Detecting Liver Injury: Drug-Induced or Not ?

1
Detecting Liver InjuryDrug-Induced or Not ?

• John R. Senior, M.D., Hepatologist
• Associate Director for Science
• Office of Pharmacoepidemiology and Statistical
  Science
• Food and Drug Administration (FDA)

2

• Material presented here is based on the
  experiences of the speaker for 20 years in
  academic hepatology and gastroenterology, 5 years
  as a senior executive in the pharmaceutical
  industry, 11 years in private consulting to
  industry. Then at the FDA, 4 years medical
  reviewer for new gastrointestinal drugs, 3 years
  senior scientific advisor for hepatology in the
  Office of Drug Safety, and 2 years as Associate
  Director for Science, Office of
  Pharmacoepidemiology and Statistical Science. The
  comments do not reflect official policies or
  positions of the Agency, but are personal
  opinions of the presenter based on the diverse
  experiences mentioned.

3

• First Ask
• Is there liver injury or disease?
• Is it progressive or serious?
• progressive getting worse or likely to do so
• serious disabling, life-threatening, fatal
• Drug-induced or some other cause?
• no pathognomonic test for DILI, including
  biopsy
• DILI may mimic any known liver disease

4

• Lets look first at some other causes
• What are they?
• acute/chronic viral, immune, vascular, metabolic
• How can they be detected?
• serum transaminases, other enzymes, bilirubin,
  INR
• How to distinguish from DILI?
• no pathognomonic test for DILI, including biopsy
• DILI may mimic any known liver disease

5
Cooperative research between the pharmaceutical
industry and FDA Adventures with a Placebo
DatabaseOctober 2001 - present

• John R. Senior, M.D., FDA
• Robert W. Tipping, M.S., Merck

6
Special thanks to

• Peter Honig, M.D., (FDA) Merck
• Harry Guess, Ph.D., (Merck) UNC
• Paul Seligman, M.D., FDA
• ...who made this work possible

7
Why study placebo participants?

• obtain data on incidence of AEs not due to drug
• fundamental assumption placebos do no harm
• they should be subtracted from those seen on
  drug
• focus on hepatic injury evidence tests,
  symptoms
• search database for cases of liver injury or
  disease
• aim to establish background rate for incidence
• determine what tests are most accurate and how
  best to make true attribution of causality
• one of the initiatives of the PM white paper
  2001

8
What to Look For and Why ?

• 1) evidence of serious or potentially serious
  liver disease
• not much interested in transient serum
  transaminase bumps
• liver is a very adaptive organ, handles
  xenobiotics well
• 2) ultimate aim - to distinguish drug-induced
  liver injury
• diagnosis of exclusion must rule out other
  causes
• 3) critical need for accurate differential
  diagnosis
• need to see serial data, time course of
  abnormal patterns
• what really is causing the abnormal pattern?
• need more information than just lab test numbers

9
AFCAPS/TexCAPS Study - 1

• carried out 1990-7, San Antonio Fort Worth TX
• 6605 participants (85 men), 3301 to placebo
• men gt45 and women gt55, up to 73 ambulatory no
  previously diagnosed cardiovascular disease
• modestly high total cholesterol, reduced
  HDL-chol
• no pre-existing liver disease, or other major
  disease
• willing and able to participate for 4-6 years
• aim show lovastatin-related reduced cardiac
  events
• results published JAMA 1998 and AmJCardiol 2001

10
AFCAPS/TexCAPS Study - 2

• 5-year observation, 20 () visits/test
  sets/participant
• visits 3 q 2wks (baseline) 8 q 6wks, 9 q 6
  mos
• each visit serum ALT, AST, ALP, TBL, CPK
• we chose PLACEBO group (3248 had 5-yr data)
• search database for cases of liver injury or
  disease
• our aim to establish background rate for
  incidence

11
Looking for Liver Disease/Injuryhow should the
search be done ?

• We looked for
• 1) any two ALTx3 (AST/ALP/TBL)x2 CPKx5 _at_ peak
• 2) confirmed ALT or AST at least 3xULN
• 3) ALT or AST 3xULN AND concurrent TBL 2xULN
• 4) symptoms, complaints, diagnoses, AE reports
• 5) clinical narratives for selected cases
• 6) review of case report forms if lab
  abnormalities

