- Email: [email protected]
Various scoring systems evaluating histologic features of chronic hepatitis C treated with interferon
Various Scoring Systems Evaluating Histologic Features of Chronic Hepatitis C Treated With Interferon HIROTSUGU IKAWA, MD, YOSHITAKE HAYASHI, MD, TOSHIAKI NINOMIYA, MD, YOSHIHIKO YANO, MD, MIYUKI NAKAJI, MD, HIDENOBU NAGANO, MD, YASUSHI SEO, MD, YOSHIKO KUMON, MD, SEITETSU YOON, MD, MASATO KASUGA, MD, HIROSHI ITOH, MD, AND CHIHO OHBAYASHI, MD Various scoring systems for chronic hepatitis have been proposed; however, there is no standard scoring system for studies of interferon (IFN) therapy in patients with chronic hepatitis C. The aims of this study were to determine the most useful system reflecting histologic changes in biopsy specimens from complete responders and predicting the efficacy of IFN therapy. Patients with chronic hepatitis C were administered IFN-␣ for 6 months. Forty-six patients were included in this study and categorized as complete responders (n ⴝ 15), partial responders (n ⴝ 24), and nonresponders (n ⴝ 7) according to viral and biochemical responses to the therapy. Biopsy specimens obtained from each patient before and after treatment were evaluated under 3 different systems: Histological Activity Index (HAI), modified HAI, and Scheuer classification. Complete responders showed considerable improvement in both grade and stage on the modified HAI and Scheuer classifications. On the HAI, a considerable improvement was observed in grade but not in stage. No significant change was observed in partial responders or nonresponders on
any system. Prediction of complete response was not possible under any system, but the pretreatment score reflecting piecemeal necrosis on any 1 of the 3 classifications and the fibrosis score on Scheuer classification were predictors of nonresponse. The modified HAI system and Scheuer classification were amply useful in evaluating histologic changes in complete responders. Scores higher than 4 of the categories reflecting piecemeal necrosis on any system and fibrosis scores of 3 or 4 on Scheuer classification predicted nonresponse to IFN therapy. HUM PATHOL 32:910-917. Copyright © 2001 by W.B. Saunders Company Key words: scoring system, histology, chronic hepatitis C, interferon therapy. Abbreviations: HAI, Histological Activity Index; IFN, interferon; ALT, alanine aminotransferase; HCV, hepatitis C virus; MU, million units; ROC, receiver operator characteristic; P-C, portal-central; P-P, portal-portal.
When the first classification for chronic hepatitis was proposed in 1968, most cases were thought to be autoimmune hepatitis.1 With the impressive progress in studies of chronic hepatitis, it is clear that the main cause of the disease in developed countries is continuous viral infection, especially with hepatitis C. Chronic viral hepatitis is now recognized as a chronologically progressive disease leading to cirrhosis, and simple classification into chronic persistent hepatitis, chronic aggressive hepatitis, and chronic lobular hepatitis is no longer enough. In the past 2 decades, some new classifications for semiquantitative assessment have been proposed to produce numeric scoring demonstrative of progression of fibrosis and necroinflammatory activity in chronic hepatitis. Histological Activity Index (HAI) by Knodell et al.2 was the first proposed scoring system used worldwide. Ishak et al3 suggested modifications and proposed the modified HAI. Scheuer4 proposed a
system in which scoring of necroinflammatory activity was somewhat simpler. Different scoring systems have been used in different clinical studies because there is no standard scoring system for chronic liver diseases. It is not clear which scoring system is the most sensitive and accurate for evaluating histologic changes in patients with chronic hepatitis C after interferon (IFN) therapy. We therefore compared HAI, modified HAI, and Scheuer classification in liver biopsy specimens of chronic hepatitis C with the goal of determining the most reliable system reflecting histologic alterations of the liver caused by viral infection and improvements after IFN treatments.
From the First Department of Pathology, Second Department of Internal Medicine, and Surgical Pathology Division, Kobe University Hospital, Kobe University School of Medicine, Kobe, Japan. Accepted for publication May 23, 2001. Address correspondence and reprint requests to Yoshitake Hayashi, MD, First Department of Pathology, Faculty of Medicine, Kobe University, 7-5-1 Kusunoki-cho, Chuo-ku Kobe 650-0017, Japan. Copyright © 2001 by W.B. Saunders Company 0046-8177/01/3209-0003$35.00/0 doi:10.1053/hupa.2001.27108
MATERIALS AND METHODS Patients and IFN-␣ Therapy Forty-seven patients (29 men and 18 women; age 23 to 66 years; mean age 47.4 ⫾ 12.0 years) with a history of abnormal serum alanine aminotransferase (ALT) levels for more than 6 months, demonstrating antibody to hepatitis C virus (HCV) antibodies (second-enzyme-linked immunosorbent assay) generation and HCV RNA (reverse-transcription polymerase chain reaction assay) in their blood samples, were treated with IFN-␣. None of the patients showed any evidence of chronic hepatitis B or other causes of hepatocellular injury. They were administered IFN-␣ by intramuscular injection at doses of 6 to 9 million units (MU) every day for 2 weeks and then 3 times a week for 22 weeks, for a total of 6 months (total
910
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
TABLE 1. Categories of Grade and Stage on the HAI, Modified HAI, and Scheuer Classification HAI (Score) Category of grade
Category of stage
Modified HAI (Score)
Periportal ⫾ bridging necrosis (0, 1, 3, 4, 5, 6, 10) Intralobular degeneration and focal necrosis (0, 1, 3, 4) Portal inflammation (0, 1, 3, 4) Fibrosis (0, 1, 3, 4)
Periportal or periseptal interface hepatitis (0-4) Confluent necrosis (0-6) Focal lytic necrosis, apoptosis, and focal inflammation (0-4) Portal inflammation (0-4) Architectural changes, fibrosis, and cirrhosis (0-6)
dose, 378 to 1,188 MU). The patients were divided into 3 groups according to their response to the therapy, as follows: the complete responder group of 16 patients (8 men and 8 women) who showed normalization of ALT levels and clearance of serum HCV RNA lasting more than 6 months after the cessation of treatment; the partial responder group of 24 patients (16 men and 8 women) who showed transient normalization of ALT levels and/or clearance of serum HCV RNA but then relapsed within 6 months after the cessation of treatment; and the nonresponder group of 7 patients (5 men and 2 women) who showed continuous abnormal ALT levels and positive serum HCV RNA. There was no significant difference between any 2 groups for age, sex, total IFN dose, or HCV genotype. The quantity of HCV RNA before treatment was significantly lower in the complete responder group than in the other groups (P ⫽ .007 and .031 compared with the partial responder and nonresponder groups, respectively, by t test). Informed consent was obtained from each patient, and the study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected by the approval of the Kobe University Human Research Committee.
