Background

Prostate cancer (PC) is the most frequent tumour in men worldwide and accounts for the third most cause for cancer-associated death in men [1]. After radical prostatectomy (RP) or radiotherapy (RT) for primary treatment of PC up to 50% of all patients develop tumour recurrence as determined by rising prostate-specific antigen (PSA) values [2]. Therefore, early and sensitive diagnosis of recurrence is crucial for patient counselling and treatment selection.

Several studies have shown that morphological imaging procedures like computed tomography or magnetic resonance imaging as well as functional positron emission tomography (PET) imaging using radiolabelled choline-derivatives are of limited used in the detection of recurrent disease and often underestimate the extent of metastatic spread [3]. The introduction of 68Ga-HBED-CC- prostate-specific membrane antigen (PSMA) and 68Ga-PSMA-I&T based PET hybrid imaging several years ago has revolutionised imaging of PC with significantly improved detection rates in comparison to tracers like 18F-choline in patients with recurrent PC after radical prostatectomy, especially at low PSA values [4,5,6,7,8,9,10,11,12,13]. With the advent of PSMA-ligand PET and the visualisation of even small soft tissue metastases salvage surgery became of increasing interest in recurrent PC. However, intraoperative detection is challenging due to inconspicuous morphology and/or atypical localization of small tissue recurrence. Above, resection might be complicated due to fibrotic changes after previous surgery and/or RT.

PSMA-radioguided surgery (PSMA-RGS)

In April 2014, PSMA-RGS was introduced at the Technical University of Munich in cooperation between the Institute of Pharmaceutical Radiochemistry, Department of Nuclear Medicine and the Department of Urology, as an individual treatment concept [14, 15]. PSMA-RGS uses the preoperative intravenous (i.v.) administration of radioactively labelled PSMA-ligands method to facilitate the intraoperative detection of small and atypically located metastatic PC lesion using a γ-probe. At first, 111Indium-labelled ligands (111In-PSMA-I&T) were used with i.v. injection of mean 150 MBq (range 86-298 MBq) followed by 111In-PSMA-RGS [14]. Later on, the protocol was adapted to a more broadly available 99mTechnetium-labelled PSMA-ligand (99mTc-PSMA-I&S) with i.v. injection of mean 570 MBq (range 221–857 MBq) [16]. After injection of the radioactively labelled PSMA-ligand combined single photon emission computed tomography/computer tomography (SPECT/CT) is performed to validate correct physiological tracer uptake and appropriate tracer uptake of tumour recurrence. Intraoperatively, suspicious lesions are localised and identified using a γ-probe with acoustic feedback (e.g. Crystal Probe, Crystal Photonics) [15, 17]. Resected tissue specimens are measured ex vivo again (rating positive or negative in comparison to background) to validate successful resection immediately. A count rate at least double the background reference is judged as positive.

111In-PSMA-I&T SPECT/CT for preoperative localization of tumour recurrence

In a study of our group comparing preoperative 111In-PSMA-I&T SPECT/CT with preoperative 68Ga-HBED-CC-PSMA PET scan in 22 patients with early biochemical recurrence, 111In-PSMA I&T SPECT/CT detected only 14 out of 29 PET-positive lesions (48.3%) with no additional lesions identified with 111In-PSMA I&T SPECT/CT [18]. A total number of 15 PET-positive lesions were not detectable in 111In-PSMA I&T SPECT/CT, neither prospectively nor retrospectively in knowledge of the PET results. There was a weak trend toward a higher detectability in 111In-PSMA I&T SPECT/CT regarding lesion size and initial PSA level, although this was not statistically significant most probably due to the small number of patients analysed. One major reason for the observed lower sensitivity of 111In-PSMA I&T SPECT/CT compared with 68Ga-PSMA HBED-CC PET has contributed to the principal technical differences between PET and SPECT imaging leading to a highly superior spatial resolution of PET. Therefore, if available, PSMA-ligand PET should be preferred to PSMA SPECT/CT for diagnostic purposes as well as to confirm indication for PSMA-RGS.

99mTc-PSMA-I&S ligands for RGS

The routine clinical application of 111In-PSMA-I&T for PSMA-RGS is limited by its suboptimal nuclear properties, high costs and restricted availability of 111InCl3. To meet the clinical need for a widely available and more cost-effective alternative with less radiation exposure to patients and medical personnel, the PSMA-I&T-based tracer concept was adapted to a 99mTc-labelled PSMA agent (99mTc-PSMA-I&S) for PSMA-RGS [19]. Therefore, in our department, 111In-labelled PSMA-I&T has been completely replaced by 99mTc-labelled PSMA-I&S. First experiences with 99mTc-PSMA-I&S ligands and subsequent RGS indicate its equal performance compared to 111In-PSMA-I&T as a probe for PSMA-RGS. As with 111In-PSMA-I&T, 99mTc-PSMA-I&S can be used as a SPECT imaging agent and might be considered as alternative PC imaging technique in institutions without access to PET. In the following, recently published data of PSMA-RGS with both ligands combined are summarised [20].

