| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Department of Radiology, Kanazawa University School of Medicine, 13-1 Takara-machi, Kanazawa 920-8641, Japan, e-mail: matsuio@med.kanazawa-u.ac.jp
Editor:
In the March 2000 issue of Radiology (1), Dr Lim and colleagues reported the results of their analysis of intranodular hemodynamics of dysplastic nodules in cirrhotic livers by using computed tomography (CT) during arterial portography (CTAP) and CT hepatic arteriography (CTHA). Because their results were substantially different from those of our study, which were published in the April 1999 issue of the American Journal of Roentgenology (2) and because our results were not cited and discussed in their report, I would like to make some comments on their article.
The histologic diagnosis of this kind of dysplastic nodule associated with liver cirrhosis is highly specialized, even for pathologists who specialize in liver assessment (3). The definite differential diagnosis between atypical adenomatous hyperplasia (or high-grade dysplastic nodule) and a well-differentiated hepatocellular carcinoma is still not easy, especially with a biopsy specimen. I do not understand how this article could be reviewed in the absence of representative histologic preparations (no photomicrograph was shown in their article). Because of the difficulty of definite histologic diagnosis due to variations of the interpretation and experience among pathologists and of the histologic heterogeneity in the nodule (especially when cases confirmed at biopsy are included), I believe that this kind of study must rely on statistical evaluation of a large number of cases, as our study did (2).
Another concern I have is that the authors retrospectively applied incidental findings in the specimen obtained after resection to the preoperative CTAP and CTHA images. In practice, it is almost impossible to precisely correlate in this retrospective fashion the actual location of an original lesion that is less than 2 cm in diameter and that is found incidentally in the resected liver. For example, in figure 4 of their article (1), the authors pointed out one dysplastic nodule that was hypoattenuating at CTAP and hyperattenuating at CTHA. However, for me, this finding represents a typical arterioportal shunt because the lesion is wedge-shaped, and a vessel is present in its center. Therefore, I believe that this lesion does not correspond to a surgically confirmed dysplastic nodule, which was probably isoattenuating at CTAP and CTHA.
As we and many other Japanese investigators have revealed, almost all dysplastic nodules (adenomatous hyperplasia) are iso- or hypovascular at CTHA relative to the surrounding regenerative nodules. The frequency of hypervascular dysplastic nodules is unusually high in the article by Lim et al (1). When a hypervascular hepatocellular nodule shows no cellular or structural atypia at histologic examination, other types of hyperplastic nodules, such as focal nodular hyperplasia–like lesions without relation to hepatocarcinogenesis, should also be considered (4).
REFERENCES
Department of Radiology, Sungkyunkwan University School of Medicine, Samsung Medical Center, 50 Ilwon-dong, Kangnam-ku, Seoul 135-710, South Korea, e-mail: jhlim@smc.samsung.co.kr
We thank Dr Matsui for his valuable comments and constructive critiques. We are sorry for not citing their article (1) in the discussion. It was not cited because the article was not published when we were preparing our manuscript. The article was well written and instructive, and the findings were based on a fairly large number of cases; however, the results were different from ours (2).
Among Japanese physicians, it is well known that dysplastic nodules (adenomatous hyperplasia) have a normal or slightly decreased portal supply and a normal or slightly decreased hepatic arterial supply; therefore, these nodules will be iso- or hypoattenuating at CTAP and CTHA relative to the surrounding regenerative nodules. However, in our experience, there have been a number of dysplastic nodules that appeared different from the described principles. Some low- and high-grade dysplastic nodules were hyperattenuating at CTHA. We discussed with pathologists the histopathologic diagnosis of those nodules, but the results were the same. This is why we conducted a study and published its findings in Radiology.
We agree with Dr Matsui that histologic diagnosis of a dysplastic nodule is not easy, and differential diagnosis between a high-grade dysplastic nodule and a well-differentiated hepatocellular carcinoma is difficult, even by liver pathologists. The pathologist at our institution specializes in liver pathology. He spent time in Japan and exchanged ideas with some Japanese pathologists; he is of the same opinion as they are, as far as the pathologic diagnosis of nodular lesions occurring with liver cirrhosis is concerned.
Regarding the argument on not presenting the histologic evidence for those nodules at CTAP and CTHA in our article, we could have presented histologic findings of all our cases, including those with 24 resected specimens and eight specimens that were examined at biopsy. We wonder how many readers of Radiology can interpret the histologic findings of dysplastic nodules or hepatocellular carcinoma. Biopsy material might represent only a part of a histologically heterogeneous nodule; therefore, diagnosis based on the biopsy material might be misleading, as we described.
The majority of the nodules in our study were proved with histologic examination of the resected specimen. The article by Hayashi et al (1) also included many cases proved at biopsy, and all of the histologic findings presented in their article were obtained at biopsy.
In our study, we analyzed the radiologic and pathologic findings independently and correlated them retrospectively. We correlated any abnormal findings at CTAP and CTHA and any nodular lesions on surgically resected or explanted livers on a one-to-one basis. We partly agree with Dr Matsui in that it is difficult to precisely correlate the actual location of small nodular lesions in the resected liver with the location of nodular lesions visualized at CT, especially in cases in which there are multiple lesions. However, we tried to correlate them precisely.
The study of the blood supply to the hepatocellular nodule should be investigated by using this retrospective method, with correlation of radiologic and pathologic findings, because many dysplastic nodules and even some hepatocellular carcinomas are isoattenuating at CT. A study of only visible nodules at CT will be misleading.
The nodule that was hypoattenuating at CTAP and hyperattenuating at CTHA in figure 4 of our article exactly coincided with the dysplastic nodule proved at pathologic examination. One may consider that the nodule has a wedge shape at CT; however, it is located in the middle of parenchyma, not in the periphery of the liver.
We do not know why the results of our study on blood supply to dysplastic nodules are different from the results of Hayashi et al. Both studies were retrospective correlations of radiologic findings with pathologic findings, the methods of CTAP and CTHA were nearly the same, and the pathologic critera were the same. The only difference was the patient population: In the study by Hayashi et al, the majority of patients (91 of 135 patients) had liver cirrhosis associated with hepatitis C, whereas the majority of patients in our study had hepatitis B–related cirrhosis. The difference in the cause of liver cirrhosis may explain the difference in the hemodynamic findings at CTAP and CTHA.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
