Fluorescein-stained confocal laser endomicroscopy versus conventional frozen section for intraoperative histopathological assessment of intracranial tumors

Abstract

Background: The aim of this clinical trial was to compare Fluorescein-stained intraoperative confocal laser endomicroscopy (CLE) of intracranial lesions and evaluation by a neuropathologist with routine intraoperative frozen section (FS) assessment by neuropathology.

Methods: In this phase II noninferiority, prospective, multicenter, nonrandomized, off-label clinical trial (EudraCT: 2019-004512-58), patients above the age of 18 years with any intracranial lesion scheduled for elective resection were included. The diagnostic accuracies of both CLE and FS referenced with the final histopathological diagnosis were statistically compared in a noninferiority analysis, representing the primary endpoint. Secondary endpoints included the safety of the technique and time expedited for CLE and FS.

Results: A total of 210 patients were included by 3 participating sites between November 2020 and June 2022. Most common entities were high-grade gliomas (37.9%), metastases (24.1%), and meningiomas (22.7%). A total of 6 serious adverse events in 4 (2%) patients were recorded. For the primary endpoint, the diagnostic accuracy for CLE was inferior with 0.87 versus 0.91 for FS, resulting in a difference of 0.04 (95% confidence interval -0.10; 0.02; P = .367). The median time expedited until intraoperative diagnosis was 3 minutes for CLE and 27 minutes for FS, with a mean difference of 27.5 minutes (standard deviation 14.5; P < .001).

Conclusions: CLE allowed for a safe and time-effective intraoperative histological diagnosis with a diagnostic accuracy of 87% across all intracranial entities included. The technique achieved histological assessments in real time with a 10-fold reduction of processing time compared to FS, which may invariably impact surgical strategy on the fly.

Keywords: brain tumor histology; confocal laser endomicroscopy; frozen section; intraoperative diagnosis; telepathology.

Figure 1.

CLE image unmarked (A) and marked (B), hematoxylin–eosin stains of FS (C), and final histopathology smear (D) of a 74-year-old female with a right paratrigonal IDH wild-type glioblastoma. The CLE image is concordant with hallmark histological features of a glioblastoma, with necrotic areas (stars), pleomorphic glial tumor cells with hyperchromatic nuclei (circles), and vascular proliferates (arrows). Abbreviations: CLE, confocal laser endomicroscopy; FS, frozen section; IDH, isocitrate dehydrogenase.

Figure 2.

CLE image unmarked (A) and marked (B), hematoxylin–eosin stains of FS (C), and final histopathology smear (D) of a 68-year-old female with a left temporal melanoma metastasis, symptomatic with grand-mal seizures. The CLE demonstrates spots of auto-fluorescence (stars) as well as denser cellular areas (circles). FS assessment was ambivalent between melanoma and glioma, rendering the in vivo CLE the more conclusive diagnosis in this case. Abbreviations: CLE, confocal laser endomicroscopy; FS, frozen section.

Figure 3.

CLE image unmarked (A) and marked (B), hematoxylin–eosin stains of FS (C), and final histopathology smear (D) of a 68-year-old female with a sphenoidal plane meningioma. The transitional meningioma features both meningothelial (lobular, pseudo-syncytial growth; circles) and fibrous phenotypes (spindle-shaped cells, collagen fibers; stars) with a high number of psammoma bodies (arrows). Abbreviations: CLE, confocal laser endomicroscopy; FS, frozen section.