Intraoperative detection of soft tissue metastases with PSMA-RGS

So far, 111In- or 99mTc-based PSMA-RGS has been performed in 63 patients with biochemical recurrent PC (median PSA: 1.29 ng/ml, range 0.19–13.9) after primary treatment (n = 61 RP, n = 1 curative intended RT, n = 1 high-intensity focused ultrasound) at our university hospital. All patients presented with a positive 68Ga-HBED-CC PSMA PET scan before PSMA-RGS proving singular or sporadic soft tissue or lymph node recurrence. Additional preoperative SPECT/CT of the pelvis was performed in all patients after administration of 111In-PSMA-I&T or 99mTc-PSMA-I&S to prove appropriate tracer uptake of tumour recurrence. A representative case of a patient with recurrent PC treated with PSMA-RGS is shown in Fig. 1.

Fig. 1
figure 1

A 67-year old patient with biochemical recurrent PC (PSA 0.39 ng/ml) after radical prostatectomy (pT3b pN1 (2/12) cM0 R0, Gleason score 4 + 3, iPSA 15.4 ng/ml) 6 years ago and salvage radiation therapy 3 years ago. During PSMA-RGS the lesion identified on 68Ga-HBED-CC-PSMA PET/CT was rated as strongly positive by ex vivo acoustic measurement and corresponded to a PSMA-positive lymph node metastasis. a Axial CT shows a small, morphologically unsuspicious lymph node in the right internal region (red arrow). b 68Ga-HBED-CC-PSMA PET/CT shows a distinct signal of this lymph node highly suspicious for metastasis. c Especially in cases of small lymph nodes and location within the depths of the pelvis guidance by γ-probe is helpful for reliable identification. d Ex vivo measurements confirm resection of the PSMA-labelled lymph node

Full size image

Direct comparison of preoperative 68Ga-HBED-CC-PSMA PET, intraoperative ex vivo measurements and histopathology was available in 60 patients. In all 60 patients, positive lesions were found by intraoperative γ-probe measurements during PSMA-RGS verified by positive histology as well. In three patients, however, suspicious lesions on preoperative 68Ga-HBED-CC-PSMA PET could not be detected during PSMA-RGS. Compared to 68Ga-HBED-CC-PSMA PET, PSMA-RGS was able to detect additional suspicious lesions in 7/60 patients (11.7%), verified by positive histopathology as well, suggesting that intraoperative measurements with PSMA-RGS is more sensitive than preoperative 68Ga-HBED-CC-PSMA PET, most probably due to reduced lesion-to-detector distance. In 14 of our 60 patients (23.3%) histology showed additional metastatic specimens in comparison to 68Ga-HBED-CC-PSMA PET suggesting that histologic evaluation is still more sensitive than preoperative 68Ga-HBED-CC-PSMA PET or measurements during PSMA-RGS. These findings demonstrate that PSMA-RGS is a valid method to track suspicious lesions and potentially identify additional lesions compared to PSMA-PET. However, microscopic lesions still might be missed which underlines the importance of complete resection of tissue adjacent to the intraoperatively detected lesions. This is concordant to a recently published study of Jilg et al. who reported a necessary short diameter of LNM to be ≥ 2.3 and ≥ 4.5 mm to reach a detection rate of 50 and 90% in 68Ga-HBED-CC-PSMA PET [21]. In three patients false positive lesions (positive with γ-probe and unobtrusive in histopathology) were detected. However, it has to be noted that most of them showed only a low to moderate activity count rate when measured with the γ-probe.

Oncological follow-up after PSMA-RGS

In 24 patients (38.1%) complications related to the salvage surgery procedure were observed. However, no specific complication related to the PSMA-RGS procedure and no adverse reaction related to tracer injection were observed. In most patients the complications were regarded as mild (grade 1 as defined by Clavien Dindo classification) such as temporary incontinence, temporary paralytic bladder with prolonged micturition or postoperative lymphedema. Six patients (9.5%) presented with high-grade complications (grade 3b according to Clavien-Dindo classification, e.g. injury of the ureter, perforation of the rectum) and further interventions were required. However, these complication rates are in concordance with results presented in literature [19]. The higher complication rates of salvage surgery procedures compared to the primary surgery setting are attributed not only to fibrotic changes after previous surgery and RT, but also reflects the often challenging location of recurrent PC lesions.

Follow-up information was available in 59/63 patients. A PSA reduction > 50% was observed in 46/59 (78%) patients and a reduction > 90% in 30/59 (51%) patients without additional treatments. In 38/59 (67%) patients a complete biochemical response with a PSA drop below 0.2 ng/ml was observed. Of these 38 patients, 17 patients (45%) are free of biochemical recurrence and without further therapy after a median follow-up of 12.3 (range 6.7–31.9) months. Duration of treatment-free survival as well as biochemical-free survival after PSMA-RGS is shown in Fig. 2.

