- Email: [email protected]
A model for predicting survival in patients undergoing elective transjugular intrahepatic portosystemic shunts (TIPS)
A1074
AASLD
ABSTRACTS
GASTROENTEROLOGY,
• T H E P R E V A L E N C E O F H E P A ' I I T I S A, B AND C IN A
M E T R O P O L I T A N H O M E L E S S POPULATION. HJ Gonzalezl,, CR Caspers2, MJ Hanson 3 AM Pheley 1 and CJ Peine!. Hermepin County Medical Center 1, Health Care for the Homeless Projectz and Memorial Blood Center 3, Minneapolis, MN. To date, studies delineating the health care needs of homeless individuals in the US are limited, though this population remains at particular risk to acquire diseases associated with significant morbidity and mortality. Exposure tO hepatitis A, B and C, as detected by serological markers, is felt to occur in 30%, 4% and 1%, respectively, of the general population. The prevalence of exposure to hepatitis A, B and C in the homeless population is not known. AIM: To determine the prevalence of Hepatitis A, B and C exposure and assess related risk factors in a metropolitan homeless population. METHODS: Clients of all 8 major homeless shelters and 3 "drop in" centers in Minneapolis participated in this study. Questionnaires were completed to obtain demographic information and assess potential risk factors (age, race, country of origin, living situation, level o f education, past known exposure to hepatitis, sexual contact with infected partners, blood transfusions, hemodialysis, tattoos, HIV exposure, 1V drug use, alcohol use, and sexual history). All participants had a 25 test general chemistry blood screen (including liver enzymes) and hepatitis serologies (antiHAV, HBsAg, anti-FIBs, HCV c-100 ELISA). Statistical analysis was performed using Chi square and Fischer's Exact Test..RESULTS: 381 of an estimated potential 1500 clients participated. Prevalence All Participants (%) Homeless.IV (%)* Hepatitis A 45.4 44.8 Hepatitis B 16.1 10.6 Hepatitis C 24.5 14.5 All Hepatitis 61.6 55.0 *Homeless-iV = those Jentified as homeless an :1not using tV drugs Only 12.7%, 13% and 30.2 % of individuals were aware of their exposure to hepatitis A, B & C , respectively. Independent risk factors associated with hepatitis (p<.05) included: hepatitis A-less than high school education, hepatitis B-tattoos, >5 sexual partners/year, hepatitis C-crack cocaine use, IV drug use, shared needles, shared bottles, and unprotected sex. ~ 1)The prevalence of hepatitis A, B, & C is significantly increased in the homeless population, even when IV drug use is taken into consideration. 2)Most clients were unaware of past exposure and 3) Independent risk factors are identified.
• A M O D E L FOR P R E D I C T I N G SURVIVAL IN PATIENTS UNDERGOING E L E C T I V E TRANSJUGULAR INTRAHEPATIC P O R T O S Y S T E M I C SHUNTS (TIPS). FD Gordon*t, M Malinchoc*. CJ Peine**, J Rank***, PS Kamath*. Mayo Clinic, Rochester. MN*; HCMC**, and U of M***. Minneapolis, MN; Deaconess Hospital, Boston, MA + We studied 167 patients who underwent elective TIPS between Sept. 1991 and June 1994 at 4 medical centers for the purpose of developmg a mathematical model for predicting survival BACKGROUND: Earlier. we demonstrated that emergency TIPS, Child's C, low albumin, high creatinine or bilirubin, prolonged prothrombin time, encephalopathy, and ascites were associated with poor survival following TIPS. METHODS: First, we used a stepwise Cox Proportional hazards regression to select variables independently predictive of survival. Second. using the selected variables, we developed a survival model from the data on 100 Minnesota patients. Third. this model was validated on 67 patients from the Deaconess Hospital, Boston. Finally,~using all patients (N=167), we re-calculated the survival model. RESULTS: Pre-TIPS loge bilirubin (P<0.001) and loge creatinine (P<0.001) were independently predictive of survival. A model using these variables accurately predicted the survival of high, moderate, and low risk Deaconess patients. The risk score, based on the total group of 167 patients. is calculated as follows: Risk score=l.26 loge creatinine mg/dL + 0.57 loge bilirubin rag/d). The predicted median survival in the high risk group (score > 1.3) is 42 days, moderate risk (score 0.8-1.3~ 367 days. and low risk (score < 0.8) well over two years. 1.o
Predicted
"'t
o.o o
