Background
Prostate cancer is the most frequent cancer and the third leading cause of death among men in developed countries [
1]. Approximately one third to one half of patients suffer from biochemical relapse (BCR) within 5–10 years after primary curative prostatectomy or radiotherapy (RT) [
2‐
4]. Salvage radiotherapy (SRT) is recommended in patients treated with radical prostatectomy who experience BCR without distant metastases [
5,
6]. The most evident benefit of SRT has been found in BCR patients with low PSA levels (≤0.5 ng/ml) [
7,
8]. Stephenson et al. reported a 6-year post-SRT progression-free probability of 48 % in patients with a PSA <0.5 ng/ml compared with 40, 28, and 18 % in patients with PSA levels of 0.51–1 ng/ml, 1.01–1.5 ng/ml, and >1.5 ng/ml, respectively [
9]. The extent of SRT (prostate bed vs. prostate bed and pelvic lymph nodes) is still a matter of controversy [
5]. Goldner et al. reported a significantly increased risk for biochemical failure after SRT in patients with a risk of lymph node involvement >15 % according to the Roach formula and SRT limited to the prostate bed [
10]. However, automatic inclusion of radiation of the pelvic lymph nodes as part of SRT is currently not recommended [
5,
6].
The role of pretherapeutic imaging in SRT planning has not been definitively assessed. The relevance of conventional imaging (e.g., computed tomography, bone scintigraphy) is low due to limited sensitivity in patients with low PSA values [
11,
12]. Molecular imaging using choline PET/CT has recently been widely used in patients with BCR.
11C-choline PET/CT proved to be sensitive in patients with a PSA level >1 ng/ml and is often used if imaging results are considered relevant for treatment planning [
5], e.g., for metastasis-directed therapy in oligometastatic disease [
13]. However, the detection rates and accuracy of choline PET/CT are limited [
14,
15]. Krause et al. reported a detection rate of <50 % in patients with a PSA level ≤2 ng/ml [
16]. A PSA level of 1.16 ng/ml was found to be an optimal cutoff value for prediction of a positive choline PET/CT scan [
17]. The American Society for Radiation Oncology (ASTRO) guidelines stated that “improved imaging techniques would help to better define appropriate therapies” [
6].
The recent introduction of PET ligands to image the expression of the prostate-specific membrane antigen (PSMA) appears to be revolutionizing prostate cancer imaging and therapy [
18,
19]. Numerous studies have reported that
68Ga-labeled peptide ligands targeting the cell surface protein PSMA are more specific and also highly sensitive [
20‐
25]. Detection rates of 74.2–89.5 % have been reported in patients with BCR even in the presence of low PSA levels (<1 ng/ml) [
20‐
24]. Van Leeuwen et al. found in a cohort of 70 patients with BCR (PSA level <1.0 ng/ml) and considered for SRT PSMA-positive lesions in 54 % of the patients [
26]. A recently published meta-analysis summarized the available studies and reported in patients with BCR an overall positivity rate of 76 % [
27]. However, only limited data are available regarding the impact of
68Ga-PSMA PET/CT on SRT planning [
28‐
30]. Change of treatment based on the
68Ga-PSMA PET/CT findings has been reported in 50.8 and 53.7 % of patients, but these cohorts included primary disease and BCR [
29,
30]. Dewes et al. found a change of TNM stage in 53.3 % and a change in RT concept in 33.3 % of cases prior to initial curative RT [
28]. Patient cohorts were heterogeneous, and only very limited data are available on patient outcome in
68Ga-PSMA PET/CT-guided RT [
31]. The aim of this study was to evaluate the impact of
68Ga-PSMA PET/CT on treatment decision-making prior to SRT and particularly on the patient outcome.
Discussion
Our data demonstrate that 68Ga-PSMA PET/CT impacted treatment in 42.2 % (19/45) of patients.
The concept of automatically combining SRT of the prostate bed with radiation of pelvic lymph nodes as part of SRT is controversial [
5,
6,
36]. In the present cohort, seven patients (15.6 %) had LNMs limited to the pelvis and would have potentially benefited from extended SRT. Interestingly, according to clinical risk stratifications for identification of patients with high risk of LNMs [
37] and therefore suitable for extended irradiation to the pelvic lymph nodes, only five of these seven patients qualified. Moreover, six patients (13.3 %) scheduled for SRT would not have benefited from radiation to the prostate bed and a general extension of SRT to pelvic lymph nodes as they had extrapelvic disease (rectal soft tissue lesions, retroperitoneal LNMs, and/or bone metastases). These data underline the clinical impact of pretherapeutic
68Ga-PSMA PET/CT on selection of the most suitable therapeutic approach.
