Hispanics Have More Aggressive Liver Disease in LA County Hepatitis Clinic
â?oMore Advanced Hepatic Fibrosis in Hispanics with Chronic Hepatitis C Infection: Role of Patient Demographics, Hepatic Necroinflammation, and Steatosisâ?
The American Journal of Gastroenterology
Volume 101 Page 1817 - August 2006
Sumita Verma, M.D., M.R.C.P.1, Maurizio Bonacini, M.D.2, Sugantha Govindarajan, M.D.3, Gary Kanel, M.D.4, Karen L. Lindsay, M.D.1, and Allan Redeker, M.D.3
1Division of Gastrointestinal and Liver Diseases, Hepatitis Research and Treatment Center, Los Angeles, California; 2Department of Transplantation, California Pacific Medical Center, San Francisco, California; 3Rancho Los Amigos Medical Center, Downey, California; and 4Department of Pathology, University of Southern California, Los Angeles, California
â?oâ?¦.From June 1994 to June 2004, consecutive patients with CHC were retrospectively recruited from the Los Angeles county hepatitis clinic by reviewing patient charts and personal summary records maintained by the senior authorâ?¦.Because of the retrospective nature and lack of serial biopsies this study had limitations. Also, the Hispanics studied were predominantly from Mexico and from the lower end of the socioeconomic spectrum. Thus, they may not have been representative of a cross section of Hispanicsâ?¦..This study confirms that Hispanic ethnicity is associated with a more aggressive course of HCV infection with more than one-third having cirrhosis at the index biopsy. Compared with NHW, Hispanic patients also tended to show more NI, have a faster rate of fibrosis progression, be older at time of exposure (20 vs 23 yrs) and biopsy, and have a higher prevalence of BT as a risk factor. Furthermore, almost 50% of the Hispanics with CHC were obese with 79% having histological evidence of hepatic steatosis. Independent predictors of fibrosis stage â?¥4 were presence of DM, grade 1-2 hepatic steatosis, AST/ALT >1, age at biopsy, NI grade, and serum bilirubinâ?¦.â?
BACKGROUND AND AIMS: We sought to assess whether Hispanics have more advanced hepatitis C virus (HCV)-related liver disease than non-Hispanic whites (NHW) and to identify contributory factors.
PATIENTS AND METHODS: Patients were recruited from the Los Angeles county hepatitis clinic. Liver fibrosis and necroinflammation (NI) were assessed by the Ishak scoring system. Hepatic steatosis was graded as 0-4.
A total of 232 patients were evaluated, 63 NHW and 169 Hispanic.
Hispanics were older and had a higher prevalence of blood transfusion (40% vs 21%), obesity (body mass index > 30) (47% vs 21%), diabetes mellitus (DM) (16% vs 5%), and hepatic steatosis (79% vs 47%), p< 0.02.
Independent predictors of hepatic steatosis were
-- Hispanic ethnicity (odds ratio [OR] 3.8, 95% CI 1.7-8.7, p = 0.001) and
-- obesity (OR 5.7, 95% CI 2.3-14.1, p = 0.0002).
Compared with NHW, Hispanics also had
-- higher fibrosis stage (3.3 Â± 2 vs 2.3 Â± 6.9, p = 0.001),
-- NI grade (6.4 Â± 1.8 vs 5.6 Â± 1.6, p = 0.002), and
-- faster fibrosis progression/yr (0.14 Â± 0.09 vs 0.09 Â± 0.07, p = 0.0002).
-- DM (OR 2.9, p = 0.02),
-- grade 1-2 hepatic steatosis (OR 2.3, p = 0.03), AST/ALT > 1 (OR 4.3, p = 0.01),
-- NI grade (OR 1.7, p< 0.0001),
-- age at biopsy (OR 1.1, p< 0.0001), and
-- serum bilirubin (OR 5.4, p< 0.0001)
were independent predictors of fibrosis stage â?¥4.
CONCLUSION: This study confirms that Hispanics have more advanced hepatic fibrosis than NHW. This is related to older age, higher NI grade, and greater prevalence of hepatic steatosis and DM.
Liver disease from chronic hepatitis C virus (CHC) has now emerged as an important etiology of cirrhosis, and hepatocellular cancer and also the most frequent indication for liver transplantation in the United States. According to the National Health and Nutrition Examination Survey, there are approximately 2.7 million Americans with chronic HCV infection (1). Race/ethnicity plays an important role in modifying the natural history of CHC and a number of studies have observed attenuated HCV disease in African Americans compared with non-Hispanic whites (NHW) (2, 3).