HEPATITIS C GENOTYPE lb IN LIVER TRANSPLANT RECIPIENTS IS ASSOCIATED WITH AN INCREASED INCIDENCE AND SEVERrFY OF RECURRENCE. F.D. Gordon. N.N. Zein, J.J. Potemcha, .J.B. Gross, A.A. Gossard, J.L. Steers. R.I-L Wiesner. D.H. Persing. Mayo Clinic, Rochester, MN. Recto'fence of hepatitis C virus (HCV) infection after liver Wansplantation (OLT) is common, The course can vatT from normal to mild chronic inflammation to the rapid development of cirrhosis. It has been suggested that in non-Wansplanted individuals, HCV genotype l b has a worse prognosis than other genotypes. The aim of this study was to assess the influence of HCV genotype on the hist01ogic severity of HCV rectrrence after OLT. MEPHODS: The study gronp comprised41 patients with HCV RNA in the serum who underwent OLT at our institution between 1985 and 1994. HCV RNA was detected by viral RNA extraction and amplification of a genomic sequence obtained from the purified amplification products. Liver biopsies were performed 1, 4-6, and 12 months after OLT and yearly thereaftex or when clinicallyindicated. Histologic severity was graded: 0 (normal liver), 1 (mild portal inflammation), 2 (moderate portal ± lobnlar inflanmmtion), and 3 (ciri-hosis). RESULTS: HCV genotype did not change after OLT in any patient. Eighteen (44%) patients were genotype la, 17 (41.5%) were lb, 3 (7.3%) were 2b, and there was one (2.4%) each of types 2a, 3a, and 4a. The l b and non- lb groups were similar in age (48.4 and 45.7 years, p--0.39) and length of follow-up (48.8 and 45.8 months, p=0.76). T h e mean time to the onset of histologic hepatitis was also similar (10.5 and 11.8 months, p=0.84). Genotype lb Non-lb p value
o.
~- 0.2
Background Hepatitis C virus (HCV) is a single-stranded RNA ~irus which replicates via an RNA-dependent RNA polymerase. This replication strategy has limited fidelity. Accordingly, HCV exists as a highly heterogeneous population of closely related genomes (quasispecies). Hypothesis HCV qnasispecies heterogeneity has clinical and biological implications. Aim To determine the significance of viral quasispecies heterogeneity in chronic HCV infection. Methods Sera from 47 patients with chronic HCV infection were studied [Male:Female 27:20; age 30-72 yrs; HCV RNA+ by RTPCR (5'UTR) 47/47; HCV RNA+ by bDNA (Chiton) 34/47; genotype 1 n=36, 2 n=9, 3 n=2; CPH:CAH:AC 7:18:22; 46/47 were subsequently treated with interferon~]. HCV quasispecies heterogeneity was determined using single-stranded conformational polymorphism (sscP). The HCV E2 hypervariable region 2 (HVR2) was amplified by RT "nested" PCR yielding an amplicon of 196 bp. 25 ng of the amplicons were denatured and electrophoresed on 12.5% polyacrylamidegels using the "Phast"system (200V, 25°C). Individual bands were identified after the gels were developed using a sensitive silver staining protocol. Technical Requirements (t) To avoid detection bias related t o viremia level, a standard quantity of amplified eDNA was loaded for each sample. (2) The primers used were derived from conserved areas but flanking a hypervariable region• (3) The amplicon was <200 bp which allowed for optimum SSCP separation• Results Validation of the SSCPMethod Standardization of the amount of amplicon loaded on the gel was critical for the precise determination of quasispecies heterogeneity. Using clones of known nucleotide sequence, the sensitivity was shown to be 1 ng/individual band (i.e. 4% of the total amplieon). Interference between multiple bands was not demonstrated in mixing experiments with characterized clones. Clinical Relevance HVR2 was amplified in 45/47 (96%) patients. The bands were enumerated under code, using standards (1, 2, and 5 ng/band) for comparison. The median number of bands/sample was 6 (range: 2-12). Increased quasispecies heterogeneity correlated with the estimated duration of HCV infection (r=0.33, p<0~03), transfusion-acqnired HCV infection (vs sporadic HCV, p<0.02), HCV viremia level (r=0.30, p<0.05), and HCV type 1 infection (p<0.02), but not with age, serum AST or ALT levels, histological diagnosis, or Knodell score. Patients who had complete and sustained response to interferon~ (n = 11) had lower pretreatment quasispecies heterogeneity compared to patients who had complete response with relapse (n= 15, p <0.04) or no response (n = 18, p<0.01). Conclusions (1) The estimated duration of HCV carriage, viral replication and HCV type 1 are three determinants for the evolution of HCV quasispeeies heterogeneity. (2) Increased HCV quasispecies heterogeneity portends a poor subsequent response to interferon-a therapy•
Survival
t~e, sk: R
~---_
"-. . . . .
