LAUSR

BackgroundNephroblastomas represent a group of heterogeneous tumours with variable proportions of distinct histopathological components.ObjectiveThe purpose of this study was to investigate whether direct comparison of apparent diffusion coefficient (ADC) measurements with post-resection histopathology subtypes is feasible and whether ADC metrics are related to histopathological components.Materials and methodsTwenty-three children were eligible for inclusion in this retrospective study. All children had MRI including diffusion-weighted imaging (DWI) after preoperative chemotherapy, just before tumour resection. A pathologist and radiologist identified corresponding slices at MRI and postoperative specimens using tumour morphology, the upper/lower calyx and hilar vessels as reference points. An experienced reader performed ADC measurements, excluding non-enhancing areas. A pathologist reviewed the corresponding postoperative slides according to the international standard guidelines. We tested potential associations with the Spearman rank test.ResultsSide-by-side comparison of MRI–DWI with corresponding histopathology slides was feasible in 15 transverse slices in 9 lesions in 8 patients. Most exclusions were related to extensive areas of necrosis/haemorrhage. In one lesion correlation was not possible because of the different orientation of sectioning of the specimen and MRI slices. The 25% ADC showed a strong relationship with percentage of blastema (Spearman rho=−0.71, P=0.003), whereas median ADC was strongly related to the percentage stroma (Spearman rho=0.74, P=0.002) at histopathology.ConclusionSide-by-side comparison of MRI–DWI and histopathology is feasible in the majority of patients who do not have massive necrosis and hemorrhage. Blastemal and stromal components have a strong linear relationship with ADC markers.