| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology and Research Institute of Radiological Science,
Yongdong Severance Hospital, Yonsei University College of Medicine, 162,
Eonju-ro, Gangnam-gu, Seoul 135-720, Korea.
2 Department of Radiology and Research Institute of Radiological Science,
Severance Hospital, Yonsei University College of Medicine, 250, Seongsanno,
Seodaemun-gu, Seoul 120-752, Korea.
Received November 11, 2007;
accepted after revision January 19, 2008.
Address correspondence to J. J. Chung
(jjchung{at}yuhs.ac).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. One hundred one cirrhotic patients (male:female = 69:32) with hypoattenuating nodules observed on initial portal or equilibrium phase CT images were retrospectively evaluated by follow-up CT performed 6–66 months after the initial CT examination. Depending on the background nodularity, patients were separated into macronodular (n = 33, 288 nodules) and micronodular (n = 68, 346 nodules) cirrhotic groups. Each nodule was categorized as category I (enlarged) or category II (stable). Nodule categories were correlated with the initial lesion size and the pattern of background cirrhosis.
RESULTS. The frequency of category I nodules was higher in patients with micronodular cirrhosis (40%) than in those with macronodular cirrhosis (27%) (p = 0.001). Category I nodules were significantly larger than category II nodules in patients with micronodular cirrhosis (p < 0.001). The doubling times of category I nodules had no statistical difference between patients with micronodular or macronodular cirrhosis (p = 0.954). Of the category I nodules in patients with micronodular cirrhosis, 8.6% showed malignant changes.
CONCLUSION. More careful attention should be paid to large nodules in patients with micronodular cirrhosis because of the potentially greater risk of malignancy, and small hypoattenuating nodules should be more often followed up in shorter intervals than large nodules.
Keywords: cirrhosis • dysplastic nodules • hepatocarcinogenesis • liver disease • liver neoplasm • MDCT • multiphase CT
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Both multistep and de novo hepatocarcinogenesis have been well described, from benign regenerative nodules to overt HCC [6, 7]. Within the spectrum of hepatocellular nodules, detection of dysplastic nodules is important because dysplastic nodules can be premalignant lesions that represent an intermediate step in the hepatocarcinogenesis pathway of the cirrhotic liver [8–10]. At the same time, it is not clear whether all dysplastic nodules inevitably become HCCs [8, 10]. The characteristic CT enhancement patterns of HCC, including hyperattenuation in the arterial phase and hypo- or isoattenuation in the delayed or equilibrium phase of multiphasic CT, have been well defined [11].
Nevertheless, in our daily practice we frequently encounter small hypoattenuating or nonenhancing nodules in the arterial phase with hypoattenuation in the portal or delayed phase of multiphasic CT. In general, we diagnose these hypoattenuating nodules as regenerative or dysplastic nodules and recommend close monitoring for the development of HCC. However, as many abdominal radiologists have experienced, some hypoattenuating nodules on contrast-enhanced portal or equilibrium phase images that lack arterial phase enhancement may show malignant characteristics in a few cases at long-term serial imaging and clinical follow-up.
The purpose of this study was to investigate the outcome and clinical implications of nonhypervascular hypoattenuating nodules depicted on either portal or equilibrium phase multiphasic CT images of the cirrhotic liver as well as the incidence of malignant transformation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The hospital information system and medical records of 101 cirrhotic
patients were retrospectively reviewed. These patients included 69 men and 32
women who were between 38 and 77 years old (mean, 56 years). Depending on the
background cirrhotic nodularity, patients were divided into two groups:
macronodular (n = 33), those in whom most cirrhotic nodules were 
3 mm, and micronodular (n = 68), those in whom most cirrhotic nodules
were < 3 mm [12]. The
nature of each nodule, as determined by follow-up imaging, was categorized as
enlarged (category I) or stable (category II), which included malignant and
potent ially malignant nodules or shrunken and unchanged benign nodules,
respectively. When a new non hypervascular hypoattenuating nodule was detected
on serial follow-up CT scans, that nodule was diagnosed as an initially
detected nodule and was closely investigated on follow-up CT images. The
nodule categories were correlated with initial lesion size and pattern of
background cirrhosis.
Liver cirrhosis was verified by gun-needle biopsy (n = 14); surgery for coexistent HCCs (n = 5); or clinical course, including serial laboratory test findings and imaging features (n = 82). The causes of liver cirrhosis were hepatitis B (n = 82, 81%), hepatitis C (n = 11, 11%), cryptogenic (n = 5, 5%), and alcohol (n = 3, 3%).