Hispanics comprise 13% of the U.S. population and are the fastest growing minority group (4). They are also the predominant ethnic group in Los Angeles county (44%) (5). Though chronic liver disease and cirrhosis was ranked as the eighth leading cause of death in Hispanics in the United States in 1998, it was not among the 10 leading causes of death in blacks or NHW. In fact for both genders, chronic liver disease was the third most common cause of death in Hispanics in the 45-64 age groups (6). Despite this, not only are there very few published data available on liver fibrosis in Hispanics with HCV infection, but available data are also conflicting. In an earlier study from our Unit we did observe faster fibrosis progression in Hispanics compared with NHW, though the sample size was small (7). A more recent study from Texas published in abstract form also observed higher prevalence of cirrhosis in Hispanics with CHC compared with NHW (8). Nonetheless, factors associated with this accelerated course in Hispanics were not analyzed in either of these two studies. In American veterans, Cheung et al. were unable to corroborate these findings and reported similar histology in Hispanics and NHW with CHC (9). We therefore designed this study to confirm whether Hispanics with chronic HCV have more advanced hepatic fibrosis compared with NHW and to identify contributing factors, with special emphasis on patient demographics, hepatic necroinflammation (NI), and hepatic steatosis.
From 1994 to 2004 there were 296 patients who were HCV RNA-positive (qualitative) by PCR, and had undergone a liver biopsy with an adequate sample (in the opinion of the reporting pathologist) available for histological analysis. All the liver biopsies had been performed as soon as possible after initial clinical evaluation. None of the patients had received interferon-based therapy prior to the biopsy. The overall patient profile (Hispanics (57%), NHW (21%), blacks (10%), Asians (9%) was consistent with the two earlier studies from our Unit (7, 10), indicating that there was no bias when selecting patients for a liver biopsy. For the purposes of the current study we included only 232 subjects (NHW, N = 63 and Hispanics, N = 169) (Table 1). The remainder (N = 64) comprising predominantly Asian and blacks were excluded. None of the 232 patients had evidence of iron overload histologically and all tested hepatitis B surface antigen (HBsAg) negative.
Table 1. Demographic and Histological Data in the Whole Group (N = 232)
Table 2 shows data in the NHW and Hispanics. Hispanics were older at HCV exposure and biopsy. They also tended to have higher ALT though the differences did not achieve statistical significance (p = 0.2). The percentage of patients with normal ALT was, however, similar among the two ethnic groups. There were no differences in serum bilirubin, albumin, and INR (data not shown). Several differences were present comparing men to women. The prevalence of obesity (BMI > 30) was greater in both NHW females (44% vs 8%, p = 0.004) and Hispanic females (54% vs 39%, p = 0.07). The prevalence of alcohol abuse was lower in women, both NHW (21% vs 57%, p = 0.01) and Hispanic (12% vs 51%, p< 0.0001) compared with men. Figure 2 shows the gender-related differences in risk factors for HCV. NHW and Hispanic males were more likely to report IDU compared with their female counterparts (86% vs 37% and 80% vs 29%, pâ?¤ 0.001, respectively). On the other hand, both NHW and Hispanic females reported BT more frequently as compared with the males (37% vs 14%, p = 0.05 and 64% vs 16%, p< 0.0001, respectively).
Table 2. Demographic and Histological Data in the Hispanics and Non-Hispanic Whites
IDU = injection drug use; BT = blood transfusion.
*BMI and HCV viral load were unavailable in 27 and 60 patients, respectively.
Among the remaining 14 patients, risk factors included high-risk sexual behavior, body piercing, tattoos, cocaine snorting, and prior surgery.
