Optical needle endoscope for safe and precise ... - OSA Publishing

4 downloads 76 Views 3MB Size Report
6Department of Translational Neurodegeneration I, German Center for Neurodegenerative Diseases e.V. (DZNE) ... Brain Tumor Primer - a comprehensive introduction to brain tumors (American Brain Tumor Association, ..... Scale bar, 50 µm.
Optical needle endoscope for safe and precise stereotactically guided biopsy sampling in neurosurgery Werner Göbel,1,* David Brucker,2 Yvonne Kienast,2 Ann Johansson,3 Gesa Kniebühler,3 Adrian Rühm,3 Sabina Eigenbrod,2 Stefan Fischer,4 Marcus Goetz,4 Friedrich-Wilhelm Kreth,5 André Ehrhardt,1,7 Herbert Stepp,3,7 Klaus-Martin Irion,1,7 and Jochen Herms2,6,7 1

Department of Research & Technology, KARL STORZ GmbH & Co. KG, Mittelstrasse 8, 78532 Tuttlingen, Germany 2 Department of Neuropathology, Ludwig-Maximilians-University Munich, Feodor-Lynen-Strasse 23, 81377 Munich, Germany 3 Laser-Research-Laboratory, Ludwig-Maximilians-University Munich, Marchioninistrasse 23, 81377 Munich, Germany 4 MRC Systems GmbH, Hans-Bunte-Strasse 8, 69123 Heidelberg, Germany 5 Department of Neurosurgery, Ludwig-Maximilians-University Munich, Marchioninistrasse 15, 81377 Munich, Germany 6 Department of Translational Neurodegeneration I, German Center for Neurodegenerative Diseases e.V. (DZNE) and Ludwig-Maximilians-University Munich, Feodor-Lynen-Strasse 23, 81377 Munich, Germany 7 shared senior authorship * [email protected]

Abstract: Proper treatment of deep seated brain tumors requires correct histological diagnosis which unambiguously necessitates biopsy sampling. Stereotactically guided sampling of biopsies is widely used but bears the danger of incorrect sampling locations and damage to intracerebral blood vessels. Here, we present a minimally invasive contact endoscopic probe that can be inserted into the tissue inside a standard biopsy needle and allows for fluorescence detection of both tumorous tissue and intracerebral blood vessels. Outer diameter of our contact probe is smaller than 1.5 mm, field-of-view in the range of several hundred microns; the optical design allows for simultaneous detection and visualization of tissue autofluorescence and selective fluorescence signals from deep seated brain tumors and vasculature as shown on in vivo animal models. We demonstrate the tumor detection capability during stereotactic needle insertion in a clinical pilot trial. Using our probe, we expect stereotactic interventions to become safer and more precise and the technology might ultimately be used also for various other kinds of applications. ©2012 Optical Society of America OCIS codes: (170.1610) Clinical applications; (170.2150) Endoscopic imaging; (170.3880) Medical and biological imaging; (170.3890) Medical optics instrumentation.

References and links 1. 2. 3. 4. 5.

Brain Tumor Primer - a comprehensive introduction to brain tumors (American Brain Tumor Association, 2010). D. Kondziolka, A. D. Firlik, and L. D. Lunsford, “Complications of stereotactic brain surgery,” Neurol. Clin. 16(1), 35–54 (1998). B. A. Flusberg, E. D. Cocker, W. Piyawattanametha, J. C. Jung, E. L. Cheung, and M. J. Schnitzer, “Fiber-optic fluorescence imaging,” Nat. Methods 2(12), 941–950 (2005). F. Helmchen, “Miniaturization of fluorescence microscopes using fibre optics,” Exp. Physiol. 87(6), 737–745 (2002). B. A. Flusberg, A. Nimmerjahn, E. D. Cocker, E. A. Mukamel, R. P. Barretto, T. H. Ko, L. D. Burns, J. C. Jung, and M. J. Schnitzer, “High-speed, miniaturized fluorescence microscopy in freely moving mice,” Nat. Methods 5(11), 935–938 (2008).

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26117

6. 7. 8.

9. 10. 11.

12. 13.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

26. 27. 28. 29.

