Abbreviations

and Acronyms

CLE ¼ confocal laserendomicroscopy

ECE ¼ extracapsular extension

FOV ¼ field of view

H&E ¼ hematoxylin and eosin

MPM ¼ multiphoton microscopy

NVB ¼ neurovascular bundle

OCT ¼ optical coherencetomography

RARP ¼ robotic assisted radicalprostatectomy

Accepted for publication November 2, 2015.No direct or indirect commercial incentive

associated with publishing this article.To view the accompanying video, please see

the online version of this article (Volume 195,Number 4) at www.jurology.com.

The corresponding author certifies that, whenapplicable, a statement(s) has been included inthe manuscript documenting institutional reviewboard, ethics committee or ethical review boardstudy approval; principles of Helsinki Declarationwere followed in lieu of formal ethics committeeapproval; institutional animal care and usecommittee approval; all human subjects providedwritten informed consent with guarantees ofconfidentiality; IRB approved protocol number;animal approved project number.

* Equal study contribution.† Correspondence: 300 Pasteur Dr., Room

S-287, Stanford, California 94305-5118 (telephone:650-858-3916; FAX: 650-849-0319; e-mail: jliao@stanford.edu).

‡ Supported by U.S. National Institutes ofHealth Grant R01 CA160986.

1110 j www.jurology.com

00

T

�

New Technology and Techniques

Intraoperative Optical Biopsy during Robotic Assisted RadicalProstatectomy Using Confocal Endomicroscopy

Aristeo Lopez,* Dimitar V. Zlatev,* Kathleen E. Mach, Daniel Bui,

Jen-Jane Liu, Robert V. Rouse, Theodore Harris, John T. Leppert

and Joseph C. Liao†,‡

From the Department of Urology (AL, DVZ, KEM, DB, JJL, JTL, JCL) and Department of Pathology (RVR),

Stanford University School of Medicine, Stanford, and Veterans Affairs Palo Alto Health Care System,

Palo Alto (AL, DVZ, KEM, DB, JJL, RVR, TH, JTL, JCL), California

Purpose: Intraoperative optical biopsy technologies may aid in the identificationof important anatomical landmarks and improve surgical outcomes of roboticassisted radical prostatectomy. We evaluate the feasibility of confocal laserendomicroscopy during robotic assisted radical prostatectomy.

Materials and Methods: A total of 21 patients with biopsy proven prostate cancerscheduled for robotic assisted radical prostatectomy were recruited. Afterintravenous administration of fluorescein 15 patients underwent in vivo intra-operative confocal laser endomicroscopy of prostatic and periprostatic structuresusing a 2.6 or 0.85 mm imaging probe. Standard robotic instruments were usedto grasp and maneuver the confocal laser endomicroscopy probes for imageacquisition. Confocal laser endomicroscopy imaging was performed ex vivo onfresh prostate specimens from 20 patients. Confocal video sequences acquiredin vivo and ex vivo were reviewed and analyzed, with additional image pro-cessing using a mosaicing algorithm. Processed confocal images were comparedwith standard hematoxylin and eosin analysis of imaged regions.

Results: Confocal laser endomicroscopy was successfully integrated with roboticsurgery, including co-registration of confocal video sequences with white lightand probe handling with standard robotic instrumentation. Intraoperativeconfocal laser endomicroscopy imaging of the neurovascular bundle before andafter nerve sparing dissection revealed characteristic features including dynamicvascular flow and intact axon fibers. Ex vivo confocal imaging of the prostaticparenchyma demonstrated normal prostate glands, stroma and prostaticcarcinoma.

Conclusions: We report the initial feasibility of optical biopsy of prostatic andperiprostatic tissue during robotic assisted radical prostatectomy. Image guid-ance and tissue interrogation using confocal laser endomicroscopy offer a newintraoperative imaging method that has the potential to improve the functionaland oncologic outcomes of prostate cancer surgery.

Key Words: prostatic neoplasms; prostatectomy; microscopy, confocal;

erectile dysfunction; surgery, computer-assisted

CANCER control and recovery of urinaryand sexual function after radical pros-tatectomy are related to surgical

22-5347/16/1954-1110/0

HE JOURNAL OF UROLOGY®

2016 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARC

quality.1 Since Walsh’s initial descrip-tion of anatomical radical prostatec-tomy there have been efforts to better

H, INC.

http://dx.doi.org/10.1016/j.juro.2015.10.182

Vol. 195, 1110-1117, April 2016

Printed in U.S.A.

