| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Letters |
University of Wisconsin School of Medicine Madison, WI 53729
WEB—This is a Web exclusive article.
40 H. Although clear take-away points such as this are always welcomed and
highly attractive, we believe this topic is much more complicated.
Our first concern is the attenuation value itself. A recent article [2] cautions about the use of universal attenuation values because of the inherent variability of these numbers being dependent on the manufacturer of the CT scanner. Thus CT scanners from different vendors will most likely result in different attenuation values. Even when using the same CT scanner, the attenuation of fat is variable because of patient size, positioning, and imaging artifacts and will actually vary within the images from a single patient. We are also concerned about the small (1-cm diameter) region of interest (ROI) used by the authors. It would be interesting to know if a larger diameter, still within the hepatic parenchyma avoiding vessels and bile ducts, might be more accurate. Theoretically, a larger ROI value will give a more accurate result.
We think that the distinction between simple steatosis from nonalcoholic steatohepatitis deserves more weight in its clinical importance and consequently in its imaging. Nonalcoholic steatohepatitis has a far poorer prognosis compared with simple steatosis, often leading to cirrhosis, liver failure, and even hepatocellular carcinoma. Nonalcoholic steatohepatitis is distinguished by the additional presence of an inflammatory and fibrotic component. The different histologic pattern of this disease could change the attenuation of liver compared with steatosis alone. The authors also point out that livers in patients with nonalcoholic steatohepatitis can have increased iron content that will alter the hepatic attenuation on CT. It should be noted that up to 40% of patients with nonalcoholic steatohepatitis may have increased iron content in the liver [3] that would affect imaging characteristics. Finally, many of these patients have increased levels of glycogen, which is well known to increase the attenuation of liver, further confounding attempts to quantify fat on the basis of attenuation.
Finally, we would like to address the issue of cost that the authors raise. The reference to relative cost of CT versus MRI dates from 1991. Of course, much has changed since then, including the relative cost of these two techniques. At our institution, the costs of CT of the abdomen and abdominal MRI are quite similar, with a difference of less than 20%. In fact, we use targeted MRI examinations to decrease the overall imaging cost for certain clinical indications. Overall, we do not think that cost is a major advantage of CT, especially compared with other imaging techniques such as MRI.
Although certain attenuation values or a range of values on an unenhanced CT scan may raise the suspicion of steatosis, we are concerned with the emphasis placed on a single attenuation value. If clinical decisions made on the basis of the CT attenuation will affect patient management (e.g., surgery or transplantation), as the authors point out, as well as overall patient prognosis, we think that the consulting radiologist must understand the limitations inherent in this approach. Although current MRI techniques will provide better sensitivity and specificity for the presence of both fat and iron, ultimately the patient's particular clinical history must be considered, and consultation with a hepatologist and perhaps liver biopsy are ultimately needed for appropriate management.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
