To compare semiquantiative MR perfusion parameters with results from intracoronary pressure derived fractional flow reserve (FFR) in the functional evaluation of known coronary artery lesions for assessment of their hemodynamic relevance

21 patients with CAD underwent MRI (Siemens Sonata, Erlangen) conventional coronary angiography (CA) with introcoronary pressure wire examination for determination of fractional flow reserve (FFR) in 23 stenotic lesions. Signal intensity curves of the first pass MR perfusion imaging (SR-turboFLASH) of the myocardium were analysed at rest and under adenosine induced hyperemia. A 0.5 mmol/kg Gd-DTPA bolus was applied at rest and during hyperemia with a flow rate of 5 ml/sec. Adenosine (140µg/kg/min) was infused over at least 3 minutes for the stress examination. Time-to-peak (TTP), maximum signal intensity (MSI) and upslope (US) values were determined using a 16 segment model. A input curve correction was performed using a linear fit. A coronary artery with lesions < 50% diameter reduction was stated as normal. A coronary lesion > 50% and FFR (PressureWire, Radi, SE) > 0.75 was called intermediate. If diameter reduction >50% and FFR < 0.75 was present, the lesion was defined as severe.

336 perfusion areas were evaluated. TTP was not significantly different between the three groups (11±3.4; 15±5.7; 13±5.1; ns). US at rest was comparable (0,08±0,17; 0.10±0,035; 0.07±0.01; n.s.). During stress US increased and was significantly different between normal coronary arteries and severe coronary stenoses (0.28±0.12 vs. 0.08±0.01; p<0.05). The ratio US at stress and rest was 3.4±1.5 (2.0-6.3) for normal coronary arteries and 1.7±0.3 (1.2-1.7) and 1.1±0.1(1.0-1.3) for intermediate and severe coronary lesions (p<0.04 and p <0.03).

TTP and US at rest were not able to discriminate between normal and severly stenosed coronary arteries. US under stress and ratio between US (stress) and US (rest) were significantly different between normal and functionally significant diseased coronary arteries. The use of these MRI parameters may therefore improve the sensitivity and specificity of noninvasive identification of hemodynamically relevant CAD.