CONFOCAL ENDOMICROSCOPY DURING ROBOTIC ASSISTED RADICAL PROSTATECTOMY 1111

understand pelvic anatomy to refine surgical tech-nique.2 Robotic assisted radical prostatectomy iscurrently the most common surgical treatment forlocalized prostate cancer in the United States.3

Technological advances of the robotic platforminclude amagnified field of view, tremor filtration andimproved surgeon ergonomics.4,5 Despite advancesin understanding pelvic anatomy and surgical tech-nologies, significant variation remains in the surgi-cal outcomes of radical prostatectomy, includingpositive surgical margins (range 6.5% to 32%)6 anderectile dysfunction (range 7% to 80%).7,8

Image guided surgery may improve intra-operative navigation and surgical outcomes. Opticalimaging technologies offer excellent spatial andtemporal resolution, are easily integrated into theoperating room and can be manipulated with in-struments commonly used in minimally invasivesurgery.9 For radical prostatectomy in vivo andex vivo feasibility studies have been reported usingnear infrared fluorescence imaging,10 OCT11e13 andMPM.14,15

Similar to OCT and MPM, confocal laser endo-microscopy is an optical biopsy technology that aimsto provide on demand, high resolution imagingreminiscent of standard histopathology.16 CLE isapproved for endoscopic applications in gastroen-terology, pulmonology and urology. CLE is based ona 488 nm laser in conjunction with fluorescein, aFood and Drug Administration approved fluo-rophore with a demonstrated safety record.17

Figure 1. Intraoperative CLE during robotic prostatectomy. CLE tower

(2.6 mm) inserted through 12 mm laparoscopic port alongside suctio

probe held by robotic needle driver (C ). TilePro functionality enab

stereoscopic view of operative field within surgeon console. Conf

inserted through 19-gauge angiocatheter (D). Confocal image shows

We have demonstrated the cystoscopic applica-tion of CLE for the optical diagnosis and grading ofbladder cancer,18,19 as well as in vivo visualizationof glandular structures in the prostatic urethra.20

We assessed the feasibility of intraoperative CLEduring RARP and evaluated potential clinical ap-plications. We developed an intraoperative confocalimaging protocol, characterized in vivo microscopicfeatures of prostatic and periprostatic anatomy, andcompared ex vivo imaging of fresh surgical prostatespecimens to histopathology.

MATERIALS AND METHODS

InstrumentationConfocal endomicroscopy was performed with Cellvizio�,and 2.6 or 0.85 mm outer diameter fiberoptic probes wereused for image acquisition (fig. 1, A). The 2.6 mm probehas a spatial resolution of 1 mm, a tissue penetrationdepth of 60 mm and FOV of 240 mm. The 0.85 mm probehas a spatial resolution of 3.5 mm, a penetration depth of50 mm and a FOV of 320 mm. Probes were sterilized beforeuse with the Sterrad� system.

Intraoperative CLE during Robotic AssistedSurgeryThe study was conducted with Stanford University insti-tutional review board and VAPAHCS (Veterans AffairsPalo Alto Health Care System) Research and Develop-ment Committee approval. Patients with clinically local-ized prostate cancer scheduled for RARP were recruited.Two surgeons (JCL and JTL) performed the operations

arrangement at head of operating room table (A). Imaging probe

n irrigator (B). Confocal imaging of NVB with 2.6 mm imaging

led simultaneous display of confocal image and white light

ocal imaging of divided bladder neck using 0.85 mm probe

vasculature of bladder lamina propria.

