The predictive value of MR cytometry in histological differentiation of rectal cancer: an exploratory study

Abstract Purpose To investigate and compare the diagnostic value of different MR cytometry methods in predicting histological differentiation of rectal tumors. Materials and methods This prospective study (ClinicalTrials.gov identifier: NCT07107815) enrolled eligible patients with rectal cancer from March 2025 to July 2025. All patients underwent rectal MRI with oscillating gradient spin-echo and pulsed gradient spin-echo sequences. Microstructural parameters were obtained from three different MR cytometry methods. Based on pathological results, rectal tumors were classified as poor differentiation and well/moderate differentiation. Intergroup comparison was conducted using Mann–Whitney U-test. The diagnostic value of imaging metrics, including microstructural parameters and apparent diffusion coefficients (ADCs), in distinguishing rectal cancers with different histological differentiation was evaluated by logistic regression analysis. Results A total of 86 patients were included (mean age: 60.5 ± 10.7 years; male proportion: 66.3%; maximal tumor diameter: 38.7 ± 11.1 mm), including 37 with poor differentiation, 49 with well/moderate differentiation. Intracellular volume fraction and cellularity were higher ( p < 0.0001), while extracellular diffusivity, water exchange rate constant, and ADC metrics were lower ( p -values from 0.01 to < 0.0001) in the poor-differentiation group. Among the classifiers based on a single imaging metric, intracellular volume fraction provided the highest areas under the receiver operating characteristic curves (AUC = 0.812). Clinical performance of the combined regression models incorporating microstructural parameters and ADC metrics (AUC = 0.883) was significantly superior to the conventional ADC measurement (AUC = 0.795). Conclusion MR cytometry provides additional information over ADC measurements in identifying histological differentiation grades of rectal cancer; the integration of MR cytometry into clinical scans may improve the diagnostic performance of rectal MRI. Key Points Question Microstructural parameters obtained from MR cytometry methods and apparent diffusion coefficients showed significant differences between rectal tumors with different histological differentiation grades. Findings The combined regression models, including both microstructural parameters and apparent diffusion coefficient metrics, provided a significantly higher value than the conventional PGSE-based ADC measurement. Clinical relevance Incorporating transcytolemmal water exchange into biophysical modeling can further improve the clinical performance over the impermeable model. Graphical Abstract