A systematic review of radiological prediction of ki 67 proliferation index of meningioma.

objectiveRadiological prediction of Ki-67 plays a crucial role in determining the clinical outcomes of meningioma patients. The aim of this review is to comprehensively review the literature to synthesize evidence on how accurate radiological prediction of the Ki-67 proliferation index is to determine the Ki-67 index's association with clinical outcomes of meningioma. MATERIALS AND

methodsA narrative synthesis systematic review followed PRISMA guidelines. The literature was searched using database-specific, relevant keywords, text words, and MeSH terms (controlled vocabulary) such as "Meningioma," "ki-67," "Ki 67 proliferation index," "proliferation index," "Radiomics," and "clinical outcomes" on electronic databases PubMed, Web of Science, and Google Scholar from 2014 to April 2024.

resultsOut of 218 publications identified initially, only 03 moderate-quality studies were included in this paper after methodological quality assessment using the Newcastle-Ottawa scale. DISCUSSION: The evidence synthesis by systematic review emphasizes the importance of radiographic approaches in predicting the Ki-67 proliferation index and their consequences for prognosis and treatment results in meningioma patients. Various radiological techniques, such as pre-operative MRI with clinical and radiomic analysis and machine learning algorithms, show promise for accurately predicting Ki-67 status, with significant associations between radiological features and clinical outcomes such as recurrence, survival, and progression-free survival. The radiomics models were compared and found that Lasso with LDA radiomic models outperformed other radiomics models in terms of accuracy, sensitivity, specificity, and predictive power. Lasso with LDA model also showed significant improvements in performance when the clinical datasets of meningiomas patients were combined to predict ki-67 levels. However, the comparison to histological evaluation emphasizes the necessity for additional validation and standardization. While these approaches provide non-invasive and possibly time-saving alternatives, drawbacks such as inherent biases, methodological limits, and practical challenges with automated selection procedures highlight future research and development topics.