Through the application of a magnetic fluorescent hybrid tracer, we were able to combine pre-operative MRI and magnetometer-guided as well as NIR fluorescence imaging guided PLND in PCa patients. We could pre-operatively identify SLNs with SPION uptake on pelvic MRI scans as well as intra-operatively detect magnetic SLNs by a handheld magnetometer probe and fluorescing SLNs by NIR fluorescence imaging in all patients. Concordance of the different techniques was 70% for pre-operative MRI vs. intra-operative magnetometer-guided PLND and 88% for intra-operative magnetic vs. fluorescent SLN detection.

We could identify SLNs with SPION uptake on pre-operative pelvic MRI scans in all patients. As shown in previous studies on breast cancer [12], PCa [13, 14], and penile cancer [19], pre-operative SLN visualisation on MRI was thus generally feasible. In our study, a relatively high number of possible SLNs per patient has been pre-operatively identified on MRI scans which is in line with our previous studies using the SPION tracer alone with medium 17.5 (IQR 12‒22.5) SLNs pre-operatively visualised and 9 (6‒12) SLNs surgically removed [13]. Studies using a radioactive SLN tracer identified median 6 (IQR 3‒9) SLNs on pre-operative pelvic SPECT/CT lymphoscintigraphy and 4 (2‒6) SLNs were intra-operatively detected with a gamma probe and surgically removed [20]. Similar pre-operative visualisation and intra-operative detection rates could be observed using the radioactive fluorescent hybrid tracer with median 3 (IQR 2‒4.5) SLNs visualised and 2 (1‒4) SLNs surgically removed [3]. Due to the fundamentally different chemical and physical properties of the different imaging agents, however, a direct comparison has to be interpreted with caution. As discussed earlier, MRI might identify more SLNs due to its high spatial resolution and its high sensitivity to very small concentrations of SPION [13]. This could possibly lead to an overestimation of the number of pelvic SLNs with SPION uptake identified on pre-operative pelvic MRI scans. Due to the high number of lymph nodes mapped, it is challenging to match individual lymph nodes identified pre- and intra-operatively on a one-to-one basis. We therefore calculated the concordance between pre- and intra-operative SLN detection based on the individual anatomical regions within our PLND template, resulting in a concordance of 70%. Considering the small number of patients included in our study, this concordance rate has to be interpreted carefully. Despite these clear limitations, our results suggest that pre-operative SLN visualisation on MRI may provide the surgeon a lymphatic roadmap for PLND comparable to SPECT/CT lymphoscintigraphy using the conventional radioactive tracer.

During PLND, we could detect magnetically active as well as fluorescent SLNs in each patient. The number of resected magnetically active SLNs did not differ from that reported in our previous studies [10, 21] suggesting that the addition of ICG to SPION did not alter lymphatic uptake and distribution properties of the tracer. Among the resected lymph nodes, there were more fluorescent nodes identified than magnetically active SLNs. It is likely that the magnetic fluorescent hybrid tracer may also contain “free” ICG which has not bound to SPION due to the non-covalent binding reaction [15]. ICG alone acts as a lymphatic agent that, because of its small hydrodynamic diameter of 1.2 nm, has the potential to migrate beyond SLNs which may result in marking of secondary lymphatic landing sites [22]. In our study, only one specimen had been identified solely by NIR fluorescence during surgery. The higher amount of lymph nodes marked by ICG than by SPION became apparent only after ex situ preparation and re-measurement of all resected specimens. Our data thus suggest that intra-operative guidance of PLND by NIR fluorescence imaging may not necessarily lead to a more extensive surgical removal of lymph nodes considering a hybrid tracer approach.

As seen in applications of the radioactive fluorescent hybrid tracer, intra-operative NIR fluorescence imaging facilitated the identification of SLNs located in lymphatic fatty tissue [7]. A more favourable signal to background ratio might furthermore improve SLN detection during surgery. Especially in periprostatic regions where strong signal originating from the prostatic injection sites may mask SLN identification using the magnetic modality alone the hybrid approach could be advantageous due to a lower tissue penetration depth of the fluorescence signal. According to our previous studies using the SPION tracer, about 4% of the pre-operatively visualised SLNs on MRI were located periprostatically [13] whereas only about 0.3% of the surgically resected SLNs stemmed from periprostatic regions pooled over more than 800 patients who underwent magnetometer-guided PLND in our centre [10]. The spatial resolution of the magnetometer probe used in our studies is about 2 mm [9] which can be helpful in identifying SLNs located in deep tissue but can be misleading in areas with strong background signal. Considering the small number of patients included in this pilot study, the additional diagnostic value of applying a magnetic fluorescent hybrid approach cannot be evaluated and should be investigated in future studies with a larger patient cohort.

There is ongoing debate on the definition of SLN diagnostics in PCa [8]. In line with our previous studies using the SPION tracer [10, 11, 13, 21], we applied a wider SLN definition than is generally used in other studies or other tumor entities using the conventional radioactive or the radioactive fluorescent hybrid tracer [3, 20]. Due to the rather complex and highly variable lymphatic drainage pattern in PCa when compared to other tumour entities, our protocol is designed to perform a lymphatic mapping of the whole prostate and to surgically resect any magnetically active node [10, 11, 13, 21]. A retrospective analysis of our data on magnetometer-guided PLND in intermediate and high risk PCa patients revealed that the magnetic activity of the resected lymph nodes did not correlate with lymph node involvement [21]. Instead, about 17% of node positive patients would have been missed when applying the 10% rule based on node magnetic activity level [21]. In this pilot study, however, no lymph node metastases could be detected and conclusions on the diagnostic accuracy of our hybrid approach cannot be drawn.

If the reliability of intra-operative NIR fluorescence guidance in combination with pre-operative MRI can be confirmed in future studies with larger sample sizes, the application of the magnetic fluorescent hybrid tracer for SLN visualisation could be especially useful in minimally invasive, robot assisted surgeries. To our knowledge, there are no magnetometer probes for laparoscopic use commercially available, yet, but only in pre-clinical testing [16, 23, 24]. The magnetic fluorescent hybrid tracer could thus be a radiation-free alternative to the radioactive fluorescent hybrid tracer for bimodal SLN mapping in PCa.