Multiphasic CT
Multiphase contrast-enhanced CT scans were obtained with either a 4-MDCT
scanner (HiSpeed Advantage, GE Healthcare) or a 16-MDCT scanner (Somatom
Sensation 16, Siemens Medical Solutions). In this study, portal phase images
were compared in dual-phase images (n = 44, 44%) and equilibrium
phase images were compared in triple-phase images (n = 57, 56%).
However, because this study was retrospective, both dual- and triple-phase CT
images were included together in the series of follow-up CT images of some
patients.
A helical CT scanner was used to perform dual-phase CT examinations after the IV administration of 150 mL of nonionic contrast material (iopromide [Ultravist 300, Bayer HealthCare]) by a power injector at a rate of 3–4 mL/s. After injection of contrast material, arterial phase CT images were obtained after a delay of 30 seconds and portal phase CT images were obtained after a delay of 65 seconds. The section thickness was 5 mm, and the incremental table speed was 5–7 mm/s.
For 16-MDCT images, triple-phase CT examinations were performed through the liver with a 5-mm collimation, a 5-mm/s table speed (pitch 1.0) during a single-breath-hold helical acquisition of 25–30 seconds (depending on the size of the liver), and a 3- to 5-mm reconstruction interval. After 150 mL of nonionic iodinated contrast material (Ultravist 300) was administered IV by an automatic injector at a rate of 3–4 mL/s, arterial and portal phase images were obtained after a delay of 30 and 60–65 seconds, respectively. Three-minute delayed equilibrium phase imaging was performed for triple-phase imaging. All patients underwent unenhanced CT for either dual- or triple-phase multiphasic CT.
Two or more separate sessions of follow-up CT (2–19 sessions; mean, 6.6 sessions) were performed with 6–66 months (mean, 20 months 3 weeks) between the first and last studies. The change of CT technologies from helical CT to 16-MDCT, with a progression of thinner slice thickness over time, was used because of the long period needed to elucidate the natural outcomes. If the initial CT images had been obtained by dual-phase CT technique and the follow-up CT images were obtained by triple-phase CT technique, we compared the portal phase images. If both initial and follow-up CT images were obtained by triple-phase CT technique, we compared the delayed phase images.
All scans were sent to a PACS (Centricity, GE Health care) for interpretation on a PACS workstation.
Image Analysis
All nonhypervascular hypoattenuating nodules observed on the initial
multiphasic CT examination were marked by electronic arrows and saved on PACS
workstations for subsequent data analyses. After synchronization of the
anatomic level of the first and last portal or equilibrium phase hepatic
images on the PACS monitors, the longest dimension of the hypoattenuating
nodules on the axial plane was measured with real-time comparison on magnified
views using electronic calipers and was recorded by two abdominal radio
logists with 14 and 6 years of experience, respectively, in reviewing
abdominal CT images. The locations of the hepatic nodules were recorded
according to Couinaud's segment classification.
To optimize objectivity, we measured the size of the same lesions on the first and last follow-up images with real-time comparisons on the PACS monitors and determined the arterial hypervascularity or hypoattenuation on either portal or equilibrium phase images by quantitative measurement of attenuation in Hounsfield units. Because size measurements and the determination of nonhypervascular hypodensity of the lesions were performed by consensus, interobserver variability could not be evaluated.
From the initial lesion size on magnified views of contrast-enhanced portal
or equilibrium phase images, an increase of 2 mm or more in the longest
dimension of the corresponding lesion on follow-up images was stipulated as
lesion growth (category I). Lesions that increased less than 2 mm, did not
change, or decreased in size were regarded as stable, shrunken, or unchanged
(cate gory II), respectively. Lesions were also divided by size as large (
10 mm) or small (< 10 mm) in both micronodular and macronodular patients to
compare the incidence of category I or II nodules with background nodularity
and lesion size.
A malignant change of the hypoattenuating nodules was diagnosed when the nodules showed contrast enhancement on arterial phase images and washout of contrast material on either portal or equilibrium phase images. The frequency of malignant changes (overt HCC) in category I nodules was also evaluated in both macronodular and patients with micronodular cirrhosis. The assessment of the enhancement pattern of hypoattenuating nodules was performed by both visual and quantitative methods for both portal and equilibrium phase images.
The volumetric doubling time of enlarging nodules was calculated as time (t) divided by 10 (log d – log d0), where the nodular diameter (d) increased from d0 to d in t days.