R. P. Barretto, T. H. Ko, J. C. Jung, T. J. Wang, G. Capps, A. C. Waters, Y. Ziv, A. Attardo, L. Recht, and M. J. Schnitzer, “Time-lapse imaging of disease progression in deep brain areas using fluorescence microendoscopy,” Nat. Med. 17(2), 223–228 (2011). T. J. Muldoon, S. Anandasabapathy, D. Maru, and R. Richards-Kortum, “High-resolution imaging in Barrett’s esophagus: a novel, low-cost endoscopic microscope,” Gastrointest. Endosc. 68(4), 737–744 (2008). K. J. Rosbach, D. Shin, T. J. Muldoon, M. A. Quraishi, L. P. Middleton, K. K. Hunt, F. Meric-Bernstam, T. K. Yu, R. R. Richards-Kortum, and W. Yang, “High-resolution fiber optic microscopy with fluorescent contrast enhancement for the identification of axillary lymph node metastases in breast cancer: a pilot study,” Biomed. Opt. Express 1(3), 911–922 (2010). W. Stummer, H. Stepp, G. Möller, A. Ehrhardt, M. Leonhard, and H. J. Reulen, “Technical principles for protoporphyrin-IX-fluorescence guided microsurgical resection of malignant glioma tissue,” Acta Neurochir. (Wien) 140(10), 995–1000 (1998). R. Ritz, G. C. Feigl, M. U. Schuhmann, A. Ehrhardt, S. Danz, S. Noell, A. Bornemann, and M. S. Tatagiba, “Use of 5-ALA fluorescence guided endoscopic biopsy of a deep-seated primary malignant brain tumor,” J. Neurosurg. 114(5), 1410–1413 (2011). K. S. Samkoe, S. L. Gibbs-Strauss, H. H. Yang, S. Khan Hekmatyar, P. Jack Hoopes, J. A. O’Hara, R. A. Kauppinen, and B. W. Pogue, “Protoporphyrin IX fluorescence contrast in invasive glioblastomas is linearly correlated with Gd enhanced magnetic resonance image contrast but has higher diagnostic accuracy,” J. Biomed. Opt. 16(9), 096008 (2011). S. Moriuchi, K. Yamada, M. Dehara, Y. Teramoto, T. Soda, M. Imakita, and M. Taneda, “Use of 5aminolevulinic acid for the confirmation of deep-seated brain tumors during stereotactic biopsy,” J. Neurosurg. 115(2), 278–280 (2011). G. Widhalm, G. Minchev, A. Woehrer, M. Preusser, B. Kiesel, J. Furtner, A. Mert, A. Ieva, B. Tomanek, D. Prayer, C. Marosi, J. A. Hainfellner, E. Knosp, and S. Wolfsberger, “Strong 5-aminolevulinic acid-induced fluorescence is a novel intraoperative marker for representative tissue samples in stereotactic brain tumor biopsies,” Neurosurg. Rev. 35(3), 381–391, discussion 391 (2012). T. Kuroiwa, Y. Kajimoto, and T. Ohta, “Development and clinical application of near-infrared surgical microscope: preliminary report,” Minim. Invasive Neurosurg. 44(4), 240–242 (2001). A. Raabe, J. Beck, R. Gerlach, M. Zimmermann, and V. Seifert, “Near-infrared indocyanine green video angiography: a new method for intraoperative assessment of vascular flow,” Neurosurgery 52(1), 132–139, discussion 139 (2003). T. J. Muldoon, M. C. Pierce, D. L. Nida, M. D. Williams, A. Gillenwater, and R. Richards-Kortum, “Subcellular-resolution molecular imaging within living tissue by fiber microendoscopy,” Opt. Express 