1112 CONFOCAL ENDOMICROSCOPY DURING ROBOTIC ASSISTED RADICAL PROSTATECTOMY

and image acquisition. Standard 5-port placement con-sisting of a 12 mm camera port, 3, 8 mm robotic ports anda 12 mm assistant port was applied. The decision fornerve sparing was based on clinical staging, technicalfeasibility and surgeon discretion. The majority of CLEimaging was performed with a 2.6 mm probe introducedthrough the 12 mm assistant port (fig. 1, B). The roboticneedle driver was used to grasp the distal metal tip forimaging (fig. 1, C ). For the 0.85 mm probe 3 strategieswere compared for intracorporeal maneuvering, including1) insertion via a standard laparoscopic cholangiogramcatheter holder operated by bedside assistant, 2) insertionvia a 5Fr angiocatheter through assistant port andgrasping using the robotic needle driver, and 3) insertionvia a 19-gauge angiocatheter introduced suprapubicallyas a needlescopic port and grasping using the roboticneedle driver (fig. 1, D).

Approximately 5 minutes before dissection of the neu-rovascular bundle, 2.5 ml 10% sodium fluorescein (Akorn,Lake Forest, Illinois) were administered intravenously.For imaging the probe tip was positioned perpendicular tothe tissue for en face contact and rinsed with irrigation asneeded to remove blood or debris. Images were acquiredas video sequences at 12 frames per second. TilePro� wasused to simultaneously view the white light stereoscopicview of the operative field and confocal imaging (fig. 1, Cand D). Prostatic and periprostatic structures, includinglevator fascia, NVB before and after the nerve sparingprocedure, prostatic capsule, bladder neck, urethralstump and pelvic floor were imaged in situ, reviewed inreal time and recorded for additional off-line analysis.

Ex Vivo CLE of Prostate SpecimensEx vivo CLE was performed within 1 hour of specimenretrieval. To optimize prostatic parenchymal staining anadditional 2.5 ml 10% fluorescein were administeredintravenously before the division of the prostatic pedicles.CLE image acquisition was performed by manual manip-ulation of the 2.6 mm probe. Imaged regions on the surfaceof the prostate included the prostatic capsule, posterolat-eral surface corresponding to the location of the NVB andapical margins. To characterize parenchymal structuresthe prostate was sectioned transversely with the assis-tance of a surgical pathologist (RVR). Each 5 mm thickprostate slice was divided into quadrants and systemati-cally imaged in a defined pattern. Additional fluoresceinwas applied topically (2 minute incubation then 7 minutesaline wash to remove excess fluorescein) to enhancevisualization. After imaging the tissues were fixed informalin and sent for hematoxylin and eosin stainingand histopathological analysis. Immunohistochemistryagainst S100 proteins was performed on select sections(Histo-Tec, Hayward, California) to identify nerves.

Data AnalysisIn vivo and ex vivo confocal video sequences werereviewed, edited and analyzed off-line using CellvizioViewer v1.6 software. A built-in mosaicing algorithm wasused to compile consecutive images into a single largercomposite image. Processed confocal images werecompared with corresponding hematoxylin and eosinstains and reviewed with a surgical pathologist (RVR).

RESULTSBetween December 2012 and March 2015, 21 pa-tients (mean age 62 years, range 49 to 69) scheduledfor RARP at VAPAHCS were recruited. Patientsunderwent bilateral (16) or unilateral (5) nervesparing RARP. In vivo CLE imaging was performedin 15 patients and ex vivo imaging was performedon 20 prostates. Patient characteristics and imagingdetails are described in the supplementary table(http://jurology.com/).

Overall 105 in vivo confocal video sequences from15 patients were collected. The average imageacquisition time was 10 minutes (range 3 to 18) perparticipant. An average of 7 video sequences (range4 to 12) was obtained per case. The average durationof imaging at each area was 91 seconds (range 6 to303). In 1 patient the metal tip of the 2.6 mm probebroke off during handling by the robotic instrumentand was removed with a laparoscopic grasperwithout complication. There were no adverse eventsrelated to fluorescein administration.

Comparison of Imaging Probes

We compared the intraoperative handling andimage quality of the 2.6 and 0.85 mm probes, testingthem in 11 and 4 participants, respectively. The2.6 mm probe was previously validated for bladdercancer imaging. The 0.85 mm probe was describedfor CLE of pancreatic cysts through a 19-gaugebiopsy needle21 and upper urinary tract throughstandard ureteroscopes.22 The smaller 0.85 mmprobe has the potential for greater flexibility forintraoperative probe deployment. While insertedthrough the 12 mm assistant port, both probes werecompatible with in-parallel insertion of additionalinstruments without significant loss of pneumo-peritoneum. To minimize trauma to the opticalfibers, the 2.6 mm probe was handled by graspingthe distal metal tip (fig. 1, C ), whereas the 0.85 mmprobe was inserted via a laparoscopic cholangio-gram instrument (1), 5Fr catheter (2) or a 19-gaugeangiocatheter for maneuvering with the roboticneedle holder (1) (fig. 1, D).

