MDCT equals MRI for RV function assessment

Dec 27, 2005

Right ventricular (RV) parameters and function are known to have major prognostic import in patients with heart disease and chronic pulmonary embolism, but RV assessment can be challenging. In particular, the position and complex geometry of the right ventricle have made volumes difficult to assess with CT, and echocardiography for that matter, compared to the MRI gold standard.

In a study presented at the 2005 RSNA meeting in Chicago, however, Dr. Patrick Hein and colleagues from Berlin's Charité Hospital found that 16-slice multidetector-row computed tomography (MDCT) was statistically equivalent to cine MRI for estimating RV volumes, ejection fraction, and other important parameters.

"So far MRA (MR angiography) is considered to be the method of choice for obtaining right ventricular parameters, but it is sometimes contraindicated and sometimes difficult to perform in the perioperative setting," Hein said in his presentation. "So the option to use the high spatial resolution that is provided by multidetector CT scanners ... would be desirable."

The study sought to validate measurements obtained from MDCT datasets in comparison with cine MRI as a reference standard.

The study population consisted of 21 patients, Hein said, "12 patients with regular right ventricular function and nine patients with impaired right ventricular function," involving either stenotic or valvular pathology.

Following injection of 100 mL of iodinated contrast at 4-5 mL/sec, all patients underwent contrast-enhanced MDCT on a 16-detector scanner at 0.5-mm collimation and a maximum gantry speed of 400 msec, resulting in an effective acquisition time of 50-200 msec per cardiac cycle using multisegmental registration, Hein said.

Within a day or two of CT, the patients also underwent steady-state free precession cine MRA on a 1.5-tesla scanner equipped with a special body phased-array coil. RV volumes, ejection fraction, and myocardial mass were calculated according to the slice summation method, according to the abstract.

CT and MRI measurements were compared using Pearson's correlation coefficient (r), Student's t-test for paired samples, and Bland-Altman analysis. The group assessed interobserver and intraobserver variability by calculating the coefficient of variability.

"The right ventricle was completely visualized and with good image quality on all MDCT and cine MR images"; there were no statistically significant differences between any CT and MRI parameters: end-diastolic volume (EDV): r = 0.94, end-systolic volume (ESV): r = 0.96, ventricular stroke volume (SV): r = 0.91, ejection fraction (EF): r = 0.95, mass: r = 0.88; p < 0.05 each, according to Hein.

Likewise, mean values of all RV measurements did not differ significantly between the modalities (EDV: 126.3 ± 39.0 versus 123.2 ± 41.0 mL; ESV: 74.3 ± 34.1 versus 69.0 ± 32.1 mL; SV: 51.9 ± 15.6 versus 54.1 ± 18.2 mL; EF: 41.3 ± 14.2 versus 43.9 ± 16.5%; mass: 58.3 ± 18.7 versus 63.2 ± 21.4 g; p > 0.05 each).

Limits of agreement were within an acceptable range (EDV: -28.6 to +29.7 mL; ESV: -17.8 to +19.5 mL; SV: -19.8 to +16.2 mL; EF: -9.6 to +7.8%; mass: -17.4 to +13.5 g). Finally, no significant differences were seen between the modalities in the variability coefficient for intra- and interobserver agreement.

"Despite its lower temporal resolution, MDCT seems to be an adequate noninvasive means for evaluating RV areas compared to MRA," Hein said. "Right ventricular examinations were independent of examination expertise and experience," as well as individual patient characteristics, he said. Compared to state-of-the-art MRI cine imaging, "MDCT is a viable option for the evaluation of right ventricular parameters ... independent of the (clinical) setting," Hein concluded.

A recent French study examined 46 patients with pulmonary embolism (PE) who underwent pulmonary angiography, CT, and echocardiography in the emergency department. The results showed that CT right/left ventricular and end-diastolic area ratio correlated with PE obstruction and echocardiography. CT RV/LV area ratios greater than 1 had sensitivity and specificity of 88% (for both) in the diagnosis of significant PE. The authors concluded that CT may be used as a triage tool to select high-risk PE patients using the calculation of the RV/LV area ratio to detect RV dysfunction (American Journal of Cardiology , May 15, 2005, Vol. 95:10, pp. 1260-1263).

By Eric Barnes
AuntMinnie.com staff writer
December 28, 2005

RV function scans integrated with chest CT, May 9, 2005

CT imaging alone may be suitable for workup of pulmonary embolism, November 22, 2004

Thoracic CT expert takes on chronic PE diagnosis, October 15, 2004

Study documents rising use of CT for suspected PE, October 11, 2004