15(25), 16413–16423 (2007). K. Irion, “US 7,662,095 (B2) - Endoscope provided with a lighting system and a combined image transmission,” (2010). A. Ehrhardt, H. Stepp, K. M. Irion, W. Stummer, D. Zaak, R. Baumgartner, and A. Hofstetter, “Fluorescence detection of human malignancies using incoherent light systems,” Med. Laser Appl. 18(1), 27–35 (2003). G. A. Wagnières, W. M. Star, and B. C. Wilson, “In vivo fluorescence spectroscopy and imaging for oncological applications,” Photochem. Photobiol. 68(5), 603–632 (1998). Y. Kienast, L. von Baumgarten, M. Fuhrmann, W. E. Klinkert, R. Goldbrunner, J. Herms, and F. Winkler, “Real-time imaging reveals the single steps of brain metastasis formation,” Nat. Med. 16(1), 116–122 (2010). F. Winkler, Y. Kienast, M. Fuhrmann, L. Von Baumgarten, S. Burgold, G. Mitteregger, H. Kretzschmar, and J. Herms, “Imaging glioma cell invasion in vivo reveals mechanisms of dissemination and peritumoral angiogenesis,” Glia 57(12), 1306–1315 (2009). S. L. Gibbs-Strauss, J. A. O’Hara, P. J. Hoopes, T. Hasan, and B. W. Pogue, “Noninvasive measurement of aminolevulinic acid-induced protoporphyrin IX fluorescence allowing detection of murine glioma in vivo,” J. Biomed. Opt. 14(1), 014007 (2009). Y. Kang, M. Choi, J. Lee, G. Y. Koh, K. Kwon, and C. Choi, “Quantitative analysis of peripheral tissue perfusion using spatiotemporal molecular dynamics,” PLoS ONE 4(1), e4275 (2009). D. N. Louis, H. Ohgaki, O. D. Wiestler, W. K. Cavenee, P. C. Burger, A. Jouvet, B. W. Scheithauer, and P. Kleihues, “The 2007 WHO classification of tumours of the central nervous system,” Acta Neuropathol. 114(2), 97–109 (2007). S. Eigenbrod, Department of Neuropathology, Ludwig-Maximilians-University Munich, R. Trabold, D. Brucker, C. Erös, B. Suchorska, R. Egensperger, G. Pöpperl, A. Rühm, W. Göbel, H. Kretzschmar, J. C. Tonn, J. Herms, A. Giese, and F. W. Kreth are preparing a manuscript to be called “Molecular stereotactic biopsy technique improves tumor classification in glioma patients.” S. M. Landau, C. Liang, R. T. Kester, T. S. Tkaczyk, and M. R. Descour, “Design and evaluation of an ultra-slim objective for in-vivo deep optical biopsy,” Opt. Express 18(5), 4758–4775 (2010). T. J. Muldoon, D. Roblyer, M. D. Williams, V. M. Stepanek, R. Richards-Kortum, and A. M. Gillenwater, “Noninvasive imaging of oral neoplasia with a high-resolution fiber-optic microendoscope,” Head Neck 34(3), 305–312 (2012). A. Roggan, J. P. Ritz, V. Knappe, C. T. Germer, C. Isbert, D. Schädel, and G. Müller, “Radiation planning for thermal laser treatment,” Med. Laser Appl. 16(2), 65–72 (2001). A. Rühm, Laser-Research-Laboratory, Ludwig-Maximilians-University Munich, W. Göbel, and H. Stepp are preparing a manuscript to be called “Fiber baser fluorescence diagnosis based on PpIX and ICG – Excitation power limitations due to thermal effects in human brain tissue.”

