Advancing intraoperative diagnostics, one (or rather three) mutations at a time

The incorporation of highly recurrent, disease-defining biomarkers into brain tumor classification has already profoundly impacted the Neuro-oncology field and promises to have far-reaching consequences on the clinical management primary brain tumors. Full realization of these advances, however, will require more facile application of crucial molecular profiling technology in the context of rapid diagnostics. Indeed, the impact of such developments would be considerable (see below). In this issue of Neuro-oncology, Maeda, Kitano, and Ohka et al. report a rapid genotyping approach for IDH1 R132H and TERT C228T/C250T mutations, enabling in most cases assignment of adult diffuse glioma samples into the three most common disease variants—astrocytoma, IDH-mutant (IDH1 R132H-positive and TERT mutation-negative); oligodendroglioma, IDH-mutant and 1p/19q codeleted (IDH1 R132H-positive and TERT mutation-positive), and glioblastoma, IDH-wildtype (mostly IDH1 R132H-negative and TERT-mutation positive)—within a 25-minute time frame.¹ Their findings represent an important milestone for a burgeoning area of study, whose significance extends beyond the specific technology implemented.

In their investigation, Maeda, Kitano, and Ohka et al. utilized a portable, real-time polymerase chain reaction (PCR) platform (GeneSoC) with microfluidic thermocycling, pairing this technology with a simplified DNA extraction methodology, consisting of boiling fresh patient tissue in ultrapure water supplemented with proteinase K. Following standardization experiments with banked frozen samples to determine fluorescence intensity detection thresholds for the different assayed mutations, plasmid reagents were used to ascertain optimal input levels of genomic DNA for each analysis, minimizing risk of false positive and negative results. These preliminary studies demonstrated that the GeneSoC-based methodology could reliably detect mutant alleles at a frequency of 5% or greater. The authors then deployed their genotyping strategy on 120 cases in the intraoperative setting, under their predetermined conditions, validating molecular findings by conventional sanger sequencing of corresponding, fully resected tumor material. They found excellent sensitivity and specificity for all assessed mutations—sensitivity: 0.985, specificity: 0.982 for IDH1 R132H; sensitivity: 1.000, specificity: 1.000 for TERT C228T/C250T—reflecting robust concordance between rapid and standard methodologies. Their large case cohort also enabled ascertainment of reasonable size standards for tissue sample inputs (4.8-48.2 mg) with good DNA yields (1.5-3.4 ug), across a range of estimated tumor cellularities (29.53%-83.87%). Finally, using multiregional sampling in a single case of oligodendroglioma, IDH-mutant and 1p/19q codeleted, they were able to detect tumor-associated mutations in one of four grossly normal peri-tumoral brain regions, validating these results with relevant post-operative histopathological and molecular analyses. These latter findings demonstrate the feasibility of implementing the GeneSoC-based rapid genotyping approach to define microscopic tumor boundaries intraoperatively.

This work is significant in multiple respects. Shifting the time frame for the acquisition of tumor-classifying biomarkers to the intraoperative setting has the potential to substantively alter current practice paradigms. Treatment standards for the different forms of adult diffuse glioma continue to diverge,^2,3 and earlier diagnosis should further enable the optimization of tumor-specific management schemes. The three-mutation genotyping approach deployed in this paper, while somewhat limited in scope, effectively delineates the majority of adult glioma cases into integrated diagnostic categories. Moreover, expansion of this methodology to incorporate additional high-yield mutations (depending on clinical context) such as H3F3A K27M and G34R/V appears relatively straight-forward. Above and beyond pure diagnostics, however, facilitating the more accurate assessment of microscopic tumor boundaries may have greater impact on the care of primary brain tumor patients. As the authors themselves assert, maximizing extent of surgical resection has been repeatedly demonstrated to improve prognosis across diffuse glioma subtypes, and supramaximal resection may confer additional survival benefit.^4-7 Improved intraoperative designation of tumor boundaries would also benefit surgery in the recurrent setting, where treatment-related effects can further complicate gross tissue assessments.

Perhaps the main limitation of this work is its reliance on a specific technological platform, likely precluding wide-spread methodological application in the short-term. Nevertheless, in a broader sense, this report further underscores the feasibility of intraoperative molecular diagnostics, building on recent literature featuring alternative technologies, such as stimulated Raman histology (SRH),⁸ DNA methylation profiling,⁹ and mass spectrometry.¹⁰ In this context, the most incisive observation of this study may in fact be its demonstration that “quick and dirty” DNA extraction provides suitable material for rapid, sophisticated molecular profiling, the individual components of which could be customized and situation-dependent, reflecting the essence of personalized medicine.

Funding

I declare no relevant funding source.

Conflict of interest statement

I declare no relevant conflict of interest.