~
_
:
R>1.3 "
2oo
4m
-~'-"-"-
n (%) 17 (41.5) 24 (58.5)
s exact
Histologic Hepatitis 16/17 14/24 I~0.015 *
Severity Score±S.D. 1.52±0.24 0.79±0.17 p<0.015**
Cirrhosis 5/17 1/24 p=0.06*
test
~
6oo
aoo
CONCLUSION: l) A model using loge bilirubin and loge creatiniue pre-. TIPS accurately predicts survival in patients undergoing elective TIPS: 2) High risk patients tscore > 1.3) should not have TIPS unless they are candidates for liver transplantation: and. 3) The model needs to be prospectively validated.
No. 4
CLINICAL IMPLICATIONS OF VIRAL QUASISPECIES HETEROGENEITY IN CHRONIC HEPATITIS C VIRUS INFECTION Gonzalez-Peralta RP, She JY, Qian KP, Davis GL, Ohno T I, Mizokami M 1, Lau JYN. Section of Hepat~bifiaryDiseases, University Of Florida, GainesviUc,FL; Nagoya City University1, Nagoya, Japan.
Observea Survival ......
-- O.B
g
•
Vol. 108,
The mean I~ne to the onset of cirrhosis in the type l b cirrhotic patients was 36.4 ± 17.2 months. The single non-lb cirrhotic patient developed cirrhosis 103 months after OLT. CONCLUSIONS: 1) HCV genotype l b is more commonly associated with histulogic recurrence of HCV than other genotypes. 2) Genotype l b is associated With a more severe histologic recto'fence of HCV. 3) Nearly all cases (83%) of post-OLT HCVassociated cirrhosis were associated with gonotype lb.
AASLD
ABSTRACTS
GASTROENTEROLOGY,
• T H E P R E V A L E N C E O F H E P A ' I I T I S A, B AND C IN A
M E T R O P O L I T A N H O M E L E S S POPULATION. HJ Gonzalezl,, CR Caspers2, MJ Hanson 3 AM Pheley 1 and CJ Peine!. Hermepin County Medical Center 1, Health Care for the Homeless Projectz and Memorial Blood Center 3, Minneapolis, MN. To date, studies delineating the health care needs of homeless individuals in the US are limited, though this population remains at particular risk to acquire diseases associated with significant morbidity and mortality. Exposure tO hepatitis A, B and C, as detected by serological markers, is felt to occur in 30%, 4% and 1%, respectively, of the general population. The prevalence of exposure to hepatitis A, B and C in the homeless population is not known. AIM: To determine the prevalence of Hepatitis A, B and C exposure and assess related risk factors in a metropolitan homeless population. METHODS: Clients of all 8 major homeless shelters and 3 "drop in" centers in Minneapolis participated in this study. Questionnaires were completed to obtain demographic information and assess potential risk factors (age, race, country of origin, living situation, level o f education, past known exposure to hepatitis, sexual contact with infected partners, blood transfusions, hemodialysis, tattoos, HIV exposure, 1V drug use, alcohol use, and sexual history). All participants had a 25 test general chemistry blood screen (including liver enzymes) and hepatitis serologies (antiHAV, HBsAg, anti-FIBs, HCV c-100 ELISA). Statistical analysis was performed using Chi square and Fischer's Exact Test..RESULTS: 381 of an estimated potential 1500 clients participated. Prevalence All Participants (%) Homeless.IV (%)* Hepatitis A 45.4 44.8 Hepatitis B 16.1 10.6 Hepatitis C 24.5 14.5 All Hepatitis 61.6 55.0 *Homeless-iV = those Jentified as homeless an :1not using tV drugs Only 12.7%, 13% and 30.2 % of individuals were aware of their exposure to hepatitis A, B & C , respectively. Independent risk factors associated with hepatitis (p<.05) included: hepatitis A-less than high school education, hepatitis B-tattoos, >5 sexual partners/year, hepatitis C-crack cocaine use, IV drug use, shared needles, shared bottles, and unprotected sex. ~ 1)The prevalence of hepatitis A, B, & C is significantly increased in the homeless population, even when IV drug use is taken into consideration. 2)Most clients were unaware of past exposure and 3) Independent risk factors are identified.
