Quantitative analysis of the cineangiogram: Why bother?

Summary

Background: Numerous studies have demonstrated that subjective analysis of cineangiograms is highly variable. While this has fostered use of quantitative methods in research projects, clinical utilization is uncommon. Because many clinical decisions in cardiology are based on an evaluation of the cineangiogram, the variability inherent in subjective analyses must also mean that the clinical decisions are subject to inconsistency. Thus, we set out to determine whether quantitative parameters obtainable from the cineangiogram were of prognostic importance. This would provide a more clinically relevant impetus for greater utilization of quantitative methods in practice in addition to the already demonstrated value in providing more reproducible results.

Methods: Baseline cineangiograms of 283 patients with at least 10 years of clinical follow-up were randomly selected from the 3,566 patients in the Coronary Artery Surgery Study Registry and subjected to quantitation of ejection fraction, regional wall motion, regional shape analysis and quantitative coronary arteriography. These parameters were considered in isolation (“quantitative coronary arteriography” and “regional wall function” predictive models), in combination (“quantitative cineangiographic” predictive model) and in combination with subjective angiographic analysis and clinical information such as age and history of prior infarction (“clinical” predictive model). Prognostic indicators of death, infarction, unstable angina and other cardiac syndromes were determined by multiple logistic regression. Event free survival curves based on the various quantitative parameters were also constructed using log-rank testing.

Results: The most important quantitative parameters were the ejection fraction and the percent diameter narrowing of left anterior descending coronary territory stenoses. Regional wall motion provided additional prognostic power beyond that of the ejection fraction in the prediction of lethal myocardial infarction. Regional shape contributed additional power in the prediction of any cardiac event and the need for bypass surgery. Surprisingly, percent diameter stenosis measurements, predominantly of the left anterior descending were even more frequently of prognostic significance. When quantitative parameters were analyzed in the “clinical” model, the factors of overriding prognostic importance were the ejection fraction and the subjective determination of the number of vessels involved with “significant” stenoses. However, even under these circumstances, quantitative coronary arteriography retained independent prognostic value. This was not the case for any of the regional wall motion or shape parameters. There were no quantitative parameters that predicted the syndrome of unstable angina.

Conclusions: Quantification of the cineangiogram, particularly with respect to the ejection fraction and severity of coronary lesions, provides objective, verifiable and reproducible parameters that are also of prognostic importance. Wider spread use of such methodologies could have the impact of promoting greater consistency in the delivery of cardiac care. Quantitative methodologies provide a clinically important mechanism for establishing greater quality control and accountability of the ever increasing number of catheterization laboratories that have been fostered by the current era of interventional cardiology.