Complete data on hepatic steatosis (i.e., presence/absence and grade) were available for 171 patients (NHW N = 45 and Hispanics N = 126). Hispanics had a significantly higher mean steatosis grade (1.2 Â± 0.9) than the NHW (0.6 Â± 0.8), p = 0.0001. The prevalence of any steatosis was 47% (NHW) and 79% (Hispanics), p< 0.0001 (Fig. 3). Hispanics were also more likely to have steatosis grade 1-2 compared with NHW (p = 0.0005). The prevalence of patients with grade 3-4 steatosis was also highest among the Hispanics (8%) compared with NHW (4%), p = ns (Fig. 3). On univariate analysis, the presence of any hepatic steatosis was associated with Hispanic ethnicity (79% vs 47%, p< 0.0001), female gender (81% vs 62%, p = 0.005), obesity (BMI > 30) (90% vs 56%, p< 0.0001), and presence of DM (20% vs 4%, p = 0.009). Those with any grade of hepatic steatosis had higher fibrosis stage (3.1 Â± 1.9 vs 2.3 Â± 1.8, p = 0.02), faster fibrosis progression (0.13 Â± 0.08/yr vs 0.10 Â± 0.07/yr, p = 0.04), higher alkaline phosphatase (120 Â± 46.7 IU/L vs 97 Â± 29 IU/L, p = 0.0007), and a lower serum albumin (3.9 Â± 0.4 g/dL vs 4.1 Â± 0.4 g/dL, p = 0.002). Those with any steatosis also had higher ALT (153 Â± 104.3 IU/L vs 132 Â± 82.2 IU/L) and grade of NI (6.3 Â± 1.9 vs 5.9 Â± 1.7) though the differences did not achieve statistical significance. We did not analyze association between hepatic steatosis and genotype 3 because only a small number (N = 15) had available data. A stepwise multiple logistic regression analysis was then performed including all variables that were significant on univariate analysis (p< 0.05). The following were identified to be independent predictors of hepatic steatosis: Hispanic ethnicity (odds ratio [OR] 3.8, 95% CI 1.7-8.7, p = 0.001) and obesity (BMI > 30) (OR 5.7, 95% CI 2.3-14.1, p = 0.0002).
Hispanics had the highest NI grades and also had the most advanced disease as regards fibrosis stage and fibrosis progression (Table 2). After adjusting fibrosis progression for age at biopsy, presence of DM, and alcohol abuse, it was still faster in the Hispanics (0.12/yr [95% CI 0.07-0.09]) versus NHW (0.08/yr [95% CI 0.11-0.14]), p< 0.0001. Table 3 shows the distribution of fibrosis stage (0-6) and grade of NI (0-18) in the three groups. There was a trend for higher prevalence of cirrhosis in Hispanics compared with NHW (p = 0.06) and 50% of the Hispanics had NI grades >6 (Table 3). There were no gender-related differences in NI grades, fibrosis stage, fibrosis progression, and cirrhosis in the two groups (data not shown).
Table 4 highlights the data in patients depending on those that had a fibrosis stage < or â?¥4 at index biopsy. Factors that were significantly associated with fibrosis stage â?¥4 on univariate analysis (pâ?¤ 0.05) were then entered into a stepwise multiple logistic regression analysis model. A separate regression model was used for the nominal and continuous variables. Independent predictors of fibrosis stage â?¥4 were age at biopsy, serum bilirubin, NI grade, presence of DM, AST/ALT >1, and presence of grade 1-2 steatosis (Table 5). Obesity (BMI > 30) was, however, not independently associated with advanced fibrosis (OR 0.9, 95% CI 0.4-1.9, p = 0.8). Because both NI grade and hepatic steatosis were independent predictors of advanced fibrosis, we analyzed probability of developing cirrhosis in those with both NI grade â?¥6 (mean grade in the whole group) and hepatic steatosis versus those with either one or neither of these factors. After 30 yr of HCV exposure, the former had a significantly higher probability of developing cirrhosis (38% vs 16%, p = 0.01). It would have been useful to analyze NI and steatosis separately, but after excluding overlapping patients and those with missing data, the sample sizes generated were too small for a meaningful statistical analysis.
Table 5. Independent Predictors of Fibrosis Stage â?¥4 on Multivariate Logistic Regression Analysis
This study confirms that Hispanic ethnicity is associated with a more aggressive course of HCV infection with more than one-third having cirrhosis at the index biopsy. Compared with NHW, Hispanic patients also tended to show more NI, have a faster rate of fibrosis progression, be older at time of exposure and biopsy, and have a higher prevalence of BT as a risk factor. Furthermore, almost 50% of the Hispanics with CHC were obese with 79% having histological evidence of hepatic steatosis. Independent predictors of fibrosis stage â?¥4 were presence of DM, grade 1-2 hepatic steatosis, AST/ALT >1, age at biopsy, NI grade, and serum bilirubin.