• A M O D E L FOR P R E D I C T I N G SURVIVAL IN PATIENTS UNDERGOING E L E C T I V E TRANSJUGULAR INTRAHEPATIC P O R T O S Y S T E M I C SHUNTS (TIPS). FD Gordon*t, M Malinchoc*. CJ Peine**, J Rank***, PS Kamath*. Mayo Clinic, Rochester. MN*; HCMC**, and U of M***. Minneapolis, MN; Deaconess Hospital, Boston, MA + We studied 167 patients who underwent elective TIPS between Sept. 1991 and June 1994 at 4 medical centers for the purpose of developmg a mathematical model for predicting survival BACKGROUND: Earlier. we demonstrated that emergency TIPS, Child's C, low albumin, high creatinine or bilirubin, prolonged prothrombin time, encephalopathy, and ascites were associated with poor survival following TIPS. METHODS: First, we used a stepwise Cox Proportional hazards regression to select variables independently predictive of survival. Second. using the selected variables, we developed a survival model from the data on 100 Minnesota patients. Third. this model was validated on 67 patients from the Deaconess Hospital, Boston. Finally,~using all patients (N=167), we re-calculated the survival model. RESULTS: Pre-TIPS loge bilirubin (P<0.001) and loge creatinine (P<0.001) were independently predictive of survival. A model using these variables accurately predicted the survival of high, moderate, and low risk Deaconess patients. The risk score, based on the total group of 167 patients. is calculated as follows: Risk score=l.26 loge creatinine mg/dL + 0.57 loge bilirubin rag/d). The predicted median survival in the high risk group (score > 1.3) is 42 days, moderate risk (score 0.8-1.3~ 367 days. and low risk (score < 0.8) well over two years. 1.o
Predicted
"'t
o.o o
HEPATITIS C GENOTYPE lb IN LIVER TRANSPLANT RECIPIENTS IS ASSOCIATED WITH AN INCREASED INCIDENCE AND SEVERrFY OF RECURRENCE. F.D. Gordon. N.N. Zein, J.J. Potemcha, .J.B. Gross, A.A. Gossard, J.L. Steers. R.I-L Wiesner. D.H. Persing. Mayo Clinic, Rochester, MN. Recto'fence of hepatitis C virus (HCV) infection after liver Wansplantation (OLT) is common, The course can vatT from normal to mild chronic inflammation to the rapid development of cirrhosis. It has been suggested that in non-Wansplanted individuals, HCV genotype l b has a worse prognosis than other genotypes. The aim of this study was to assess the influence of HCV genotype on the hist01ogic severity of HCV rectrrence after OLT. MEPHODS: The study gronp comprised41 patients with HCV RNA in the serum who underwent OLT at our institution between 1985 and 1994. HCV RNA was detected by viral RNA extraction and amplification of a genomic sequence obtained from the purified amplification products. Liver biopsies were performed 1, 4-6, and 12 months after OLT and yearly thereaftex or when clinicallyindicated. Histologic severity was graded: 0 (normal liver), 1 (mild portal inflammation), 2 (moderate portal ± lobnlar inflanmmtion), and 3 (ciri-hosis). RESULTS: HCV genotype did not change after OLT in any patient. Eighteen (44%) patients were genotype la, 17 (41.5%) were lb, 3 (7.3%) were 2b, and there was one (2.4%) each of types 2a, 3a, and 4a. The l b and non- lb groups were similar in age (48.4 and 45.7 years, p--0.39) and length of follow-up (48.8 and 45.8 months, p=0.76). T h e mean time to the onset of histologic hepatitis was also similar (10.5 and 11.8 months, p=0.84). Genotype lb Non-lb p value
o.