Histologic Evaluation and Classification Biopsy specimens from each patient were evaluated by hematoxylin and eosin staining and Masson trichrome staining before and after IFN-␣ therapy. Each specimen was evaluated blindly by 3 individuals (H.I., Y.Y., and Y.H.) under
Scheuer Classification (Score) Portal/periportal activity (0-4) Lobular activity (0-4)
Fibrosis (0-4)
each of the 3 scoring systems: HAI, modified HAI, and Scheuer classification. Each score was determined after discussion among the 3 reviewers. One female patient in the complete responder group was excluded from the study because the 1 portal tract detected in the posttreatment biopsy specimen was too small to evaluate scores. The HAI scoring system comprises four categories: (I) periportal ⫾ bridging necrosis (score 0, 1, 3, 4, 5, 6, 10); (II) intralobular degeneration and focal necrosis (score 0, 1, 3, 4), (III) portal inflammation (score 0, 1, 3, 4); (IV) fibrosis (score 0, 1, 3, 4). The HAI score is calculated as the sum of the 4 scores. This is apt to be misleading however, in that it combines the grade of necroinflammatory activity with the effects of such activity, that is, the stage of fibrosis.4 Therefore, in this study, the HAI score was divided into 2 scores: the grading score (the sum of the scores of the first 3 categories), and the staging score (the score of the fourth category; Table 1). The modified HAI score includes 4 categories of necroinflammatory activity: (A) periportal or periseptal interface hepatitis (score 0 to 4); (B) confluent necrosis (score 0 to 6), which includes zone 3 necrosis and bridging necrosis; (C) focal (spotty) lytic necrosis, apoptosis, and focal inflammation (score 0 to 4); (D) portal inflammation (score 0 to 4). The sum of these 4 scores constitutes the grading score. Architectural changes, fibrosis, and cirrhosis make up the category for staging (score 0 to 6; Table 1). In Scheuer classification, inflammatory grade is scored by (a) portal/periportal activity (score 0 to 4) and (b) lobular activity (score 0 to 4). Fibrosis
TABLE 2. Scores of the Categories Reflecting Piecemeal Necrosis on the HAI, Modified HAI, and Scheuer Classification Scoring System Category Score 0 1 2 3 4 5 6 10
HAI Periportal ⫾ Bridging Necrosis
Modified HAI Periportal or Periseptal Interface Hepatitis
Scheuer Classification Portal/Periportal Activity
None Mild piecemeal necrosis — Moderate piecemeal necrosis (⬍50% of portal tracts) Marked piecemeal necrosis (⬎50% of portal tracts) Moderate piecemeal necrosis ⫹ bridging necrosis Marked piecemeal necrosis ⫹ bridging necrosis Multilobular necrosis
Absent Mild (focal, few portal areas) Mild/moderate (focal, most portal areas) Moderate (continuous around ⬍50% of tracts) Severe (continuous around ⬎50% of tracts) —
None or minimal Portal inflammation (CPH) Mild piecemeal necrosis (mild CAH) Moderate piecemeal necrosis (moderate CAH) Severe piecemeal necrosis (severe CAH) —
—
—
—
—
NOTE. Under the HAI system, scores of 2 and 7 through 9 are omitted. The periportal ⫾ bridging necrosis category on the HAI includes both piecemeal and bridging necrosis. Abbreviations: CPH, chronic persistent hepatitis; CAH, chronic active hepatitis.
911
HUMAN PATHOLOGY
Volume 32, No. 9 (September 2001)
and cirrhosis constitute the staging category (score 0 to 4; Table 1). Scores of the categories reflecting piecemeal necrosis under the 3 classifications are shown in Table 2.
Statistical Analysis Comparisons of scores of biopsy specimens between preand post-IFN treatment were done by the exact method of the Wilcoxon signed rank test for paired sample. Comparisons of the relationship between the pretreatment scores of each category and the result of the IFN treatment were done by the exact method of the Wilcoxon rank sum test. A P value of ⬍.05 was considered statistically significant. Receiver operator characteristic (ROC) analysis was used to compare predictive superiority among the 3 classifications.5
RESULTS Changes of Histologic Score of Each Classification After Therapy The grading and staging scores for each patient before and after IFN treatment under the 3 classifications are shown in Table 3. Each biopsy specimen contained more than 5 portal areas and 4 central veins that were sufficient for applying any of the 3 classifications. In the complete responder group, significant decreases were observed in grading under the 3 classifications after therapy compared with those before therapy. In staging, a significant decrease was observed only under the modified HAI. The P value of staging
TABLE 3. Grading and Staging for Each Case Before and After IFN Therapy on the HAI, Modified HAI, and Scheuer Classification HAI Score Grade
Modified HAI Score Stage
Grade
Scheuer Classification
Stage
Grade
Stage
PrePostPrePostPrePostPrePostPrePostPrePostAge Response Case Sex (yr) to IFN* treatment treatment treatment treatment treatment treatment treatment treatment treatment treatment treatment treatment 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
M M M M M F M F M F F F F M F F M M M F M M M F M M M F F M M M M F M F F M M M F M M M F M
49 60 34 25 47 59 53 59 35 46 56 50 46 30 36 56 48 66 37 43 36 66 51 54 55 56 30 23 53 32 46 65 51 40 30 56 62 23 50 54 62 61 56 43 58 47
CR CR CR CR CR CR CR CR CR CR CR CR CR CR CR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR NR NR NR NR NR NR NR
10 11 3 5 7 5 8 9 10 7 8 8 11 8 5 8 8 11 6 3 8 3 7 11 8 8 8 8 5 5 6 9 8 8 8 10 8 7 8 8 11 7 10 10 8 11
10 8 5 8 5 5 5 10 5 8 7 5 5 5 2 8 5 12 5 5 5 8 5 10 9 9 6 11 5 8 6 9 7 8 5 9 8 5 5 8 11 8 7 11 10 8
3 3 1 1 1 1 3 3 3 1 1 1 3 1 3 1 1 3 1 1 3 1 1 3 3 3 3 3 1 1 1 1 1 1 1 3 1 1 1 1 3 1 3 3 3 1
3 3 1 1 1 1 1 3 1 1 1 1 1 1 1 1 1 4 1 1 1 3 1 3 3 3 3 3 1 3 1 1 3 1 1 3 1 1 1 1 3 1 1 3 3 1
11 10 4 6 7 5 9 7 8 8 9 8 11 9 7 8 8 11 4 4 9 3 7 11 7 9 8 8 6 6 7 11 8 9 9 9 7 7 9 7 11 7 9 9 9 12
*CR, complete response; PR, partial response; NR, nonresponse.
912
10 6 6 9 5 5 5 7 5 9 7 4 4 4 2 9 5 9 5 4 7 9 7 9 10 8 7 9 5 8 7 11 8 10 6 10 8 5 7 9 10 8 6 9 10 9
5 4 1 1 2 1 3 4 3 2 2 2 3 2 3 2 2 5 2 1 4 1 2 5 3 4 3 5 2 2 2 2 2 2 2 5 2 1 2 1 5 2 4 4 4 2
4 3 2 2 2 1 1 3 1 2 2 1 1 2 1 2 1 6 2 1 2 4 2 5 5 4 3 4 1 4 2 2 3 2 1 5 2 1 2 2 4 2 2 5 5 2
5 6 4 5 5 4 5 5 5 5 5 5 6 5 4 5 5 6 4 4 5 3 5 6 4 5 5 5 4 3 4 6 5 5 5 6 5 5 5 5 6 5 6 5 6 6
5 4 5 5 3 4 4 5 4 5 5 4 3 4 3 5 3 6 4 4 4 5 4 5 6 6 4 6 4 4 4 6 5 5 3 5 5 3 4 5 6 5 4 6 6 5
3 2 1 1 1 1 2 3 2 1 1 1 2 1 2 1 1 3 1 1 2 1 1 2 2 3 2 3 1 1 1 1 1 1 1 3 1 1 1 1 3 1 3 3 3 1
3 2 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 4 1 1 1 2 1 3 3 3 2 3 1 3 1 1 2 1 1 3 1 1 1 1 3 1 1 3 3 1
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
TABLE 4. Medians of Pretreatment and Posttreatment Grading and Staging and Median Change After IFN Treatment on the HAI, Modified HAI, and Scheuer Classification Grade
Response to IFN* HAI CR PR NR Modified HAI CR PR NR Scheuer classification CR PR NR
Stage
n
Pretreatment
Posttreatment
Median Change
P†
Pretreatment
Posttreatment
Median Change
15 24 7
8 8 10
5 7.5 8
⫺2 0 0
.048 .385 .875
1 1 3
1 1 1
0 0 0
.125 .688 1.000
15 24 7
8 8 9
5 8 9
⫺2 0 0
.017 .320 .719
2 2 4
2 2 2
⫺1 0 0
.043 .905 1.000
15 24 7
5 5 6
4 4.5 5
⫺1 0 0
.027 .362 .750
1 1 3
1 1 1
0 0 0
.063 .234 1.000
P†
*CR, complete response; PR, partial response; NR, nonresponse. †Exact method of the Wilcoxon signed rank test. The decrease in grade in the CR group was statistically significant under the 3 classifications. The decrease in stage in the CR group was significant under the modified HAI. Under the Scheuer classification, the staging score tended to decrease in the CR group, although not significantly. No significant change was observed in any score in either the PR group or the NR group under any system.
Prediction of Efficacy of IFN Therapy on Each Classification
change under Scheuer classification was close to that under the modified HAI but was not statistically different. Under the HAI, no significant change was observed in staging. Of the 3 systems, the modified HAI was the most sensitive to improvement with the lowest P values for both grade and stage (Table 4). One case (case 3, Table 3) on the 3 classifications, 2 cases (cases 4 and 10, Table 3) on the HAI and the modified HAI systems, and 1 case (case 8, Table 3) on the HAI system showed an increase in grading after therapy. In the partial responder and nonresponder groups, no significant differences between pretreatment and posttreatment were observed in grading or staging under any of the 3 classifications (Table 4). However, even in the nonresponder group, 1 case (case 43, Table 3) on the 3 classifications, 1 case (case 46, Table 3) on the HAI and the modified HAI, and 1 case (case 41, Table 3) on the modified HAI showed improvement in grading that was not reflective of viral clearance.
