Background

Prostate cancer (PCa) represents the third-ranked highest cause of death in Europe with very high associated economic costs and morbidity [1]. The need for effective and early curative treatment options is substantial. Defining the presence and extent of PCa metastases is of paramount importance in a curative surgical setting. Radical prostatectomy (RP) with or without lymph node excision is the current standard of curative surgical approaches to localised PCa. Lymph node excision is performed in roughly one-third of RP patients [2]. Predictive nomograms are used to define the indication for a nodal dissection [3]. Due to a complex lymphatic drainage pattern for PCa the golden standard of nodal dissection has been extended pelvic lymph node dissection (ePLND) for patients with a > 5% risk of nodal metastases based on externally validated preoperative nomograms such as the Briganti nomogram [4]. This dissection entails excision of the nodes located at the external iliac artery, within the obturator fossa, at the internal iliac artery, and the nodes overlying the common iliac artery and vein up to the ureteral crossing according to the European Association of Urology (EAU) guidelines [5].

Sentinel lymph node biopsy (SLNB) was first introduced for prostate cancer in 1999 by Wawroschek et al. [6]. A sentinel lymph node (SLN) refers to any lymph node that receives direct lymphatic drainage from the primary injected site (i.e., tumor). For malignancies such as melanoma, breast cancer and penile cancer, sentinel node biopsy is an established diagnostic modality for detection of nodal metastases while avoiding undesirable side effects of extended lymphadenectomies [3].

The lymphatic drainage pattern of the prostate is complex with pelvic lymphatic drainage pathways draining to a myriad of anatomical regions, including the external iliac group, the obturator fossa, the internal iliac group, anterior to the bladder and lastly via a presacral route anterior to the sacrum, coccyx and the perirectal lymphatic plexus. This last route can result in metastases as far up as the medial chain of the common iliac nodes [7].

These lymphatic drainage routes represent a challenge in PCa due to the multiple possible sites for tumour dissemination, hereby giving rise to the possibility of multiple sentinel nodes. A recent systematic review presented SLNB as having a comparable diagnostic accuracy to ePLND with substantially lower number of removed lymph nodes [8]. Furthermore the SLNB paradigm allows for the identification of sentinel nodes outside of the ePLND template [9]. In combination with ePLND, SLNB could thus improve staging and potentially aid in improving biochemical recurrence-free survival [7]. The presence of nodal metastasis has been associated with a significantly worse prognosis in prostate cancer [10].

The correlation between preoperative imaging and intraoperative findings is key for streamlining minimally invasive laparoscopic surgery. Apart from the location of the detected sentinel nodes, preoperative imaging may provide additional information on the signal intensity that could be expected during the intra-operative in vivo probe measurement. The aim of this study was therefore to correlate tracer uptake of sentinel lymph nodes as measured with single-photon emission computed tomography (SPECT/CT) and intra-operative probe measurements in patients who participated in the UZ Leuven cohort of a prospective multicentre clinical trial evaluating the SENSEI® drop-in gamma probe for minimally-invasive SLN dissection in prostate cancer (NCT04632251) [12].

Methods

From August 2021 to December 2021, nine patients with histologically proven prostate cancer categorised from intermediate to high risk (EAU risk group) underwent a radical prostatectomy with extended pelvic lymph node dissection as part of a prospective, open-label, multicentre clinical trial evaluating SENSEI® for SLND in PCa patients (ClinicalTrials.gov: NCT04632251) [12]. Ethical approval for this retrospective analysis was granted by the Ethics Committee Research UZ / KU Leuven (S66503).

The day before surgery (2-day protocol), 240 ± 10 MBq 99mTc-nanocolloid was administered transrectally under ultrasound guidance by an experienced uro-radiologist. The tracer with a colloid particle concentration of 0.5 mg 99mTc-nanocolloid per 1.0 ml was injected systemically in the basal, middle and apical part of both prostate lobes. In total 1.0 mL of tracer was administered to every prostate (0.5 mL per lobe). Following administration of the tracer, anterior and lateral planar lymphoscintigraphy images (LS) were taken of the lower abdomen 15–30 min post-injection. A SPECT/CT scan of the lower abdomen was acquired 2 h after injection (Symbia Intevo Bold, Siemens, Erlangen, Germany). Two nuclear medicine physicians trained in the interpretation of SLN images identified the number and location of the SLNs. The maximum counts of each SLN were measured with a spherical VOI using MIM Encore (MIM Software Inc., Ohio, USA). An example of SPECT/CT imaging is displayed in Fig. 1. SLNs on LS and SPECT/CT were defined as all nodes that appear first in each drainage basin as seen on the early images and nodes identified in new basins on late images that were not yet seen on the early images.

