Abstract
To provide surgeons with optimal guidance during interventions, it is crucial that the molecular imaging data generated in the diagnostic departments finds its way to the operating room. Sentinel lymph node (SLN) biopsy provides a textbook example in which molecular imaging data acquired in the department of nuclear medicine guides the surgical management of patients. For prostate cancer, in which SLNs are generally located deep in the pelvis, procedures are preferably performed via a (robot-assisted) laparoscopic approach. Unfortunately, in the laparoscopic setting the senses of the surgeon are reduced. This topical review discusses technologic innovations that can help improve surgical guidance during SLN biopsy procedures.
Metastasis in pelvic lymph nodes (LNs) is considered an important prognostic factor in prostate cancer. Prostate-specific antigen levels, pathologic stage, and Gleason score are predictors for LN involvement; the higher these factors are, the greater is the chance of nodal involvement. Postoperative (histo)pathologic examination of tissue samples obtained during (extended) pelvic lymphadenectomy is considered the gold standard in assessing metastatic spread. With an increasing LN dissection template, the prognosis of both N0 and N1 groups increases (Will Rogers phenomenon). Unfortunately, (extended) pelvic lymphadenectomy also increases the chance of postoperative complications such as lymphoceles, injuries to the obturator nerve or the ureter, and lymphedema of the lower extremity. Such complications can lead to a decrease in the patient’s quality of life.
Sentinel LN (SLN) biopsy focuses on the identification, subsequent minimally invasive excision, and pathologic and histopathologic evaluation of the LNs that drain directly from the primary tumor. Assuming the orderly spread of tumor cells through the lymphatic system, SLN biopsy can be used for LN staging. After staging, therapeutic follow-up can be decided on.
The potential of SLN biopsy for detecting LN metastasis has been validated in several studies. The Augsburg group validated the SLN biopsy procedure in more than 2,000 patients with prostate cancer and reported a high sensitivity and an overall false-negative rate of 5.9% (1). Moreover, SLN biopsy allows the identification of SLNs outside the pelvic lymphadenectomy field (2–4). Recently, Joniau et al. showed that 44% of SLNs were located outside the extended pelvic lymphadenectomy field; in 6% of patients, a positive LN was located exclusively in the presacral or paraaortic region (2).
Ideally, a surgeon is able to identify and excise the preoperatively identified SLNs in a minimally invasive manner, with a high sensitivity and specificity. This topical review discusses technologic improvements that may help improve the different aspects involved in (robot-assisted) laparoscopic SLN biopsy for prostate cancer; SLN biopsy for the prostate is often performed in combination with laparoscopic radical prostatectomy. Potential improvements can be found in (hybrid) tracers that are radioactive and fluorescent, the injection procedure, preoperative SLN identification and planning of the surgical procedure, translation of the preoperatively acquired imaging data to the operating room (e.g., via navigation), and intraoperative imaging for SLN identification. A schematic overview of these points is given in Figure 1. Similar technologies are also expected to help improve guidance for other SLN indications and in the future may even help enable tumor-specific resections.
Schematic overview of the integrated hybrid SLN biopsy procedure. On presentation of the patient, a hybrid SLN tracer (1) is injected into the prostate (2). Preoperative imaging is performed to identify the SLNs (3). Preoperatively acquired images can be directly translated to operating room—for example, via augmented reality-based navigation (4)—to provide both radio- and fluorescence-based surgical guidance toward SLNs (5). FLU = fluorescence imaging; λem = emission wavelength of the fluorophore; λex = excitation wavelength of the fluorophore; γ = γ-signal coming from the radioisotope; RA = radioactivity-based detection (i.e., γ-imaging/tracing).
(HYBRID) TRACERS
Despite the success of radioguided surgery, to provide optical guidance blue dyes are often injected before the start of the operation. However, for prostate cancer, in which SLNs are generally localized deep within the tissue, blue dye is of little value. An alternative optical detection technology can be found in near-infrared (NIR) fluorescence imaging, which allows for real-time optical detection of lesions less than 10 mm deep (5). The requirement of preoperative SLN mapping data for surgical planning, however, dictates that fluorescence guidance has to be used in combination with the more common radioguided procedures.
