Oliva et al., 2009[41] Retrospective 45 64 1,5 Tesla MRI pRCC, ccRCC Rosenkrantz et al., 2010[55] Retrospective 41 67 1,5 Tesla MRI ONC, chRCC Hindman NM et al., 2012[61] Retrospective 23 56 CT/MRI mcRCC Hindman N et al. 2012[65] Retrospective 108 59 (ccRCC)
54 (AML) MRI ccRCC, AML Gupta et al., 2012[26] Clinicopathologic analysis 52 55/22 Various cdRCC/mRCC Zhu et al., 2013[32] Retrospective 20 52 CT, mpMRI cdRCC Cornelis et al. 2014[45] Retrospective 90 64,1 mpMRI ccRCC/pRCC/chRCC/ONC/AML Murray et al., 2016[6] Retrospective 64 62,2/57,3 mpMRI pRCC/AML Jeong et al. 2016[71] Retrospective 152 N/A CT/MRI ccRCC, pRCC, chRCC Canvasser et al., 2017[48] Retrospective 110 57 mpMRI ccRCC/pRCC/chRCC/benign Zhang et al.,
2016 [14] Prospective 36 58 mpMRI ccRCC/pRCC/chRCC/AML Park and Kim, 2017[69] Retrospective 56 54,4 (AML) 55,7 (RCC) mpMRI AML/RCC Kay et al., 2018[42] Retrospective 103 56,7 mpMRI ccRCC/pRCC/chRCC/ONC/AML Vendrami et al., 2018[52] Retrospective 47 56/60 1,5–3 Tesla mpMRI pRCC Type 1/2 Johnson et al., 2019[47] Retrospective 57 61.7 mpMRI (ccLS) ccRCC/pRCC/chRCC/ONC/benign Zhu et al., 2021[58] Retrospective 33 52,1 CT/MRI mcRCC, cdRCC Steinberg 2021 [46] Retrospective 434 60 mpMRI ccRCC/pRCC/chRCC De Silva et al., 2021[62] Retrospective 66 N/A 3 Tesla MRI ccRCC/pRCC/chRCC/ONC/AML Dunn et al., 2022[63] Retrospective 102 56. 1,5 Tesla mpMRI ccRCC/pRCC/chRCC/ONC/AML Beek et al., 2023[44] Retrospective 6 12 1,5 Tesla MRI MiT-RCC: 2 patients (33%); ccRCC: 2 patients (33%); Other types: 2 patients. Wang et al., 2024[56] Retrospective 105 62 mpMRI ccRCC, pRCC, chRCC, cdRCC mRCC Study T1 characteristics T2 characteristics Diffusion restriction Enhancement pattern Main findings Oliva et al., 2009[41] No T1 signal intensity ratio difference pRCC vs ccRCC pRCC: T2 hypointense ccRCC: T2 hyperintense N/A N/A T2 imaging aids RCC differentiation pRCC (T2 hypointense), ccRCC (T2 hyperintense) Rosenkrantz et al., 2010[55] Hypointense Heterogeneous lipid noted in some chRCC. Peripheral, well-circumscribed, no fat/vein invasion, segmental enhancement inversion in 13.3–42.9%. Central Scar: 50–60.7% in ONC, 33.3–40% in chRCC. Hindman NM et al., 2012[61] N/A N/A N/A N/A McRCC lesions behaved benignly, with no metastasis or recurrence. Hindman N et al. 2012[65] Signal loss on opposed-phase imaging showed no significant difference AML and ccRCC AML: Low SI relative to cortex strongly associated; ccRCC: High SI more frequent. N/A Necrosis and cystic degeneration were significantly associated with ccRCC Opposed-phase imaging lacked reliability, but small size and low T2 SI strongly predicted AML. Gupta et al., 2012[26] Hypo to isointense heterogeneous hyperintense Heterogeneity and restricted diffusion cdRCC: Heterogeneous mRCC: Rapid, high vascularity cdRCC: Aggressive, metastatic, desmoplastic stroma, infiltrative margins.
mRCC: Highly aggressive, advanced stages, linked to sickle cell anemia. Zhu et al., 2013[32] Isointense Iso-or hypointense N/A lower enhancement Medullary; poorly defined, often solid with cystic/necrotic changes, may include calcifications, show higher radiodensity on CT, lower enhancement and isointensity on T1/T2 MRI. Cornelis et al. 2014[45] pRCC: Slow and low enhancement. ONC: Early and strong enhancement. pRCC: Low T2WI chRCC: Intermediate T2WI
ONC: T2WI signal is similar to parenchyma. pRCC: Low ADC ratio (ADCr<54.2). ccRCC: Moderate ADC ratio. ONC: Higher ADC values. pRCC: Low WiI1 (<30.9). ONC: High WiI2 (>257). chRCC: Delayed WoI2 (>− 8.8). pRCC: Low T2WI signal, low ADC.
