA level at the time of PSMA-based imaging in their study was substantially higher (2.2 ng/mL) compared to the median PSA in the present cohort (0.8 ng/mL). Therefore, it is likely that the proportion of positive PSMA scans was substantially higher in their study [13].
Another study that assessed the impact of 68 Ga-PSMA-11 PET/CT on treatment plan is that of Calais et al. [18]. A change of management in 53% of included patients (54/ 101) was reported, especially in patients with positive 68 Ga-PSMA-11 PET/CT scans. Similar to the study of Albisinni et al. [26], Calais et al. reported on 9 different treatment options, including active surveillance. In addition, patients with BCR were scanned at a median PSA level of 1.7 ng/mL (range, 0.05-140 ng/mL), again a substantially higher level than that in our study (0.8 ng/mL). Moreover, the management options consisted of the preferred treatment before the PSMA scan and the actually implemented treatment for the patient. Since Calais et al. applied the actual implemented management as outcome instead of the intended treatment, which was used in our study, other factors such as patient preferences might have influenced the results.  (1) miTr (3) Locoregional (2) 38 miN1 (5) MDT (4) miM1a-miM1c (3) Systemic (4) Multiple locations (1) RARP + SRT Systemic (41) NED (12) Locoregional (4) 13/41=32% miTr (3) MDT (9) 41 miN1 (12) Systemic (28) miM1a-miM1c (2) Multiple locations (12) EBRT + PSA ≥2.0 ng/mL Systemic (42) NED (4) Local (7) 17/42=40% miTr (10) Locoregional (6) 42 miN1 (8) MDT (4) miM1a-miM1c (5) Systemic (25) Multiple locations (15) BCR biochemical recurrence, RARP robot-assisted laparoscopic radical prostatectomy, PSA prostate-specific antigen, NED no evidence of disease, MDT metastasis-directed radiation therapy, SRT salvage radiation therapy, EBRT external beam radiation therapy Moreover, two recent studies evaluated the impact of 68 Ga-PSMA-11 PET/CT on radiotherapeutic management, both at initial staging and at BCR. Koerber et al. [27] found a change of radiotherapeutic management in 50.8% of patients, whereas Sterzing et al. [28] observed a change of management in 56.3% of patients with BCR after RARP. In both studies, the PSA level at the time of the scan was higher (1.1 ng/mL and 2.8 ng/mL, respectively) compared to that in the present study (0.8 ng/mL), which may explain the slightly higher percentages of management changes in those studies. Next to these studies, Sonni et al. [29] confirmed the impact of 68 Ga-PSMA PET/CT on staging and management of PCa patients outside of the two main classical indications (BCR and presurgical staging).
Only one previous study assessed the impact of the 18 F-DCFPyL tracer on the change of management in patients with BCR after prostatectomy or radiotherapy. Song et al. [19] found a change in management in 60% of cases (43/72 patients) on BCR. Interestingly, in 59 patients, conventional imaging results, such as those from bone scan, CT, or MRI, were also available, and these were compared to 18 F-DCFPyL findings. From the cases who underwent any of the other staging imaging modalities as well as PSMA PET/CT, 17 out of 43 had lesion localization on 18 F-DCFPyL PET only, despite negative results on conventional imaging. The most important methodological difference between that study and the present study was that Song et al. reported on the actual implemented management post 18 F-DCFPyL PET/CT, which may have influenced their rate of change of management. Furthermore, the median PSA at the time of 18 F-DCFPyL PET/CT was 3.0 ng/ mL, which was substantially higher than in the present study, probably resulting in a higher percentage positive PSMA PET/CT scans, which may lead to a higher percentage of change of management.
To assess whether clinical, biochemical, or pathological parameters could predict intended management changes in patients with BCR, a multivariable analysis was performed. A positive 18 F-DCFPyL PET/CT scan (miTr, miN1, miM1a-miM1a c), positive pathological lymph node status (pN1), and a negative surgical margin status (R0) were significantly associated with an increased odds of intended treatment management change based on the 18 F-DCFPyL PET/CT scan. Patients with a negative pathological lymph node status (pN0) and a positive surgical margin status (R1) were less likely to have a management change based on 18 F-DCFPyL PET/CT findings. Possibly, patients in whom 18 F-DCFPyL PET/CT findings did not result in an intended management change may be withheld an 18 F-DCFPyL PET/CT. Future studies on which patients to withhold PSMA imaging are warranted to assess this issue.
It needs to be addressed that a treatment change due to the application of modern imaging modalities does not necessarily translate into improved oncological outcomes. Through the earlier detection of recurrences and metastases, the Will Rogers phenomenon is likely to occur, i.e., the improvement of clinical outcome in separate staging groups, whereas the prognosis in the entire group is not changed [17]. The concept of the Will Rogers phenomenon and its pitfalls was reported for different other malignancies and might well be observed in the pro-PSMA trial [14]. In this prospective, randomized clinical trial assessing the diagnostic accuracy of PSMA PET/CT imaging in the diagnostic setting, including 302 patients with high-risk PCa, Hofman et al. assigned patients at random to conventional imaging with CT and bone scanning or 68 Ga-PSMA-11 PET/CT. PSMA-based imaging proved to be more accurate than conventional imaging regarding the detection of pelvic nodal and distant metastatic disease, and treatment changes were more frequently observed with modern PET imaging. Despite this improved diagnostic accuracy, diseasefree survival or overall survival was not (yet) assessed in the trial. Although PSMA PET is increasingly implemented as the "standard of care," the effect of increased diagnostic accuracy on oncological outcome should be better clarified.
Several limitations need to be addressed. Firstly, all included patients underwent an 18 F-DCFPyL PET/CT in three different hospitals, with different PET scanners, different scan protocols, and different nuclear medicine physicians reporting. This might have resulted in heterogeneous results. Secondly, the inherent personal interpretation of the European and Dutch urology guidelines by both urologists may result in different preferred treatment options inter-individually and from country to country. MDT, for instance, accounting for 13% of the intended management changes, is still considered experimental in the EAU guidelines for patients with oligometastatic disease [20]. Moreover, all patients included in this study underwent 18 F-DCFPyL PET/CT as a restaging modality. Consequently, no conventional imaging techniques were used. It could be that a selection of lesions visualized by PSMA PET/CT would have been found by conventional imaging techniques. Lastly, due to the retrospective nature of this analysis, a bias may have occurred. To truly determine the impact of PSMA PET/CT imaging on management decisions, future randomized trials are warranted.

Conclusion
This study showed a significant impact of 18 F-DCFPyL PET/ CT on intended management of patients with biochemically recurrent hormone-sensitive PCa. A positive 18 F-DCFPyL PET/CT scan, a positive pathological lymph node status, and a negative surgical margin status were significantly associated with increased odds of having a change of management based on 18 F-DCFPyL PET/CT findings. Whether treatment changes by modern PSMA-based imaging result in an improvement in oncological outcomes is a question that needs to be answered within well-designed prospective trials. In conclusion, this study demonstrates that 18 F-DCFPyL PET/ CT might be a helpful tool to obtain the best management decisions in patients with biochemically recurrent PCa.
Funding Open Access funding provided by Amsterdam UMC (Vrije Universiteit Amsterdam).

Declarations
Ethics approval and consent to participate Approval of the institutional review board of VUmc (VUmc2020.048) and NCI (IRBd19-182) was obtained for this study, waiving the need to receive informed consent. All patients included from NWZ have given written informed consent.

Conflict of interest The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.