In the pretreatment score of any category under any classification, the complete responder group showed no significant difference from the other groups. Thus, no category on the 3 classifications was a predictor of complete response. However, a significant difference was observed between the responder (complete or partial) and nonresponder groups in the pretreatment scores of categories reflecting piecemeal necrosis: (1) the periportal ⫾ bridging necrosis category on the HAI classification, (2) the periportal or periseptal interface hepatitis category on the modified HAI, and (3) the portal/periportal activity category on Scheuer classification (Table 5). The most reasonable cutoff score was 4 on any 1 of the 3 systems with true-positive and false-positive rates, respectively, of 0.57 and 0.13 on the HAI and the
TABLE 5. Predictive Value of Categories Reflecting Piecemeal Necrosis on the 3 Classifications for Response to IFN Therapy HAI Score 1 R NR
9 0
True-positive rate for NR† False-positive rate for NR†
1 1
3 25 3 P ⫽ .019* 1 0.77
Scheuer Classification Score
Modified HAI Score 4
1
5 4
4 0
0.57 0.13
1 1
2 5 0
3
25 3 P ⫽ .015* 1 1 0.90 0.77
4
2
5 4
9 0
0.57 0.13
1 1
3 24 3 P ⫽ .023* 1 0.77
4 6 4 0.57 0.15
NOTE. The score of 4 on the 3 classifications was the optimal cuttoff point with a relatively high true-positive rate (sensitivity) and a low false-positive rate (1 ⫺ specificity). Abbreviations: R, responders (complete or partial responders); NR, nonresponders. *Exact method of the Wilcoxon rank sum test. The score of the category reflecting piecemeal necrosis under any classification had predictive value for responders and was statistically significant. †True-positive rate for NR and false-positive rate for NR show those for prediction of nonresponse when the score above is the cuttoff point.
913
HUMAN PATHOLOGY
Volume 32, No. 9 (September 2001)
FIGURE 1. ROC of categories reflecting piecemeal necrosis on the 3 classifications for prediction of nonresponse to IFN. ROC is a graph of true-positive rate (sensitivity) versus falsepositive rate (1 ⫺ specificity) because the cutoff is varied. A model with no predictive power would be a 45° line. The greater the predictive power, the more inclined the curve toward the upperleft corner. Three curves of the categories reflecting piecemeal necrosis under the 3 classifications in this figure all incline toward the upperleft corner and almost overlap, thus showing almost the same predictive efficiency for nonresponse.
modified HAI and 0.57 and 0.15 on Scheuer classification (Table 5). The advantage of any system over the other 2 was then analyzed by the ROC curves of the 3 systems for prediction of nonresponse. The ROC curve is a graph constructed to correlate the true-positive rate (sensitivity) and the false-positive rate (1 ⫺ specificity) for a series of cutoff points. The greater the predictive power, the more inclined the curve toward the upper left corner. The 3 curves of the category reflecting piecemeal necrosis under the 3 classifications almost overlapped, showing almost the same predictive efficiency for nonresponse (Fig 1).
FIGURE 2. ROC of stage on the 3 classifications for prediction of nonresponse to IFN. The ROC curves indicating stage of fibrosis under the 3 classifications on prediction of nonresponse are somewhat different. The predictive efficiency seems to be highest for the Scheuer classification, then the modified HAI; the HAI shows limited efficiency.
The difference in pretreatment stages on any system was not statistically significant between the responder (complete or partial) and nonresponder groups (Table 6). In the ROC analysis of pretreatment stages under the 3 systems for prediction of nonresponse, Scheuer classification was the most discriminating, then the modified HAI; the HAI system was the least discriminating (Fig 2). The most reasonable cutoff score under Scheuer classification was 3, with a truepositive rate of 0.57 and a false-positive rate of 0.15 (Table 6). The pretreatment scores of all other categories under the 3 systems showed no significant differences between the responder (complete or partial) and nonresponder groups.
TABLE 6. Predictive Value of the Pretreatment Stage on the 3 Classifications for Response to IFN Therapy HAI Score 1 R NR True-positive rate for NR† False-positive rate for NR†
3
24 15 3 4 P ⫽ .424* 1 0.57 1 0.38
Scheuer Classification Score
Modified HAI Score 1 6 1 1 1
2 18 2 0.86 0.85
3 6 0 P ⫽ .406* 0.57 0.38
4
5
4 3
5 1
0.57 0.23
0.14 0.13
1 24 3 1 1
2 9 0 P ⫽ .135* 0.57 0.38
3 6 4 0.57 0.15
Abbreviations: R, responders (complete or partial responders); NR, nonresponders. *Exact method of the Wilcoxon rank sum test. †True-positive rate for NR and false-positive rate for NR show those for prediction of nonresponse when the score above is the cutoff point. Scheuer classification was the most discriminating between R and NR among the 3 systems, although the difference was not statistically significant. Under the Scheuer classification, the score of 3 was the optimal cutoff point with a true-positive rate (sensitivity) of 0.57 and a false-positive rate (1 ⫺ specificity) of 0.15.
914
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
DISCUSSION Semiquantitative evaluation of histologic features is important in estimating the subtle changes reflected by treatment. A scoring system is essential in therapeutic trials for treatment of chronic hepatitis. Chronic hepatitis C, the most common viral form in developed countries, is sometimes fatal when it leads to cirrhosis and hepatocellular carcinoma.6 IFN therapy, the best strategy for chronic hepatitis C, has been discussed from the viewpoint of virology; however, histologic distortion and its amelioration after treatment are also important in assessing the efficacy of IFN therapy. Therefore, selection of the scoring system most conducive to assessment of histology of patients with chronic hepatitis C under IFN therapy is crucial. In our study, 4 of 15 complete responders had an increase in necroinflammatory activity after IFN therapy under some classifications. On the other hand, 3 of 7 nonresponders showed improvement in necroinflammatory activity under some classifications that was not reflective of viral elimination. In general, however, the complete responders showed histologic improvement after IFN treatment in both grade and stage, and their posttreatment histologic improvement of necroinflammatory activity was well reflected in grading scores under the 3 systems; their staging scores of fibrosis showed the greatest improvement under the modified HAI, then under the Scheuer system, and least under the HAI. The most essential difference in fibrosis score between the first 2 systems and the last was that the score for portal-central (P-C) bridging fibrosis was higher than that for portal-portal (P-P) bridging fibrosis under the modified HAI and Scheuer classification, whereas no distinction between P-P and P-C bridging fibrosis was made under the HAI system. P-C bridging fibrosis is thought to have a more severe pathogenetic significance than P-P bridging fibrosis. Compared with P-P bridging fibrosis, P-C bridging fibrosis could make blood flow alternation more severe, leading to architectural distortion. Therefore, it seems reasonable to assign a higher score for P-C bridging fibrosis than for P-P bridging fibrosis. That P-C bridging fibrosis could be improved after IFN therapy in complete responders was an important finding, and this improvement was reflected in the staging scores under the modified HAI and Scheuer’s classification (Fig 3). Thus these 2 systems would be more sensitive to subtle changes after therapy. It is well known that the prognosis of patients demonstrating elimination of the virus by IFN therapy is very good; therefore, the goal of therapy has been complete viral response to IFN. Unfortunately, in this study, no category under the 3 classifications was predictive for complete response to IFN. However, some recent studies have shown that partial responders without sustained elimination of HCV RNA also show a more favorable prognosis than nonresponders or those not undergoing IFN therapy.7-11 These studies have led us to believe that the prediction for partial responders is also important. Therefore, we investigated whether
some categories were predictive for complete or partial response. Viral factors, such as genotype and quantity of HCV RNA, are the most significant variants affecting the response to IFN therapy12; moreover, some of the pretreatment histologic scores were helpful in predicting complete or partial response. In this study, the score of each category reflecting piecemeal necrosis (score of periportal ⫾ piecemeal necrosis under HAI, periportal or periseptal interface hepatitis under modified HAI, and portal/periportal activity under Scheuer classification) before therapy showed a significant difference between the complete or partial responder group and the nonresponder group. The cutoff score of 4 seemed useful under the 3 systems with a relatively high true-positive rate and a low false-positive rate. These findings are simply the results of examining the present data; therefore, further independent confirmation is needed. A score of 4 in the categories reflecting piecemeal necrosis on the modified HAI and Scheuer classification is extreme and compatible with our result, whereas on the HAI it is fairly modest, showing a discrepancy with our results. This is because the HAI system assesses piecemeal necrosis and bridging necrosis together in one category and yields a higher score when bridging necrosis is present. Under the HAI system, bridging necrosis is assessed as the extreme of periportal necrosis. On the other hand, under the modified HAI system and Scheuer classification, it is assessed as the result of confluent necrosis; therefore, it is included in the category