The concept of intraoperative labeling of tumorous tissue or vascular structures has intrigued surgeons for decades. The emergence of fluorescence-based imaging techniques highlighting tumors or vascular structures during surgery has improved intraoperative decision-making. In principle, a clear distinction is provided to discern the lesion to be resected from surrounding viable tissue, which can guide through the most delicate process of tumor resection. This holds particularly true for neurosurgery, where manipulation or resection of tissue is done at a microsurgical level. Currently, the methods employed often rely on surgical microscopes equipped with appropriate light sources and filters to visualize the fluorophore. To visualize microstructures at high magnification, this principle has been refined even further, resulting in the development of the confocal laser endomicroscope. In essence, this technique can deliver in vivo images in real time without the need for tissue extraction. Based on the application of a fluorescent agent, a scanner probe through gentle contact with the tissue surface reveals optical details at the subsurface level. The device emits focused laser light via a probe into the tissue and recaptures emanating light from the focal plane while rejecting light from outside the focused spot.1 This enables visualization of an area of interest on a specific focal plane in a tissue, hence “confocal.” Images can be obtained not only in the x- and y-axes but also in the z-axis as z-stacks (an automated recording of a user-defined range of images relative to the current focal plane).
The miniaturization and technical refinement of this technique have resulted in a device readily available for intraoperative use in neurosurgery (CONVIVO, Carl Zeiss AG). The technical evolution of fluorescence-guided surgery in neurosurgery has produced crucial innovations in the recent decade.2 Among the popular agents, 5-aminolevulinic acid (5-ALA) was clinically validated in a pivotal study by Stummer et al.3 Its approval, however, is limited for use in patients with high-grade gliomas. In other entities such as brain metastases, the results have varied.4 Fluorescein sodium (FNa) has recently reemerged as a promising fluorophore, with retrospective and prospective data suggesting that FNa improves visualization and resection of many CNS tumors.5–11 A prerequisite for staining of CNS lesions is a pathological alteration of the blood-brain barrier, which equally applies for 5-ALA and FNa. In the case of FNa, the presumed mechanism of action is passive staining of the extracellular space in areas with a blood-brain barrier breakdown, corresponding to gadolinium uptake on MRI.12 The safety of FNa has been demonstrated through decades of experience in ophthalmology, where its properties as a vascular dye have been exploited.13
Intraoperative diagnosis in neurosurgery is traditionally provided by histopathological frozen section analysis to instantly assess the tumor entity and monitor surrounding tissue for tumor-free margins as the gold standard. This technique relies on the examination of extracted tissue samples with different staining methods using benchtop microscopes. The diagnostic yield is dependent on the technique used, and, through sampling errors and artifacts, the accuracy can be as low as 50%.14,15 Furthermore, waiting times for the results prolong the time to surgery. Confocal laser endomicroscopy (CLE) offers the possibility for fluorophore-dependent in vivo microscopy to neuropathologists, potentially ameliorating these shortcomings. Thus far, there is a scarcity of scientific evidence for CLE in neurosurgery, with only a few studies exploring its feasibility and safety.16,17 As with conventional neuropathological assessment, in CLE, staining with a fluorescent dye is needed for sufficient image contrast, which is provided by FNa in the case of CONVIVO (Fig. 1).
Intraoperative probe view under white light (left) and YE560 filter view (right).
CLE with FNa has provided a real-time intraoperative assessment of suspected tumor and healthy tissue (digital biopsies),18 omitting the need for time-consuming tissue transfer, fixation, and interpretation to improve the diagnostic yield and shorten sampling times.16,19 All in all, through real-time tissue evaluation, this technique has a true potential to further refine surgical decision-making and strategy.20 As the CONVIVO device has just recently been introduced, at this time the existing data are not sufficient to compare the quality of frozen section analysis with CLE findings.18
Here, we present our first experience with FNa and CLE in different CNS tumors. We provide intraoperatively acquired CLE images with matching frozen section analysis, microscope photographs, and intraoperative navigation. This user report offers information on the feasibility, safety, and potential applications of this technique.