~- 0.2
Background Hepatitis C virus (HCV) is a single-stranded RNA ~irus which replicates via an RNA-dependent RNA polymerase. This replication strategy has limited fidelity. Accordingly, HCV exists as a highly heterogeneous population of closely related genomes (quasispecies). Hypothesis HCV qnasispecies heterogeneity has clinical and biological implications. Aim To determine the significance of viral quasispecies heterogeneity in chronic HCV infection. Methods Sera from 47 patients with chronic HCV infection were studied [Male:Female 27:20; age 30-72 yrs; HCV RNA+ by RTPCR (5'UTR) 47/47; HCV RNA+ by bDNA (Chiton) 34/47; genotype 1 n=36, 2 n=9, 3 n=2; CPH:CAH:AC 7:18:22; 46/47 were subsequently treated with interferon~]. HCV quasispecies heterogeneity was determined using single-stranded conformational polymorphism (sscP). The HCV E2 hypervariable region 2 (HVR2) was amplified by RT "nested" PCR yielding an amplicon of 196 bp. 25 ng of the amplicons were denatured and electrophoresed on 12.5% polyacrylamidegels using the "Phast"system (200V, 25°C). Individual bands were identified after the gels were developed using a sensitive silver staining protocol. Technical Requirements (t) To avoid detection bias related t o viremia level, a standard quantity of amplified eDNA was loaded for each sample. (2) The primers used were derived from conserved areas but flanking a hypervariable region• (3) The amplicon was <200 bp which allowed for optimum SSCP separation• Results Validation of the SSCPMethod Standardization of the amount of amplicon loaded on the gel was critical for the precise determination of quasispecies heterogeneity. Using clones of known nucleotide sequence, the sensitivity was shown to be 1 ng/individual band (i.e. 4% of the total amplieon). Interference between multiple bands was not demonstrated in mixing experiments with characterized clones. Clinical Relevance HVR2 was amplified in 45/47 (96%) patients. The bands were enumerated under code, using standards (1, 2, and 5 ng/band) for comparison. The median number of bands/sample was 6 (range: 2-12). Increased quasispecies heterogeneity correlated with the estimated duration of HCV infection (r=0.33, p<0~03), transfusion-acqnired HCV infection (vs sporadic HCV, p<0.02), HCV viremia level (r=0.30, p<0.05), and HCV type 1 infection (p<0.02), but not with age, serum AST or ALT levels, histological diagnosis, or Knodell score. Patients who had complete and sustained response to interferon~ (n = 11) had lower pretreatment quasispecies heterogeneity compared to patients who had complete response with relapse (n= 15, p <0.04) or no response (n = 18, p<0.01). Conclusions (1) The estimated duration of HCV carriage, viral replication and HCV type 1 are three determinants for the evolution of HCV quasispeeies heterogeneity. (2) Increased HCV quasispecies heterogeneity portends a poor subsequent response to interferon-a therapy•
Survival
t~e, sk: R
~---_
"-. . . . .
~
_
:
R>1.3 "
2oo
4m
-~'-"-"-
n (%) 17 (41.5) 24 (58.5)
s exact
Histologic Hepatitis 16/17 14/24 I~0.015 *
Severity Score±S.D. 1.52±0.24 0.79±0.17 p<0.015**
Cirrhosis 5/17 1/24 p=0.06*
test
~
6oo
aoo
CONCLUSION: l) A model using loge bilirubin and loge creatiniue pre-. TIPS accurately predicts survival in patients undergoing elective TIPS: 2) High risk patients tscore > 1.3) should not have TIPS unless they are candidates for liver transplantation: and. 3) The model needs to be prospectively validated.
No. 4
CLINICAL IMPLICATIONS OF VIRAL QUASISPECIES HETEROGENEITY IN CHRONIC HEPATITIS C VIRUS INFECTION Gonzalez-Peralta RP, She JY, Qian KP, Davis GL, Ohno T I, Mizokami M 1, Lau JYN. Section of Hepat~bifiaryDiseases, University Of Florida, GainesviUc,FL; Nagoya City University1, Nagoya, Japan.
Observea Survival ......
-- O.B
g
•
Vol. 108,
The mean I~ne to the onset of cirrhosis in the type l b cirrhotic patients was 36.4 ± 17.2 months. The single non-lb cirrhotic patient developed cirrhosis 103 months after OLT. CONCLUSIONS: 1) HCV genotype l b is more commonly associated with histulogic recurrence of HCV than other genotypes. 2) Genotype l b is associated With a more severe histologic recto'fence of HCV. 3) Nearly all cases (83%) of post-OLT HCVassociated cirrhosis were associated with gonotype lb.