Assessment of Gliomas' Grade of Malignancy and Extent of Resection Using Intraoperative Flow Cytometry

Abstract

Background: Intraoperative Flow Cytometry (iFC) is a novel technique for the assessment of the grade of malignancy and the diagnosis of tumor type and resection margins during solid tumor surgery. Herein, we set out to analyze the role of iFC in the grading of gliomas and the evaluation of resection margins.

Material and methods: iFC uses a fast cell cycle analysis protocol (Ioannina Protocol) that permits the analysis of tissue samples within 5-6 min. Cell cycle analysis evaluated the G0/G1 phase, S-phase, mitosis, and tumor index (S + mitosis phase fraction) and ploidy status. In the current study, we evaluated tumor samples and samples from the peripheral borders from patients with gliomas who underwent surgery over an 8-year period.

Results: Eighty-one patients were included in the study. There were sixty-eight glioblastoma cases, five anaplastic astrocytomas, two anaplastic oligodendrogliomas, one pilocytic astrocytoma, three oligodendrogliomas and two diffuse astrocytomas. High-grade gliomas had a significantly higher tumor index than low grade gliomas (median value 22 vs. 7.5, respectively, p = 0.002). Using ROC curve analysis, a cut-off value of 17% in the tumor index could differentiate low- from high-grade gliomas with a 61.4% sensitivity and 100% specificity. All low-grade gliomas were diploid. From the high-grade gliomas, 22 tumors were aneuploid. In glioblastomas, aneuploid tumors had a significantly higher tumor index (p = 0.0018). Twenty-three samples from glioma margins were evaluated. iFC verified the presence of malignant tissue in every case, using histology as the gold standard.

Conclusion: iFC constitutes a promising intraoperative technique for glioma grading and resection margin assessment. Comparative studies with additional intraoperative adjuncts are necessary.

Keywords: flow cytometry; glioma; grade of malignancy; intraoperative.

Figure 1

DNA analysis with intraoperative Flow Cytometry in a case of an oligodendroglioma (WHO grade II). Histograms represent DNA content distributions. Markers M1, M2, and M3 correspond to cells in phase G0/G1, S, and G2/M, respectively. Left: distribution of peripheral blood mononuclear cells/PBMCs (solid green in the right overlay), middle: distribution of cancer cells (presented in red in the right overlay). The presented case is diploid, with a DNA index = 1 and a tumor Index of ~8%; right: overlay of two previous histograms.

Figure 2

(A). DNA analysis with intraoperative Flow Cytometry in a case of a glioblastoma (WHO grade IV). Markers M1, M2, and M3 correspond to cells in phase G0/G1, S, and G2/M, respectively. Left histogram: distribution of peripheral blood mononuclear cells/PBMCs (in green), middle histogram: distribution of malignant cells (in red). The presented case is diploid, with a DNA index = 1 and a tumor index of ~18%; right histogram: a tumor margin (in orange) with index of ~11%. (B). I. Histological section of a case of glioblastoma (Hematoxylin & Eosin stain, original magnification × 100). II. Neoplastic cells infiltrate normal glial tissue at the peripheral borders of the tumor (Hematoxylin & Eosin stain, original magnification ×40). Scale bars (100 μM) are present in lower right part of each figure.

Figure 3

Tumor characterization and margin evaluation in a case of high-grade glioma using intraoperative flow cytometry (iFC). Following PI-staining, DNA was quantified in a flow cytometer. DNA quantity in cell populations of peripheral blood mononuclear cells (PBMCs), cancer, and margins are presented in respective plots. Marker M1 in PBMCs sample denotes cells in G0/G1 cell cycle phase. In cancer sample, Markers M2, M3, and M4 correspond to G0/G1, S, and G2/M cell cycle phases. Tumor index and DNA index quantified, based on the results obtained by iFC and presented in the upper right in each respective fluorescence plot (*** represents absence of cells in the respective marker). In our case, the tumor index has been calculated as ~22% (proportion of cells in S and G2/M). DNA index, a measure of the DNA content and ploidy status, having normal G0/G1 of PBMCs as a reference, was calculated as ~1.75, meaning that the tumor is hyperploid. The gradual reduction in tumor index and/or DNA index in margin samples denotes margin status. The margins are sorted from positive to negative. The lower panel presents overlays of different margin samples regarding normal cells. Margin samples are represented in red in the overlay with green sample representing normal PBMCs.

Figure 4

DNA index in low-grade versus high-grade gliomas: DNA index from individual cases has been quantified by iFC as the geometric mean of G0/G1 in cancer cells divided by that of normal diploid cells. DNA index is presented as blue or red dots, for low- and high-grade gliomas, respectively. Median DNA index is shown as a horizontal line in each group. A DNA index of ≠1 has been found only in high-grade gliomas.

Figure 5

The percentage of cells in G0/G1 (A) and the respective tumor index (B) in low-grade (blue dots) versus high-grade gliomas (red dots), as quantified by iFC. Tumor index has been quantified in individual samples as the cumulative percentage of cells in S and G2/M cell cycle phases. Median percentages are shown as horizontal lines in each group.

Figure 6

Receiver operating characteristic (ROC) curve analysis of tumor index for glioma grading. The ROC curve illustrates the sensitivity and specificity of iFC-derived tumor index values in differentiating low-grade from high-grade gliomas. The area under the curve (AUC) was 0.876. The optimal cut-off value for tumor index was determined to be 17%. This cut-off value yielded a sensitivity of 61.4% and a specificity of 100% for distinguishing low-grade from high-grade gliomas. The diagonal line on the ROC curve represents a random classifier, while the curve represents the performance of the iFC-derived tumor index in distinguishing between low- and high-grade gliomas. The closer the curve is to the top-left corner, the better the test performs.