Methods
Patients and Radiological Evaluation
In this study, we included 12 nonconsecutive patients who were operated on at our institution in July 2020. Inclusion criteria were suspected intracranial tumor confirmed by gadolinium-enhanced MRI, informed consent about the off-label use of FNa, and no known allergy to FNa. All patients underwent preoperative and prompt postoperative (within 72 hours after surgery) contrast-enhanced MRI. A review of preoperative imaging–determined localization and contrast-enhancing behavior of the tumor was conducted. The retrospective study was approved by our institutional review board.
Surgical Protocol
Fluorescein Sodium
FNa (Alcon) 10% injection solution is frequently utilized in ophthalmology and optometry for the diagnosis of corneal abrasions, corneal ulcers, herpetic corneal infections, and retinal angiography.21 Moreover, since its discovery, FNa has been found to be useful in a myriad of scientific and civil areas and is listed in the WHO List of Essential Medicines.22 Adverse reactions are rare and have been extensively studied and are therefore well known. The administration of FNa for the purpose of tumor visualization is currently an off-label use. Following earlier reports,6,18,23 a weight-adjusted dose (5 mg/kg body weight) of 10% FNa was injected intraoperatively via a central venous line. To observe changes in enhancement, the time of injection varied (Table 1).
Patient characteristics
| Imaging Position & FNa Fluorescence | ||||||
|---|---|---|---|---|---|---|
| Case No. | Diagnosis | Time Btwn FNa & Imaging (mins) | Tumor Contrast Enhancement on MRI | I | II | III |
| 1 | Lung cancer metastasis | 60 | Homogeneous | + | + | + |
| 2 | Glioblastoma WHO grade IV | 30 | Homogeneous | + | + | + |
| 3 | Oligodendroglioma WHO grade II | 120 | Inhomogeneous | + | + | − |
| 4 | Glioblastoma WHO grade IV | 30 | Homogeneous | + | + | + |
| 5 | Neurocytoma WHO grade II | 60 | Inhomogeneous | + | + | − |
| 6 | Melanoma brain metastasis | 30 | Rim-like | + | + | + |
| 7 | Brain metastasis of a carcinoma, not otherwise specified | 60 | Rim-like | + | + | + |
| 8 | Gliosis/recurrent hemangiopericytoma | 45 | Homogeneous | + | + | − |
| 9 | Colon cancer metastasis | 10 | Rim-like | + | + | − |
| 10 | Glioblastoma WHO grade IV | 20 | Rim-like | + | + | + |
| 11 | Glioblastoma WHO grade IV | 45 | Rim-like | + | − | + |
| 12 | Melanoma brain metastasis | 65 | Rim-like | + | + | − |
+ = positive; – = negative.
Intraoperative CLE Imaging and Diagnostic Sampling
After craniotomy and durotomy, on contact with the suspected tumor, CLE imaging commenced, after which a snapshot of the position according to neuronavigation findings was obtained (still picture acquisition through the microscope), and the selected tissue was removed. Before imaging, the CLE microscope probe is dressed in a sterile drape and is then ready to be inserted into the resection cavity. During routine tumor microsurgery, these steps were conducted at different locations: at the border or capsule of the tumor, referred to as the tumor capsule (I) (Fig. 2); at the tumor center (II) (Fig. 3); and at the potential infiltration or edema zone where the fluorescence signal started to become faint, referred to as the perilesional zone (III) (Fig. 4).
Case 4. H & E–stained slide (left; original magnification ×200) and CLE image (right) obtained during biopsy of a glioblastoma, imaging position I (tumor border). Microscopic examination revealed a highly cellular, diffuse infiltrating glioma with small, hyperchromatic nuclei and hemorrhages.
Case 6. H & E–stained slide (left; original magnification ×200) and CLE image (right) obtained during biopsy of a metastatic malignant melanoma, imaging position II (tumor center). The malignant tumor is characterized by densely packed epithelioid cells with a high nuclear-to-cytoplasmic rate, round nuclei, and incidental large nucleoli. The blood vessels captured in this image enclosed activated, prominent endothelial cells.