Overall the flexibility of the fiberoptic probesenabled efficient access to various pelvic anatomicallandmarks for imaging. Given its higher spatialresolution, the image quality from the 2.6 mm probewas significantly better than that from the 0.85 mmprobe and, thus, the 2.6 mm probe was used exclu-sively after the fourth case (see supplementarytable, http://jurology.com/). The image quality ofthe probes did not decrease noticeably with repeatedsterilization.

NVB Imaging

Prior ex vivo studies indicate that NVBs are locatedposterolateral of the prostate and enclosed in

CONFOCAL ENDOMICROSCOPY DURING ROBOTIC ASSISTED RADICAL PROSTATECTOMY 1113

lateral pelvic fascia.23,24 Intraoperative CLE of re-gions corresponding to the NVB was performedbefore and after nerve sparing dissection. Charac-teristic confocal features of the NVB include parallelthin dark lines corresponding to axonal fibers,bordered by dark cells consistent with adipocytes,and interspersed with vessels with flowing eryth-rocytes (fig. 2 and supplementary video, http://jurology.com/). Generally, confocal features ofNVBs were not visualized until the lateral pelvicfascia was incised. NVBs were visualized with the0.85 mm (fig. 2, A) and 2.6 mm imaging probes(fig. 2, B-E ). The mosaicing algorithm was appliedoff-line to generate a wide field view of the NVB(fig. 2, G). In vivo CLE identified the NVB in 11 of 15patients. In 1 case residual nerve tissues wereobserved on the prostatic capsule after initialdissection, prompting additional dissection and CLEconfirmation of NVB separation from the prostate.

Identification of Prostatic and Periprostatic

Structures

Representative in vivo confocal images of prostaticand periprostatic structures are shown in figure 3.The prostatic capsule, bladder neck margin, ure-thral stump, levator ani and obturator nerve wereimaged. Imaging of the prostatic capsule demon-strated striated fibrous tissue with occasional smallcaliber vasculature. Given the relatively small FOVof confocal laser endomicroscopy, the prostaticcapsule was not comprehensively imaged in vivo. Nodiscernible prostatic parenchymal features such asglandular structures were observed in vivo. Confocalimaging of the bladder neck mucosa showed normalurothelium with umbrella and intermediate cells as

Figure 2. CLE images of NVB. Nerve axons visualized with 0.85 mm p

(B) and after (C, D) NVB dissection. Residual nerve structures were p

Intact NVB seen ex vivo on nonsparing prostate specimen (F ). Panoimages obtained during in vivo CLE, with erythrocytes within blood v

well as the underlying vasculature of the laminapropria, consistent with previous bladder imaging.20

Ex Vivo Confocal Imaging of Fresh Prostate Tissue

To further characterize the confocal imaging fea-tures with hematoxylin and eosin correlation, freshprostate specimens were imaged ex vivo. A total of259 imaging sequences were collected from 20 sub-jects. The prostate was imaged intact to visualizethe capsular features (fig. 4, A), followed by imagingof prostate sections to visualize the parenchymalstructures (fig. 4, B). In patients who underwent anonnerve sparing procedure residual NVBs wereobserved on the prostate specimen, and confirmedby hematoxylin and eosin and immunohistochem-istry staining of myelin specific antigen S100 (fig. 2, Eand supplementary figure, http://jurology.com/).

While most prostate cancer arises from theperipheral zone,25 given the 60 mm penetrationdepth of CLE, we did not expect to visualize stromaland glandular structures through an intact capsuleunless there were positive surgical margins orextracapsular extension. In specimen 7 with pT3Bdisease CLE imaging along the lateral prostaticcapsule showed glandular structures distinct fromthe surrounding fibrous capsule (fig. 4, C andsupplementary table, http://jurology.com/). This pa-tient was confirmed to have multifocal ECE onpathological examination. On prostate sectionsbenign prostatic glands were characterized bylobular structures with a rim of increased sur-rounding fluorescence (fig. 4, E ). Benign featuressuch as corpora amylacea were easily identifiedas round circumscribed structures within glands(fig. 4,G). Prostatic glands in tissues found to contain

robe (A) and 2.6 mm probe (B-G). Nerves were visualized before

resent on prostatic capsule after neurovascular dissection (E ).

ramic image of NVB generated with mosaicing algorithm from

essels on left and nerve fibers on right (G).