Conventional SLN mapping is performed using 20- to 600-nm radiocolloids. Because of their size, these exogenous compounds are recognized by the immune system, leading to accumulation in the SLN (6). We found that premixing of the clinically approved NIR dye indocyanine green and an albumin-based radiocolloid (99mTc-nanocolloid) yields the noncovalent indocyanine green–99mTc-nanocolloid complex. This complex has migratory properties similar to the parental 99mTc-nanocolloid (7). Other hybrid nanoparticles also have the potential to guide SLN biopsy (8). A recent preclinical example of a hybrid SLN tracer can be found in 99mTc-tilmanocept labeled with the NIR dye Cy7 (9). Alternatively, Cerenkov imaging of positron-emitting radionuclides has been proposed as a hybrid imaging technology; Thorek et al. demonstrated that LNs could be detected after a subdermal injection of 18F-FDG in the tail of a mouse (10).
Ideally, for more accurate LN staging, direct identification of nodal metastases would be preferred. Research is now focusing on the introduction of hybrid tracers that specifically target tumor tissue (11,12)—for example, by targeting prostate-specific biomarkers such as prostate-specific membrane antigen or the gastrin-releasing peptide receptor. In this light, a hybrid prostate-specific membrane antigen tracer, labeled with 111In and the NIR dye CW800, was shown to facilitate radioactivity- and fluorescence-based detection of prostate-specific membrane antigen–overexpressing tumors in mice (13).
INJECTION PROCEDURE
Before SLN biopsy, transrectal ultrasound guidance is used to direct the tracer deposition toward the peripheral zone of the prostate (Fig. 2A). Nevertheless, a recent study showed that in only 53% of patients was the tracer actually deposited in this area (14). Interestingly, the same study also suggested that the location of tracer deposition influences the lymphatic drainage pattern. The main question is whether, in order to identify the true tumor-draining SLNs, the tracer should be injected randomly in the peripheral zone of the prostate. It might be better to aim for peri- or intratumoral tracer deposition as is common in, for example, breast cancer and melanoma.
Preoperative SLN mapping. After tracer injection in the peripheral zone of prostate (A), planar lymphoscintigraphic images are acquired to identify the SLNs (B). SPECT/CT imaging allows identification of the anatomic location of the SLNs and, in some cases, identification of SLNs outside extended pelvic lymphadenectomy field; here, a paravesical SLN (white arrow) not clearly detectable on the lymphoscintigraphic image was identified after SPECT/CT imaging (C and D). In this patient, 5 SLNs were detected, with 2 being located outside extended pelvic lymphadenectomy field (white and pink arrows). 3D-VR = 3D volume rendering.
Multiparametric MR imaging (T2-weighted, contrast-enhanced, and diffusion weighted) was shown to be promising in the identification of localized prostate cancer (15). Integrating such MR imaging information with real-time–acquired contrast-enhanced transrectal ultrasound may allow MR imaging–based navigation of injection needles toward the intraprostatic tumor foci.
PREOPERATIVE SLN IDENTIFICATION AND PLANNING OF THE SURGICAL PROCEDURE
After tracer injection, obtaining sequential anterior (Fig. 2B) and posterior lymphoscintigraphic images is recommended in order to differentiate early draining SLNs from higher-echelon nodes (16). Via the introduction of SPECT/CT imaging, the 3-dimensional (3D) distribution of the radiocolloid can be directly placed in the anatomic context provided by the CT component (Figs. 2C and 2D). Moreover, with SPECT/CT imaging, SLNs not seen on lymphoscintigraphic images, such as those in the presacral region, can be identified (17). As such, preoperative identification of the lymphatic drainage patterns allows surgeons to decide beforehand on the optimal and least invasive surgical approach.
PET/CT, PET/MR imaging, or multiparametric MR imaging may, in the future, also be of value in planning surgical procedures. For example, LN mapping can be performed using radiocolloids suitable for PET imaging, via the direct identification of LN metastasis using targeted PET tracers or using ultrasmall superparamagnetic iron oxide particles or targeted dendrimers suitable for MR imaging.
TRANSLATION OF THE PREOPERATIVELY ACQUIRED IMAGING DATA TO THE OPERATING ROOM
Ideally, preoperatively acquired 2-dimensional (2D) and 3D imaging data can be directly translated into the operating room to help navigate the surgeon to the areas of interest. Improvement is especially desired in localization of SLNs near vital structures. Most straightforward is 2D navigation provided by a portable γ-camera (18). By placing a 125I-seed on a laparoscopic γ-probe and performing dual-isotope γ-imaging, it is possible to surgically navigate a laparoscopic γ-probe toward the SLNs (Fig. 3C) (18).
