| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, University of Pittsburgh Medical Center, 200 Lothrop
St., Pittsburgh, PA 15213.
2
Department of Transplantation Medicine, University of Pittsburgh Medical
Center, Pittsburgh, PA 15213.
Received January 17, 2000;
accepted after revision April 13, 2000.
Address correspondence to S. Katyal.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Helical CT findings for 57 patients with hepatocellular carcinoma were classified into one of three patterns of vascularity on the basis of the degree of tumor or liver enhancement during the hepatic arterial phase. Cases in which hypervascular lesions predominated were classified as a type 1 pattern. Cases in which hypovascular lesions predominated were classified as a type 2 pattern. Patients were classified as responders or nonresponders on the basis of the changes of tumor size revealed on CT after three transcatheter arterial chemoembolization treatments.
RESULTS. We classified the 57 patients as 37 responders (65%) and 20 nonresponders (35%). A statistically significant correlation between the type 1 hypervascular pattern and response to transcatheter arterial chemoembolization was seen; conversely, the type 2 hypovascular pattern correlated with nonresponse to transcatheter arterial chemoembolization (chi-square = 7.85, p = 0.02). Patients classified as responders lived significantly longer than those classified as nonresponders with 12-, 24-, and 36-month survival rates of 90%, 67%, and 36%, respectively, for responders and 70%, 17%, and 10%, respectively, for nonresponders.
CONCLUSION. We found that patients who responded to transcatheter arterial chemoembolization had prolonged survival (p < 0.01). Response correlated closely with tumor vascularity as shown on hepatic arterial phase helical CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Commonly used forms of treatment for patients with unresectable hepatocellular carcinoma include percutaneous ethanol ablation, intraarterial chemotherapy, or intraarterial chemoembolization [4,5,6,7]. The liver has a dual blood supply with the portal vein supplying 80% of blood to the normal liver. By contrast, the hepatic artery provides at least 80% of the blood supply to hepatocellular carcinoma [8]. This permits selectively directed chemotherapy to the tumor via the hepatic artery, with relative sparing of underlying liver. Hepatocellular carcinoma is a hypervascular tumor and the direct intraarterial infusion of chemotherapeutic agents with or without hepatic artery occlusion has resulted in improved tumor response rates compared with systemic IV chemotherapy [9]. The impact on survival is still the subject of debate. Many institutions perform transcatheter arterial chemoembolization with a variety of agents for the chemotherapy and several vascular occluding agents such as gelatin particles (Gelfoam; Pharmacia-Upjohn, Peapack, NJ), polyvinyl alcohol, or iodinized oil.
Transcatheter arterial chemoembolization therapy is invasive and is associated with high financial costs and patient morbidity. The ability to predict which patients are likely to benefit in terms of tumor response and survival from transcatheter arterial chemoembolization would be useful to both the oncologist and the patient. A variety of studies have attempted to predict, with both clinical and laboratory parameters, the prognosis of patients with unresectable hepatocellular carcinoma who are being treated with transcatheter arterial chemoembolization [10,11,12]. Hepatocellular carcinomas with increased vascularity, as determined on catheter angiography, have also been shown to be more responsive to treatment with intraarterial chemoembolization. Patients in whom tumor shows increased vascularity have been claimed by some to have improved survival rates compared with those with relatively decreased tumor vascularity [13]. Although catheter angiography can be used to evaluate hepatocellular carcinoma blood supply, a noninvasive method to assess tumor vascularity is preferable. With the advent of helical CT, it is now possible to image the entire liver during the hepatic arterial dominant phase of enhancement. Images obtained during this phase allow noninvasive assessment of tumor vascularity based on the degree of lesion-to-liver enhancement [3].
Therefore, this retrospective analysis was undertaken to determine whether the enhancement pattern (vascularity) of hepatocellular carcinoma during hepatic arterial phase helical CT could be used to predict which patients with unresectable hepatocellular carcinoma would show a response to transcatheter arterial chemoembolization therapy. In addition, the prognostic value of being classified as a "responder" to transcatheter arterial chemoembolization in terms of patient survival was also investigated.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 43 men and 14 women with biopsy-proven hepatocellular carcinoma ranged in age from 17 to 82 years (mean age, 56 years). The cause of the associated liver disease included cryptogenic cirrhosis (n = 18), hepatitis B (n = 15), hepatitis C (n = 11), alcohol (n = 5), hemochromatosis (n = 3), hepatitis B and C (n = 2), hepatitis C and alcohol (n = 1), primary biliary cirrhosis (n = 1), and autoimmune hepatitis (n = 1). The patients were classified according to the Child's classification [14]. Forty-four patients had Child's A cirrhosis, 12 patients had Child's B cirrhosis, and one patient had Child's C cirrhosis.