Figure 3. In vivo CLE of prostatic and periprostatic structures with corresponding stereoscopic views from robotic prostatectomy as

insets. Confocal characteristics of fibrous prostatic capsule (A), urothelium of bladder neck margin (B), levator ani muscle fibers (C )

and axons of obturator nerve (D).

1114 CONFOCAL ENDOMICROSCOPY DURING ROBOTIC ASSISTED RADICAL PROSTATECTOMY

carcinoma were characterized by smaller, less reg-ular lobular structures without a surrounding rim offluorescence (fig. 4, I ).

DISCUSSIONWe report the initial feasibility of in vivo CLE dur-ing RARP. We demonstrated the ease of integratingCLE with robotic surgery, including co-registrationof confocal video sequences with white light imag-ing, probe handling with standard robotic instru-mentation and tremor-free image acquisition. Wecharacterized in vivo imaging features of clinicallyrelevant prostatic and periprostatic anatomicallandmarks, particularly the NVB. IntraoperativeCLE was performed successfully with 2.6 and 0.85mm probes, with the 2.6 mm probe offering superiorimage quality. Dynamic imaging of intact NVBsdemonstrated parallel axonal fibers lined by adipo-cytes and small caliber vessels. In vivo microscopyfeatures of the NVB were confirmed with ex vivo

CLE and standard hematoxylin and eosin in pros-tate samples where nerve sparing was notperformed.

Erectile dysfunction is a complication of radicalprostatectomy that can be minimized by preserva-tion of the NVB. Since components of the NVB havevariable distribution and location23 and cannot bevisualized directly during surgery, nerve sparingtechniques are based on gross inspection and mini-mizing thermal energy use near the presumed NVBlocation. Intraoperative visualization of microscopicfeatures may better guide nerve sparing surgeryand provide real-time feedback for adequatedissection. Our results suggest that CLE may beused to map NVB location. Dynamic characteriza-tion of the intact NVB after dissection may serve asa marker of successful preservation of the NVB.

Positive surgical margin status is an adverseoncologic outcome of radical prostatectomy thatmight be improved by image guided identification ofECE at surgical margins. Ex vivo CLE of prostatic

Figure 4. Ex vivo CLE imaging of prostatic tissue with corresponding H&E. CLE probe application to intact prostate specimen (A). CLEprobe application to transverse cut section of prostate (B). ECE of carcinoma, arrows point to region of ECE in background of striated

pattern capsule marked by asterisk with corresponding H&E, reduced from �50 (C, D). Lobular structure of benign prostatic glands

(E, F ). Corpora amylacea within glands (G, H ). Prostate cancer glands with Gleason 3þ3 pattern (I, J ). Corresponding H&E in parts

F, H and J, reduced from �100. Panoramic image of normal prostate generated with mosaicing algorithm that increased FOV by

approximately fourfold (K ).

CONFOCAL ENDOMICROSCOPY DURING ROBOTIC ASSISTED RADICAL PROSTATECTOMY 1115

sections revealed benign and cancerous glandularstructures. While most of the patients in this serieshad organ confined disease, in the patients withpT3b disease we were able to detect apparent ECE ofcarcinoma. CLE could be used in conjunction withpreoperative magnetic resonance imaging for tar-geted intraoperative imaging of areas concerning forECE. Identification of any glandular structures atsurgical margins would prompt the surgeon toredirect the plane of dissection.