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26118

30. W. Göbel, J. N. Kerr, A. Nimmerjahn, and F. Helmchen, “Miniaturized two-photon microscope based on a flexible coherent fiber bundle and a gradient-index lens objective,” Opt. Lett. 29(21), 2521–2523 (2004). 31. A. Johansson, G. Palte, O. Schnell, J. C. Tonn, J. Herms, and H. Stepp, “5-Aminolevulinic acid-induced protoporphyrin IX levels in tissue of human malignant brain tumors,” Photochem. Photobiol. 86(6), 1373–1378 (2010).

1. Introduction Deep seated brain tumors constitute one the most devastating medical conditions with incidence rates of 3/100.000, affecting patients from early childhood on at all ages alike [1]. Despite recent advances in various therapeutic modalities including surgery, chemotherapy and radiotherapy, prognosis remains still poor and individual therapy inherently relies on correct histological diagnosis and molecular characterization of the tumor tissue. As one of the standard procedures for effectively diagnosing brain tumors, stereotactically guided biopsy sampling is widely used (Fig. 1). During the stereotactic procedure, a small diameter hollow needle is inserted into the putative tumorous tissue along a trajectory that was carefully planned and evaluated, typically using CT/MRI data, to ensure placing of the needle’s distal end tip within the tumorous tissue and to avoid damage of intracerebral blood vessels. Despite careful planning, misguidance of the needle - e.g. due to brain shift after surgical opening of the skull - potentially leads to false negative diagnostic findings and ultimately inappropriate further treatment. Even more severely, in 1/200 cases a damage of intracerebral blood vessels leads to severe complications such as speech impairment and paralysis which may eventually even necessitate immediate surgical intervention bearing the ultimate risk of the patient’s death [2].

Fig. 1. Stereotactically guided biopsy sampling in neurosurgery. A hollow stereotactic needle is attached to the stereotactic frame firmly connected to the patient’s skull and inserted into the brain along a predefined trajectory. The distal tip is positioned within the putative tumor region. Damage to blood vessels is to be unambiguously avoided.

While new treatment options might become available in the near future, novel technologies for intraoperative feedback during biopsy sampling are urgently needed. Imaging technologies based on fiber optics solutions are widely explored for imaging in the intact brain within animal studies and humans [3, 4]. Combined with fluorescence labeling, these techniques allow detailed studies of intracerebral microvasculature [5], cellular morphology and functional signaling in animal models of disease [6]. One problem with transferring animal model data directly to applications on human patients is among others the lack of regulatory approval of many of the fluorescent dyes that are commonly used in animal experiments. For that reason, only few studies have shown imaging in human patients, mostly confined to easily accessible organs, e.g. oral mucosa [7, 8]. Administration of 5aminolaevulinic acid (5-ALA) leads to a specific accumulation of red fluorescing #176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26119

protoporphyrin IX (PpIX) in highly proliferating cells such as tumor cells [9]. Fluorescence detection and visualization of PpIX is used in clinical practice for tumor demarcation, including neurosurgery [9, 10]. As PpIX is only synthesized by vital cells, necrotic parts of the tumor do not show PpIX-fluorescence. PpIX-fluorescence therefore indicates vital tumor tissue with high specificity [11] and has been used for tumor delineation by means of intraoperative fluorescence observation on excised tissue specimens during stereotactic biopsy sampling [12, 13], in the latter case in combination with co-registered pre- and postoperative imaging (MRI, CT, PET) using intraoperatively set markers. In the neurosurgical field, tumor labeling based on 5-ALA induced PpIX fluorescence has been shown to significantly enhance completeness of resection and thereby postoperational survival time for the patient. On the other hand, the near infrared fluorescence marker indocyanin green (ICG) is getting commonly used for fluorescence visualization of the vascular system following direct intravenous injection [14, 15]. While novel fluorescence technologies for visualization of tumors and the vascular system are evolving, suitable techniques for minimally invasive and high-precision intraoperative detection of tumors are mostly lacking. Such techniques are especially relevant for the field of biopsy sampling during stereotactic interventions where small diameter instruments are to be inserted into intact brain tissue. Here, we describe a fiber optics based miniature endoscopic probe that can be integrated into standard biopsy needles and is optimized to simultaneously visualize vital tumor tissue and blood vessels based on interstitial PpIX and vascular ICG fluorescence detection. We show the optical performance of our device on ex vivo and in vivo data and demonstrate its principal usefulness for detection of deep seated brain tumors on an animal model, and we finally report the first clinical use of our device. 2. Setup and optical design Major challenges in designing our endoscopic probe were severalfold: first, the geometrical constraints of standard stereotactic needles needed to be taken into account. Typical hollow needles used in neurosurgery have an inner diameter of 1.6 mm, which required our probe to have an outer diameter of < 1.5 mm. Second, the distal end face of the probe is in direct tissue contact during insertion so that a contact endoscopic probe needed to be designed that can reveal high resolution images when inserted into tissue. Lastly, two fluorescence modes for tumor and blood vessel detection had to be realized in a way that allows simultaneous visualization of healthy tissue, tumor regions and blood vasculature. For size reduction, we used a scheme where excitation and detection light is guided through the same optical channel [16, 17] as opposed to standard endoscopes that typically employ fiber optic light guides for illumination and a separate detection path for imaging (Fig. 2(a)). Here, we used semiflexible coherent fiber bundles with several tens of thousands of individual fiber cores, an outer bundle diameter of ~1000 µm, an active area of ~700 µm diameter and a numerical aperture (NA) of 0.35 for the individual fibers within the bundle. The fiber-to-fiber spacing within the honeycomb structure of the bundle is approximately 4 µm. Standard endoscopic light sources typically employ Xenon arc lamps or light emitting diodes. Despite their general good suitability for fluorescence applications [18], these high NA emitters are not ideal for coupling light into small diameter fiber optics. Here, we used two laser diodes emitting at 405 nm and 785 nm wavelength, tuned to the absorption maxima of PpIX and ICG, respectively (Fig. 2(b)). The spectrally combined two-color laser beam was deflected using a custom dual band dichroic mirror and coupled into the image guide using an endoscopic eye piece (KARL STORZ). The proximal end face of the image guide was imaged onto a 3-chip CCD camera (Tricam SL II PDD, KARL STORZ) exhibiting a second maximum for near infrared light in the blue channel (Fig. 2(c)). Remitted fluorescence light was separated from the illumination path by the dual band dichroic mirror; an additional longpass and notch filter in front of the camera objective selectively blocked residual, backscattered excitation light at 405 nm and 785 nm.