Case 4. H & E–stained slide (left; original magnification ×200) and CLE image (right) obtained during biopsy of a glioblastoma, imaging position III (perilesional zone). The infiltration zone is characterized by an increased cell density of gray and subcortical white matter. In the cortical areas, perineuronal satellitosis is evident and signals the formation of secondary structures.
Tumor Resection
In between these steps, tumorous tissue was removed with the aid of FNa-induced fluorescence visualized with an additional filter on the operating microscope (KINEVO 900 or PENTERO 900, ZEISS Meditec). This filter (YELLOW 560) is tailored to the excitation and emission wavelength of FNa.24 When applicable, the Cavitron Ultrasonic Surgical Aspirator (Integra), intraoperative monitoring, and intraoperative ultrasound were used. Typically, the lesions exhibit bright yellow staining through the microscope filter when approached (Fig. 1 right). Resection was stopped when the yellow-green staining of the enhancing tissue became faint and pinkish nonenhancing tissue appeared at the circumference of the tumor. Unless continuation of surgery was deemed unsafe, for instance, because of venturing into eloquent areas, surgical intervention was finished after removal of all fluorescing tissue as confirmed by the filter view. The addition of CLE imaging to the surgical workflow did not change the resection in any case.
Intraoperative Fluorescence Characteristics and Side Effects
Surgical reports were screened for subjective evaluation of the grade of fluorescent staining of the targeted lesion. The screening was conducted for any reference to the degree of fluorescent staining: bright versus medium and heterogeneous versus effectively no fluorescence. Furthermore, medical reports were evaluated for any possible adverse effect or allergic reaction to FNa.
Results
Fluorescence
Homogeneous or heterogeneous yellow-green fluorescent macroscopic staining of the tumor tissue was observed in all patients (n = 12; 100%). According to the surgical report, the staining was considered helpful guidance in all cases. Data on demographic patient characteristics, histopathological classification of the disease, timing, and dosing for the injection of FNa were available for all patients (Table 1).
CLE Imaging
Tumor surgery was performed using standard microsurgical techniques depending on the tumor location and type of tumor expected and adapted to the intraoperative situation. CLE could be safely performed, as the small and versatile probe mirrors any other microsurgical instrument that is held in one hand and inserted into the cavity without traumatizing healthy tissue (Fig. 1).18,20,25 A movable monitor mounted on top of the device serves as a viewer and as a user interface via touchscreen capabilities. Furthermore, a foot control panel can be connected and allows surgeon-controlled image acquisition.
Adjustable parameters, which are set first, are a choice of filters. The best image quality is obtained with the green filters, of which the bandpass filter produced better interpretable images than the longpass filter. Another baseline adjustment is gain, which was set at 2400 with good results. Laser power and brightness can be adjusted on the fly on a scale from 0% to 100%; ideally, the laser was set at 50% or below and the brightness adjusted during imaging. Additionally, an autobrightness mode can be activated and speed up the process of image acquisition. After sterile draping of the probe, the 0 position of the window has to be determined to find the optimal depth for initial imaging. In the viewing mode, there are two imaging speeds, fast (0.44 sec/frame) and lower resolution or slow (1.29 sec/frame) and higher resolution (1920 × 1080 pixel). Images can be recorded continuously as a time series or sequentially as single frames. Another feature is a z-stack option, imaging preset layers sequentially along the z-axis to a depth of up to 30 μm below the window of the sterile cover. There are 3 zoom options: ×1, ×1.4, and ×2. We found the ×1 zoom the clearest option with a field of view of 475 × 267 μm (width by height).
CLE was integrated seamlessly in the surgical flow. Representative images of different tumor entities and different intraoperative areas are depicted in Figs. 2–4. As the interpretation of CLE images needs further systematic evaluation, CLE images in this case series did not alter surgical tactics or the procedure.
Timing of FNa Administration
In all patients, a weight-adapted dose of 5 mg/kg FNa was administered intravenously prior to imaging. The timing varied. In essence, a shorter elapsed time resulted in more assessable images.
Adverse Events
We did not encounter any morbidity or mortality attributable to the use of FNa. Furthermore, no major side effects related to fluorescein throughout the observation period, apart from yellow-colored urine and, in some patients, slight yellow discoloration of the skin, were documented.