Fig. 2
figure 2

PC-specific treatment-free survival and biochemical-free survival (PSA < 0.2 ng/ml) with 95% confidence interval after PSMA-RGS

Full size image

Prerequisites and appropriate patient selection for PSMA-RGS

First of all, a positive 68Ga-HBED-CC PSMA PET scan with preferably only singular soft tissue or lymph node recurrence after primary treatment is required. Here, standardisation of PET imaging protocols as well as interdisciplinary (urology, nuclear medicine and radiology) side-by-side analysis of 68Ga-HBED-CC PSMA PET can help to improve and standardise its evaluation and interpretation. A study of Rausch et al. has recently identified significant variations in FDG-PET/CT protocol parameters and system performance among different PET/CT users [22]. Subsequently, efforts need to be put in place to further standardise imaging protocols. At a minimum clinical PET/CT operations should ensure compliance with existing guidelines. Therefore, the recently published joint first EANM and SNMMI procedure guidelines for PC imaging can offer assistance to improve the quality of 68Ga-HBED-CC PSMA PET examinations [23].

Although PSMA-RGS seems to reduce tumour burden effectively reflected by the high drop of postoperative PSA level and might, therefore, delay additional systemic treatment, it represents still an individual and not guideline-conform treatment concept without proven benefit on overall surveillance or cancer-specific survival. Therefore, the morbidity and the potential surgical complications of PSMA-RGS have to be weighted against the potential delay of systemic treatment. Individual assessment of each patient prior to PSMA-RGS with critical discussion with each patient is mandatory. Recently, several factors identifying suitable candidates for salvage lymphadenectomy have been defined in order to achieve acceptable cancer control rates and to avoid unnecessary morbidity including a preoperative PSA value < 4 ng/ml, a primary Gleason score < 8 and metastases limited to the pelvis [24, 25]. In a study with patients undergoing secondary lymphadenectomy, it was observed that the 5-year clinical recurrence-free survival was higher for patients with preoperative PSA values below 4 ng/ml compared to patients with a PSA ≥ 4 ng/ml (48 versus 13%) [25, 26]. Further, 5-year clinical recurrence-free survival was lower for patients with positive LN in the retroperitoneum compared to patients revealing positive LN only in the pelvis (11 vs 53%; p < 0.001) [25, 26].

Correspondingly, all of our 17 patients presenting without biochemical recurrence (PSA < 0.2 ng/ml) so far showed a preoperative PSA < 4 ng/ml (median 1.14 ng/ml, range 0.28–3.29) and no extra-pelvic tumour manifestation. Interestingly, 6 of those 17 patients presented with tumour manifestation in more than one localization (e.g. lymph node metastases adjacent to the external iliac artery on both sides). Therefore, constraining salvage surgery procedures only to patients with singular metastases within the pelvis might not be necessary in the era of PSMA-PET imaging. Jilg et al. found a primary Gleason score of 8–10 to be an independent predictor of clinical progression [27]. However, 7 of our 17 recurrence-free patients presented with an initial Gleason score of 8 or 9. Thus, further studies are necessary to confirm relevant clinical parameters for appropriate patient selection.

Adjuvant radiotherapy after PSMA-RGS?

So far, it remains uncertain whether patients can profit from additional adjuvant RT after PSMA-RGS. Rischke et al. evaluated retrospectively the use of adjuvant RT after salvage lymphadenectomy versus salvage lymphadenectomy alone in patients with recurrent PC restricted to pelvic/retroperitoneal lymph node metastases after primary therapy. Adjuvant RT after salvage lymphadenectomy resulted in overall significantly improved 5-year relapse-free survival compared to patients who received salvage lymphadenectomy alone (5-year relapse-free rate 70.7 versus 26.3%) [28]. However, in this study salvage surgery was based on conventional imaging or choline-based PET examinations. Thus, currently, it remains unclear which parameters should trigger adjuvant radiation therapy.

Future vision of robotic PSMA-RGS with drop-in γ-probe

Currently PSMA-RGS procedures at our institution are performed in an open surgical fashion. Reasons for this are oftentimes challenging intraoperative conditions precluding the use of available laparoscopic γ-probes with limited manoeuverability and the lack of γ-probes for robotic intervention. Recently, van Oosterom et al. described a drop-in γ-probe to be inserted via a trocar that can be picked up and manoeuvered by a robotic grasper [29]. However, evaluation of this medical device for intraoperative use is warranted to determine its usability for complex PSMA-RGS procedures requiring subtle and extensive resections.

In summary, PSMA-RGS proved to be of high value for intraoperative detection and identification of small soft tissue metastases. In patients with biochemical recurrent PC, salvage surgery using PSMA-RGS has the potential to delay further disease progression and the need for further systemic treatment (e.g. ADT, chemotherapy). However, identification of suitable patients on the basis of PSMA-PET as well as clinical parameters is crucial to obtain satisfactory results.