Intraoperative confocal microscopy is an emerging technology that enables the cytoarchitecture of live tissue to be visualized with spatial resolution on a cellular level (Becker et al. 1996; Khoshyomn et al. 1998; Tadrous 2000). Until recently, the size of requisite apparatus limited the use of confocal microscopy to examination of excised tissue samples or isolated cells in a laboratory. The latest incarnation of this technology, however, features fiber-optic and microscopic miniaturization, substantially expanding its portability and in vivo applicability in a clinical setting (Delaney et al. 1994; Flusberg et al. 2005, 2008; Helmchen 2002). A single optical fiber can be used as both the illumination point source and detection pinhole to acquire high-resolution images and to combine them with miniaturized scanning and optical systems (Hoffman et al. 2006).

Consequently, this device has recently been integrated into the distal tip of conventional video endoscopes and combined with intravenous fluorophores to screen the gastrointestinal mucosa for cancer (Kiesslich et al. 2005; Polglase et al. 2005). The bladder mucosa, skin, and eye have also been studied with in vivo confocal microscopy (Brezinski et al. 1998; Bussau et al. 1998; Koenig et al. 2001; Papworth et al. 1998). We have recently used a glioma model to study the capacity of handheld confocal imaging to discern microvasculature, the grey-white junction, and tumor margins in the rodent cortex (Eschbacher et al. 2012). This initial feasibility analysis of intraoperative confocal microscopy establishes its potential value during the microsurgical resection of both intra- and extra-axial tumors.

For a variety of tumor histologies, including gliomas, meningiomas, hemangioblastomas, and central neurocytomas, this handheld device generates a real-time, fluorescein-enhanced pathological image that is of sufficient resolution for a neuropathologist to establish a preliminary diagnosis. Mounting evidence supports the value of maximizing the extent of resection of many intracranial tumors, including gliomas (Sanai and Berger 2008). Consequently, several intraoperative techniques have been developed to help achieve this goal: intraoperative ultrasonography (LeRoux et al. 1989), intraoperative MR imaging (Alexander et al. 1997), optical spectroscopy (Toms et al. 2005), and fluorescent dyes such as 5-aminolevulinic acid (Stummer et al. 2006). However, none of these technologies can detect infiltrating tumor margins at a microscopic level. In contrast, our preliminary experience with intraoperative confocal microscopy not only suggests a correlation between imaging and tumor grade but has demonstrated a capacity to distinguish tumor margins from adjacent parenchyma.

Tumor heterogeneity and biopsy sampling error remain considerable sources of inaccuracy and primary causes of undergrading gliomas. The ease with which a handheld confocal microscope can generate numerous optical biopsies during the course of tumor exposure and resection may effectively neutralize such diagnostic imprecision. Furthermore, the ability of the device to scan as deeply as 500 μm through the visible tissue surface permits analysis of an even broader spectrum of tissue. This unique capability also raises the possibility of examining subependymal regions through an intraventricular corridor, allowing intraoperative confocal microscopy to detect pockets of subependymal tumor migration that are radiographically undetectable and remote from the primary tumor site. It has been postulated that subependymal spread is both a negative prognostic sign and a primary route of glioma migration (Lim et al. 2007). The ability to detect its occurrence at an early stage may permit anticipatory and focused interventions as well as lead to insights into basic mechanisms of gliomagenesis (Sanai et al. 2005).

Routine neuropathological diagnostic analysis depends on excision and processing of tumor tissue, which can be altered both mechanically and chemically during the analysis itself. Use of intraoperative confocal microscopy circumvents these limitations with real-time in situ visualization of tumor cytoarchitecture and microvasculature. The opportunity to observe active blood flow through tumor capillaries also provides a unique visual dimension to tissue assessment. The probe can quickly survey the visible extent of a tumor with the potential to detect disconnected, cell-dense islands that may portend worse histologic grade and heterogeneity. Future pilot studies will not only enhance our familiarization with this technique, but they may demonstrate a correlation with frozen section findings that could allow confocal microscopy to replace conventional intraoperative diagnostic methods. This modern approach to intraoperative diagnosis would also lend itself to remote, internet-based analyses, expanding the accessibility of diagnostic expertise beyond centers with dedicated tumor neuropathologists.

Future versions of the device may benefit from an auto-irrigation mechanism, similar to the endoscope. Furthermore, while fluorescein contrast allows detailed assessment of overall tissue cellularity, the ability to discern cellular cytoplasm remains limited, as does its specificity for nuclear morphology. Since tumor cell nucleus-to-cytoplasm ratios are critical to histopathological diagnosis, other contrast agents should be investigated. One candidate agent is 5-aminolevulinic acid (5-ALA), which allows highly specific localization of tumor cells. Currently, however, the technique is only available for clinical use in Europe.