Records of all patients with hepatocellular carcinoma were reviewed by the liver tumor board. All patients were judged to have unresectable carcinoma either because of tumor location and stage or because of poor hepatocellular reserve, but all patients were thought to be able to tolerate transcatheter arterial chemoembolization and were offered this option.
All CT examinations were performed using a commercially available CT scanner (HiSpeed Advantage; General Electric Medical Systems, Milwaukee, WI). Although study this is retrospective, because our institution is a large referral center for liver disease and hepatocellular carcinoma, a standard prospective protocol was used for imaging patients with cirrhosis and hepatocellular carcinoma. Unenhanced axial CT of the liver was performed using a 5-mm collimation and a 3-mm interscan gap. Biphasic helical CT protocols (7-mm collimation and a pitch of 1.0-1.7) were performed in all patients. The remainder of the abdomen and pelvis was then scanned with contrast-enhanced axial CT (5- to 10-mm-thick sections and a 0- to 5-mm intersection gap).
Each patient received 150 mL of either iothalamate meglumine (Conray 60; Mallinckrodt Medical, St. Louis, MO) or ioversol (Optiray 320; Mallinkrodt Medical) IV via a mechanical power injector (model OP 100; Medrad, Pittsburgh, PA). The IV contrast material was injected at rates of from 2.5 to 5.0 mL/sec. For rates of 2.5-3.5 mL/sec, 28- and 70-sec scan delays were used for arterial and portal venous phase imaging, respectively. For rates of 4.0-5.0 mL/sec, scan delays of 20 and 60 sec were used for arterial and portal venous phase imaging, respectively. The slower injection rates were used for patients imaged during our initial experience with biphasic helical CT.
Chemoembolization
After informed consent was obtained from the patient, selective superior
mesenteric and common hepatic digital arteriography was performed via femoral
catheterization with a 5-French angiography catheter. Nineteen patients were
treated with Spherex (Pharmacia-Upjohn) and 38 patients were treated with
Gelfoam embolization. All patients were treated by a single oncologist and
received cisplatin in doses ranging from 125 to 150 mg/m2.
Treatments were repeated every 4-8 weeks until progression or disease
stability, as determined using CT tumor measurements. No therapy was given to
any of the patients in addition to the chemoembolization protocol. In all
patients, the main vessel feeding the tumor or tumors was embolized by
selective (right or left hepatic artery) or subselective catheterization. This
procedure involved the slowing of the blood flow—but never the complete
occlusion of the arterial flow—to preserve hepatic function. In two
patients with extensive bilobar tumor, the proper hepatic artery was
embolized. Gelfoam embolization was performed initially before chemotherapy
infusion to decrease the washout of the tumor's blood supply. After
chemotherapy was administered, embolization with gelatin sponge particles was
again performed in an attempt to cause stagnation of the arterial supply to
the tumor.
CT Analysis
Hepatocellular carcinoma has a variety of appearances on CT. The typical
appearance is a lesion that is less dense than the adjacent liver on
unenhanced CT and that shows hypervascular enhancement during the hepatic
arterial phase (enhancing to a greater degree than the adjacent liver
parenchyma). However, hepatocellular carcinoma lesions differ in their degree
of hypervascularity and may be homogeneously hypervascular (>90% of the
lesion enhances more than the adjacent liver parenchyma) or heterogeneously
hypervascular (>50% of the lesion enhances more than the adjacent liver
parenchyma). Hepatocellular carcinoma lesions with less than 50%
hypervascularity and those that enhance less than or equal to the adjacent
liver parenchyma during the hepatic arterial phase are characterized as
hypovascular. Each individual hepatocellular carcinoma lesion can vary in the
degree of vascularity (enhancement) during the hepatic arterial phase.
