Correlation of apparent diffusion coefficient values with tumor cellularity and grade in brain neoplasms: a prospective study

Abstract

While conventional MRI provides excellent anatomical delineation of brain tumors, it remains fundamentally limited in predicting cellularity or histological grade preoperatively. Apparent diffusion coefficient (ADC) values, derived from diffusion-weighted imaging (DWI), quantify microscopic water mobility, serving as a reliable surrogate marker for tissue microstructure. This study investigates how accurately ADC measurements track histopathological features across the principal intracranial neoplasms. Sixty-one patients (31 males, 30 females; 15–65 years) with brain tumors were imaged on a standard 1.5 T MRI. DWI sequences were acquired using b-values of 0, 500, and 1000 s/mm2. Tumoral and peritumoral ADCmean and ADCmin were recorded via a standardized region-of-interest protocol. Histopathological correlation was definitively obtained in 47 surgically resected cases. Multi-group comparisons utilized ANOVA with Bonferroni-corrected post hoc testing, while two-group comparisons employed unpaired Student t-tests. Diagnostic performance for glioma grade was evaluated via ROC analysis. The ADC–cellularity relationship was assessed using Pearson correlation, and measurement reproducibility was verified via intraclass correlation coefficients (ICC). Low-grade gliomas demonstrated significantly higher ADCmin than high-grade gliomas (1161.57 ± 330.96 vs. 755.50 ± 170.25 × 10⁻⁶ mm2/s; p = 0.0008). ROC analysis for tumoral ADCmin yielded an AUC of 0.87 (95% CI: 0.74–0.99) with an optimal cutoff of 934 × 10⁻⁶ mm2/s (sensitivity 83.3%, specificity 85.7%). Meningioma cellularity revealed a strong inverse correlation with ADCmean (r = −0.67, p < 0.01). Conversely, neither tumoral (p = 0.6491) nor peritumoral ADCmin (p = 0.8503) differentiated high-grade gliomas from metastases. Densely cellular tumors—lymphomas and medulloblastomas—consistently exhibited the lowest ADC values. Intraobserver and interobserver ICC demonstrated excellent reproducibility at 0.93 and 0.89, respectively. ADC reliably correlates with glioma grade and meningioma cellularity at the group level, though considerable inter-group overlap tempers its standalone diagnostic utility. Furthermore, diffusion parameters failed to accurately distinguish high-grade gliomas from metastatic deposits. Because this cohort lacked molecular profiling (utilizing the 2016 WHO classification), these findings establish an important pre-molecular baseline. Ultimately, ADC remains a highly valuable supplementary biomarker, best interpreted alongside conventional MRI, molecular pathology, and direct tissue diagnosis.