Consensus remains elusive for best left ventricle test

January 1, 2007

Researchers may agree that cardiac MR is the modality of choice for predicting left ventricular remodeling, but they remain split on which contrast-enhanced CMR technique produces the most accurate prediction.

Researchers may agree that cardiac MR is the modality of choice for predicting left ventricular remodeling, but they remain split on which contrast-enhanced CMR technique produces the most accurate prediction.

Two studies presented in scientific sessions at the 2006 RSNA conference supported using measurements of myocardial infarction size with delayed-enhancement imaging. Two other papers concluded that first-pass measurement of microvascular obstruction of the myocardium or a no-flow phenomenon caused by microvascular obstruction is a better test.

LV remodeling following acute MI compromises the architectural integrity of the heart from the formation of fibrosis, dilatation, hypertrophy, and scarring. These weaken the pumping power of the heart and increase the likelihood of repeat MI or heart failure.

Arguing for infarct size as the better measure, Dr. Alexander Stork of the University Clinic Hamburg-Eppendorf in Germany found in a study of 55 patients with a first reperfused MI that the degree of LV remodeling is directly related to infarct size measured with baseline imaging. Contrast MRI was performed five days and then eight months after acute MI. Infarct size was calculated as a percentage of left ventricular area using a threshold method that included only enhanced myocardium with a signal intensity that was at least two standard deviations greater than normal myocardium.

Remodeling appeared in the follow-up scans of 13 patients and was defined as a 20% or greater increase in end-diastolic volume compared with baseline imaging. Regression analysis found that infarct size was more powerful than microvascular obstruction for predicting LV remodeling, Stork said. Baseline delayed-enhancement imaging was 92% sensitive, 92% specific, and 93% accurate in its prediction of LV remodeling when the infarct involved more than one-quarter of the LV myocardium.

"Delayed enhancement might serve as a tool for risk stratification for remodeling," Stork said.

Dr. Guido Ligabue of the University of Modena in Italy determined that transmural extension index, a ratio of infarct size to total myocardial mass, can predict the probability of LV remodeling (see figure). His prospective study involved 61 patients who were imaged within two weeks of an MI and angioplasty. Follow-up imaging was performed six months later.

Dr. Stijnje D. Roes of the University of Leiden in the Netherlands found merit in both imaging approaches. She reported from her study of 27 patients at high-risk for LV modeling that microvascular obstructions detected with first-pass perfusion MRI more reliably predicted the onset of LV remodeling than did microvascular obstruction measured with delayed enhancement.

Dr. Luigi Natale of the Catholic University in Rome discovered that a no-reflow finding can be used to predict LV remodeling when considered with measurements of edema and infarction transmurality. The no-reflow phenomenon results from microvascular obstructions such as swollen endothelium, endothelial protrusions, and edema following infarction. These obstructions compress myocardial capillary vessels, thereby limiting blood flow even after recanalization.

Natale examined 32 post-infarction patients with first-pass MRI to measure no-reflow. Triple-inversion recovery fast spin-echo imaging established the extent and distribution of edema, and delayed enhancement MRI delineated the size and location of infarction. A score for edema, no-flow, and hyperenhancement was calculated in each segment of a 17-segment LV model based on the presence of each and transmural extension.

From these calculations, Natale detected no-flow in 26 of 32 MI patients following revascularization. Remodeling was evident from follow-up imaging of 14 patients. Significant differences between patients who ultimately do and do not develop remodeling were observed in the scores for edema, microvascular obstruction, and infarctions.

Given the lack of consensus, moderator Dr. Albert De Roos, a professor of radiology at the University of Leiden, recommended considering evidence from both tests to get an accurate prediction of remodeling.

"The bottom line is that MRI is an excellent tool to use, and it is probably a good idea to use both techniques," he said.