Our findings of more advanced hepatic fibrosis in Hispanics with CHC compared with NHW is in variance with the recent study by Cheung et al. (9). They observed similar NI grades, fibrosis stage, and fibrosis progression in the two groups despite the former having a significantly higher prevalence of HIV coinfection (20.4% vs 3.9%). There could be a number of reasons for these differences. First, only 30% of the cohort reported by Cheung et al. underwent a liver biopsy and it was unclear as to whether they were representative of the study population. Second, in contrast with Cheung's study, our Hispanics were older and had higher BMI than the NHW. Third, because of the multicenter nature of Cheung et al.'s study, the histology was interpreted by local pathologists and therefore subject to interobserver and tissue sample variability. Ours was a single-center study, only those with adequate tissue were included, and all the biopsies had been examined by two experienced hepatopathologists (SG, GK). Most importantly, the study by Cheung described American veterans, who were therefore predominantly men, and presumably healthy enough to serve in the Armed Forces, introducing a potential bias. Ours on the other hand, was a community-based study that was conducted in a publicly funded health-care facility in which the patient population was highly representative of Hispanics in Los Angeles county (26.2% of adults aged 18-64 is uninsured, and 40% of the uninsured is Hispanic) (13).
Presence of both grade 1-2 steatosis and NI grade was independent predictor of advanced fibrosis and those who had both NI grade â?¥6 and hepatic steatosis were twice as likely to develop cirrhosis (36% vs 18%, p = 0.01) compared with those with either one or neither of these features. These observations are not surprising as both hepatic steatosis and NI have been associated with fibrosis in earlier studies (14-16). Fartoux et al. showed that regardless of genotype, steatosis was a major determinant in fibrosis progression in patients with mild hepatitis C (14). Castera et al. further observed that worsening of hepatic steatosis was an independent predictor of fibrosis progression in untreated patients with CHC (15). The overall prevalence of obesity in our cohort was high (40% [82 of 205]) with 84% (173 of 205) being overweight (BMI > 25). Hispanics had the highest prevalence (47%) of obesity, this being especially true for the Hispanic females (54%). Therefore, not surprisingly hepatic steatosis was observed in 79% of the Hispanics, with almost 30% having â?¥grade 2 steatosis. Others have also reported a higher prevalence of nonalcoholic fatty liver disease in Hispanics than other ethnic groups (17) and therefore the advanced hepatic fibrosis in Hispanics with CHC may be related to this factor rather than to a direct viral effect. However, we did observe that those with advanced fibrosis were less likely to have grade 3-4 steatosis (Table 4). This may be as a result of reduction or disappearance of steatosis as fibrosis progresses (18).
HCV infection is associated with increased production of inflammatory cytokines (16) and this process may be accentuated in Hispanics, resulting in more severe hepatic NI. We and others (19) have in fact observed higher NI grades in Hispanics compared with NHW. NI can worsen hepatic fibrosis as stellate cells are activated around NI lesions (20). In a study of 70 patients who had had a mean of 3.9 biopsies over an 8-yr period, Yano et al. found that cirrhosis developed within 17 yr in all patients with a high grade (>5/8) of NI on initial biopsy, whereas only 30% of those with low-level inflammation (<3.4/8) developed cirrhosis after 13 yr (16). Furthermore, there may also be an association between NI and steatosis (21). This could be explained by the fact that the HCV core protein leads to oxidative stress (22) and steatosis could result from the inflammatory changes associated with this oxidative stress (23). In this study, hepatic steatosis was associated with a higher grade of NI (6.3 vs 5.9), though the results did not achieve statistical significance.
Because of the retrospective nature and lack of serial biopsies this study had limitations. Also, the Hispanics studied were predominantly from Mexico and from the lower end of the socioeconomic spectrum. Thus, they may not have been representative of a cross section of Hispanics. Finally, the female preponderance among the Hispanics could have reflected better compliance in this group, thereby introducing a selection bias. However, this probably would not have influenced the results as no gender-related differences were observed in the histological parameters.
In conclusion, Hispanics with CHC have more advanced hepatic fibrosis compared with NHW and this appears to be related largely due to host factors such as older age and higher prevalence of obesity (47%), DM (16%), and hepatic steatosis (79%). Hispanics also have higher NI grades and whether this is also related to host characteristics or a direct viral effect is unclear. Because age at biopsy, NI grade, presence of DM, and grade 1-2 steatosis were independent predictors of advanced fibrosis in this study, it is imperative that these variables be considered when designing a treatment and management regimen for Hispanic patients with histologically mild HCV disease. Nonetheless, this important and controversial issue can only be unequivocally resolved by prospective studies recruiting a well-represented Hispanic population.
