| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 All authors: Department of Radiology, Medical University of South Carolina, 169 Ashley Ave., Rm. 297, PO Box 250322, Charleston, SC 29425.
Received July 25, 2007;
accepted after revision March 26, 2008.
Address correspondence to J. G. Ravenel
(ravenejg{at}musc.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Viewing of cardiac CT scans obtained only at a limited field of view can result in missing more than 67% of nodules larger than 1 cm and more than 80% of nodules smaller than 1 cm.
Keywords: cardiac CT • incidental findings • lung cancer • pulmonary nodule
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although it is true that extracardiac findings are made with relative frequency (10–35%) at cardiac CT [1, 5–11], the incidental finding of pulmonary neoplasm is made at rates as high as 1.2% [1]. Because reports of most studies include only full-FOV data, it is difficult to discern whether opening the FOV aids in detection of pulmonary lesions. Because most screening-detected neoplasms are peripheral, we hypothesized that many suspicious lesions and pulmonary neoplasms detected at cardiac CT would be missed with only limited-FOV reconstruction. In this study, we retrospectively reviewed all cardiac CT reports of pulmonary nodules over a 3-year period to determine the percentage that would be missed with viewing only images obtained with a limited FOV.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The technique used depended on the indication for the study. All studies were performed on a 64-MDCT scanner (Somatom Sensation 64 Cardiac or Somatom Definition, Siemens Medical Solutions). For calcium scoring, studies were performed with prospective gating from below the transverse aortic arch through the base of the heart with 1.2-mm detector collimation, 120 kV, 450 effective mAs, and 0.33-second gantry rotation time. For calcium scoring, images were reconstructed at 3-mm intervals for limited FOV images and 5-mm intervals for full-FOV evaluation of the thorax. Coronary CTA and pulmonary venous stenosis studies were performed with retrospective gating from below the transverse aortic arch through the base of heart with 0.6-mm detector collimation, 120 kV, 900 effective mAs (Sensation 64 scanner) or 380 effective mAs (Definition scanner) and 0.33-second gantry rotation time. Images were reconstructed at 0.75-mm intervals for cardiac interpretation of limited-FOV images and 3-mm intervals for full-FOV evaluation of the thorax. For CABG evaluation, the study protocol was similar to the CTA protocol, but the z-axis was extended cephalad to include the subclavian arteries. For CTA, pulmonary venous stenosis, and CABG evaluation, prescan delay was determined with a test bolus. Injection of IV contrast medium (iopamidol, Isovue 370, Bracco) at a rate of 5–6 mL/s with total volume calculated on the basis of scanning time was followed by a 30- to 50-mL saline flush.
In each case, limited-FOV images were reconstructed from the original data obtained. No additional scans were obtained. The full FOV extended from outer rib to outer rib and encompassed the entire lung parenchyma within the imaged portion of the thorax. By definition, our limited FOV encompassed the heart and 1 cm from the farthest anterior, posterior, and lateral extents of the cardiac chambers, typically approx imately 17–20 cm. All 202 cases in which a nodule was detected, excluding calcified nodules, were chosen for further review. If more than one nodule was present, results were considered on a per-patient basis. Therefore, one visible nodule was considered sufficient to trigger follow-up, and the patient was considered to have a nodule within the limited FOV. The following characteristics of each nodule larger than 1 cm in diameter were recorded: density (solid, mixed, or ground-glass), margin (smooth, ill-defined, spiculated), and visibility (partially visible, not visible) in the limited FOV. We also made an effort to determine the final pathologic type of each lesion larger than 1 cm if that information was present in the medi cal record.
We selected a subset (n = 55) of 20% of the studies showing a nodule smaller than 1 cm and all studies showing a nodule larger than 1 cm to determine the relative percentage volume of limited to full FOV. For both limited FOV and full FOV, the volume of a cylinder was calculated from the diameter and z-axis extent and expressed as a percentage.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Figures 1, 2A, 2B, 3A, 3B show examples of nodules in each category. Final pathologic results were available for six of the 15 patients with nodules larger than 1 cm, and all six results were positive for neoplasm (five primary lung neoplasms and one metastatic lesion). Only two pathologically proven neoplasms were visible in the limited FOV. Pathologic findings were unavailable for nine patients, who were undergoing follow-up at our institution or elsewhere. Compared with full FOV, the average percentage of volume covered by the limited FOV was 55% ± 5%.
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most of the published data on extracardiac findings are based solely on findings on cardiac CT scans obtained with an expanded FOV. With regard to calcium scoring, in two studies [5, 6] suspicious pulmonary nodules were found at 65 (4.9%) of 1,326 and 113 (8.3%) of 1,356, respectively, imaging examinations. Of 65 nodules in the study by Horton et al. [5], 12 were larger than 1 cm. The investigators attempted follow-up with the first 29 patients who had incidental findings. Those 29 patients were the first and only group to undergo follow-up in the study. Fifteen of the 29 patients had suspicious nodules, and one of the nodules was diagnosed as lung cancer. Of 113 nodules described in the report by Schragin et al. [6], 46 were considered suspicious, but the authors did not report the number of nodules diagnosed as lung cancer or the number of nodules larger than 1 cm.