To reduce intraobserver variation in detection and size measurements of hypoattenuating nod ules, all the case images were reviewed twice at an interval of 2 weeks by the same radiologists. Depending on the measured values, changes in both lesion size and hypoattenuation of the lesions were determined by consensus of the same radiologists. For all of the measured values, the mean value of two separate measurements was used for data description and subsequent statistical analysis.
Statistical Analysis
The frequency of lesions with malignant potential was analyzed according to
the background nodular size (macronodular or micronodular) by the Pearson's
chi-square test. The Student's t test was used to compare the mean
size of potentially malignant (category I) and benign (cat egory II)
hypoattenuating nodules. For enlarged lesions, the mean doubling time of
potentially malignant and benign lesions was also compared using a Student's
t test. For all tests, a p value of < 0.05 was considered
statistically sign ificant. Fisher's exact test was used to analyze the
frequency of category I nodules according to background nodular size in
individual patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Of the 634 hypoattenuating nodules, 218 were category I nodules (34%) and 416, category II nodules (66%). In patients with macronodular cirrhosis (n = 288), there were 79 category I nodules (27%) and 209 category II nodules (73%). In patients with micronodular cirrhosis (n = 346), there were 139 category I nodules (40%) and 207 category II nodules (60%). The frequency of category I nodules was higher in those with micronodular cirrhosis (Fig. 1A, 1B) (139/346, 40%) than in those with macronodular cirrhosis (79/288, 27%) (p = 0.001).
|
|
-fetoprotein—was detected. The hypovascular category I nodules
were not expected to respond to transcatheter arterial chemoembolization.
Among the 101 patients, 90 patients (89%) had one or more enlarged nodules and
11 patients (11%) had no enlarged nodules. Seventy-five patients (74%) had up
to eight nodules (n = 354, 56%) and 26 patients (26%) had more than
eight (n = 280, 44%).
The mean size of category I nodules was 11.8 mm (range, 5.6–50.2 mm) and that of category II nodules was 10.2 mm (range, 5.1–24.4 mm) (p < 0.001). The mean size of category I and II nodules was comparable in patients with macronodular cirrhosis (11.5 and 11.1 mm, respectively; p = 0.286), but category I nodules were significantly larger (mean, 11.9 ± 4.9 mm) than category II nodules (mean, 9.4 ± 2.8 mm) in patients with micronodular cirrhosis (p < 0.001).
Lesions were also divided by size as 10 mm or < 10 mm. In patients
with micronodular cirrhosis, 30% (58/192) of the lesions < 10 mm and 53%
(81/154) of lesions 10 mm were classified as category I nodules
(p < 0.001). However, for patients with macronodular cirrhosis,
27% (31/113) of lesions < 10 mm and 27% (48/175) of lesions 10 mm were
category I nodules (p = 1.000) (Fig.
2A,
2B,
2C). Therefore, the malignant
potential of nonhypervascular hypoattenuating nodules on either portal or
equilibrium phase images is higher for large nodules ( 10 mm) found in
patients with micronodular cirrhosis than in those with macronodular
cirrhosis.
|
|
|
10 mm (n = 129) were 428 ±
282 days and 607 ± 445 days, respectively (Student's t test,
p = 0.001). Of the 139 category I nodules in micronodular patients, 12 nodules (8.6%) underwent malignant changes associated with overt HCC with contrast enhancement on arterial phase images and washout of contrast material on portal or equilibrium phase images (Figs. 3A, 3B, 3C, 3D and 4A, 4B, 4C, 4D); three of these nodules displayed a nodule-in-nodule pattern (Fig. 5A, 5B, 5C, 5D). The mean volume-doubling time of these 12 overt HCCs was 375 days (range, 103–535 days), which was statistically significant (p < 0.05) compared with that of other category I nodules in patients with micronodular cirrhosis. Of the 79 category I nodules in macronodular patients, however, only three nodules (3.8%) underwent malignant changes associated with overt HCC on the follow-up images.
|
|
|
|
|
|
|
|
|
|
|
|
In terms of individual patients, at least one category I nodule was found in 91% (30/33) of the patients with macronodular cirrhosis and 88% (60/68) of those with micronodular cirrhosis. There were no enlarged nodules in 9% (3/33) of the patients with macronodular cirrhosis and 12% (8/68) of the patients with micronodular cirrhosis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These hypoattenuating nodules may develop through the stages of multistep hepatocarcinogenesis from low-grade dysplastic nodules, to high-grade dysplastic nodules, to early HCC or nodule-in-nodule HCC, and finally to overt HCC [7, 16]. However, many hypoattenuating nodules on either portal or equilibrium phase images without arterial phase enhancement have shown no malignant transformation during relatively long-term follow-up imaging studies [10, 17, 18].
