ADC Is a More Accurate Biomarker of Tumor Burden Than T1, T2, or FLAIR in Primary Glioma

Introduction: Limitations of radiographic markers for delineating tumor burden contribute to poor prognosis and recurrence in glioma. In IDHwt glioma, CE – the standard marker for GTR – is limited by its lack of specificity and its binary properties. In IDHmut glioma, no canonical marker exists to guide GTR. Accurate tumor representation is vital for successful surgical resection, diagnostic biopsies, and radiation field mapping. To establish the best noninvasive biomarker for tumor burden, we assessed the relationship between tumor burden and standard imaging.

Methods: 71 IDHwt and 64 IDHmut MRI localized biopsies were collected at the time of resection with all patients receiving standard-of-care preoperative imaging, including diffusion-weighted imaging and the apparent diffusion coefficient (ADC). For each biopsy, imaging intensity values were compared to tumor burden (SOX2) and other histologic markers representing the brain tumor microenvironment (NeuN, CD68, Neuropil).

Results: IDHwt gliomas showed a significant negative parametric relationship between ADC and tumor cell density; this relationship was positive in IDHmut gliomas. Importantly, there was no significant relationship between tumor burden and CE or T2-FLAIR values. Tumor margins generated using ADC values establish a larger tumor area than CE or T1hypointensity. The ADC generated tumor margins were smaller but more accurate than FLAIR.

Conclusion : By improving the precision of imaging-based assessments of tumor margins and burden, surgeons can make more informed decisions regarding the onco-functional costs associated with different surgical margins. This study highlights the utility of the ADC to reliably delineate tumor margins and quantitatively estimate tumor burden, thereby redefining GTR parameters and introducing a potential novel standard for tumor assessment. As a parametric measure, ADC enables the creation of a heatmap representing tumor burden, effectively producing a detailed map of virtual biopsies, which can inform personalized treatment strategies for optimal patient outcomes. Future studies should explore the potential significance of tumor type differences.