0.5 Tesla cine-magnetic resonance angiography is a poor screening test for atheromatous renal artery stenosis in chronic renal failure

SUMMARY: A prospective comparative study of 0.5 Tesla cine-magnetic resonance arteriography (MRA) versus standard arteriography (SA) was performed in 42 patients with clinical suspicion of atheromatous renal artery stenosis (ARAS), all of whom had chronic renal failure (average creatinine for 42 patients was 269.2±103.4 μmol/L). MRA was performed on a Philips 0.5 Tesla T5 release III, (London, UK) using T1 gradient echo cine, and 3D phase contrast. SA was performed as an aortic flush, with or without selective renal studies, via the femoral artery. Intravenous digital subtraction angiography was performed (instead of a flush procedure) in two patients with severe femoral atheroma. One radiologist reported the MRA's, another reported the SA's; both were blinded to the results of other scans. MRA correctly identified the number of renal arteries in 31 of the 42 patients (75%); accessory arteries could not be visualized. Using SA as the gold standard', MRA had a sensitivity of 90% and a specificity of 54%, with regard to detection of significant (>50%) ARAS lesions. MRA had a negative predictive value of 70%, and a positive predictive value of 82%. MRA was well tolerated by all patients, oral sedation being needed for just three. We conclude that 0.5 Tesla MRA has only limited usefulness as a non-invasive screening test for ARAS, demonstrating fair sensitivity, comparable to other screening methods, in the context of significant renal impairment. However, specificity was poor, due to seven false positives. Improving the signal to noise ratio by using signal-enhancing media, or using a more powerful magnet, are likely to yield more accurate information.