of confluent necrosis or lobular activity. Bridging necrosis should be assessed in the category of lobular necroinflammation because it has a pathogenetic significance separate from that of piecemeal necrosis.13 The P value of the category reflecting piecemeal necrosis on the Scheuer classification for prediction of nonresponse was relatively higher than those on the other 2 systems. We believed this difference was seen because periportal activity and portal inflammation were assessed in 1 category on the Scheuer classification. The categories of portal inflammation under the HAI and the modified HAI had little value in predicting the response (P ⫽ .856 and .961 by the exact method of the Wilcoxon rank sum test, respectively), whereas the category of piecemeal necrosis alone showed great predictive value. Thus, the category of periportal activity should be scored independently of the category of portal inflammation. Studies on fibrosis as a predictive factor have shown that progression of fibrosis before therapy results in poor response to IFN therapy.14-16 In our study, the fibrosis score on the Scheuer classification was the most discriminating between responders and nonresponders, followed by the modified HAI. The HAI was the least discriminating of the 3 systems. Under the HAI system, no distinction was made between P-P and P-C bridging fibrosis, and the scores before IFN therapy were only 1 and 3 and thus were not statistically significant because patients with cirrhosis before treatment were not included in our study. Scoring for fibrosis on the Scheuer classification was simpler than that on the
915
HUMAN PATHOLOGY
Volume 32, No. 9 (September 2001)
FIGURE 3. Biopsy specimens obtained from a complete responder (case 8, Table 1) before and after IFN therapy. (A) Pretreatment biopsy specimen showing P-C bridging fibrosis with moderate lobular inflammation and focal lobular distortion. (B) Posttreatment biopsy specimen showing P-P bridging fibrosis. P-C bridging fibrosis disappeared, and portal inflammation persisted. The modified HAI scoring system and the Scheuer classification reflected well the improvement in bridging fibrosis as decreases in fibrosis scores. (Hematoxylin and eosin; original magnification ⫻100.)
modified HAI system but amply reliable in predicting IFN efficacy, although data on cirrhotic patients could not be inferred. A fibrosis score of 3 or 4 on the Scheuer classification tended to indicate an outcome of nonresponse to IFN therapy. Two cases in the complete responder group showed pretreatment cutoff scores of 4 in the categories reflecting piecemeal necrosis on the 3 systems (cases 2 and 13, Table 3). In case 2, the quantity of HCV RNA before treatment was the smallest of all the patients. In case 13, the total IFN dose administrated was the highest of all the patients. On the other hand, 3 cases in the nonresponder group showed pretreatment scores of 3 in the categories reflecting piecemeal necrosis under the 3 systems (cases 40, 42, and 44, Table 3). In case 42, the quantity of HCV RNA before treatment was the third largest of all patients, and the total IFN dose administrated was relatively low. In case 44, pretreatment fibrosis was advanced. In 2 cases of the complete responder group, the pretreatment staging score of 3 under the Scheuer
classification was predictive of nonresponse (cases 1 and 8, Table 3). Favorable factors, if any, other than staging scores were not demonstrated histopathologically, virally, or by total IFN dose. On the other hand, 3 cases in the nonresponder group showed a pretreatment staging score of 1 under the Scheuer classification (cases 40, 42, and 46, Table 3). In case 46, unfavorable factors other than stage were scores of 4 in the categories reflecting piecemeal necrosis under the 3 systems. Unfavorable factors in case 42 are discussed in the preceding paragraph. We must emphasize that there are many significant variants affecting the response to IFN therapy: viral factors, such as genotype and quantity of HCV RNA; therapeutic factors, such as total dose and duration of IFN administration; and host factors, such as age and duration of infection. However, the role of the histology of biopsy specimens should not be underestimated because it yields important information as part of the host factors when it is assessed by the discriminating features of a reliable system (the aim of this study).
916
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
In predicting the efficacy of IFN therapy, evaluation of piecemeal necrosis on any system and of the progression of fibrosis on the Scheuer classification was useful. Our study showed that the modified HAI system and the Scheuer classification were both reliable in estimating subtle histologic changes in patients with chronic hepatitis C treated with IFN and that the score reflecting piecemeal necrosis on any system and the fibrosis score on the Scheuer classification were valuable in predicting response to IFN therapy. The Scheuer classification is the simplest of the 3 systems and shows considerably better reproducibility than the HAI, which is simpler than the modified HAI,17 although there are no data on the reproducibility of the Scheuer classification compared with that of the modified HAI in the literature. Thus, we suggest that (1) the Scheuer system, relatively easy to use, is the best for routine clinical use in IFN treatment of patients with chronic hepatitis C, and (2) in clinical studies aimed at evaluating the histology of liver biopsy specimens of hepatitis C patients treated with IFN, a combination of grade on the modified HAI, which is the most sensitive to a subtle improvement of necroinflammation, and stage on the Scheuer classification, which is a little less sensitive for reflecting improvement of fibrosis than the modified HAI but much more predictive for response to IFN therapy, would constitute a superior system.
REFERENCES 1. De Groote J, Desmet VJ, Gedigk P, et al: A classification of chronic hepatitis. Lancet 2:626-628, 1968 2. Knodell RG, Ishak KG, Black WC, et al: Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1:431435, 1981 3. Ishak K, Baptista A, Bianchi L, et al: Histological grading and staging of chronic hepatitis. J Hepatol 22:696-699, 1995
4. Scheuer PJ: Classification of chronic viral hepatitis: A need for reassessment. J Hepatol 13:372-374, 1991 5. Griner PF, Mayewski RJ, Mushlin AI, et al: Selection and interpretation of diagnostic tests and procedures. Principles and applications. Ann Intern Med 94:557-592, 1981 6. Niederau C, Lange S, Heintges T, et al: Prognosis of chronic hepatitis C: Results of a large, prospective cohort study. Hepatology 28:1687-1695, 1998 7. Yoshida H, Shiratori Y, Moriyama M, et al: Interferon therapy reduces the risk for hepatocellular carcinoma: National surveillance program of cirrhotic and noncirrhotic patients with chronic hepatitis C in Japan. IHIT Study Group. Inhibition of Hepatocarcinogenesis by Interferon Therapy. Ann Intern Med 131:174-181, 1999 8. Morisco F, Marmo R, Iasevoli P, et al: Clinical outcome of chronic hepatitis C in patients treated with interferon: Comparison between responders and non-responders. Ital J Gastroenterol Hepatol 31:454-458, 1999 9. Okanoue T, Itoh Y, Minami M, et al: Interferon therapy lowers the rate of progression to hepatocellular carcinoma in chronic hepatitis C but not significantly in an advanced stage: A retrospective study in 1148 patients. Viral Hepatitis Therapy Study Group. J Hepatol 30:653-659, 1999 10. Shindo M, Ken A, Okuno T: Varying incidence of cirrhosis and hepatocellular carcinoma in patients with chronic hepatitis C responding differently to interferon therapy. Cancer 85:1943-1950, 1999 11. Kasahara A, Hayashi N, Mochizuki K, et al: Risk factors for hepatocellular carcinoma and its incidence after interferon treatment in patients with chronic hepatitis C. Osaka Liver Disease Study Group. Hepatology 27:1394-1402, 1998 12. Zeniya M, Toda G: Case selection for interferon treatment of hepatitis C. J Gastroenterol Hepatol 15:117-122, 2000 (suppl E) 13. Cooksley WGE, Bradbear RA, Robinson W, et al: The prognosis of chronic active hepatitis without cirrhosis in relation to bridging necrosis. Hepatology 6:345-348, 1986 14. Jenkins PJ, Cromie SL, Bowden DS, et al: Chronic hepatitis C and interferon alpha therapy: Predictors of long term response. Med J Aust 164:150-152, 1996 15. Morisco F, Marmo R, Iasevoli P, et al: Clinical outcome of chronic hepatitis C in patients treated with interferon: Comparison between responders and non-responders. Ital J Gastroenterol Hepatol 31:454-458, 1999 16. Kim SR, Hayashi Y, Yoon S, et al: Prediction of efficacy of interferon treatment of chronic hepatitis C by multivariate analysis and a new classification. Pathol Int 48:215-220, 1998 17. Goldin RD, Goldin JG, Burt AD, et al: Intra-observer and inter-observer variation in the histopathological assessment of chronic viral hepatitis. J Hepatol 25:649-654, 1996
917
any system. Prediction of complete response was not possible under any system, but the pretreatment score reflecting piecemeal necrosis on any 1 of the 3 classifications and the fibrosis score on Scheuer classification were predictors of nonresponse. The modified HAI system and Scheuer classification were amply useful in evaluating histologic changes in complete responders. Scores higher than 4 of the categories reflecting piecemeal necrosis on any system and fibrosis scores of 3 or 4 on Scheuer classification predicted nonresponse to IFN therapy. HUM PATHOL 32:910-917. Copyright © 2001 by W.B. Saunders Company Key words: scoring system, histology, chronic hepatitis C, interferon therapy. Abbreviations: HAI, Histological Activity Index; IFN, interferon; ALT, alanine aminotransferase; HCV, hepatitis C virus; MU, million units; ROC, receiver operator characteristic; P-C, portal-central; P-P, portal-portal.