Up to now, only limited data are available on the impact of
68Ga-PSMA PET/CT on RT planning [
28‐
30]. Changes in treatment due to
68Ga-PSMA PET/CT prior to RT have been reported in previous studies; however, the reported rates were higher compared with our results [
29,
30]. Shakespeare found an alteration of treatment in 53.7 % of patients due to
68Ga-PSMA PET/CT performed in addition to conventional imaging [
29]. This study included an inhomogeneous group of 54 patients including some with PSA relapse after definitive or post-prostatectomy RT and some in whom curative doses of RT were planned (definitive or post-prostatectomy) [
29]. Sterzing et al. reported an alteration in the therapeutic approach in 29 (50.8 %) of a cohort of 57 patients with primary and recurrent prostate cancer [
28]. Of these patients, 62.1 % received a boost to pelvic LNMs, while in 27.5 %, the irradiation field was extended to retroperitoneal LNMs and 13.8 % received a systemic treatment [
28]. A radiation boost to PET-positive lesions seems to be justified due to the high PPV of
68Ga-PSMA PET/CT [
28]. However, it is crucial to exclude false-positive interpretations due to PSMA-positive celiac ganglia [
38].
Pretherapeutic imaging is gaining importance with the emergence of individualized treatment strategies to replace the use of systemic therapies such as ADT for the treatment of all patients with metastatic disease irrespective of the disease extent, thereby avoiding the latter’s various side effects [
29,
31,
39,
40]. In the present study, patients with oligometastatic disease (≤5 lesions) were treated with SRT guided by
68Ga-PSMA PET/CT. Recently published studies have limited the number of metastases to ≤3 lesions [
34,
40]. A consistent definition of the oligometastatic tumor stage is warranted for future trials to allow for comparison of results. The strategy of metastasis-directed treatment in the management of prostate cancer patients with oligometastatic disease as an individualized therapeutic approach is still controversial. Oligometastatic disease seems to be an intermediate and unique clinical state in prostate cancer [
40]. Patients with oligometastatic disease may have a superior survival rate [
41,
42] and may benefit from metastasis-directed treatment, with delayed clinical progression and postponement of systemic therapy or even cure [
34,
40]. Pretherapeutic imaging with a high accuracy is crucial to the use of these novel salvage strategies. Molecular imaging has been found to outperform conventional imaging for assessment of disease extent in cases of BCR [
11,
12,
14,
15]. Picchio et al. reported that
11C-choline PET/CT is a valuable tool for guidance of RT of
11C-choline-positive LNMs and that PET/CT-guided RT resulted in a high early biochemical response rate [
39]. Casamassima et al. reported a 3-year local control rate of 90 % in 25 patients treated with RT for LNMs [
43]. However, previous studies and meta-analyses reported that choline PET/CT offers a low detection rate in patients with early BCR [
14,
16,
44,
45], i.e., the patient cohort recommended for SRT [
5,
7,
8].
68Ga-PSMA PET/CT, on the other hand, has been reported to achieve high detection rates, superior to those of choline PET/CT, even in patients with low PSA levels [
20‐
24]. In addition, it has been found that more lesions with a higher SUV
max and tumor-to-background ratio are detected by
68Ga-PSMA PET/CT compared with choline PET/CT [
21,
46]. Therefore,
68Ga-PSMA PET/CT is regarded as the preferred method for guidance of RT, particularly in patients with low PSA values [
47].
In our total patient cohort,
68Ga-PSMA PET/CT-based treatment resulted in biochemical response in all but one patient. PSA values significantly decreased by 78 % compared with pretherapeutic PSA levels. Five of nine patients were treated with a dose escalation to PET-positive LNMs and/or bone metastases using an IMRT technique (oligometastatic concept); the PSA level decreased on average to 15 % of pretherapeutic values. However, three patients had not completed RT as the PSA level was rising during therapy, indicating disease progression, and in one patient, no follow-up was available. Recently, Henkenberens et al. [
31] also reported a significant decrease in PSA values in 29 patients followed up for a similar period and also treated according to an oligometastatic concept.
68Ga-PSMA PET/CT seems to be a highly sensitive method for detection of oligometastatic disease and guidance of metastasis-directed treatment; however, the oligometastatic concept has to be investigated in further studies.
The current study has several limitations: First, the analysis is a retrospective analysis of consecutive patients referred for
68Ga-PSMA PET/CT prior to SRT. Thus, referral bias cannot be excluded. The highly selected nature of the cohort may be the reason for the relatively low detection rate (53.3 %) compared with previous studies investigating
68Ga-PSMA PET/CT in large patient cohorts with BCR (74.2–89.5 %) [
20,
23,
24] or prior to radiotherapy (73.5 and 96.6 %) [
30,
31]. Second, the number of included patients is low, but it is within the range of or superior to the size of patient populations in previously published studies. Third, to date, no long-term follow-up is available for the patients. Fourth, the reason for the treatment failure with increasing post-therapeutic PSA levels in one patient could not be assessed as the patient refused further imaging. Fifth, no histopathological confirmation was available in the present study; however, only few studies are available with histopathologic correlation of the results of
68Ga-PSMA PET/CT [
27]. Thus, only limited knowledge about the negative predictive value of
68Ga-PSMA PET/CT is available. In the present study, also, patients without any PSMA-positive lesion responded to SRT with decreasing PSA values suggesting small lesions may be missed by
68Ga-PSMA PET/CT.
Acknowledgements
This publication was supported by Open Access Publication Fund of the University of Wuerzburg.