Additionally, a patient may have multiple hepatocellular carcinomas, with each
appearing differently on imaging. Therefore, a more useful approach to
evaluating tumor vascularity in a patient is to classify the predominant
pattern of vascularity for the entire tumor burden within the treated hepatic
lobe.
Biphasic CT scans obtained before initial transcatheter arterial chemoembolization treatment were retrospectively reviewed by two experienced gastrointestinal radiologists. The predominant enhancement pattern during the hepatic arterial phase of all lesions in the treated lobe were classified as one of the following types: type 1, predominantly hypervascular (Fig. 1); type 2, predominantly isovascular to hypovascular (Fig. 2); and type 3, mixed vascularity with an equal number of hypervascular and hypovascular enhancing lesions (Fig. 3). Patients were included in the type 1 group if a predominant pattern of hypervascular lesions (either homogeneously or heterogeneously hypervascular) was present during the hepatic arterial phase. Patients with a predominant pattern of hypovascular lesions were included in the type 2 group.
|
|
|
The type 1 (predominantly hypervascular pattern) group was further subclassified into patients with all hypervascular lesions (type 1A) and those with a predominant number of hypervascular lesions (type 1B). In addition to classifying the predominant enhancement pattern of the hepatocellular carcinoma lesions in the treated lobe, we recorded the size of each lesion (maximal anteroposterior and transverse diameter) and the intrahepatic tumor stage according to the TNM classification.
Follow-Up
Transcatheter arterial chemoembolization procedures were performed at 4- to
8-week intervals, with biphasic CT scans obtained the same day of each
treatment. Tumor measurements of the index lesions (identified on the CT scans
obtained before transcatheter arterial chemoembolization) were recorded from
the CT examination just before the fourth transcatheter arterial
chemoembolization treatment, and change in lesion size was expressed as a
percentage of the initial size of the tumor. An assessment was made to
document whether the overall pattern of the lesions followed that of the index
lesions. If the overall lesion pattern matched that of the index lesions at
follow-up, then response was determined on the basis of the index lesions. If
the pattern differed from that of the index lesions (most of the tumor burden
increased with stable or decreased index lesions), the patient was classified
as a nonresponder. Survival, in months, from the time of diagnosis was
recorded for each patient.
Treatment Response
The standard oncologic definition of a partial response, a 50% or greater
decrease in the product of two perpendicular measurements of tumor size, was
used in this study. Patients with increases in tumor size over the treatment
interval were classified as nonresponders with progressive disease
("nonre-sponders-progressors"). Patients with a decrease in the
product of two perpendicular measurements of tumor size that was less than 50%
but had no increase in tumor size were classified as nonresponders with stable
disease ("nonresponders-stable").
Statistical Analysis
Chi-square and Aspin-Welch (for unequal variances) t tests were
used for statistical analysis of discrete and continuous two-sample results,
respectively. Survival analysis was performed with the Kaplan-Meier product
limit method; the Gehans-Wilcoxon statistic was used for comparative analysis
[15]. A p value of
less than 0.05 was considered to be significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Thirty-seven patients were classified as responders (65%) and 20 as nonresponders (35%) after three treatments of transcatheter arterial chemoembolization. Figure 4A,4B shows an example of a treatment response. Figure 5A,5B shows an example of a treatment failure (nonresponder-progressor). Table 1 shows the response rate by CT vascularity pattern. Chi-square analysis showed that there was a statistically significant correlation between the type 1 hypervascular pattern and transcatheter arterial chemoembolization responses; conversely, the type 2 hypovascular pattern correlated with nonresponse to transcatheter arterial chemoembolization. Although the response rate for the patients with mixed (type 3) pattern of vascularity (4 responders and 1 nonresponder) was similar to the response rate for type 1 patients, no statistical analysis was performed on the patients with type 3 vascularity because of the small number of patients in this category.
|
|
|
|
In addition to the determination that hypervascular arterial phase CT enhancement was associated with a statistically significant increased likelihood of a positive response to transcatheter arterial chemoembolization therapy by tumor shrinkage, the survival benefit to the patients classified as responders was determined and compared with the survival rates of nonresponders. The Kaplan-Meier product limit survival plot showed a statistically significant improvement in patient survival for the responder group compared with that of the nonresponder group (Fig. 6). The 12-, 24-, and 36-month survival rates were 90%, 67%, and 36%, respectively, for the responders and 70%, 17%, and 10%, respectively, for the nonresponders (p<0.01).