When the first classification for chronic hepatitis was proposed in 1968, most cases were thought to be autoimmune hepatitis.1 With the impressive progress in studies of chronic hepatitis, it is clear that the main cause of the disease in developed countries is continuous viral infection, especially with hepatitis C. Chronic viral hepatitis is now recognized as a chronologically progressive disease leading to cirrhosis, and simple classification into chronic persistent hepatitis, chronic aggressive hepatitis, and chronic lobular hepatitis is no longer enough. In the past 2 decades, some new classifications for semiquantitative assessment have been proposed to produce numeric scoring demonstrative of progression of fibrosis and necroinflammatory activity in chronic hepatitis. Histological Activity Index (HAI) by Knodell et al.2 was the first proposed scoring system used worldwide. Ishak et al3 suggested modifications and proposed the modified HAI. Scheuer4 proposed a
system in which scoring of necroinflammatory activity was somewhat simpler. Different scoring systems have been used in different clinical studies because there is no standard scoring system for chronic liver diseases. It is not clear which scoring system is the most sensitive and accurate for evaluating histologic changes in patients with chronic hepatitis C after interferon (IFN) therapy. We therefore compared HAI, modified HAI, and Scheuer classification in liver biopsy specimens of chronic hepatitis C with the goal of determining the most reliable system reflecting histologic alterations of the liver caused by viral infection and improvements after IFN treatments.
From the First Department of Pathology, Second Department of Internal Medicine, and Surgical Pathology Division, Kobe University Hospital, Kobe University School of Medicine, Kobe, Japan. Accepted for publication May 23, 2001. Address correspondence and reprint requests to Yoshitake Hayashi, MD, First Department of Pathology, Faculty of Medicine, Kobe University, 7-5-1 Kusunoki-cho, Chuo-ku Kobe 650-0017, Japan. Copyright © 2001 by W.B. Saunders Company 0046-8177/01/3209-0003$35.00/0 doi:10.1053/hupa.2001.27108
MATERIALS AND METHODS Patients and IFN-␣ Therapy Forty-seven patients (29 men and 18 women; age 23 to 66 years; mean age 47.4 ⫾ 12.0 years) with a history of abnormal serum alanine aminotransferase (ALT) levels for more than 6 months, demonstrating antibody to hepatitis C virus (HCV) antibodies (second-enzyme-linked immunosorbent assay) generation and HCV RNA (reverse-transcription polymerase chain reaction assay) in their blood samples, were treated with IFN-␣. None of the patients showed any evidence of chronic hepatitis B or other causes of hepatocellular injury. They were administered IFN-␣ by intramuscular injection at doses of 6 to 9 million units (MU) every day for 2 weeks and then 3 times a week for 22 weeks, for a total of 6 months (total
910
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
TABLE 1. Categories of Grade and Stage on the HAI, Modified HAI, and Scheuer Classification HAI (Score) Category of grade
Category of stage
Modified HAI (Score)
Periportal ⫾ bridging necrosis (0, 1, 3, 4, 5, 6, 10) Intralobular degeneration and focal necrosis (0, 1, 3, 4) Portal inflammation (0, 1, 3, 4) Fibrosis (0, 1, 3, 4)
Periportal or periseptal interface hepatitis (0-4) Confluent necrosis (0-6) Focal lytic necrosis, apoptosis, and focal inflammation (0-4) Portal inflammation (0-4) Architectural changes, fibrosis, and cirrhosis (0-6)
dose, 378 to 1,188 MU). The patients were divided into 3 groups according to their response to the therapy, as follows: the complete responder group of 16 patients (8 men and 8 women) who showed normalization of ALT levels and clearance of serum HCV RNA lasting more than 6 months after the cessation of treatment; the partial responder group of 24 patients (16 men and 8 women) who showed transient normalization of ALT levels and/or clearance of serum HCV RNA but then relapsed within 6 months after the cessation of treatment; and the nonresponder group of 7 patients (5 men and 2 women) who showed continuous abnormal ALT levels and positive serum HCV RNA. There was no significant difference between any 2 groups for age, sex, total IFN dose, or HCV genotype. The quantity of HCV RNA before treatment was significantly lower in the complete responder group than in the other groups (P ⫽ .007 and .031 compared with the partial responder and nonresponder groups, respectively, by t test). Informed consent was obtained from each patient, and the study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected by the approval of the Kobe University Human Research Committee.
Histologic Evaluation and Classification Biopsy specimens from each patient were evaluated by hematoxylin and eosin staining and Masson trichrome staining before and after IFN-␣ therapy. Each specimen was evaluated blindly by 3 individuals (H.I., Y.Y., and Y.H.) under
Scheuer Classification (Score) Portal/periportal activity (0-4) Lobular activity (0-4)
Fibrosis (0-4)
each of the 3 scoring systems: HAI, modified HAI, and Scheuer classification. Each score was determined after discussion among the 3 reviewers. One female patient in the complete responder group was excluded from the study because the 1 portal tract detected in the posttreatment biopsy specimen was too small to evaluate scores. The HAI scoring system comprises four categories: (I) periportal ⫾ bridging necrosis (score 0, 1, 3, 4, 5, 6, 10); (II) intralobular degeneration and focal necrosis (score 0, 1, 3, 4), (III) portal inflammation (score 0, 1, 3, 4); (IV) fibrosis (score 0, 1, 3, 4). The HAI score is calculated as the sum of the 4 scores. This is apt to be misleading however, in that it combines the grade of necroinflammatory activity with the effects of such activity, that is, the stage of fibrosis.4 Therefore, in this study, the HAI score was divided into 2 scores: the grading score (the sum of the scores of the first 3 categories), and the staging score (the score of the fourth category; Table 1). The modified HAI score includes 4 categories of necroinflammatory activity: (A) periportal or periseptal interface hepatitis (score 0 to 4); (B) confluent necrosis (score 0 to 6), which includes zone 3 necrosis and bridging necrosis; (C) focal (spotty) lytic necrosis, apoptosis, and focal inflammation (score 0 to 4); (D) portal inflammation (score 0 to 4). The sum of these 4 scores constitutes the grading score. Architectural changes, fibrosis, and cirrhosis make up the category for staging (score 0 to 6; Table 1). In Scheuer classification, inflammatory grade is scored by (a) portal/periportal activity (score 0 to 4) and (b) lobular activity (score 0 to 4). Fibrosis
TABLE 2. Scores of the Categories Reflecting Piecemeal Necrosis on the HAI, Modified HAI, and Scheuer Classification Scoring System Category Score 0 1 2 3 4 5 6 10
HAI Periportal ⫾ Bridging Necrosis
Modified HAI Periportal or Periseptal Interface Hepatitis
Scheuer Classification Portal/Periportal Activity
None Mild piecemeal necrosis — Moderate piecemeal necrosis (⬍50% of portal tracts) Marked piecemeal necrosis (⬎50% of portal tracts) Moderate piecemeal necrosis ⫹ bridging necrosis Marked piecemeal necrosis ⫹ bridging necrosis Multilobular necrosis
Absent Mild (focal, few portal areas) Mild/moderate (focal, most portal areas) Moderate (continuous around ⬍50% of tracts) Severe (continuous around ⬎50% of tracts) —
None or minimal Portal inflammation (CPH) Mild piecemeal necrosis (mild CAH) Moderate piecemeal necrosis (moderate CAH) Severe piecemeal necrosis (severe CAH) —
—
—
—
—
NOTE. Under the HAI system, scores of 2 and 7 through 9 are omitted. The periportal ⫾ bridging necrosis category on the HAI includes both piecemeal and bridging necrosis. Abbreviations: CPH, chronic persistent hepatitis; CAH, chronic active hepatitis.
