Cardiovascular MRI yields a high number of noncardiac findings, more than one for every two cases, with 5 percent of these findings leading to major changes in patient management, according to a study published in the April issue of the American Journal of Roentgenology.
Debate continues to swell on the significance of incidental findings, with a high level of scrutiny among radiologists and other physicians who point to the rareness of important ancillary findings and the extensive costs and stress associated with the large number of "incidentalomas."
Previous studies have examined the prevalence and significance of incidental findings on cardiac CT, among other tests. However, few data exist regarding the occurrence and clinical importance of ancillary findings on cardiovascular MRI, an exam that is being increasingly incorporated into clinical practice as a modality that offers superior structural and functional assessment of the heart, according to Faisal Khosa, MD, and colleagues from Beth Israel Deaconess Medical Center and Harvard University in Boston.
Khosa and co-authors examined the prevalence of incidental findings among 495 consecutive cardiovascular MR studies, classifying the ancillary results as benign, indeterminate or worrisome, the latter two typically resulting in follow-up. The authors also reported the relevance of different MR sequences as well as patient age to the clinical management of their findings.
Khosa and colleagues, excluding the radiologist reading the images, had information on follow-up for all exams included in the study.
Out of the 495 consecutive cardiovascular MRIs sampled during 2006, the reviewer identified 295 noncardiac findings present in 212 different studies. Half of these incidental findings were considered benign by the radiologist, while 45 percent were identified as indeterminate and 5 percent as worrisome.
Referring to medical record follow-up, 14 incidental findings were found to result in major changes to the management of 11 patients, an incidence of 3 percent of all cardiovascular MRIs. The authors also noted that approximately half of the 247 indeterminate and worrisome findings were not observed at the time of clinical reading (before the present retrospective study).
Significantly more noncardiac incidental findings were observed in older patients (56 years versus 48 years), though this difference was principally attributable to benign gynecomastia, which was the most common incidental finding among all patients. Other common results included hiatal hernia, pleural effusion and renal mass.
Lung nodules, which were present in 11 individuals, were the most commonly found worrisome finding.
The most frequent type of review for cardiovascular MRIs was a combined readout by a radiologist and a cardiologist, which occurred in 78 percent of the studies. This method was also associated with significantly more noncardiac findings: 42 percent in combined readouts versus 15 percent for serial readings.
The authors also found that most incidental findings were observed on single-shot fast steady-state free precession (SSFP) scout images and axial T1-weighted fast spin-echo images. Together these exams were responsible for 95 percent of all noncardiac findings and 99 percent of all potential management-changing findings.
The authors acknowledged that relying on one radiologist to review all 495 cases in this study might have limited their findings, given that radiologists commonly disagree when reading the same image.
“In summary, noncardiac findings on clinical cardiovascular MRI are common although only a small minority of studies contain management-changing noncardiac findings. The vast majority of management-changing noncardiac findings are visualized on thoracic SSFP scout and axial T1-weighted thoracic fast spin-echo images.”
Thus, the authors concluded, “Cardiovascular MRI practitioners and trainees should focus attention on these images for noncardiac findings.”