|
To further investigate the survival benefit for patients with the type 1 hypervascular pattern, we calculated and compared the survival rates of the patients with only hypervascular lesions (type 1A) with the survival rates of the remaining patients with a predominant number of hypervascular lesions (type 1B). Thirty patients had lesions with purely hypervascular type 1A enhancement and four patients had a predominant number of lesions with hypervascular type 1B enhancement. The survival rates at 12, 24, and 30 months are 93%, 58%, and 38%, respectively, for the purely type 1A hypervascular group compared with 50%, 25%, and 25%, respectively, for the type 1B predominantly hypervascular group (p < 0.03). The improved survival benefit for the type 1A group occurred during the first 18 months of intraarterial chemoembolization treatment (Fig. 7). After this time, the slopes of the two survival curves became almost identical.
|
The responder and nonresponder categories were then analyzed separately to determine whether the predominant vascular pattern seen during the hepatic arterial phase of helical CT had any impact on survival within the specific context of a response or nonresponse. The nonresponder group (n = 20) was subclassified into patients with type 1 hypervascular lesions and patients with type 2 hypovascular lesions. Eight nonresponders had type 1 hypervascular lesions and 11 nonresponders had type 2 hypovascular lesions. One nonresponder had mixed type 3 enhancing lesions and was excluded from statistical analysis. The 12- and 24-month survival rates were 86% and 15%, respectively, for the hypervascular nonresponder group and 47% and 11%, respectively, for the hypovascular nonresponder group (p = 0.05). The greatest survival benefit occurred during the first 13 months of transcatheter arterial chemoembolization treatment. After this time, the slopes of the two curves became almost identical (Fig. 8).
|
Given the improved survival for the nonresponder group compared with historical controls (mean survival, 1.6 months) [1], this group was further subclassified into patients with progressive disease and those with stable disease. Eleven nonresponders had stable disease and nine nonresponders had progressive disease over the treatment interval. Survival for the patients with stable disease was increased, with 12- and 24-month survival rates of 90% and 45%, respectively, compared with survival rates of 50% and less than 5%, respectively, for patients with progression of disease (p = 0.05) (Fig. 9).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study, responders lived significantly longer than nonresponders, with 2-year survival rates of 67% and 17%, respectively (Fig. 6). Even the nonresponder group with progressive disease, however, had improved survival rates when compared with historical controls of patients with untreated hepatocellular carcinoma [1]. The majority (93%) of the patients in our study had advanced stage III or IV tumor (TNM classification). This finding suggests that although responders to transcatheter arterial chemoembolization live significantly longer than nonresponders, transcatheter arterial chemoembolization treatment may offer a survival benefit to all patients with unresectable hepatocellular carcinoma.
The ability to predict which patients will respond to transcatheter arterial chemoembolization therapy should not be used to screen patients from receiving transcatheter arterial chemoembolization because all patients with unresectable hepatocellular carcinoma who can tolerate treatment should receive transcatheter arterial chemoembolization. The value in distinguishing responders from nonresponders on the basis of arterial phase CT enhancement findings before treatment is in the prognostic information provided to both the oncologist and the patient. For the oncologist, optimization of treatment dose and frequency, based on degree of response, has been shown to sustain treatment responses and to improve survival (Carr BI et al., presented at the Proceedings of American Society of Clinical Oncology, May 1988). Ernst et al. [16] found that the efficacy of transcatheter arterial chemoembolization increases if used only when necessary on the basis of CT findings obtained after treatment. For the patient, the subclassification of responder versus nonresponder allows more accurate estimation of survival rate.
To explain the survival benefit given to our type 1 group of patients (those with a predominant pattern of hypervascular arterial phase enhancing lesions) treated with transcatheter arterial chemoembolization, we hypothesized that the significant advantage of transcatheter arterial chemoembolization is in treating hypervascular lesions. We made the initial assumption that the degree of tumor hypervascularity shown by histology or angiography would correlate with the degree of hypervascular enhancement on arterial phase CT. For the type 1 group we separated the patients with only hypervascular lesions during hepatic arterial phase helical CT from the remaining patients with only a predominant number of hypervascular lesions. Interestingly, we found that patients with only purely hypervascular lesions also showed improved survival compared with patients with a predominant number of hypervascular lesions (Fig. 7). This survival benefit, however, extended for only the first 18 months of transcatheter arterial chemoembolization treatment.