911
HUMAN PATHOLOGY
Volume 32, No. 9 (September 2001)
and cirrhosis constitute the staging category (score 0 to 4; Table 1). Scores of the categories reflecting piecemeal necrosis under the 3 classifications are shown in Table 2.
Statistical Analysis Comparisons of scores of biopsy specimens between preand post-IFN treatment were done by the exact method of the Wilcoxon signed rank test for paired sample. Comparisons of the relationship between the pretreatment scores of each category and the result of the IFN treatment were done by the exact method of the Wilcoxon rank sum test. A P value of ⬍.05 was considered statistically significant. Receiver operator characteristic (ROC) analysis was used to compare predictive superiority among the 3 classifications.5
RESULTS Changes of Histologic Score of Each Classification After Therapy The grading and staging scores for each patient before and after IFN treatment under the 3 classifications are shown in Table 3. Each biopsy specimen contained more than 5 portal areas and 4 central veins that were sufficient for applying any of the 3 classifications. In the complete responder group, significant decreases were observed in grading under the 3 classifications after therapy compared with those before therapy. In staging, a significant decrease was observed only under the modified HAI. The P value of staging
TABLE 3. Grading and Staging for Each Case Before and After IFN Therapy on the HAI, Modified HAI, and Scheuer Classification HAI Score Grade
Modified HAI Score Stage
Grade
Scheuer Classification
Stage
Grade
Stage
PrePostPrePostPrePostPrePostPrePostPrePostAge Response Case Sex (yr) to IFN* treatment treatment treatment treatment treatment treatment treatment treatment treatment treatment treatment treatment 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
M M M M M F M F M F F F F M F F M M M F M M M F M M M F F M M M M F M F F M M M F M M M F M
49 60 34 25 47 59 53 59 35 46 56 50 46 30 36 56 48 66 37 43 36 66 51 54 55 56 30 23 53 32 46 65 51 40 30 56 62 23 50 54 62 61 56 43 58 47
CR CR CR CR CR CR CR CR CR CR CR CR CR CR CR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR PR NR NR NR NR NR NR NR
10 11 3 5 7 5 8 9 10 7 8 8 11 8 5 8 8 11 6 3 8 3 7 11 8 8 8 8 5 5 6 9 8 8 8 10 8 7 8 8 11 7 10 10 8 11
10 8 5 8 5 5 5 10 5 8 7 5 5 5 2 8 5 12 5 5 5 8 5 10 9 9 6 11 5 8 6 9 7 8 5 9 8 5 5 8 11 8 7 11 10 8
3 3 1 1 1 1 3 3 3 1 1 1 3 1 3 1 1 3 1 1 3 1 1 3 3 3 3 3 1 1 1 1 1 1 1 3 1 1 1 1 3 1 3 3 3 1
3 3 1 1 1 1 1 3 1 1 1 1 1 1 1 1 1 4 1 1 1 3 1 3 3 3 3 3 1 3 1 1 3 1 1 3 1 1 1 1 3 1 1 3 3 1
11 10 4 6 7 5 9 7 8 8 9 8 11 9 7 8 8 11 4 4 9 3 7 11 7 9 8 8 6 6 7 11 8 9 9 9 7 7 9 7 11 7 9 9 9 12
*CR, complete response; PR, partial response; NR, nonresponse.
912
10 6 6 9 5 5 5 7 5 9 7 4 4 4 2 9 5 9 5 4 7 9 7 9 10 8 7 9 5 8 7 11 8 10 6 10 8 5 7 9 10 8 6 9 10 9
5 4 1 1 2 1 3 4 3 2 2 2 3 2 3 2 2 5 2 1 4 1 2 5 3 4 3 5 2 2 2 2 2 2 2 5 2 1 2 1 5 2 4 4 4 2
4 3 2 2 2 1 1 3 1 2 2 1 1 2 1 2 1 6 2 1 2 4 2 5 5 4 3 4 1 4 2 2 3 2 1 5 2 1 2 2 4 2 2 5 5 2
5 6 4 5 5 4 5 5 5 5 5 5 6 5 4 5 5 6 4 4 5 3 5 6 4 5 5 5 4 3 4 6 5 5 5 6 5 5 5 5 6 5 6 5 6 6
5 4 5 5 3 4 4 5 4 5 5 4 3 4 3 5 3 6 4 4 4 5 4 5 6 6 4 6 4 4 4 6 5 5 3 5 5 3 4 5 6 5 4 6 6 5
3 2 1 1 1 1 2 3 2 1 1 1 2 1 2 1 1 3 1 1 2 1 1 2 2 3 2 3 1 1 1 1 1 1 1 3 1 1 1 1 3 1 3 3 3 1
3 2 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 4 1 1 1 2 1 3 3 3 2 3 1 3 1 1 2 1 1 3 1 1 1 1 3 1 1 3 3 1
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
TABLE 4. Medians of Pretreatment and Posttreatment Grading and Staging and Median Change After IFN Treatment on the HAI, Modified HAI, and Scheuer Classification Grade
Response to IFN* HAI CR PR NR Modified HAI CR PR NR Scheuer classification CR PR NR
Stage
n
Pretreatment
Posttreatment
Median Change
P†
Pretreatment
Posttreatment
Median Change
15 24 7
8 8 10
5 7.5 8
⫺2 0 0
.048 .385 .875
1 1 3
1 1 1
0 0 0
.125 .688 1.000
15 24 7
8 8 9
5 8 9
⫺2 0 0
.017 .320 .719
2 2 4
2 2 2
⫺1 0 0
.043 .905 1.000
15 24 7
5 5 6
4 4.5 5
⫺1 0 0
.027 .362 .750
1 1 3
1 1 1
0 0 0
.063 .234 1.000
P†
*CR, complete response; PR, partial response; NR, nonresponse. †Exact method of the Wilcoxon signed rank test. The decrease in grade in the CR group was statistically significant under the 3 classifications. The decrease in stage in the CR group was significant under the modified HAI. Under the Scheuer classification, the staging score tended to decrease in the CR group, although not significantly. No significant change was observed in any score in either the PR group or the NR group under any system.
Prediction of Efficacy of IFN Therapy on Each Classification
change under Scheuer classification was close to that under the modified HAI but was not statistically different. Under the HAI, no significant change was observed in staging. Of the 3 systems, the modified HAI was the most sensitive to improvement with the lowest P values for both grade and stage (Table 4). One case (case 3, Table 3) on the 3 classifications, 2 cases (cases 4 and 10, Table 3) on the HAI and the modified HAI systems, and 1 case (case 8, Table 3) on the HAI system showed an increase in grading after therapy. In the partial responder and nonresponder groups, no significant differences between pretreatment and posttreatment were observed in grading or staging under any of the 3 classifications (Table 4). However, even in the nonresponder group, 1 case (case 43, Table 3) on the 3 classifications, 1 case (case 46, Table 3) on the HAI and the modified HAI, and 1 case (case 41, Table 3) on the modified HAI showed improvement in grading that was not reflective of viral clearance.