Discussion
Most surgeries of different lesions of the CNS still solely rely on tactile differentiation and visual cues to guide tumor resection. Technical advances have aimed at bridging that gap. In the past decades, efforts to safely aid maximal extent of resection have added the operating microscope, intraoperative MRI, ultrasound, and neuronavigation to the neurosurgeon’s armamentarium. Fluorescence-guided surgery has been part of this evolution. In that respect, FNa has regained interest among the neurosurgical community. In numerous tumors of the CNS, including high-grade gliomas, metastases, and hemangioblastomas, FNa has been shown to safely guide tumor resection.5–9,26 While FNa fluorescence can be observed by the naked human eye at high doses,27 the addition of a specific filter allows for a significant dose reduction, lowering the odds for dose-related complications. Furthermore, good image quality that depicts anatomy in high detail allows for continued microsurgical work in filter mode, an advantage not seen with other fluorescent agents such as indocyanine green or 5-ALA.
Here, we report our experience with CLE in 12 patients in whom contrast-enhancing lesions were removed with FNa guidance. We did not see any major differences in terms of image quality with respect to the timing of FNa injection. Data by Folaron et al. have suggested that peak fluorescence is induced by FNa at 15–30 minutes, and the best viewing opportunities occur within 90 minutes of injection.28 In our series, most CLE images were obtained within that time frame.
Several CLE systems are being utilized in medicine, mainly in gastrointestinal interventions and research settings, such as Cellvizio (Mauna Kea Technologies), EC3870 K (Pentax), or the CONVIVO. During gastrointestinal interventions such as colonoscopy, CLE can guide diagnosis and polypectomy.29
The emergence of a handheld probe-based CLE system has facilitated proof-of-concept studies in animals. Belykh et al. found high diagnostic accuracy in a mouse glioma model, with a specificity of 86% and a sensitivity of 96% when neuropathologists had to differentiate between unaffected and tumorous tissue based on CLE-generated images.30 These promising findings extended into ex vivo and in vivo human studies. Snuderl et al. obtained images of different tumor tissues and normal brain ex vivo in a multimodal fashion with CLE and were able to render diagnoses in a comparable fashion to those rendered by H & E slides.31 Sanai et al. utilized an earlier version of the CONVIVO in 31 patients for FNa-based in vivo CLE, obtaining images of different brain tumors, and found a high concordance with routine diagnostics.32 More recently, in a prospective study with 74 consecutive patients, Martirosyan et al. showed sensitivities and specificities, respectively, of 94% and 91% in gliomas and 93% and 97% in meningiomas when in vivo and ex vivo CLE images were compared with a standard frozen section for diagnosis.33 Therefore, CLE with FNa could improve intraoperative decision-making and safety in neurosurgery.34 The potential applications of CLE in neurosurgery are manifold. Potentially, it could improve the extent of resection, equip the surgeon with a tool to precisely define necessary resection margins, shorten the surgery, and provide rapid intraoperative diagnosis, which instantly benefits patients. In the future, with the establishment of a large CLE image data collection, deep learning could offer automated classification.35
Thus far, comprehensive evaluation of CLE for neurosurgical purposes is pending. FNa applied for the purpose of tumor visualization is still an off-label use. Since CLE offers an insight into a living system, we will have to learn what we see. The optimal application of CLE remains unclear as well as the timing and dose of the FNa contrast medium.
In our study, FNa was utilized to help with visualization and maximize resection. Our results were obtained using standard image guidance and microscopy with FNa fluorescence. We did not encounter any side effects attributable to the use of FNa intraoperatively or during the hospital stay. Neither did we encounter any adversities to the surgical workflow or ergonomics.
Conclusions
FNa and CONVIVO CLE are safe and feasible tools. Further prospective evaluations are needed to help neurosurgeons determine the validity of CLE and to confirm its promising potential.
Disclosures
Julius Höhne and Karl-Michael Schebesch have received speaker’s fees from Carl Zeiss Meditec, Oberkochen, Germany.