In the current study, 8.6% (12/139) of category I nodules in patients with
micronodular cirrhosis revealed malignant changes into overt HCC, three of
which displayed a nodule-in-nodule pattern. Thus, hypoattenuating nodules
10 mm in patients with micronodular cirrhosis should be closely observed using
either portal or equilibrium phase CT scans for potential progression toward
malignancy. The mean volume-doubling time of these 12 overt HCCs discussed
here was 375 days, which was statistically significant (p < 0.05),
compared with that of other category I nodules in the patients with
micronodular cirrhosis. These observations are comparable with those described
by Yu et al. [14] in a
previous report.
Takayasu et al. [17] asserted that once a hypoattenuating lesion develops to the hyperin-hypoattenuating (nodule-in-nodule) type, the speed of progression to overt HCC might accelerate. Similar findings have been reported in studies conducted with MRI [16] and with a combination of CT during hepatic arteriography and CT during arterial portography [19]. Takayasu et al. [17] also suggested that more attention should be paid to attenuation changes than to changes in lesion size, particularly in patients with positive results for hepatitis C viral antibody and a relatively large lesion on initial CT. In our study, however, most patients had hepatitis B virus (n = 82, 81%) instead of hepatitis C virus (n = 11, 11%). Lim et al. [20] also suggested that initial tumor size might influence the conversion of small low-attenuating hepatocellular nodules into hyperattenuating lesions.
When suspicious malignant change in a dysplastic nodule was newly detected on follow-up images, most physicians and patients usually wanted the quick treatment, especially in cases of an inoperable condition without invasive histologic study for confirmation. Therefore, transcatheter arterial chemoembolization (TACE) or radiofrequency ablation was used as a noninvasive treatment.
There is no current consensus about the management of hypoattenuating nodular lesions in hepatitis virus–related chronic liver disease [21], and the critical time for treatment of the growing hypoattenuating nodules using surgical resection [22], percutaneous ethanol injection [22, 23], or radiofrequency ablation [24] has not been established. Most surgically resected large or growing hypoattenuating nodules are histologically diagnosed as low- or high-grade dysplastic nodules or as early HCC [13, 25, 26]. However, histologic diagnosis by percutaneous needle biopsies of hypoattenuating nodules is not easy because of sampling error for small lesions and because of intratumoral variation in histologic components [13]. Nevertheless, because of their potential malignancy, large or growing hypoattenuating nodules have been treated at times without confirmation of malignancy by needle biopsy, obscuring the definition of the true features of hypoattenuating nodules.
The incidence of dysplastic nodules, which are currently considered premalignant lesions, is 15–25% at autopsy or in explanted livers [26]. Dysplasia can be diagnosed when the hepatic nodules show distinct low attenuation during all three phases of helical CT or when the nodule is seen only as low attenuation on equilibrium phase images [5, 10], even though regenerative nodules and well-differentiated HCCs can have a similar appearance. Most dysplastic nodules are isoattenuating or slightly low attenuating compared with the surrounding cirrhotic liver on arterial, portal, and equilibrium phase images, making them difficult to differentiate using multiphasic CT [27, 28]. Lim et al. [10] reported that helical dynamic triple-phase CT depicted only 10% of pathologically proven dysplastic nodules and, therefore, is not sufficiently sensitive for the detection of dysplastic nodules in cirrhotic livers.
According to a previous study, HCC develops at a rate of approximately 3–10% per year in patients with chronic viral hepatitis and cirrhosis [29, 30]. The progression rate of HCC from high-grade dysplastic nodules is significantly high, and the annual HCC development rate exceeds 30% in the first 2 years [5]. Sakamoto and Hirohashi [22] reported that 72% and 6% of 18 low-grade dysplastic nodules in their study developed to early and overt HCC, respectively, and that 33% of early HCCs progressed to overt HCC. In our study, some category II nodules grew after the imaging follow-up period. Such delayed growth is probably only a minor problem that would not change the prognosis of patients with advanced cirrhosis, because there is always a high risk of new HCC development in other hepatic locations.
Our inclusion criteria included cases with a time interval between the first and last CT images of 6 months or longer. Six months could be insufficient for dysplastic nodules to progress to carcinoma in situ or to overt HCC. Taouli et al. [31], however, reported that the tumor volume-doubling time of HCC was 17.5–541.4 days (mean, 127 days). For cases of small HCC, it has taken approximately 5 months to grow from 1 to 3 cm in diameter. Thus, physicians recommend a regular 6-month follow-up by sonography in patients with cirrhotic liver, and we included the cases with the time interval of 6 months or longer.