In the pretreatment score of any category under any classification, the complete responder group showed no significant difference from the other groups. Thus, no category on the 3 classifications was a predictor of complete response. However, a significant difference was observed between the responder (complete or partial) and nonresponder groups in the pretreatment scores of categories reflecting piecemeal necrosis: (1) the periportal ⫾ bridging necrosis category on the HAI classification, (2) the periportal or periseptal interface hepatitis category on the modified HAI, and (3) the portal/periportal activity category on Scheuer classification (Table 5). The most reasonable cutoff score was 4 on any 1 of the 3 systems with true-positive and false-positive rates, respectively, of 0.57 and 0.13 on the HAI and the
TABLE 5. Predictive Value of Categories Reflecting Piecemeal Necrosis on the 3 Classifications for Response to IFN Therapy HAI Score 1 R NR
9 0
True-positive rate for NR† False-positive rate for NR†
1 1
3 25 3 P ⫽ .019* 1 0.77
Scheuer Classification Score
Modified HAI Score 4
1
5 4
4 0
0.57 0.13
1 1
2 5 0
3
25 3 P ⫽ .015* 1 1 0.90 0.77
4
2
5 4
9 0
0.57 0.13
1 1
3 24 3 P ⫽ .023* 1 0.77
4 6 4 0.57 0.15
NOTE. The score of 4 on the 3 classifications was the optimal cuttoff point with a relatively high true-positive rate (sensitivity) and a low false-positive rate (1 ⫺ specificity). Abbreviations: R, responders (complete or partial responders); NR, nonresponders. *Exact method of the Wilcoxon rank sum test. The score of the category reflecting piecemeal necrosis under any classification had predictive value for responders and was statistically significant. †True-positive rate for NR and false-positive rate for NR show those for prediction of nonresponse when the score above is the cuttoff point.
913
HUMAN PATHOLOGY
Volume 32, No. 9 (September 2001)
FIGURE 1. ROC of categories reflecting piecemeal necrosis on the 3 classifications for prediction of nonresponse to IFN. ROC is a graph of true-positive rate (sensitivity) versus falsepositive rate (1 ⫺ specificity) because the cutoff is varied. A model with no predictive power would be a 45° line. The greater the predictive power, the more inclined the curve toward the upperleft corner. Three curves of the categories reflecting piecemeal necrosis under the 3 classifications in this figure all incline toward the upperleft corner and almost overlap, thus showing almost the same predictive efficiency for nonresponse.
modified HAI and 0.57 and 0.15 on Scheuer classification (Table 5). The advantage of any system over the other 2 was then analyzed by the ROC curves of the 3 systems for prediction of nonresponse. The ROC curve is a graph constructed to correlate the true-positive rate (sensitivity) and the false-positive rate (1 ⫺ specificity) for a series of cutoff points. The greater the predictive power, the more inclined the curve toward the upper left corner. The 3 curves of the category reflecting piecemeal necrosis under the 3 classifications almost overlapped, showing almost the same predictive efficiency for nonresponse (Fig 1).
FIGURE 2. ROC of stage on the 3 classifications for prediction of nonresponse to IFN. The ROC curves indicating stage of fibrosis under the 3 classifications on prediction of nonresponse are somewhat different. The predictive efficiency seems to be highest for the Scheuer classification, then the modified HAI; the HAI shows limited efficiency.
The difference in pretreatment stages on any system was not statistically significant between the responder (complete or partial) and nonresponder groups (Table 6). In the ROC analysis of pretreatment stages under the 3 systems for prediction of nonresponse, Scheuer classification was the most discriminating, then the modified HAI; the HAI system was the least discriminating (Fig 2). The most reasonable cutoff score under Scheuer classification was 3, with a truepositive rate of 0.57 and a false-positive rate of 0.15 (Table 6). The pretreatment scores of all other categories under the 3 systems showed no significant differences between the responder (complete or partial) and nonresponder groups.
TABLE 6. Predictive Value of the Pretreatment Stage on the 3 Classifications for Response to IFN Therapy HAI Score 1 R NR True-positive rate for NR† False-positive rate for NR†
3
24 15 3 4 P ⫽ .424* 1 0.57 1 0.38
Scheuer Classification Score
Modified HAI Score 1 6 1 1 1
2 18 2 0.86 0.85
3 6 0 P ⫽ .406* 0.57 0.38
4
5
4 3
5 1
0.57 0.23
0.14 0.13
1 24 3 1 1
2 9 0 P ⫽ .135* 0.57 0.38
3 6 4 0.57 0.15
Abbreviations: R, responders (complete or partial responders); NR, nonresponders. *Exact method of the Wilcoxon rank sum test. †True-positive rate for NR and false-positive rate for NR show those for prediction of nonresponse when the score above is the cutoff point. Scheuer classification was the most discriminating between R and NR among the 3 systems, although the difference was not statistically significant. Under the Scheuer classification, the score of 3 was the optimal cutoff point with a true-positive rate (sensitivity) of 0.57 and a false-positive rate (1 ⫺ specificity) of 0.15.
914
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
DISCUSSION Semiquantitative evaluation of histologic features is important in estimating the subtle changes reflected by treatment. A scoring system is essential in therapeutic trials for treatment of chronic hepatitis. Chronic hepatitis C, the most common viral form in developed countries, is sometimes fatal when it leads to cirrhosis and hepatocellular carcinoma.6 IFN therapy, the best strategy for chronic hepatitis C, has been discussed from the viewpoint of virology; however, histologic distortion and its amelioration after treatment are also important in assessing the efficacy of IFN therapy. Therefore, selection of the scoring system most conducive to assessment of histology of patients with chronic hepatitis C under IFN therapy is crucial. In our study, 4 of 15 complete responders had an increase in necroinflammatory activity after IFN therapy under some classifications. On the other hand, 3 of 7 nonresponders showed improvement in necroinflammatory activity under some classifications that was not reflective of viral elimination. In general, however, the complete responders showed histologic improvement after IFN treatment in both grade and stage, and their posttreatment histologic improvement of necroinflammatory activity was well reflected in grading scores under the 3 systems; their staging scores of fibrosis showed the greatest improvement under the modified HAI, then under the Scheuer system, and least under the HAI. The most essential difference in fibrosis score between the first 2 systems and the last was that the score for portal-central (P-C) bridging fibrosis was higher than that for portal-portal (P-P) bridging fibrosis under the modified HAI and Scheuer classification, whereas no distinction between P-P and P-C bridging fibrosis was made under the HAI system. P-C bridging fibrosis is thought to have a more severe pathogenetic significance than P-P bridging fibrosis. Compared with P-P bridging fibrosis, P-C bridging fibrosis could make blood flow alternation more severe, leading to architectural distortion. Therefore, it seems reasonable to assign a higher score for P-C bridging fibrosis than for P-P bridging fibrosis. That P-C bridging fibrosis could be improved after IFN therapy in complete responders was an important finding, and this improvement was reflected in the staging scores under the modified HAI and Scheuer’s classification (Fig 3). Thus these 2 systems would be more sensitive to subtle changes after therapy. It is well known that the prognosis of patients demonstrating elimination of the virus by IFN therapy is very good; therefore, the goal of therapy has been complete viral response to IFN. Unfortunately, in this study, no category under the 3 classifications was predictive for complete response to IFN. However, some recent studies have shown that partial responders without sustained elimination of HCV RNA also show a more favorable prognosis than nonresponders or those not undergoing IFN therapy.7-11 These studies have led us to believe that the prediction for partial responders is also important. Therefore, we investigated whether
some categories were predictive for complete or partial response. Viral factors, such as genotype and quantity of HCV RNA, are the most significant variants affecting the response to IFN therapy12; moreover, some of the pretreatment histologic scores were helpful in predicting complete or partial response. In this study, the score of each category reflecting piecemeal necrosis (score of periportal ⫾ piecemeal necrosis under HAI, periportal or periseptal interface hepatitis under modified HAI, and portal/periportal activity under Scheuer classification) before therapy showed a significant difference between the complete or partial responder group and the nonresponder group. The cutoff score of 4 seemed useful under the 3 systems with a relatively high true-positive rate and a low false-positive rate. These findings are simply the results of examining the present data; therefore, further independent confirmation is needed. A score of 4 in the categories reflecting piecemeal necrosis on the modified HAI and Scheuer classification is extreme and compatible with our result, whereas on the HAI it is fairly modest, showing a discrepancy with our results. This is because the HAI system assesses piecemeal necrosis and bridging necrosis together in one category and yields a higher score when bridging necrosis is present. Under the HAI system, bridging necrosis is assessed as the extreme of periportal necrosis. On the other hand, under the modified HAI system and Scheuer classification, it is assessed as the result of confluent necrosis; therefore, it is included in the category of confluent necrosis or lobular activity. Bridging necrosis should be assessed in the category of lobular necroinflammation because it has a pathogenetic significance separate from that of piecemeal necrosis.13 The P value of the category reflecting piecemeal necrosis on the Scheuer classification for prediction of nonresponse was relatively higher than those on the other 2 systems. We believed this difference was seen because periportal activity and portal inflammation were assessed in 1 category on the Scheuer classification. The categories of portal inflammation under the HAI and the modified HAI had little value in predicting the response (P ⫽ .856 and .961 by the exact method of the Wilcoxon rank sum test, respectively), whereas the category of piecemeal necrosis alone showed great predictive value. Thus, the category of periportal activity should be scored independently of the category of portal inflammation. Studies on fibrosis as a predictive factor have shown that progression of fibrosis before therapy results in poor response to IFN therapy.14-16 In our study, the fibrosis score on the Scheuer classification was the most discriminating between responders and nonresponders, followed by the modified HAI. The HAI was the least discriminating of the 3 systems. Under the HAI system, no distinction was made between P-P and P-C bridging fibrosis, and the scores before IFN therapy were only 1 and 3 and thus were not statistically significant because patients with cirrhosis before treatment were not included in our study. Scoring for fibrosis on the Scheuer classification was simpler than that on the
915
HUMAN PATHOLOGY
Volume 32, No. 9 (September 2001)
FIGURE 3. Biopsy specimens obtained from a complete responder (case 8, Table 1) before and after IFN therapy. (A) Pretreatment biopsy specimen showing P-C bridging fibrosis with moderate lobular inflammation and focal lobular distortion. (B) Posttreatment biopsy specimen showing P-P bridging fibrosis. P-C bridging fibrosis disappeared, and portal inflammation persisted. The modified HAI scoring system and the Scheuer classification reflected well the improvement in bridging fibrosis as decreases in fibrosis scores. (Hematoxylin and eosin; original magnification ⫻100.)
