| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ospedale Belcolle I-01100 Vitebo, Italy
In the November 2004 issue of the AJR, Giorgio et al. [1] compare second-generation contrast-enhanced sonography with contrast-enhanced helical CT in the study of 74 biopsy-confirmed hepatocellular carcinomas (HCCs) in cirrhotic livers, including 28 small nodules that were less than or equal to 20 mm in diameter. They report "sensitivity [rates] in characterizing HCC" of 97.9% (sonography) and 91.9% (CT) and conclude that the two methods are of similar value in "detecting HCC hypervascularity." Some points in this report seem unclear or misleading.
First of all, the sensitivity figures reported in Table 4 simply reflect the percentage of tumors that could be visualized with the two methods, i.e., not only those displaying hypervascularity, which is the criterion for imaging diagnosis of HCC [2], but also those that appeared hypovascular or avascular. The actual diagnostic sensitivities of the two methods (i.e., the rates of true positivity defined in terms of concordance with the gold standard biopsy diagnosis of HCC) are much less impressive, particularly for small nodules: 77% for sonography (91.3% for large nodules, 53.6% for those < 20 mm) and 63.5% for CT (76.1% for large nodules, 42.9% for small ones).
Second, CT and sonography yielded discordant results for 17.6% of the HCCs. This difference was reported as nonsignificant, but there is an evident trend here. In fact, 10 nodules that were hypervascular on sonography were classified as hypovascular (6 cases) or missed altogether (4 cases) on CT, but there were no cases in which hypervascularity noted on CT was missed on the sonographic study. This suggests that sonography might be slightly more sensitive than CT for detecting HCC hypervascularity, but it is impossible to evaluate this hypothesis in the absence of an appropriate gold standard evaluation of nodule vascularization (e.g., angiography).
The fact is that, while arterial hypervascularity displays high specificity for imaging diagnosis of HCC [2], its sensitivity is less satisfactory. During hepatocarcinogenesis, the intranodal arterial blood supply initially decreases [3] as the normal hepatic arterial vasculature undergoes degeneration [4]. However, neoplastic angiogenesis soon gives rise to an abnormal arterial vasculature that rapidly replaces the normal hepatic arterial supply [3] and, in advanced stages, completely obliterates the intranodular portal veins [4]. The presence of hypervascularity reliably distinguishes HCCs from a variety of similar focal liver lesions (e.g., regenerating and dysplastic nodules, non-Hodgkin's lymphomas, hepatocellular adenomas, or liver metastases), but it may not be demonstrable (or indeed present) in early HCCs (i.e., those characterized by hepatic artery degeneration and preserved portal veins). Given the risk of false-negative findings of this type (i.e., hypovascularity or avascularity), biopsy rather than imaging is recommended for diagnosing nodules of less than 20 mm [2]. The relatively high rate in the study by Giorgio et al. of small HCCs that were hypovascular or avascular (sonography, 35.7%; CT, 46.4%) might thus be related to their degree of differentiation. This correlation was not investigated, but it might shed useful light on the significance of the findings that emerged from this study.
 Top References References |
|---|
D. Cotugno Hospital Naples 80131, Italy
We want to thank Caturelli et al. for their interest in our work. They raise an important point regarding the sensitivity of contrast-enhanced sonography in detecting and characterizing hepatocellular carcinoma (HCC) smaller than 20 mm, which depends on the vascular supply to the nodule. We agree that although arterial hypervascularity displays high specificity for imaging diagnosis of HCC, its sensitivity is less satisfactory for small HCCs. This is particularly true for HCCs that were avascular or missed in our series [1]. In such cases, biopsy is mandatory because these nodules do not have a specific pattern on both imaging techniques (sonography and CT) and also because the majority of avascular or missed HCC nodules are smaller than 20 mm.
The slightly better results of sonography compared with CT in the assessment of tumor vascularity are probably because sonography offers the advantage of continuous real-time imaging in the detection of a fugacious hypervascularity that could instead be missed in a technique such as CT, which depends on interval-delay imaging acquisition. We agree with Caturelli et al. that to confirm this hypothesis an appropriate gold standard evaluation of nodule vascularization such as angiography is needed. Nevertheless, the aim of our study was to investigate the sonographic appearance of HCC using a nondestructive contrast agent by comparing the technique to contrast-enhanced helical CT, which is the most commonly used method for the characterization and staging of HCC.
Early HCCs are usually well differentiated, and as they increase in size, the number of portal tracts apparently decreases and intratumoral arterioles develop [2]. In a recent prospective study [3] undertaken to investigate the characterization of small HCCs in cirrhotic patients by means of sonography and comparing the results to dynamic gadolinium-enhanced MRI, we ascertained that the sensitivity of sonography in detecting nodules smaller than or equal to 10 mm is poor; eight out of 11 (72.7%) HCC nodules were missed. Moreover, in lesions smaller than 10 mm, sonography failed in differentiating benign hyperechoic lesions such as angioma from hyperechoic angioma-like HCC; both were missed on sonography.
The findings of small hyperechoic lesions on cirrhotic liver indicate HCC in almost 50% of cases [4]. These small echoic lesions are indistinguishable from the typical hepatic hemangioma, and they are a clinical challenge, especially when an angioma-like nodule is found at the initial sonographic examination during the surveillance. In our series, five out of eight (62.5%) hyperechoic nodules smaller than or equal to 10 mm detected on conventional sonography were HCCs at histological diagnosis. In these cases, fine-needle biopsy can be safely performed and should not be delayed until the nodules become larger than 10 mm, as the guidelines of Barcelona-2000 EASL Conference suggest [5].
|
TopReferencesReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