Intraoperative SLN identification. Via patient and tool tracking (yellow square and circle, respectively) (A), preoperatively acquired images can be translated into the operating room via 3D virtual reality navigation (B). A portable γ-camera allows visualization of the radioactive hot spots in 2D whereby a 125I-seed placed on the tip of the laparoscopic γ-probe (white circle) can be used for navigation (C). Hybrid tracers allow SLNs to be acoustically traced using a γ-probe (D) and optically detected via fluorescence imaging (E). RARP = robot-assisted radical prostatectomy.
The freehand-SPECT technology, which is based on real-time tracking of both the patient and a γ-probe, enables the generation of intraoperative 3D SPECT data that can be viewed in augmented or virtual reality (i.e., mixed reality) as an improvement over 2D imaging (19). Alternatively, virtual reality images can be generated using segmentation of SPECT scans (Fig. 3B), CT scans (20) or using transrectal ultrasound images (21). For robot-assisted procedures, one can load virtual reality images and/or the preoperatively acquired 3D images into the TilePro function of the da Vinci robot (Intuitive Surgical, Inc.). By attaching a 3D motion controller, it even becomes possible to manually manipulate these images (22).
For mixed reality–based navigation in soft tissue, organ deformation and movement of organs or the cameras can be a serious problem. Hence, one must compensate for such movements by placing internal (20) or external (23) navigation aids (Fig. 3A). A disadvantage of that approach is that it is difficult to correct for motion and organ deformation. The fluorescence signature of a hybrid tracer, in combination with the tracking of a fluorescence endoscope, can potentially be used to compensate for navigation errors of less than 10 mm (23).
INTRAOPERATIVE SLN IDENTIFICATION
Intraoperative SLN identification is traditionally facilitated via acoustic γ-tracing (Fig. 3D). With the introduction of robot-assisted laparoscopic procedures, not only are the senses of the surgeon reduced but also new challenges are faced for intraoperative γ-tracing. For example, the reduction in the movement of the γ-probe reduces the spatial accuracy of this technology even further. This is particularly problematic in areas near the injection site, where the high background signal hinders SLN identification. With regard to a 2-d protocol, the natural decay of the radioactive signal over time can reduce the detection sensitivity, which is already relatively low for the prostate.
Hybrid tracers such as indocyanine green–99mTc-nanocolloid can be used to extend the conventional radioguided surgical procedure with the benefits that NIR fluorescence imaging has to offer (24). With this tracer, γ-tracing can be used to obtain a rough localization of the SLN while the increased spatial resolution provided by the fluorescent signature enables accurate delineation of the SLN (Fig. 3E). An excellent example of the spatial information that fluorescence imaging provides during laparoscopic surgery was demonstrated by Jeschke et al., who showed lymphatic tracts draining from the prostate to the SLN, as well as the SLN itself (25).
PATIENT BENEFIT
We envision that symbiosis between the above-mentioned surgical guidance technologies may, in the future, provide patient benefit. Diagnostic images may help surgeons to select the least invasive surgical approach. Such planning should result in minimization of the exploration time. A one-to-one correlation between pre- and intraoperatively generated images helps validate complete excision of lesions. Finally, more accurate surgical identification of diseased and anatomic structures may result in further reduction of complications associated with nodal dissection.
CONCLUSION
Optimal use of interventional molecular imaging techniques is expected to lead to new surgical treatment paradigms for indications such as the SLN biopsy procedure for prostate cancer. One of the major challenges in the wide implementation of such technologies is their clinical translation. The first proof-of-concept studies, however, can provide a clinical basis for further improvements in this research field.
DISCLOSURE
The costs of publication of this article were defrayed in part by the payment of page charges. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734. This work was supported in part by a Dutch Cancer Society translational research award (PGF 2009-4344), a National Research Council NWO-STW-VIDI grant (STW BGT11272), and an Intuitive Surgical, Inc., clinical research grant. No other potential conflict of interest relevant to this article was reported.
Footnotes
Published online Mar. 14, 2013.
- © 2013 by the Society of Nuclear Medicine and Molecular Imaging, Inc.
REFERENCES
- Received for publication December 10, 2012.
- Accepted for publication February 21, 2013.