modified HAI system but amply reliable in predicting IFN efficacy, although data on cirrhotic patients could not be inferred. A fibrosis score of 3 or 4 on the Scheuer classification tended to indicate an outcome of nonresponse to IFN therapy. Two cases in the complete responder group showed pretreatment cutoff scores of 4 in the categories reflecting piecemeal necrosis on the 3 systems (cases 2 and 13, Table 3). In case 2, the quantity of HCV RNA before treatment was the smallest of all the patients. In case 13, the total IFN dose administrated was the highest of all the patients. On the other hand, 3 cases in the nonresponder group showed pretreatment scores of 3 in the categories reflecting piecemeal necrosis under the 3 systems (cases 40, 42, and 44, Table 3). In case 42, the quantity of HCV RNA before treatment was the third largest of all patients, and the total IFN dose administrated was relatively low. In case 44, pretreatment fibrosis was advanced. In 2 cases of the complete responder group, the pretreatment staging score of 3 under the Scheuer
classification was predictive of nonresponse (cases 1 and 8, Table 3). Favorable factors, if any, other than staging scores were not demonstrated histopathologically, virally, or by total IFN dose. On the other hand, 3 cases in the nonresponder group showed a pretreatment staging score of 1 under the Scheuer classification (cases 40, 42, and 46, Table 3). In case 46, unfavorable factors other than stage were scores of 4 in the categories reflecting piecemeal necrosis under the 3 systems. Unfavorable factors in case 42 are discussed in the preceding paragraph. We must emphasize that there are many significant variants affecting the response to IFN therapy: viral factors, such as genotype and quantity of HCV RNA; therapeutic factors, such as total dose and duration of IFN administration; and host factors, such as age and duration of infection. However, the role of the histology of biopsy specimens should not be underestimated because it yields important information as part of the host factors when it is assessed by the discriminating features of a reliable system (the aim of this study).
916
SCORING SYSTEMS IN IFN THERAPY FOR HEPATITIS C (Ikawa et al)
In predicting the efficacy of IFN therapy, evaluation of piecemeal necrosis on any system and of the progression of fibrosis on the Scheuer classification was useful. Our study showed that the modified HAI system and the Scheuer classification were both reliable in estimating subtle histologic changes in patients with chronic hepatitis C treated with IFN and that the score reflecting piecemeal necrosis on any system and the fibrosis score on the Scheuer classification were valuable in predicting response to IFN therapy. The Scheuer classification is the simplest of the 3 systems and shows considerably better reproducibility than the HAI, which is simpler than the modified HAI,17 although there are no data on the reproducibility of the Scheuer classification compared with that of the modified HAI in the literature. Thus, we suggest that (1) the Scheuer system, relatively easy to use, is the best for routine clinical use in IFN treatment of patients with chronic hepatitis C, and (2) in clinical studies aimed at evaluating the histology of liver biopsy specimens of hepatitis C patients treated with IFN, a combination of grade on the modified HAI, which is the most sensitive to a subtle improvement of necroinflammation, and stage on the Scheuer classification, which is a little less sensitive for reflecting improvement of fibrosis than the modified HAI but much more predictive for response to IFN therapy, would constitute a superior system.
REFERENCES 1. De Groote J, Desmet VJ, Gedigk P, et al: A classification of chronic hepatitis. Lancet 2:626-628, 1968 2. Knodell RG, Ishak KG, Black WC, et al: Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1:431435, 1981 3. Ishak K, Baptista A, Bianchi L, et al: Histological grading and staging of chronic hepatitis. J Hepatol 22:696-699, 1995
4. Scheuer PJ: Classification of chronic viral hepatitis: A need for reassessment. J Hepatol 13:372-374, 1991 5. Griner PF, Mayewski RJ, Mushlin AI, et al: Selection and interpretation of diagnostic tests and procedures. Principles and applications. Ann Intern Med 94:557-592, 1981 6. Niederau C, Lange S, Heintges T, et al: Prognosis of chronic hepatitis C: Results of a large, prospective cohort study. Hepatology 28:1687-1695, 1998 7. Yoshida H, Shiratori Y, Moriyama M, et al: Interferon therapy reduces the risk for hepatocellular carcinoma: National surveillance program of cirrhotic and noncirrhotic patients with chronic hepatitis C in Japan. IHIT Study Group. Inhibition of Hepatocarcinogenesis by Interferon Therapy. Ann Intern Med 131:174-181, 1999 8. Morisco F, Marmo R, Iasevoli P, et al: Clinical outcome of chronic hepatitis C in patients treated with interferon: Comparison between responders and non-responders. Ital J Gastroenterol Hepatol 31:454-458, 1999 9. Okanoue T, Itoh Y, Minami M, et al: Interferon therapy lowers the rate of progression to hepatocellular carcinoma in chronic hepatitis C but not significantly in an advanced stage: A retrospective study in 1148 patients. Viral Hepatitis Therapy Study Group. J Hepatol 30:653-659, 1999 10. Shindo M, Ken A, Okuno T: Varying incidence of cirrhosis and hepatocellular carcinoma in patients with chronic hepatitis C responding differently to interferon therapy. Cancer 85:1943-1950, 1999 11. Kasahara A, Hayashi N, Mochizuki K, et al: Risk factors for hepatocellular carcinoma and its incidence after interferon treatment in patients with chronic hepatitis C. Osaka Liver Disease Study Group. Hepatology 27:1394-1402, 1998 12. Zeniya M, Toda G: Case selection for interferon treatment of hepatitis C. J Gastroenterol Hepatol 15:117-122, 2000 (suppl E) 13. Cooksley WGE, Bradbear RA, Robinson W, et al: The prognosis of chronic active hepatitis without cirrhosis in relation to bridging necrosis. Hepatology 6:345-348, 1986 14. Jenkins PJ, Cromie SL, Bowden DS, et al: Chronic hepatitis C and interferon alpha therapy: Predictors of long term response. Med J Aust 164:150-152, 1996 15. Morisco F, Marmo R, Iasevoli P, et al: Clinical outcome of chronic hepatitis C in patients treated with interferon: Comparison between responders and non-responders. Ital J Gastroenterol Hepatol 31:454-458, 1999 16. Kim SR, Hayashi Y, Yoon S, et al: Prediction of efficacy of interferon treatment of chronic hepatitis C by multivariate analysis and a new classification. Pathol Int 48:215-220, 1998 17. Goldin RD, Goldin JG, Burt AD, et al: Intra-observer and inter-observer variation in the histopathological assessment of chronic viral hepatitis. J Hepatol 25:649-654, 1996
917