Author Contributions
Conception and design: Höhne, Riemenschneider, Schmidt. Acquisition of data: Höhne, Schebesch, Zoubaa, Riemenschneider. Analysis and interpretation of data: Höhne, Zoubaa, Schmidt. Drafting the article: Höhne. Critically revising the article: Schebesch, Proescholdt. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Höhne. Statistical analysis: Riemenschneider, Schmidt.
References
- 1↑
Nabi Z, Reddy DN. Optical biopsy in gastroenterology: focus on confocal laser endomicroscopy. Indian J Gastroenterol. 2019;38(4):281–286.
- 2↑
Schebesch KM, Hoehne J, Hohenberger C, et al. Fluorescein sodium-guided resection of cerebral metastases—experience with the first 30 patients. Acta Neurochir (Wien). 2015;157(6):899–904.
- 3↑
Stummer W, Pichlmeier U, Meinel T, et al. Fluorescence-guided surgery with 5-aminolevulinic acid for resection of malignant glioma: a randomised controlled multicentre phase III trial. Lancet Oncol. 2006;7(5):392–401.
- 4↑
Kamp MA, Munoz-Bendix C, Mijderwijk HJ, et al. Is 5-ALA fluorescence of cerebral metastases a prognostic factor for local recurrence and overall survival? J Neurooncol. 2019;141(3):547–553.
- 5
Acerbi F, Broggi M, Schebesch KM, et al. Fluorescein-guided surgery for resection of high-grade gliomas: a multicentric prospective phase II study (FLUOGLIO). Clin Cancer Res. 2018;24(1):52–61.
- 6↑
Schebesch KM, Proescholdt M, Höhne J, et al. Sodium fluorescein-guided resection under the YELLOW 560 nm surgical microscope filter in malignant brain tumor surgery—a feasibility study. Acta Neurochir (Wien). 2013;155(4):693–699.
- 7
Rey-Dios R, Cohen-Gadol AA. Intraoperative fluorescence for resection of hemangioblastomas. Acta Neurochir (Wien). 2013;155(7):1287–1292.
- 8
Höhne J, Hohenberger C, Proescholdt M, et al. Fluorescein sodium-guided resection of cerebral metastases-an update. Acta Neurochir (Wien). 2017;159(2):363–367.
- 9
Acerbi F, Cavallo C, Schebesch KM, et al. Fluorescein-guided resection of intramedullary spinal cord tumors: results from a preliminary, multicentric, retrospective study. World Neurosurg. 2017;108:603–609.
- 10
Akçakaya MO, Göker B, Kasımcan MO, et al. Use of sodium fluorescein in meningioma surgery performed under the YELLOW-560 nm surgical microscope filter: feasibility and preliminary results. World Neurosurg. 2017;107:966–973.
- 11
Hamamcıoğlu MK, Akçakaya MO, Göker B, et al. The use of the YELLOW 560 nm surgical microscope filter for sodium fluorescein-guided resection of brain tumors: our preliminary results in a series of 28 patients. Clin Neurol Neurosurg. 2016;143:39–45.
- 12↑
Nduom EK, Yang C, Merrill MJ, et al. Characterization of the blood-brain barrier of metastatic and primary malignant neoplasms. J Neurosurg. 2013;119(2):427–433.
- 13↑
Novotny HR, Alvis DL. A method of photographing fluorescence in circulating blood in the human retina. Circulation. 1961;24:82–86.
- 14↑
Mat Zin AA, Zulkarnain S. Diagnostic accuracy of cytology smear and frozen section in glioma. Asian Pac J Cancer Prev. 2019;20(2):321–325.
- 15↑
Sharifabadi AH, Haeri H, Zeinalizadeh M, et al. Intraoperative consultation of central nervous system lesions. Frozen section, cytology or both? Pathol Res Pract. 2016;212(3):179–184.
- 16↑
Charalampaki P, Nakamura M, Athanasopoulos D, Heimann A. Confocal-assisted multispectral fluorescent microscopy for brain tumor surgery. Front Oncol. 2019;9:583.