Patients & Methods
From June 1994 to June 2004, consecutive patients with CHC were retrospectively recruited from the Los Angeles county hepatitis clinic by reviewing patient charts and personal summary records maintained by the senior author (AR). There have already been two studies from our Unit assessing the demographic and biochemical differences in Hispanics with HCV infection (7, 10). Hence the main focus of this study was to assess the histological severity of CHC in this ethnic group. Only those subjects who tested HCV RNA-positive (qualitative) by PCR and had had a liver biopsy were therefore included. In this clinic, at the first visit a detailed history is taken from every patient and relevant information is recorded in the clinic summary card and patient charts including age, ethnic group, alcohol consumption, risk factors, body mass index (BMI), and presence of diabetes mellitus (DM). Pertinent labs and radiographic data are recorded at entry and at each subsequent visit. Liver biopsy results are also accordingly noted. All the liver biopsies had been assessed by either one of the two hepatopathologists (SG, GK) and the NI grade (0-18) and stage of fibrosis (0-6) were obtained using Ishak scoring system (11). Hepatic steatosis was graded depending on the percentage of hepatocytes containing fat: 0 = no fat, 1+ = <5%, 2+ = 6-33%, 3+ = 34-66%, 4+ = 67-100%. Those in whom no grade of steatosis was available but had "scanty/mild steatosis" were presumed to have 1+ steatosis. Similarly those with "mild-moderate/moderate steatosis" were graded as 2+, "moderate-severe" 3+, and severe as 4+. The estimated fibrosis progression was calculated by dividing the fibrosis stage at index biopsy with the estimated disease duration (12). HCV disease duration (in yr) was calculated as the interval from age at exposure to age at liver biopsy. Age at exposure was defined as the earliest year of (a) blood transfusion (BT), (b) initiating injection drug use (IDU), (c) surgical procedure, (d) other high-risk behavior (body piercing, tattoos).
HCV antibody status was determined by HCV-EIA 2.0 (Abbott Laboratories, Chicago, IL). HCV RNA analysis (qualitative) was performed by nested PCR (in-house method) with lower limit of 10 copies. Quantitative HCV analysis was performed using COBAS Amplicor HCV Monitor v2.0 (Roche Molecular Systems, Pleasanton, CA) from October 1999 and subsequently using TaqMan HCV Test (Focus Laboratories, Cypress, CA). Genotyping for HCV was performed by Line-probe assay (INNO-LiPA HCV II, Immunogenetics, Ghent, Belgium).
Ethnic status was classified (according to initial clinic self-classification) as NHW or Hispanics. Blacks, Asians, and patients with other racial/ethnic backgrounds were excluded from the study. Because it is difficult to accurately quantify alcohol consumption retrospectively, we only recorded presence or absence of alcohol abuse rather than the actual amount of alcohol consumed. For the purpose of this study alcohol abuse was defined as consumption of â?¥40 g of alcohol per day. Study exclusion criteria were unidentifiable risk factors, presence of liver disease of another etiology (active hepatitis B infection, iron overload, autoimmune hepatitis), HCV RNA-negative by PCR, liver biopsy unavailable, coinfection with HIV, and prior treatment with interferon-based therapy.
Data are presented as mean Â± SD except the HCV viral load which is presented as median (range). Differences between continuous and nominal variables were calculated using the Mann-Whitney and Ï?2 test/Fisher's exact test, respectively. Multiple linear regression analysis was used to adjust fibrosis progression for age at biopsy, presence of DM, and alcohol abuse. Stepwise multiple logistic regression analysis was performed to determine independent predictors of fibrosis stage â?¥4 and hepatic steatosis. Kaplan-Meier survival analysis was used to assess probability of developing cirrhosis depending on NI grade (â?¥6 vs<6) and presence/absence of hepatic steatosis and differences between the groups were compared using log rank test (Fig. 1). The statistical software programs used were Stat View (version 5.1) and SPSS, versions 8.2 and 13.
NATAP nataphcv mailing list -- nataphcv@...
This is an annoucement-only mailing list. Do not reply.
To unsubscribe: send a blank email to nataphcv-request@...
with a subject of unsubscribe.
For more information, see http://seven.pairlist.net/mailman/listinfo/nataphcv
[Non-text portions of this message have been removed]