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26120

For contact endoscopy, the bare distal end face of the image guide was brought into direct tissue contact so that the focal plane in such a widefield illumination scheme coincided with the distal end face (Fig. 2(d) and (e)). For most types of biological tissue, autofluorescence at 405nm excitation occurs predominantly within the green wavelength region [19] which potentially enables the distinction of induced autofluorescence within healthy tissue from red PpIX fluorescence originating from putative tumor tissue.

Fig. 2. Scheme of setup and optical layout for minimally invasive tumor biopsy needle endoscope. (a) Schematic view of the entire system. Light from two laser diodes emitting at 405 nm and 785 nm wavelength is combined with a dichroic mirror and coupled into a coherent image guide using a dual-band dichroic mirror and an eye piece lens. Fluorescence is separated by the dual-band dichroic mirror and imaged onto a 3-chip CCD camera. Two emission filters selectively block excitation wavelengths 405 nm and 785 nm. (b) Filter characteristics of the excitation wavelengths (blue and dark brown curves), dichroic mirrors (green curve) and emission filters (red curve). Expected fluorescence modalities are indicated (PpIX, ICG - external marker fluorescence; AF - autofluorescence). (c) Spectral characteristics of the 3-chip camera system. Interestingly, near infrared fluorescence light is visualized in the blue color channel due to a second sensitivity maximum of the blue-channel CCD chip in the near infrared region. (d) Example autofluorescence contact images of felt (top) and skin with sweat duct (bottom). (e) Example image of near-infrared ICG fluorescence at edge of a fluorescing region of an ICG test object.

3. Fluorescence imaging of tumors and vasculature We used an animal model of cerebral tumorigenesis to evaluate visualization of cerebral tumors and blood vessels as described in detail elsewhere [20, 21]. In brief, human glioma cells were implanted into superficial layers of neocortex or deep seated brain regions of immune deficient mice. Above the putative tumor region, a craniotomy was performed. 3 hours before imaging, mice were intraperitonally injected a solution of 5-ALA (Gliolan, Medac GmbH) dissolved in PBS at a concentration of about 200 mg/kg body weight leading to similar PpIX concentrations as used for human neurosurgery [9, 22]. For vasculature

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26121

visualization a dose of 200 mg/kg body weight ICG (ICG-Pulsion, Pulsion Medical Systems SE) was intravenously injected right before starting the imaging experimentation [23]. Animal surgery, cranial window preparation and imaging were performed as previously described [20]. All animal procedures were carried out according to the guidelines of the Ludwig-Maximilians-University Munich and were approved by the local authorities. For image recordings, the distal tip of the probe was lowered onto the exposed brain surface using a manual 3-axis micromanipulator stage (Luigs & Neumann). The distal light power was adjusted to be in a range of 1 – 10 mW for both of the two excitation modes. The camera system was adjusted in a fluorescence mode that allows long integration times of up to 1/15 s, manual color balance and optimized gain settings. After each experiment, brain tissue was extracted and analyzed to check for the presence/absence of tumorous tissue within the tissue regions that underwent optical imaging with our biopsy endoscope.

