| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Commentary |
1 Both authors: Department of Radiology, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030.
Received May 7, 2005; accepted after revision May 25, 2005.
Each month the American Journal of Roentgenology will republish
online one of the 100 most-cited articles from its first century. A
corresponding commentary in the print journal by a contemporary radiologist
will provide a current perspective. For a full list of these articles see page
3 of the January 2006 issue of
AJR or
www.ajronline.org.
Keywords: cancer • CT • lung
Although CT has revolutionized the radiologic diagnosis of lung diseases, its importance was not initially appreciated. This was because the natural contrast within the lung facilitated the imaging of lung disease by conventional radiography and linear tomography, obviating the pursuit of improved imaging methods. Thus, although CT was embraced by those imaging the brain and the abdominal organs, its usefulness for imaging the chest was not intuitively obvious. In addition, despite anecdotal reports of CT's ability to detect small pulmonary nodules [1], it was claimed that, overall, the accuracy of CT of the chest did not surpass that of conventional radiography [2].
Schaner et al. [3] were the first to systematically compare in a prospective trial the pulmonary nodule detection rates of chest radiography, linear tomography, and CT with the detection rate of thoracotomy in patients undergoing pulmonary metastectomy. The authors were also the first to show the much greater sensitivity of CT for lung nodule detection. As a result of this observation, their study marked a turning point in the workup of candidates for pulmonary metastectomy. The goal of this surgical procedure is complete resection of the lesions with maximal preservation of the normal lung parenchyma because, in the absence of extrathoracic metastases, the complete resection of the pulmonary metastases is associated with increased survival regardless of the histologic characteristics of the metastases [4]. Thus, the greater sensitivity of CT had direct clinical implications for patients with pulmonary metastases. In fact, the use of CT for the detection of nodules would later be incorporated into an evidence-based medicine approach to pulmonary metastases [5].
Perhaps of even greater importance than this finding from the work of Schaner et al. [3] are two questions that we are still struggling to answer today in this era of widespread screening by CT. With the increased sensitivity of CT, the first question is "What is the gold standard for pulmonary nodule detection?" In the study by Schaner et al., CT was able to detect nodules missed at thoracotomy that were proven to exist on follow-up CT scans. Better detection of these nodules by CT should become even more common as the technology continues to improve. In fact, a study has already shown that CT can detect nodules not detected on autopsy [6].
The second question this study raised is that of specificity—whether CT has an increased ability to detect pulmonary metastases versus benign nodules. Despite their methodical research on the increased sensitivity of CT in lung nodule detection and their excitement over this imaging technique, Schaner et al. [3] refrained from recommending CT as a routine method of screening for early pulmonary metastases. More recent studies of CT as a screening tool for the detection of lung cancer have shown that, depending on the geographic location in which the CT scan is performed, CT can detect indeterminate pulmonary nodules in as much as 73% of the patients studied [7-9]. Although only a small fraction (3.6%) of the patients in a screening trial will require surgical removal of highly suspicious pulmonary nodules, such a procedure is associated with operative complications in 27% of cases and a 1.7% mortality rate when the operations are performed in a tertiary center [9].
As we struggle today with the clinical implications of new imaging techniques and the creation of guidelines in preventive medicine, this landmark article serves as an important guide in two respects. First, prospective evaluation of new imaging techniques is essential for establishing their sensitivity and specificity before replacing well-established imaging techniques. Second, we should remain cautious before recommending the widespread use of new techniques to the general population, in particular to a healthy population, and should strive to show improved patient outcomes in randomized trials before the widespread implementation of new imaging techniques.
We as radiologists should actively push for such randomized trials to test each new technology advancement in our field. In addition, we should do so in a timely fashion because there is a brief window of opportunity to perform such trials after the introduction of a new imaging technique when ethics committees will allow 50% of the participating patients to be denied the supposedly better test [10].
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Related articles in AJR:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