Fig. 1
figure 1

Overview of preoperative imaging and intraoperative SENSEI probe measurement. (A, B,C) Axial, sagittal and coronal SPECT images taken approximately 2 h after ultrasound guided transrectal intraprostatic injection of 240 MBq 99mTc-nanocolloid tracer. (D, E,F) Axial, sagittal and coronal fusion SPECT/CT images allowing anatomical location of a sentinel lymph node. The red arrow denotes the sentinel lymph node located in the left internal iliac region. (G) Surgical view of robot-assisted radioguided SLN detection in the left hemipelvic region. Integration of the gamma probe signal in the surgical view can be seen in the bottom right corner

Full size image

The following day a radical prostatectomy with extended pelvic lymph node dissection was performed as per standard of care. A urologist specialised in robot-assisted surgery performed all of the procedures using a four-armed surgical robotic da Vinci system (Xi, Intuitive Surgical, Sunnyvale, USA). For SLN detection, the SENSEI® drop-in gamma probe for minimally invasive SLN dissection (Lightpoint, a Telix company, London, UK) was used. An in vivo background measurement was recorded with the probe directed away from the injection site and the bladder, followed by a systematic bilateral scan of the pelvic area to identify areas of increased activity (an example of the surgical view is provided in Fig. 1). Intra-operatively, SLNs were identified as any lymph node specimen in the pelvic area that had been either identified as a SLN on the preoperative image and which had an in vivo signal above background, or any lymph node not previously identified as an SLN but with an in vivo signal greater or equal to 10% of the SLN specimen with the highest signal. The anatomical location and standardised in vivo count rate (counts per second; cps) of each SLN was recorded (single measurement). The SLN were excised based on the surgeon’s discretion. After completion of the SLNB procedure a standard of care ePLND was performed using the EAU ePLND template definition. Standardised ex vivo measurements with the SENSEI probe were performed after removal of the SLN (single measurement). SLN detection and in vivo and ex vivo count rate measurements were completed within 24 h of the original injection of the tracer.

Patient characteristics/categorical data were presented as absolute values and percentage frequencies. The number of SLNs removed and their intraoperative locations were gathered. Paired t-testing and Pearson correlation coefficient was used to compare the two types of uptake measures. P-values less than 0.05 were considered as significant.

Results

Patient characteristics

From August 2021 to December 2021, nine patients with histologically proven prostate cancer categorised from intermediate to high risk (EAU risk group) were prospectively included in the trial in UZ Leuven [12]. Table 1 shows the baseline characteristics of these nine patients. Median age and BMI at surgery was 64 years (IQR 60.5–67.5) and 24.6 (IQR 22.9–26.3), respectively. Median PSA at diagnosis was 8 ng/mL (IQR 6–10). The patients were classified by the International Society of Urological Pathology (ISUP) grade group 3 to 5 and had a clinical T stage of T2a to T3a.

Table 1 Baseline patient characteristics
Full size table

Sentinel node procedure, imaging and surgery

The mean 99mTc-nanocolloid activity administered was 237 ± 6.1 MBq. Planar lymphoscintigraphy and SPECT/CT imaging was performed in all 9 patients. Planar lymphoscintigraphy resulted in the visualisation of nodes in only one patient (2 LNs, located in the presacral and para-aortic region). SPECT/CT however was able to identify at least 1 SLN in all 9 patients. Bilateral detection was seen in 4 patients with SPECT/CT and in 4 with SENSEI gamma-probe. Also using the SENSEI® drop-in gamma probe for minimally invasive SLN dissection, at least 1 SLN was identified in all patients (success rate 100%).