ONC: High wash-in, low wash-out. Minimal-fat AMLs: High T2WI signal in non-fat saturated sequences. Murray et al., 2016[6] Chemical shift T1WI MRI distinguishes pRCC T2WI alone can’t differentiate pRCC and AML N/A N/A Chemical shift MRI aids pRCC distinction but lacks sufficient sensitivity alone. Jeong et al. 2016[71] Similar for AML (0.97) and RCC (0.89) Lower in AML (0.75) compared to RCC (1.21) N/A N/A Fat-invisible AML is best differentiated from RCC by tumor-to-cortex ratios on T2WI MRI and unenhanced CT, while chemical-shift MRI shows poor accuracy. Canvasser et al., 2017[48] ccRCC: Heterogeneous, microscopic fat. Mostly high signal. ccRCC: Intense contrast uptake in cortical regions. ccRCC: Cortical contrast uptake. ccRCC: 78% sensitivity, 80% specificity (ccLS 4–5); ccLS 1–2 indicates benign/non-ccRCC. Zhang et al.,
2017 [14] ccRCC: Iso-to hyperintense; pRCC: Hypointense ccRCC: Hyperintense; pRCC: Hypointense ccRCC: Variable; pRCC: Minimal ccRCC: Heterogeneous with high Ktrans and Kep; pRCC: Lower Ktrans and Kep ASL correlates with DCE; ccRCC: heterogeneous, pRCC: low perfusion Park and Kim, 2017[69] No intensity difference AML versus RCC AML: Predominantly low T2WI intensity RCC had lower ADC N/A ADC predicted RCC versus AML, with higher accuracy when combined with male sex; minimal-fat AMLs had higher ADC, while T2WI metrics lacked differentiation. Kay et al., 2018[42] Signal intensity in the corticomedullary phase. ccRCC: High T2 signal; pRCC and benign lesions: Low T2 signal. N/A ccRCC: High enhancement; pRCC: Low enhancement; ONC: Segmental enhancement inversion. Diagnosis accuracy: 81% for ccRCC, 91% for pRCC. Vendrami et al., 2018[52] Type 1: 54% iso, 23% hypo, 23% hyperintense Type 2: 56% iso, 31% hypo, 13% hyperintense Type 1: Homogeneous (36%); Heterogeneous (64%) – Type 2: Homogeneous (12%); Heterogeneous (88%) Type 2 tumors had lower mean ADCs Type 1: Predominantly homogeneous (65%) Type 2: Predominantly heterogeneous (75%) Type 2 pRCC shows more heterogeneity, necrosis, and benefits from texture analysis for differentiation. Johnson et al., 2019[47] High intravoxel fat signal Heterogeneous signal in ccRCC, low signal in pRCC Significant diffusion restriction in ccRCC Heterogeneous enhancement in ccRCC; homogeneous low enhancement in pRCC. ccLS 4–5 scored ccRCC at 84% accuracy, ccLS 1–2 scored non-ccRCC at 100%. Zhu et al., 2021[58] Hypointense mcRCC: Hyperintense; cdRCC: Hypointense N/A mcRCC: Thickened enhancing internal septations and mural soft-tissue nodules mcRCC better-defined boundaries, exogenous growth, and excellent survival, cdRCC infiltrative growth, renal pelvis/ureter involvement, and poor prognosis with high metastasis and mortality. Steinberg 2021 [46] Assessed per ccLS using intensity patterns Assessed per ccLS using intensity patterns B800 diffusion-weighted images. Heterogeneous, moderate ccLS1–2: mostly benign; ccLS5: 93% ccRCC De Silva et al.,
2021 [62] Isointense Hypointense Moderate restriction Homogeneous, mild ONCs the highest ADC (max), pRCC the lowest ADC, ccRCC has higher ADC than pRCC and chRCC Dunn et al., 2022[63] ccRCC: Higher signal intensity ccRCC: Typically T2WI hyperintense N/A ccRCC: >75% enhancement; ADER aids subtype differentiation ccLS: 85% sensitivity, 82% specificity, 83% accuracy; ccLS ≥4 strongly predicts ccRCC Beek et al., 2023[44] Mostly isointense. Mostly hypointense Median ADC: 0.70–1.20×10−3 mm2/s; lower in MiT-RCC. homogeneous strong enhancement MiT-RCC: T2-hypointense, well-defined pseudocapsules (4/6), median volume 393 cm≥, lobulated shape (4/6). Wang et al., 2024[56] Intensity variations, pseudocapsules, and necrosis. T2WI hyperintense signals. Less hypointense signals in sarcomatoid components. Lower ADC values indicate higher cellular density and aggressiveness. Lower TCEI in adverse pathology. Male gender, high RENAL score, necrosis, irregular margins, and low ADC predict adverse pathology.