| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary |
1 Department of Imaging Science, University of Southern California, LAC/USC Imaging Science Center, 1744 Zonal Ave., Los Angeles, CA 90033.
Received April 28, 2008; accepted after revision May 1, 2008.
Address correspondence to P. M. Colletti.
Keywords: cardiac imaging • field of view • incidental findings • pulmonary nodules
Incidental findings are common in radiology practice, so it is not surprising that lesions are found incidentally during cardiac imaging examinations. This phenomenon is well illustrated in cardiac SPECT, in which, for example, Shih et al. [1] reviewed 566 99mTc-tetrofosmingated cardiac SPECT examinations looking specifically for abdominal abnormalities. Those investigators found 234 abnormalities, including mostly enterogastric and gastroesophageal reflux and abnormal bone marrow uptake in vertebrae. McKenna et al. [2] reported extracardiac findings in 107 (81%) of 132 elderly (mean age, 74 years) persons undergoing cardiac MRI; 63 (48%) of the patients had multiple findings. A total of 224 incidental findings were visualized, including at least one potentially significant lesion in 23 (17%) and one moderately significant finding in 43 (33%) of the subjects. Horton et al. [3] examined 1,326 screening electron-beam CT studies and found significant extracardiac pathologic findings in 7.8% of the studies. In a similar study of 1,812 electronbeam CT examinations, Hunold et al. [4] found that 34% of the patients had visible extracardiac abnormalities. Although those findings included a number of clinically insignificant abnormalities, 9.3% of 2,055 total extracoronary findings necessitated further diagnostic testing. Schragin et al. [5] found that 278 (20.5%) of 1,356 patients undergoing electron-beam CT had one or more noncardiac findings.
There is a consistent theme: The larger the field of view (FOV) and the more views obtained, the greater are the variety and number of innocent and significant lesions detected [6–12]. In this issue of AJR, Northam et al. [13] specifically address this topic. They found that during coronary CT angiography (CTA), if only limited-FOV images were viewed, more than 66.7% of pulmonary nodules larger than 1 cm in diameter and more than 80% of nodules smaller than 1 cm would be missed. Their findings raise the question: Should we look for, report, and follow up on incidental findings identifiable during cardiac imaging examinations? According to the findings of Northam et al. and others [7–12], it may seem obvious that we should. More health information is better. Patients and their physicians can decide how to proceed with the additional data identified.
For now, let us consider five counterarguments proposed by Budoff et al. [14] to the review of full-FOV data: "(a) > 50% of participants may have at least one noncalcified nodule; (b) the increased costs and radiation exposure associated with the resulting follow-up CT scans; (c) the cost and the morbidity of follow-up, including further testing, as well as biopsy or resection of benign noncalcified nodules (at least 25% of such procedures in several trials); (d) a small but difficult to quantify potential risk of cancer associated with multiple follow-up CT scans; and (e) a potential for increased anxiety of both the patient and the physician about nonsignificant pathology."
Budoff et al. [14] concluded: "We have reviewed all the relevant literature and sought to determine the potential benefits and harms of specifically overreading CTA for noncardiac pathology. The weight of the evidence suggests that it is most prudent to not specifically reconstruct and re-read CTA scans for lung nodules. If a noncardiac abnormality is visualized by the primary interpreter of the cardiac CT, appropriate referral or follow-up is prudent."Supporting these conclusions, radiation oncologists Smitt and Mehta [15] found that although incidental findings were identified on 20% of 132 radiation-planning CT scans, only three patients had important previously unknown findings, and only one patient with neck adenopathy had potential outcome benefit from discovery.
The findings by Northam et al. [13] balance the compelling arguments made by Budoff et al. [14]. Combined, the results of Onuma et al. [6] and Northam et al. show that without review of available full-FOV CT data, as many as 67% of noncalcified nodules would not be detected. The moral implications are clear; legal liability for cardiac CTA may be less so. Raw CT data are generally stored for a short time, typically days or weeks. If full-FOV data are never formatted and reviewed or stored before reuse by overwriting, evidence of the presence of detectable lesions may be absent. If full-FOV data are available, the likelihood of discovery remains low. One possible scenario may be that a patient has a nodule detected during a future CT examination and is then asked whether a previous examination was performed. The patient may recall undergoing CT scan of the chest, perhaps even in the same scanner. At that point, the consulting radiologist may review full-FOV data electronically and discover a previously unreported lesion, opening the Pandora's box described by both Budoff et al. [14] and Berlin [16]. What is it like to have a confident, smug radiologist as an adversarial witness? All physicians interpreting medical images should read the testimony of the plaintiff's expert radiologist in Berlin's "Defending the `Missed' Radiographic Diagnosis" [17].
|
A summary of the literature on incidental findings detected with cardiac imaging is presented in Table 1. Limited follow-up data are available. It is likely that depending on the cohort, one in 100 to fewer than one in 1,000 patients may benefit from serendipitous discovery of noncardiac lesions.
|
The issue of incidental findings in medical imaging will always be with us. Budoff et al. [14] reminded us of the cost of pursuing unrequested information. Northam et al. [13] found potential benefit. We may debate whether to perform an imaging examination. Once an examination is performed, the noblest approach is to view and evaluate all available data, to apply appropriate judgment, and to proceed in the best interest of the patient and society.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  W. B. Hall, S. G. Truitt, L. P. Scheunemann, S. A. Shah, M. P. Rivera, L. A. Parker, and S. S. Carson The Prevalence of Clinically Relevant Incidental Findings on Chest Computed Tomographic Angiograms Ordered to Diagnose Pulmonary Embolism Arch Intern Med, November 23, 2009; 169(21): 1961 - 1965. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Budoff Ethical Issues Related to Lung Nodules on Cardiac CT Am. J. Roentgenol., March 1, 2009; 192(3): W146 - W146. [Full Text] [PDF]
|
|
|
|
|
|
P. M. Colletti Reply Am. J. Roentgenol., March 1, 2009; 192(3): W147 - W148. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