On preoperative SPECT/CT imaging, 32 SLNs were visualized (median 3 per patient, range 2–6), while using the SENSEI® drop-in gamma probe 29 SLNs were detected (median 3, range 2–5, p = 0.5). The anatomical locations of visualised SLNs on a patient basis are outlined in Table 2. In four patients, there was a mismatch between preoperative imaging and intraoperative findings. In three patients, SLNs were identified on preoperative SPECT/CT imaging (4 SLNs in total) that were not resected due to localisation outside the ePLND template and related operative surgical risk. In one patient two adjacent resection specimens could be matched to one preoperative imaging result. Seven of the 29 lymph nodes identified by the SENSEI® drop-in gamma probe were located outside of the ePLND template.

Table 2 Overview of SLN detection per patient using SPECT/CT and SENSEI® drop-in gamma probe
Full size table

The maximal counts on SPECT/CT had a median value of 5550 (range 501–37042), while the median signal using the in vivo gamma probe measurement was 103 cps (range 27–323) and the ex vivo gamma probe measurement 84 cps (range 0.3–335). The correlation between SPECT/CT counts and standardised in vivo and ex vivo probe measurements was significant but moderate (Pearson r = 0.57, p = 0.002 and r = 0.64, p = 0.0003, respectively). Figure 2 illustrates the correlation between SPECT/CT counts and in vivo and ex vivo SENSEI gamma probe measurements, respectively.

Fig. 2
figure 2

Correlation of counts of SLNs on SPECT/CT and intraoperative SENSEI drop-in gamma probe measurements. The comparison shows a moderate linear correlation (black line)

Full size image

All patients underwent RP with ePLND. In two of the surgeries, the RP was performed before SLND and ePLND. In 7 patients the lymph node dissection was carried out first. The prostate count rate was measured ex vivo in 8/9 patients with a median of 627 cps (range 235–1651).

Two patients exhibited lymph node metastases after ePLND, totalling 3 positive lymph nodes on histopathological analysis. Of the 29 lymph nodes excised as part of the SLNB, none were shown to be positive for prostate cancer. The first patient had metastases located at the left sided external iliac artery (8 mm diameter) and obturator fossa (4 mm diameter). The second patient had a 3 mm lymph node metastasis located in the left obturator fossa. In both patients, preoperative imaging and intraoperative probe measurements did not demonstrate tracer uptake at these locations.

Discussion

In this study, we have shown the good intraoperative detection rate of SLN using the SENSEI® drop-in gamma probe for minimally invasive SLN dissection in the subgroup of patients that were treated in our centre as part of the recently completed multicentre clinical trial.12 The multicentre clinical trial demonstrated that the SENSEI® drop-in gamma probe was able to detect SLNs in all 27 patients included (100%) [12]. In our single centre subset of 9 patients (7 of which were included in the multicentre trial), we also found a 100% success rate, identifying at least 1 SLN in each patient. There was a positive correlation between preoperative SPECT/CT findings and intraoperative SENSEI® drop-in gamma probe measurements, opening up new horizons for the intraoperative use of drop-in gamma probes for sentinel lymph node detection. Recently, SLN detection was also shown to be safe and feasible in patients with early-stage cervical cancer using the SENSEI® drop-in probe [13]. The potential applications for the use of drop-in gamma probes, such as the SENSEI probe in minimally invasive surgical settings are rapidly increasing.

Our study showed a moderate correlation between SPECT/CT and intraoperative probe measurements. This result is most likely accounted for due to the significant time difference (18 to 20 h) between measurements. The lymphatic drainage pattern for prostate cancer is complex and certainly not static between measurements, with a potential for both tracer washout from nodes with high uptake, and delayed tracer supply to nodes with low uptake at the time of SPECT/CT. To our knowledge, this is the first correlation analysis performed between preoperative SPECT/CT measurements and intraoperative probe measurements for sentinel lymph node biopsy in prostate cancer or other tumour types. Intraoperative gamma probe performance is dependent on target to background ratios, which can be influenced by clearance kinetics, probe design and the choice of radiopharmaceutical. Further studies are necessary to better understand the factors influencing the correlation between preoperative SPECT/CT and intraoperative probe measurements. In our centre, sentinel lymph node detection appeared to be independent of the timing of radical prostatectomy, with no negative effect of having the prostate (with high tracer concentration) in situ during ePLND. Indeed, in 7/9 patients, the SLND and subsequent ePLND was performed before radical prostatectomy. This good performance in the presence of high background activity could be attributed to the increased manoeuvrability of the SENSEI probe allowing accurate positioning of the probe when high background signals are present. Since the majority of the operations were performed with the prostate in situ, this exemplifies the directional sensitivity and shielding of the SENSEI gamma probe. Lymph nodes adjacent to the prostate, such as obturator nodes, could still be distinguished with a higher than background signal.