Haller et al. [1] made extracardiac findings in 27.4% of patients who underwent coronary artery MDCT. Those incidental findings included two (1.2% of 166) pulmonary nodules larger than 1 cm, which were determined to be lung cancer. In addition to the extracardiac findings, Haller et al. commented on limited versus full FOV. They determined that at standard coronary artery MDCT, only 35.5% of the total chest volume (lung apices through lung bases) was displayed, whereas 70.3% of the total chest volume was displayed with the full FOV. Our findings were similar to those of Haller et al. in that the limited FOV was approximately 50–55% of the full FOV for cardiac CT. Haller et al. did not, however, comment on whether the two malignant lesions were seen in the limited FOV.
Onuma et al. [8] reported that at MDCT, 12 (2.4%) of 503 patients had the incidental finding of pulmonary nodules larger than 1 cm and that 49 (9.7%) of 503 patients had the incidental finding of noncalcified pulmonary nodules smaller than 1 cm. Eight of 12 nodules larger than 1 cm were thought to necessitate follow-up. Four (0.8%) of the 503 patients had asymptomatic malignant tumors detected during 6 months of follow-up: two patients had adenocarcinoma of the lung; one, a malignant breast tumor; and one, angiosarcoma. Onuma et al. also found that only one of four malignant tumors would have been detected in a limited FOV.
Like Onuma et al. [8], we found that most pulmonary nodules larger than 1 cm would have been missed had full FOV not been used. In addition, we found that more than 80% of nodules smaller than 1 cm would have been missed if only limited FOV had been used. Because FOV can be adjusted with relative ease and without the consequence of increased radiation exposure, it seems reasonable to require full FOV for cardiac CT interpretation.
Our results showed that 1% of patients had the incidental finding of neoplasm or suspicious pulmonary nodule larger than 1 cm necessitating close follow-up. It can be argued that our numbers do not seem significant, but in a comparison of our numbers with those reported in lung cancer screening trials, similarities would be found. At minimum, six (0.3%) of 1,764 patients undergoing cardiac CT had pulmonary neoplasms. This number can be compared with the results of a lung cancer screening trial [13] in which primary lung neoplasms occurred with a frequency of 0.5–3.2%. The difference presumably exists because our population included patients without risk factors for lung cancer.
It is an ongoing debate in cardiology whether radiologists should read cardiac CT scans in their entirety for noncardiac findings. Budoff et al. [4] argued that the risks of follow-up may outweigh the benefit of detection of important findings. This difference in approach is apparent in a comparison of the American College of Radiology guidelines [14] for performance of cardiac CT with the guidelines of the American College of Cardiology. The American College of Radiology requires that interpreters of cardiac CT scans also meet guidelines for interpreting diagnostic CT scans or have experience with at least 100 thoracic CT or CTA examinations and assess for and document important extracardiac findings in a diagnostic report. The American College of Cardiology [3] makes no implicit or explicit competence requirement for evaluation of extracardiac structures.
Combining our findings with those of Onuma et al. [8] shows that fewer than one third of pulmonary neoplasms would be diagnosed with use of only a limited FOV. The benefit to patients in whom neoplasms are found should not be so easily dismissed. These malignant tumors can be likened to other nodules expeditiously evaluated and diagnosed after being incidentally detected on CT performed for other reasons. Patients with these tumors need not be caught up in the debate over the population-based validity of lung cancer screening. Furthermore, the ultimate results for an individual nodule do not diminish the importance of detection. Guidelines are already in place for management of indeterminate nodules [12].
We must acknowledge several limitations. First, we relied on written reports to find candidate nodules. It is possible that we excluded nodules that existed but were not detected. Second, because we did not have long-term follow-up findings on most of the nodules, we cannot state with certainty the absolute percentage of lung neoplasms that would be excluded with use of a limited FOV. Finally, we focused on one aspect of extracardiac findings, pulmonary nodules, and therefore did not encompass the full spectrum of findings that can lie outside a limited FOV.
Most pulmonary nodules (80%) and pulmonary neoplasms (67%) can go undetected at cardiac CT without the acquisition and interpretation of images at full FOV.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  P. M. Colletti Reply Am. J. Roentgenol., March 1, 2009; 192(3): W147 - W148. [Full Text] [PDF]
|
|
|
|
|
|
P. M. Colletti Incidental Findings on Cardiac Imaging Am. J. Roentgenol., September 1, 2008; 191(3): 882 - 884. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