The reasons why the improved survival benefit for patients with purely hypervascular lesions lasted over only the first 18 months of transcatheter arterial chemoembolization treatment are not completely understood. No significant differences in treatment dose, frequency, or response to therapy exist between the two groups. The differences in survival may be because of the biologic and pharmacokinetic behavior of hypervascular tumors in response to transcatheter arterial chemoembolization treatment. Tumors with greater hypervascularity may be more susceptible to sudden and repeated occlusion of their rich vascular blood supply compared with tumors with less abundant vascular supply. In fact, given multiple lesions (hypervascular and hypovascular) in a treated lobe or segment of liver, hypervascular lesions may have both increased delivery and uptake of chemoembolic agents compared with the less vascular lesions. The greatest effect of transcatheter arterial chemoembolization treatment is probably in the initial sessions with increased severity and impact of embolization on the hypervascular tumors. With repeated treatments, these "hypervascular" tumors may be converted into relatively "hypovascular" tumors. Once this change occurs, these tumors behave like other hypovascular lesions with similar treatment responses and survival curves.
This may explain the improved survival rates for nonresponders with type 1 hypervascular lesions compared with nonresponders with type 2 hypovascular lesions during hepatic arterial phase helical CT. As shown in Figure 8, the increased survival benefit for the nonresponders with type 1 lesions is statistically significant and occurs over approximately the first 13 months of transcatheter arterial chemoembolization therapy. During this period, the hypervascular lesions, although not fulfilling the criterion for an oncologic response, may still have a treatment-delivery advantage compared with other hypovascular lesions. Again, this enhanced survival lasts only over a certain time interval of treatment (similar to the 18-month advantage for the purely hypervascular group compared with the predominantly hypervascular group).
We recognize that this study was a retrospective analysis and that selection bias may have affected the results of the study. However, the entry criteria were clearly defined at the start of the study and no patients in the study population were excluded on the basis of treatment toxicity. Patients were excluded only if helical CT scans with arterial phase imaging to assess for vascularity had not been obtained or if the patients were lost to follow-up before a sufficient time interval (3 transcatheter arterial chemoembolization treatments) to assess for treatment response. No differences in patient characteristics including age, sex, Child's classification, or tumor stage were found between the responder and nonresponder groups. This investigation is a preliminary study of the potentially valuable role of arterial phase CT in the prediction of response to transcatheter arterial chemoembolization. A prospective study with more patients is needed to further evaluate the impact of tumor vascularity, as determined by hepatic arterial phase helical CT, on response to transcatheter arterial chemoembolization therapy.
In summary, we found that patients with unresectable hepatocellular carcinoma who have a predominant pattern of hypervascular enhancing lesions (type 1) on hepatic arterial phase helical CT are more likely to show a response to transcatheter arterial chemoembolization and, more important, that these responders are statistically more likely to live longer. In addition, the greatest survival benefit is offered to patients with only purely hypervascular enhancing lesions on hepatic arterial phase helical CT when compared with patients with predominantly hypervascular or hypovascular enhancing lesions. Our results correlate well with those of Yamashita et al. [13] who found better response and survival rates for patients with hypervascular tumors as determined on angiography. We recommend that arterial phase CT vascular enhancement be used before transcatheter arterial chemoembolization therapy to predict which patients are more likely to respond to therapy. This can provide prognostic information to the patient (improved survival rates) and allow the oncologist the ability to better tailor the dose and frequency of chemoembolization treatments.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  I. R. Kamel, E. Liapi, and E. K. Fishman Incidental nonneoplastic hypervascular lesions in the noncirrhotic liver: diagnosis with 16-MDCT and 3D CT angiography. Am. J. Roentgenol., September 1, 2006; 187(3): 682 - 687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. V. Pandharipande, G. A. Krinsky, H. Rusinek, and V. S. Lee Perfusion Imaging of the Liver: Current Challenges and Future Goals Radiology, March 1, 2005; 234(3): 661 - 673. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