This study has a number of limitations. Firstly, the number of patients was limited. This is most clearly evident in the histopathological analysis. Only two patients exhibited lymph node metastases. In both patients preoperative imaging and intraoperative probe measurements did not demonstrate tracer uptake at these locations. This apparent low per node sensitivity is reflected in a recent narrative review providing an overview of the diagnostic accuracy measures and the oncological outcomes of SNLB in prostate cancer [14]. Rossin et al. showed overall, median per patient sensitivity was 92.5% (IQR 82.8–100.0; 24 studies included with 2090 patients) compared to 65.4% per node sensitivity (IQR 51.5–74.2; 8 studies included with 642 patients). The low per node sensitivity exhibited in our subset and the larger multicentre trial could possibly be explained by the use of a non-tumour specific tracer. Intravenous injection of 99mTc-PSMA would bypass the necessity of transrectal injections, thereby simplifying the workflow and increasing patient comfort and reducing background activity in the prostate. Furthermore, the signal would directly reflect tumour burden and not mere lymphatic drainage. Tumour targeted procedures also lend themselves towards combination with preoperative PSMA PET/CT. PSMA-targeted radiopharmaceuticals in PET imaging are increasingly being used in a primary staging and restaging after biochemical recurrence in high risk prostate cancer [15]. Mauer et al. described PSMA radioguided surgery in the setting of primary treatment and salvage [16]. Gondoputro et al. has shown that 99mTc-PSMA radioguided surgery was safe and aided in the resection of PSMA- avid (out of template) lymph nodes and residual disease at the prostate bed in a primary treatment setting [17]. The lymph drainage pattern for the prostate is complex. The combination of ePLND and SLNB may however enable more accurate staging in patients exhibiting aberrant, out of template drainage pathways, as has been suggested in a significant number of patients [18].

This study reaffirms intraoperative detection of suspected out of template lymph nodes and notably provides ex vivo confirmation of successful resection of preoperatively visualised SLNs on SPECT/CT. Seven of the 29 (24%) identified lymph nodes by the SENSEI gamma probe were located outside of the ePLND template. This result is comparable to a recent analysis by Dell’Oglio et al. showing 17% out of template nodes [19]. Our study also demonstrates the value of ex vivo confirmation of successful resection of preoperatively visualised SLNs on SPECT/CT by illustrating a moderate concordance between in vivo, ex vivo and preoperative SPECT/CT counts.

Lastly, the use of planar lymphoscintigraphy was shown not to be useful, considering it failed to demonstrate lymph nodes in 8/9 patients. This could be due to the scan time being 15–30 min after injection, possibly not allowing sufficient time for tracer uptake in the lymph nodes. Perhaps if planar imaging was repeated 2 h after injection, it would have been more successful. This result mirrors the experience at the other 2 centres in the larger multicentre trial where the anatomical location of the SLNs for 16/27 patients could not be determined precisely on planar lymphoscintigraphy. SPECT/CT was however able to identify lymph nodes in all 9 patients. Due to the higher resolution and anatomical localisation with SPECT/CT, we would recommend performing SPECT/CT where available in lieu of planar LS.

Conclusions

The SENSEI drop-in gamma probe for minimally invasive SLN dissection is a valuable tool in the intraoperative detection of SLNs. The counts measured in SLNs on SPECT/CT had a moderate concordance with the uptake measured in vivo using the SENSEI® drop-in gamma probe. Furthermore this study reaffirms intraoperative detection of suspected out of template lymph nodes and notably provides ex vivo confirmation of successful resection of preoperatively visualised SLNs on SPECT/CT.