Double-echo gradient chemical shift MR imaging fails to differentiate minimal fat renal angiomyolipomas from other homogeneous solid renal tumors.

2015 
Abstract Objectives The purpose of this retrospective study was to evaluate the diagnostic performance of double-echo gradient chemical shift (GRE) magnetic resonance (MR) imaging for the differentiation of angiomyolipomas with minimal fat (mfAML) from other homogeneous solid renal tumors. Methods Between 2005 and 2010 in two institutions, all histologically proven homogenous solid renal tumors imaged with computed tomography and MR imaging, including GRE sequences, have been retrospectively selected. A total of 118 patients (mean age: 61 years; range: 20–87) with 119 tumors were included. Two readers measured independently the signal intensity (SI) on GRE images and calculated SI index (SII) and tumor-to-spleen ratio (TSR) on in-phase and opposed-phase images. Intra- and interreader agreement was obtained. Cut-off values were derived from the receiver operating characteristic (ROC) curve analysis. Results Twelve mfAMLs in 11 patients were identified (mean size: 2.8 cm; range: 1.2–3.5), and 107 non-AML tumors (3.2 cm; 1–7.8) in 107 patients. The intraobserver reproducibility of SII and TSR was excellent with an intraclass correlation coefficient equal to 0.99 [0.98–0.99]. The coefficient of correlation between the readers was 0.99. The mean values of TSR for mfAMLs and non-mfAMLs were −7.0 ± 22.8 versus −8.2 ± 21.2 for reader 1 and −6.7 ± 22.8 versus −8.4 ± 20.9 for reader 2 respectively. No significant difference was noticed between the two groups for SII ( p  = 0.98) and TSR ( p  = 0.86). Only 1 out of 12 mfAMLs and 11 of 107 non-AML tumors presented with a TSR inferior to −30% ( p  = 0.83). Conclusion In a routine practice, GRE sequences cannot be a confident tool to differentiate renal mfAMLs from other homogeneous solid renal tumors.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    12
    References
    18
    Citations
    NaN
    KQI
    []