CLE differs from other optical biopsy technolo-gies. Compared to clinical OCT systems,11,12 CLEoffers a higher spatial resolution but a lower pene-tration depth. MPM offers spatial resolution similar

to CLE and improved depth of penetration. Howev-er, current studies using MPM are limited to ex vivohuman specimens and in vivo animal studies.14,15

While OCT and MPM do not require the adminis-tration of exogenous dye, fluorescein is inexpensive,has a proven safety profile17 and may be coupledwith targeting agents for molecular imaging usingCLE to further improve optical diagnostics.26,27

The small sample size and pilot nature of thisstudy precluded diagnostic accuracy assessment ofCLE imaging of the prostate. Furthermore, theimpact that CLE may have on long-term functionaloutcomes is unknown as CLE was not used to directsurgical guidance. A larger sample size and defined

1116 CONFOCAL ENDOMICROSCOPY DURING ROBOTIC ASSISTED RADICAL PROSTATECTOMY

clinical end points will be necessary to assess theclinical benefits of CLE during RARP. Visualizationof intact nerves does not equate to functionalnerves. Future integration of CLE with molecularimaging agents28 or nerve stimulator29 may providephysiological feedback of nerve function. Othertechnical limitations include the small FOV of CLE,which makes intraoperative surveying of largesurface areas impractical. This may not negativelyimpact the usefulness of CLE for nerve sparingprocedures as it was possible to scan the length ofthe NVB within 90 seconds. However, it may limitthe use of CLE for the detection of incidental ECE.While CLE imaging is optimal within 20 minutes offluorescein administration, we demonstrated thefeasibility of fluorescein re-dosing. Future investi-gation of topical contrast administration with anendoscopic spray catheter may offer alternativestrategies for intraoperative CLE without therequirement of intravenous fluorescein.30

CLE is a promising technology for microscopicimaging during RARP. CLE optical biopsy of livetissue may provide a new method for the intra-operative identification of the NVB with spatial andtemporal resolutions not previously described.Additional experience is required to assess theusefulness of CLE in detecting surgical marginstatus, and to evaluate if this promising imagingtechnique will translate to improved oncologic andfunctional outcomes.

CONCLUSIONSCLE of the prostate and NVB is feasible duringRARP. Nerve fibers can be visualized and differen-tiated from vessels and connective tissue. Ex vivoCLE can be used to identify prostatic glandularstructures and ECE. Additional prospective anal-ysis is required to assess the clinical benefits of CLEguided nerve sparing RARP.

REFERENCES

1. Vickers A, Savage C, Bianco F et al: Cancercontrol and functional outcomes after radicalprostatectomy as markers of surgical quality:analysis of heterogeneity between surgeons at asingle cancer center. Eur Urol 2011; 59: 317.

2. Walsh PC: Anatomic radical prostatectomy:evolution of the surgical technique. J Urol 1998;160: 2418.

3. Tsui C, Klein R and Garabrant M: Minimallyinvasive surgery: national trends in adoption andfuture directions for hospital strategy. SurgEndosc 2013; 27: 2253.

4. Sood A, Jeong W, Peabody JO et al: Robot-assisted radical prostatectomy: inching towardgold standard. Urol Clin North Am 2014; 41: 473.

5. Volont�e F, Pugin F, Buchs NC et al: Console-integrated stereoscopic OsiriX 3D volume-rendered images for da Vinci colorectal roboticsurgery. Surg Innov 2013; 20: 158.

6. Yossepowitch O, Briganti A, Eastham JA et al:Positive surgical margins after radical prosta-tectomy: a systematic review and contemporaryupdate. Eur Urol 2014; 65: 303.

7. Alemozaffar M, Regan MM, Cooperberg MRet al: Prediction of erectile function followingtreatment for prostate cancer. JAMA 2011; 306:1205.

8. Johansson E, Bill-Axelson A, Holmberg L et al:Time, symptom burden, androgen deprivation,and self-assessed quality of life after radicalprostatectomy or watchful waiting: the Ran-domized Scandinavian Prostate Cancer GroupStudy Number 4 (SPCG-4) clinical trial. Eur Urol2009; 55: 422.

9. Hsu M, Gupta M, Su LM et al: Intraoperativeoptical imaging and tissue interrogation duringurologic surgery. Curr Opin Urol 2014; 24: 66.

10. KleinJan GH, van den Berg NS, Brouwer ORet al: Optimisation of fluorescence guidanceduring robot-assisted laparoscopic sentinel nodebiopsy for prostate cancer. Eur Urol 2014;66: 991.