The progression of dysplastic nodules to HCC can take from several months to a few years based on follow-up CT scans [32]. Takayama et al. [32] reported that 50% of low-grade dysplastic nodules were transformed into overt HCC in a mean of 21 months. Kaji et al. [33] reported that 36% of high-grade dysplastic nodules monitored clinically for 12–24 months transformed into overt HCC, but no cases of low-grade dysplastic nodules transformed into HCC during the same period. Clinically, the detection of a hypervascular spot in a hypovascular nodule predicts the prognosis of the nodule because this kind of nodule always progresses to entirely hypervascular classic HCC and a hypervascular spot may indicate an abnormal arterial blood supply resulting from tumor angiogenesis [17].
Marchiano et al. [34] reported that an increased concentration of iodine improved liver-to-lesion contrast and might improve detection of HCC in cirrhotic patients. Pozzi Mucelli et al. [35] also showed that using a contrast material with a high iodine concentration progressively increased the sensitivity and positive predictive value when passing from early arterial phase to late arterial phase to double arterial phase images; however, the difference in sensitivity between late arterial phase and double arterial phase was not statistically significant. In our study, we used a contrast material with a relatively low iodine concentration (300 mg I/mL) and a relatively low injection rate (3–4 mL/s), probably resulting in the possibility of decreased detectability of focal liver lesions on dual- or triple-phase multiphasic CT scans. However, we used a relatively large amount of contrast material, at 150 mL, to compensate for the low iodine concentration.
In the current study, there was no statistical difference in the
volume-doubling times of category I nodules between patients with macronodular
cirrhosis and those with micronodular cirrhosis. In contrast, the
volume-doubling time of nodules < 10 mm in diameter was shorter than that
of nodules 10 mm (p = 0.001), indicating that smaller
hypoattenuating nodules might grow more rapidly regardless of nodule size in
the background cirrhotic liver (Fig.
3A,
3B,
3C,
3D). Therefore, small
hypoattenuating nodules should be carefully assessed within a relatively short
interval. In fact, Yu et al.
[14] reported that the mean
volume-doubling time for each enlarged hyperintense nodule on unenhanced
T1-weighted gradient-echo MR images was 410.7 days, which is comparable with
that of small nodules in our study.
In our study, category I nodules were significantly larger (11.9 ± 4.9 mm) than category II nodules (9.4 ± 2.8 mm) in patients with micronodular cirrhosis (p < 0.001). Therefore, depending on the size of the hypoattenuating nodule, the frequency of potentially malignant category I nodules increased prominently in patients with micronodular cirrhosis, whereas this was not the case in those with macronodular cirrhosis.
Nonhypervascular hypoattenuating nodules should be differentiated from thrombosed hemangiomas, small-volume cysts, focal fibrosis, benign regenerative nodules, and infarcted regenerative nodules on contrast-enhanced multiphasic CT images. Hypovascular well-differentiated HCC was also included together in the differential diagnosis [15].
There were some limitations to this study. First, because of the limitations of this type of clinical study, no lesions were histopathologically proven by needle biopsy when they were first diagnosed as hypoattenuating. Second, CT technologies changed over time, with progression to a thinner slice thickness because of the long period needed to elucidate the natural outcomes. Third, possible errors in the measurement of lesion size by electronic calipers could not be avoided. Fourth, a contrast material with a higher iodine concentration, such as 370 or 400 mg I/mL, and an injection rate of 5 mL/s or higher were not used in this study. If a higher iodine concentration and a higher rate were used, the detection rate of hypervascular lesions would probably increase. Fifth, only 19% of patients (19/101) had histologically proven liver cirrhosis as assayed by gun-needle biopsy (n = 14) or surgery (n = 5). Finally, this study was retrospective, and therefore both dual- and triple-phase CT images could be simultaneously included in the series of follow-up CT images of some patients; a prospective study with a larger number of patients is needed.
We conclude that the malignant potential of nonhypervascular hypoattenuating nodules observed on either portal or equilibrium phases of a multiphasic CT is higher for large nodules found in patients with micronodular cirrhosis; therefore, careful attention should be paid to these nodules on follow-up CT images. In addition, small hypoattenuating nodules should be more often followed up in a shorter interval than large nodules because the volume-doubling time of small nodules < 10 mm in diameter was significantly shorter than that of large nodules.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