Fig. 3. In vivo fluorescence visualization of superficial tumors and blood vessels. (a) Fluorescence images of healthy tissue (top), tumor center with bright red PpIX fluorescence (middle), and tumor margin (bottom). (b) Bleaching of PpIX fluorescence signal over time. Top: exemplary decay of red fluorescence over time; bottom: example images at beginning and end of recording at time points indicated as red vertical lines in the decay graph. (c) Blood vessel visualization in pure ICG mode at 785 nm excitation (top), in pure autofluorescence mode at 405 nm excitation (middle), and simultaneous autofluorescence and ICG visualization (bottom).

Healthy tissue showed a relatively unspecific greenish autofluorescence with blood vessels silhouetted against dark reddish structures (Fig. 3(a), top panel). In the tumor center, a typical bright red PpIX fluorescence was observed (Fig. 3(a), middle panel) that showed a time-dependent intensity decay due to photobleaching effects (Fig. 3(b)). Interestingly, a rather clear demarcation of tumor margins could be visualized when placing the distal tip of the image guide at the putative tumor margin (Fig. 3(a), bottom panel). As expected, nearinfrared ICG fluorescence appeared inside blood vessels under 785 nm excitation (Fig. 3(c), top and bottom panel), while under 405 nm excitation the blood vessels effectively absorb both excitation light and autofluorescence and thus appeared rather dark in the pure autofluorescence image (Fig. 3(c), middle panel). Combining both excitation wavelengths allows tissue visualization and at the same time a clear visualization of the vasculature in front of the needle tip (Fig. 3(c), bottom panel). In principle, the relative weighting of ICG fluorescence and autofluorescence could be fine-tuned to highlight one or the other feature. In some cases, blood flow could be directly visualized in the dynamic video data.

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26122

Next, we aimed at detection of structures deep within tissue during insertion of the endoscopic probe. We simulated detection of blood vessels during probe progression using blood vessel dummies made from thin silicone tubes (600 µm inner diameter) filled with blood diluted with ICG at physiological concentration. These artificial blood vessels were embedded ex vivo into porcine brain tissue, and the distal tip of the probe was inserted into the tissue and lowered directly onto the artificial vessel (Fig. 4(a)). A diffuse, bluish fluorescence signal from the ICG contained inside the artificial vessel could be detected at a distance of approximately 1 mm before touching the vessel (Fig. 4(b) and (c)). As expected, higher illumination intensities at the distal end face of our probe resulted in longer maximal distances at which the ICG fluorescence signal originating from a blood vessel could be detected. However, presumably due to a concomitant increase in stray light, higher light intensities also lead to a higher signal baseline level. Thus, a higher illumination intensity allowed for blood vessel detection at longer distances albeit at the expense that the contrast from baseline fluorescence level to actual blood vessel fluorescence was smaller and it might be advantageous to use medium intensity levels for efficiently detecting ICG fluorescence originating from blood vessels.

Fig. 4. Detection of blood vessels inside tissue. (a) Photo of artificial blood vessels filled with blood/ICG solution and embedded into brain tissue for simulating blood vessel detection. The black-coated bare fiber probe is also visible. (b) Signal decay in the blue/infrared camera channel as a function of distance of the distal probe tip from the blood vessel dummy for three different light intensities. The remaining signal at distances above 2000 µm results from scattered excitation light detection. Signals were analyzed across the active area of the image guide and normalized to the maximum intensity value of 255. Intensity levels between 1 – 5 mW seem optimal for blood vessel detection. (c) ICG fluorescence images at various distances from the blood vessel dummy (top: direct contact; middle: 750µm distance; bottom: 1500µm distance). Bluish ICG fluorescence was detectable at a distance of ~1 mm before touching the artificial vessel.

