Quadruple inversion-recovery b-SSFP MRA of the abdomen: initial clinical validation

Abstract

The purpose of this study is to assess the image quality and diagnostic accuracy of non-contrast quadruple inversion-recovery balanced-SSFP MRA (QIR MRA) for detection of aortoiliac disease in a clinical population. QIR MRA was performed in 26 patients referred for routine clinical gadolinium-enhanced MRA (Gd-MRA) for known or suspected aortoiliac disease. Non-contrast images were independently evaluated for image quality and degree of stenosis by two radiologists, using consensus Gd-MRA as the reference standard. Hemodynamically significant stenosis (≥50%) was found in 10% (22/226) of all evaluable segments on Gd-MRA. The sensitivity and specificity for stenosis evaluation by QIR MRA for the two readers were 86%/86% and 95%/93% respectively. Negative predictive value and positive predictive value were 98%/98% and 63%/53% respectively. For stenosis evaluation of the aortoiliac region QIR MRA showed good agreement with the reference standard with high negative predictive value and a tendency to overestimate mild disease presumably due to the flow-dependence of the technique. QIR MRA could be a reasonable alternative to Gd-MRA for ruling out stenosis when contrast is contraindicated due to impaired kidney function or in patients who undergo abdominal MRA for screening purposes. Further work is necessary to improve performance and justify routine clinical use.

Figure 1

QIR MRA sequence diagram. The initial non-selective IR pulse inverts the entire volume of interest, while the subsequent slice-selective IR (first SS IR pulse) reinverts the spins that are located inside the aorta. In this way, during the inversion time, TI, aortic spins travel to the femoral arteries to permit large FOV coverage. The second SS IR pulse is used to null signal from venous blood inferior to the imaging slab. The final inversion pulse, a STIR pulse, improves vascular contrast by suppressing background signal from tissue with short T₁.

Figure 2

Non-contrast MRA with excellent image quality in agreement with contrast-enhanced MRA in a 82 y.o. male referred for suspected peripheral arterial disease. Hemodynamically significant disease of the aortoiliac region was not observed. Right renal artery aneurysm (arrows) was identified on both Gd-MRA and QIR MRA.

Figure 3

QIR MRA and Gd-MRA in a 63 y.o. female patient with severe disease. Incomplete filling of the distal iliac arteries was observed bilaterally due to near complete occlusion of the common iliac arteries (arrows).

Figure 4

QIR MRA with ECG-triggering (free breathing), QIR MRA with respiratory-triggering, and Gd-MRA in a 75 y.o. female referred for suspected PAD. Gd-MRA was rated with excellent image quality, while both non-contrast images had fair image quality ratings (score 2). Mild left external iliac stenosis (grade 1) was identified on Gd-MRA. The lesion was graded as mild stenosis (grade 1) on the ECG-triggered scan in agreement with the reference scan. However, both readers scored the stenosis as significant (grade 2) on the respiratory-triggered image, presumably due to the signal attenuation observed distal to the stenosis. ECG-triggering likely minimized the variation in the timing of peak systole relative to the readout, translating into better inflow during the TI, and thereby offered more accurate depiction of the stenotic lesion.

Figure 5

QIR MRA angiograms obtained in a patient (male, age 75) with endovascular AAA stent, renal transplant and low eGFR (29.6 mL/min/1.73m²). The standard coronal QIR MRA scan (a) exhibited poor image quality due to: (1) susceptibility artifacts caused by the stent and (2) insufficient inflow to the transplant region. QIR MRA was repeated (b) with axial FOV, positioned over the transplant region (white box). Patency of the transplant renal artery was successfully evaluated on the repeat scan.

Figure 6

QIR MRA patient with low GFR (25.7 mL/min/1.73m²).80 y.o. female referred for suspected AAA. Mild stenotic narrowing of the main renal arteries and atherosclerotic plaque of the suprarenal and infrarenal aorta were described by both readers.