ADC histogram analysis for differentiating high-grade and low-grade meningiomas: a systematic review and meta-analysis.

AIM: Accurate meningioma grading is essential in neurosurgical practice as World Health Organization grade influences prognosis, recurrence, and treatment strategies. This meta-analysis evaluating ADC histogram analysis for preoperative meningioma grading represents a major methodological advancement by extracting multiple radiomic parameters, overcoming limitations of conventional single mean ADC approaches. This enables paradigmatic shift from histopathology-dependent to preoperative radiomic-based grading, transforming neurosurgical practice from reactive to predictive medicine.

MATERIALS AND

METHODS

We conducted a systematic review and meta-analysis following preferred reporting items for systematic reviews and meta-analysis, searching PubMed, Embase, and Cochrane Library databases. Studies evaluating preoperative ADC histogram analysis diagnostic performance in intracranial meningioma were included. Primary outcomes were sensitivity, specificity, and overall accuracy. Data were pooled using random-effects models with bivariate-derived area under the curve analysis.

RESULTS

Six studies comprising 533 patients diagnosed with intracranial meningioma were analysed. Sensitivity and specificity analysis showed pooled sensitivity of 0.71 (95% CI: 0.55-0.83; I2 = 68.5%, P=.0072) and pooled specificity of 0.72 (95% CI: 0.55-0.85; I2 = 86.3%, P<.0001). Overall diagnostic accuracy was 0.76 (95% CI: 0.60-0.84). A moderate threshold effect was observed (correlation coefficient = 0.42), with no significant publication bias detected (P=0.27). Heterogeneity was substantial for specificity but moderate for sensitivity.

CONCLUSION

ADC histogram analysis shows potential for preoperative meningioma grading despite moderate heterogeneity. In the era of minimally invasive medicine, findings suggest enhanced surgical decision-making and personalised patient management. Future research should focus on standardisation and multicenter validation.