During stereotactic interventions a correct positioning of the needle’s distal end tip within the putative tumorous tissue is crucial. Therefore, one of the major goals of this project was to show the feasibility of visualizing and thus exactly localizing deep seated tumors during needle propulsion along the biopsy channel. Deep seated tumor detection during needle insertion was evaluated by implanting tumor cells into deep brain regions of mice. After 5ALA administration and ICG injection, the endoscopic probe was attached to the micromanipulator and positioned on the brain surface above the expected deep seated tumor region. During needle propulsion, we first observed greenish autofluorescence signals

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26123

presumably originating from healthy tissue regions. In several experiments on mice (n = 3) we consistently detected clearly red fluorescing tissue regions and the respective tissue margins at depths expected from tumor cell implantation (Fig. 5(a)). At still deeper imaging depths, red tumor fluorescence was no longer present since the needle's distal tip had crossed the lower tumor margin, resulting in greenish autofluorescence visualization. Note that several insertion trials were necessary until the probe could be guided into red fluorescing tumor tissue and that the bluish color in many of the images in Fig. 5(a) results primarily from ICG fluorescence due to blood vessel leakage. We repeated this intervention in another animal with preoperative 5-ALA administration but without tumor implantation and found no sign of red fluorescence at any of several insertion sites (Fig. 5(b)). From these investigations we conclude that our miniature needle-like endoscopic probe can be used for simultaneous fluorescence visualization of tumor regions and blood vessels during successive propulsion of biopsy needles such as the ones used in stereotaxy.

Fig. 5. Images obtained during probe insertion towards, into and through deep seated brain tumors based on red PpIX fluorescence. The image series demonstrate the probe’s ability to detect tumor margins. (a) Series of 4 consecutive images obtained during 5 probe insertions through a tumorous region. Top row: pial surface; second row: autofluorescence during insertion before reaching tumor region (note pial blood vessels being dragged until pial rupture during some of the insertion trials); third row: vital tumor region as indicated by strong red PpIX fluorescence; fourth row: autofluorescence below/behind tumor. (b) Series of 4 consecutive images as in (a), but obtained on a control animal with previous 5-ALA administration but without implanted tumor. Image series at same depth regimes as in (a). The bluish background seen in many images in the left panel is due to ICG leakage from damaged blood vessels into the tissue, which occurred during consecutive trials to hit the vital, deep seated tumor region.

In a final step, we devised optical probes with dimensions equivalent to mandrins conventionally employed to facilitate the needle’s tissue penetration (Fig. 6(a)). Needleshaped mandrins with dimensions matched to the lumen of the hollow stereotactic needle are used inside the stereotactic needle during its insertion into the tissue. The mandrin is removed when the stereotactic needle is in its desired axial position. Subsequently, biopsies are sampled, typically using small diameter biopsy forceps. Mechanical properties of our optical probes such as stiffness, diameter, distal length and tip shape were nearly identical to those of a mandrin. In conformance with federal laws of Germany for usage of medical devices without CE marking, this type of probe was subsequently used in a clinical pilot trial (‘Heilversuch’) within a stereotactic intervention for a patient diagnosed a malignant brain tumor in the lower temporal lobe. Based on written informed consent of the patient, 5-ALA (Gliolan) was administered at a concentration of 20 mg/kg body weight 6h before the intervention [9]. Remarkably, no significant difference was encountered in mechanical terms by the medical doctors when using our optical probe as compared to the standard mandrin. A strong, red fluorescence was readily detected, apparently originating from vital tumor regions,

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26124

whereas regions outside the vital part of the tumor showed only a faint autofluorescence signal (Fig. 6(b), left panels). The correlation between the presence of vital tumor tissue and enhanced red fluorescence was verified by subsequent histological analysis of biopsies sampled at the respective imaging positions (Fig. 6(b), right panels) [24, 25]. In conclusion, we expect a rather straight-forward usability of our needle-like endoscope during stereotactic interventions.

Fig. 6. Probe design and imaging results obtained during a clinical pilot trial. (a) Photograph of the optical probe in comparison to the conventional mandrin and the hollow needle. Mandrin and optical needle’s end tip share nearly identical mechanical properties (see Fig. 4(a) for comparison with a bare optical fiber bundle). (b) Example images (left) from putative necrotic (top panels) and vital tumor region (bottom panels), respectively. The two images on the left were acquired at different depths along the axial position of the biopsy channel during probe propulsion. Histological sections of biopsies from the same tissue locations are shown on the right, correlating well with the imaging results. Scale bar, 50 µm.