- 17↑
Lee S, Park WY, Chang H, et al. Fast and sensitive delineation of brain tumor with clinically compatible moxifloxacin labeling and confocal microscopy. J Biophotonics. 2020;13(1):e201900197.
- 18↑
Belykh E, Miller EJ, Carotenuto A, et al. Progress in confocal laser endomicroscopy for neurosurgery and technical nuances for brain tumor imaging with fluorescein. Front Oncol. 2019;9:554.
- 19↑
Xue Z, Kong L, Pan CC, et al. Fluorescein-guided surgery for pediatric brainstem gliomas: preliminary study and technical notes. J Neurol Surg B Skull Base. 2018;79(4)(suppl 4):S340–S346.
- 20↑
Belykh E, Patel AA, Miller EJ, et al. Probe-based three-dimensional confocal laser endomicroscopy of brain tumors: technical note. Cancer Manag Res. 2018;10:3109–3123.
- 22↑
WHO Model List of Essential Medicines (19th List). World Health Organization. Published 2015. Accessed November 23, 2020. https://www.who.int/groups/expert-committee-on-selection-and-use-of-essential-medicines/essential-medicines-lists
- 23↑
Ferroli P, Acerbi F, Albanese E, et al. Application of intraoperative indocyanine green angiography for CNS tumors: results on the first 100 cases. Acta Neurochir Suppl. 2011;109:251–257.
- 24↑
Rey-Dios R, Cohen-Gadol AA. Technical principles and neurosurgical applications of fluorescein fluorescence using a microscope-integrated fluorescence module. Acta Neurochir (Wien). 2013;155(4):701–706.
- 25↑
Tvrdik P, Kalani MYS. In vivo imaging of microglial calcium signaling in brain inflammation and injury. Int J Mol Sci. 2017;18(11):E2366.
- 26↑
Höhne J, Acerbi F, Falco J, et al. Lighting up the tumor-fluorescein-guided resection of gangliogliomas. J Clin Med. 2020;9(8):E2405.
- 27↑
Okuda T, Kataoka K, Yabuuchi T, et al. Fluorescence-guided surgery of metastatic brain tumors using fluorescein sodium. J Clin Neurosci. 2010;17(1):118–121.
- 28↑
Folaron M, Strawbridge R, Samkoe KS, et al. Elucidating the kinetics of sodium fluorescein for fluorescence-guided surgery of glioma. J Neurosurg. 2018;131(3):724–734.
- 29↑
Belderbos TDG, van Oijen MGH, Moons LMG, Siersema PD. Implementation of real-time probe-based confocal laser endomicroscopy (pCLE) for differentiation of colorectal polyps during routine colonoscopy. Endosc Int Open. 2017;5(11):E1104–E1110.
- 30↑
Belykh E, Miller EJ, Patel AA, et al. Diagnostic accuracy of a confocal laser endomicroscope for in vivo differentiation between normal injured and tumor tissue during fluorescein-guided glioma resection: laboratory investigation. World Neurosurg. 2018;115:e337–e348.
- 31↑
Snuderl M, Wirth D, Sheth SA, et al. Dye-enhanced multimodal confocal imaging as a novel approach to intraoperative diagnosis of brain tumors. Brain Pathol. 2013;23(1):73–81.
- 32↑
Sanai N, Eschbacher J, Hattendorf G, et al. Intraoperative confocal microscopy for brain tumors: a feasibility analysis in humans. Neurosurgery. 2011;68(2 Suppl Operative):282–290.
- 33↑
Martirosyan NL, Eschbacher JM, Kalani MY, et al. Prospective evaluation of the utility of intraoperative confocal laser endomicroscopy in patients with brain neoplasms using fluorescein sodium: experience with 74 cases. Neurosurg Focus. 2016;40(3):E11.
- 34↑
Fotteler M, Holl F, Käsbach C, et al. Using confocal endomicroscopy for digital biopsy during brain surgery: presentation of a study protocol. Stud Health Technol Inform. 2019;258:237–238.
- 35↑
Aubreville M, Knipfer C, Oetter N, et al. Automatic classification of cancerous tissue in laserendomicroscopy images of the oral cavity using deep learning. Sci Rep. 2017;7(1):11979.