11. Aron M, Kaouk JH, Hegarty NJ et al: Secondprize: preliminary experience with the Nirisoptical coherence tomography system duringlaparoscopic and robotic prostatectomy.J Endourol 2007; 21: 814.

12. Muller BG, de Bruin DM, van den Bos W et al:Prostate cancer diagnosis: the feasibility ofneedle-based optical coherence tomography.J Med Imaging (Bellingham) 2015; 2: 037501.

13. Rais-Bahrami S, Levinson AW, Fried NM et al:Optical coherence tomography of cavernousnerves: a step toward real-time intraoperativeimaging during nerve-sparing radical prostatec-tomy. Urology 2008; 72: 198.

14. Durand M, Jain M, Aggarwal A et al: Real-timein vivo periprostatic nerve tracking using multi-photon microscopy in a rat survival surgerymodel: a promising pre-clinical study forenhanced nerve-sparing surgery. BJU Int 2015;116: 478.

15. Tewari AK, Shevchuk MM, Sterling J et al:Multiphoton microscopy for structure identifica-tion in human prostate and periprostatic tissue:implications in prostate cancer surgery. BJU Int2011; 108: 1421.

16. Chen SP and Liao JC: Confocal laser endomi-croscopy of bladder and upper tract urothelial

carcinoma: a new era of optical diagnosis?Curr Urol Rep 2014; 15: 437.

17. Wallace MB, Meining A, Canto MI et al:The safety of intravenous fluorescein forconfocal laser endomicroscopy in the gastro-intestinal tract. Aliment Pharmacol Ther 2010;31: 548.

18. Chang TC, Liu JJ and Liao JC: Probe-basedconfocal laser endomicroscopy of the urinarytract: the technique. J Vis Exp 2013; 71: e4409.

19. Sonn GA, Jones SN, Tarin TV et al: Opticalbiopsy of human bladder neoplasia with in vivoconfocal laser endomicroscopy. J Urol 2009; 182:1299.

20. Wu K, Liu JJ, Adams W et al: Dynamic real-timemicroscopy of the urinary tract using confocallaser endomicroscopy. Urology 2011; 78: 225.

21. Konda VJ, Meining A, Jamil LH et al: A pilotstudy of in vivo identification of pancreatic cysticneoplasms with needle-based confocal laserendomicroscopy under endosonographic guid-ance. Endoscopy 2013; 45: 1006.

22. Bui D, Mach KE, Zlatev DV et al: A pilot study ofin vivo confocal laser endomicroscopy of uppertract urothelial carcinoma. J Endourol 2015; 29:1418.

23. Kiyoshima K, Yokomizo A, Yoshida T et al:Anatomical features of periprostatic tissue andits surroundings: a histological analysis of 79radical retropubic prostatectomy specimens.Jpn J Clin Oncol 2004; 34: 463.

24. Tewari AK, Srivastava A, Huang MW et al:Anatomical grades of nerve sparing: a risk-stratified approach to neural-hammock sparing

CONFOCAL ENDOMICROSCOPY DURING ROBOTIC ASSISTED RADICAL PROSTATECTOMY 1117

during robot-assisted radical prostatectomy(RARP). BJU Int 2011; 108: 984.

25. McNeal JE, Redwine EA, Freiha FS et al:Zonal distribution of prostatic adenocarci-noma. Correlation with histologic pattern anddirection of spread. Am J Surg Pathol 1988;12: 897.

26. Nguyen QT and Tsien RY: Fluorescence-guidedsurgery with live molecular navigationea

new cutting edge. Nat Rev Cancer 2013;13: 653.

27. Pan Y, Volkmer JP, Mach KE et al: Endoscopicmolecular imaging of human bladder cancer using aCD47 antibody. Sci Transl Med 2014; 6: 260ra148.

28. Boyette LB, Reardon MA, Mirelman AJ et al:Fiberoptic imaging of cavernous nerves in vivo.J Urol 2007; 178: 2694.

29. Burnett AL, Teloken PE, Briganti A et al:Intraoperative assessment of an implantableelectrode array for cavernous nerve stimulation.J Sex Med 2008; 5: 1949.

30. Hsiung PL, Hardy J, Friedland S et al:Detection of colonic dysplasia in vivo usinga targeted heptapeptide and confocal micro-endoscopy. Nat Med 2008; 14: 454.