4. Discussion By using the same fiber optics for both illumination and image detection we devised a miniature, versatile probe for minimally invasive fluorescence examinations that enables a clear demarcation of healthy tissue, tumorous tissue and vasculature based on tissue autofluorescence, PpIX fluorescence, and ICG fluorescence. By combining these signals we achieved a clear demarcation of deep seated brain tumors in a neurooncological mouse model. Our device is simple, robust and can be easily integrated into standard biopsy needles. After reaching compliance with the respective regulatory guidelines, we expect a rather straight forward transfer to the intended clinical application. Similar fiber optics based solutions as the one we present in this study have been reported previously for biopsy guidance in the case of breast cancer [8, 26] and for detection of oral mucosal lesions using topical application of fluorescent dyes [27]. The chosen illumination wavelengths, however, are not suitable for additional autofluorescence imaging and allow fluorescence visualization of a single fluorescence signal only. Here, the combination of two illumination wavelengths by using easily focusable laser diodes enables distinct, simultaneous visualization of tissue disease state based on 5-ALA induced PpIX fluorescence and of vasculature based on ICG fluorescence. Moreover, autofluorescence detection provides

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26125

further information during needle insertion into healthy tissue. While blood vessels can be discerned at high contrast in pure autofluorescence mode under the condition of direct tissue contact, this approach does not allow an early detection of blood vessels some distance ahead of the distal end of the probe, which appears highly favorable in order to reduce the risk of vessel perforation. Using ICG as a contrast agent, we could detect blood vessels – albeit at lower contrast compared to pure autofluorescence when in direct tissue contact – at distances below 1 mm, which is presumably sufficient to slow down propulsion of the stereotactic needle by the surgeon and eventually recapitulate the propulsion path. Regulatory hurdles might prevent widespread clinical use of fluorescence technologies in standard clinical settings. However, in neurosurgery, 5-ALA induced PpIX fluorescence is progressively finding a widespread use for tumor labeling during resection and has been shown to significantly increase completeness of tumor resection thereby augmenting postoperational survival time for patients [10]. Likewise, using 5-ALA during stereotactic tumor biopsy seems straight forward and the additional use of ICG can potentially be easily accomplished. For imaging, we employed light intensity levels of 10 mW and less at the distal probe end. Numerical simulations using the LITCIT software package [28] have shown that this is far below the limit for inducing destructive thermal effects on brain tissue [29]. Ultimately, a pure optical biopsy would be favorable. For this purpose, however, a still higher spatial resolution is required to resolve individual cell shapes and intracellular compartments such as cell nuclei. Here, we have presented a contact endoscopic probe by omitting a distal objective system. In such an arrangement, the spatial resolution is intrinsically limited by the pixel-to-pixel spacing of the fiber bundle, in our case ~4µm. Combination of image guides with miniature gradient-index lens objectives has been shown to provide resolutions in the 1 µm range [30], which is potentially sufficient for detection of intracellular compartments. Moreover, as a further possible extension of the probe concept presented here, a quantitative spectroscopic analysis of the obtained fluorescence signals could enable quantitative measurements of fluorophore concentrations and thus even staging of tumor grades and subtypes [31]. In summary, we have introduced a simple, yet quite effective minimally invasive optical probe for tumor biopsy in neurosurgery. Potential applications, however, are presumably not restricted to neurosurgery, but could also include diverse fields of clinical endoscopy where small diameter, minimally invasive optical feedback for guiding biopsies is required. Acknowledgments We thank Thomas Hinding for expert technical assistance, Thomas Pongratz for help with the ICG experimentation, and Martin Leonhard for comments on the manuscript. This study was supported by the German Ministry of Education and Research (BMBF, project Neurotax, grant 13N101-69/70/71/72) and we would like to thank Hans-Joachim Schwarzmaier for project guidance.

#176216 - $15.00 USD (C) 2012 OSA

Received 13 Sep 2012; revised 23 Oct 2012; accepted 24 Oct 2012; published 5 Nov 2012 19 November 2012 / Vol. 20, No. 24 / OPTICS EXPRESS 26126