Drug Induced Liver Injury Network (DILIN) Prospective Study: Initial Results
Naga Chalasani, MD for the Drug Induced Liver Injury Network
Background
Although DILI is a rare clinical event, it is the most common cause of acute liver failure in the US DILI is one of the most common causes for medication withdrawal or lack of approval by the FDA A significant barrier for drug development
Background
Its epidemiology, etiopathogenesis, diagnosis, and natural history are poorly understood DILIN is a federally funded consortium of 5 clinical centers and DCC
DILIN
Duke Clinical Research Data Coord. Ctr. Institute (PI – Rochon) U. Conn (PI - Bonkovsky) U. Mich (PI - Fontana) Clinical Centers IU (PI- Chalasani) UNC (PI- Watkins) UCSF (PI-Davern) NIDDK (Serrano, Seeff, Hoofnagle)
DILIN: Scope of Recruitment
U Mich UCSF U Conn
IU
UNC
~12.8 million lives
Aim
To report the initial findings of the etiology and clinical characteristics of patients with drug induced liver injury (DILI) enrolled into the “DILIN Prospective Study”
DILIN Prospective Study
Multicenter, prospective, and observational study Patients with suspected idiosyncratic DILI with no competing etiology occurring within 6-months are eligible DILI due to acetaminophen were not included
Eligibility Criteria
Children ≥ 2 years and adults Pre-defined biochemical criteria - AST or ALT > 5 ULN twice consecutively - Alk Phos > 2 ULN twice consecutively - Bilirubin ≥ 2.5 mg/dl - Specific criteria for patients with preexisting liver disease
Methods
Patients were seen in the GCRC Patients followed up to 24 months depending on their biochemistries at 6 months after enrollment Extensive baseline evaluations to exclude competing etiologies Samples of serum, urine, DNA, and peripheral lymphocytes
Causality Assessment
A panel of study hepatologists systematically assessed the strength of causal relationship between implicated agent and liver injury RUCAM, DILIN-specific causality grading, data completeness forms are filled by 3 hepatologists Causality finally assessed based on consensus
Chronic DILI
Persistent biochemical abnormalities at 6 months following the onset of acute DILI Histological or radiological or clinical evidence of chronic liver disease at 6 months Patients with known HCV or HBV or cirrhosis were not assessed for chronic DILI
Total Enrollment (9/04 – 8/06)
Total U Conn Indiana UCSF Michigan UNC Total 39 54 35 52 53 233
Although 233 patients were enrolled, data from 169 patients were available for this presentation
Demographics
Age (mean ± S.D.) Females African-Americans BMI (Kg/m2, mean ± S.D.) Children Pre-existing liver disease 48.3 ± 18.6 60% 13% 27 ± 7 9% 16%
Pattern of Liver Injury
Percent
Hepatocellular Cholestatic Mixed
53% 22% 25%
Implicated DILI Drugs / CAM
% Single Rx Drug 74
Single CAM Multiple Drugs / CAM
4
22
Drug Class for Single Implicated Agent
Drug Class
Antimicrobials Anticonvulsants Antineoplastics Antidepressants NSAIDs
%
40 7 4 4 4
Drug Class
%
Gen.anesthestics
Choles-lowering Peptide-biologics
3
5 5
CAM Products
Others
6
22
Implicated Antimicrobials – Single agent DILI
Bactrim (9) Augmentin (17) Macrobid (14) Ketek (5) Levaquin (5) Anti-retroviral (9) Anti TB (12)
Signs and Symptoms
%
Jaundice 66 Vomiting Rash Itching % 36 27 49 23
Nausea Anorexia Dark Urine
Fever Abdominal Pain
59 51 68
31 48
Δ Mental Status Any symptoms
96
Signs and Symptoms
%
Hepatomegaly 8
Lymphadenopathy Extrahepatic Manifestations
Liver Biopsy:
2
10
50
Steroid treatment
18
Biochemistries (mean ± s.d.)
Onset AST (IU/L) ALT (IU/L)
Alk Pho (IU/L) Bilirubin (mg/dl)
Peak 962 ± 2116 907 ± 980
407 ± 412 10.5 ± 10
652 ± 831 722 ± 835
287 ± 256 5.3 ± 5.5
Final Causality Assessment (n=68)
Percent
Definite
Very Likely
28%
43%
Probable
Possible
13%
15%
Unlikely
1.5%
Outcome Reported to the DCC Thus Far
Death (within 6 months)
Liver Transplant (event related up to 6 months) Chronic DILI
12.7%
2%
20%
Conclusions (1)
Our prospective study represents a systematic effort to recruit patients with clinically important DILI in a robust fashion Antimicrobials and anticonvulsants are the most common classes of agents to cause DILI
Conclusions (2)
The incidence of chronic DILI is higher than previously anticipated Extensive clinical data and biosamples are available for conducting clinical and mechanistic ancillary studies including genetic analysis to predict risk factors and outcome
DILIN Team of Investigators
UNC NIDDK
Jose Serrano, MD Leonard Seeff, MD Jay Hoofnagle, MD Jim Rochon, PhD John McHutchison, MD Don Rockey, MD Katherine Berezny Herb Bonkovsky, MD Bob Rosson, MD Jim Freston, MD Laura Glynn
Paul Watkins, MD Paul Hiyashi, MD Susan Pusek
IU
Naga Chalasani, MD Larry Lumeng, MD Audrey Corne, RN
DCC
UCSF
Tim Davern, MD M. Bonacini, MD Dalia Mowad
U. Conn
U. Mich
Bob Fontana, MD H. Conjeevaram, MD Suzanne Welch