12
Distribution of Abnormalitiespeak values among
3301 studied for 5 years

13
The First 44 Cases
trt sex age ALTx3 ASTx2 ALPx2 TBLx2 CPKx5 P M 6
1 2.45 2.35 5.59 7.0 0.72 P M 52 1.50 2.19 0.50
0.8 10.83 P M 70 9.60 3.54 2.42 2.9 0.68 P F 6
5 5.00 2.59 0.50 0.8 0.55 P F 56 4.35 3.30 1.45
0.6 1.45 P M 59 3.15 7.95 6.65 6.7 4.10 P M 55
3.90 3.03 0.55 0.9 1.12 P M 57 50.25 40.76 1.38
8.8 0.84 etc. to 44 cases

14
Case M 61, placebo
Day ALTx ASTx ALPx TB Lx CPKx -34 0.45 0.38 0.47
0.7 0.44 -14 0.53 0.41 0.48 0.7 0.36 1 0.50 0.38
0.45 0.7 0.23 43 0.38 0.32 0.49 0.8 0.28 85 0.5
0 0.41 0.51 1.0 0.22 127 0.38 0.30 0.42 0.5 0.30
169 0.53 0.38 0.40 0.7 0.27 211 0.48 0.43 0.36 0
.6 0.27 253 0.40 0.35 0.40 0.8 0.46 295 0.40 0.4
1 0.43 0.7 0.35 337 0.38 0.41 0.42 0.8 0.72 421
0.40 0.35 0.42 0.8 0.41 547 0.45 0.43 0.40 0.7 0.
54 729 0.75 0.62 0.58 0.4 0.71 839 2.20 1.97 2.7
1 0.9 919 2.45 2.35 5.59 7.0 0.36

15

16

17

18

19

20

21

22

23
The Next 87 Cases - 2

• Liver function abnormality (no symptom) 27
• Cholecystitis, cholelithiasis, or both 37
• 3 acute, 1 gangrenous, 1 perforated, 1
  pancreatitis
• Pruritus 9 Fatty Liver 7 Cholesteatoma 2
• Cholangiocarcinoma, Hepatitis, Liver Cyst,
  Cholestasis, Jaundice 1 each

24
Sensitivity-Specificity for 6 of 3248

25
Conclusions - so far

• Serum transaminase elevations not disease
• often may represent transient adaptations
• Requiring confirming tests may miss cases
• unless done very promptly within a few days
• Additional information beyond lab test scores
  needed for making true causal attribution
• AST elevations dont add much to ALTs, (see in
  alcoholic hepatitis, cirrhosis, muscle)
• Concurrent total bilirubin elevation suggests
  that serum ALT gt3xULN may be serious
• Hys Rule may become validated by data

26
Rich Findings in Placebo Data

• I. Concurrent bilirubin rise adds specificity
  to ALT testing, without losing sensitivity
• II. Serum transaminase activities vary greatly,
  CPK even more, but ALP less so

27
Where Do Elevated Serum Transaminases Come From ?

• John R. Senior, M.D., FDA
• Robert W. Tipping, M.S., Merck

28
The First 44 Cases
trt sex age ALTx3 ASTx2 ALPx2 TBLx2 CPKx5 P M
61 2.45 2.35 5.59 7.0 0.72 P M 52 1.50 2.19 0.5
0 0.8 10.83 P M 70 9.60 3.54 2.42 2.9 0.68 P
F 65 5.00 2.59 0.50 0.8 0.55 P M 59 3.15 7.95
6.65 6.7 4.10 P M 55 3.90 3.03 0.55 0.9 1.12
P M 57 50.25 40.76 1.38 8.8 0.84 etc. to
44 cases

29
But, no evidence of liver disease

trt sex age ALTx3 ASTx2 ALPx2 TBLx2 CPKx5 P M 5
2 1.50 2.19 0.50 0.8 10.8 So, why the rises in
transaminases?
30

31
AST ALT and CPK Rises
sort data by ascending CPK values

32

33
Two questions

1) What is the source of the elevated serum
transaminase activities? 2) Does CPK gt10xULN
really indicate muscle disease (myopathy)?
34
Organ/Serum Activity Ratios

• 
  muscle liver
• alanine aminotransferase (ALT) 7501
  76001
• aspartate aminotransferase (AST) 52001
  90001
• lactate dehydrogenase LDH) 14001
  14001
• pyruvate kinase (PK)
  62001 14001
• creatine phosphokinase (CK) 200001
  3001
• Geigy Scientific Tables, 1984 Volume 3, page 169

35
Body Composition(Geigy Scientific Tables, 1993
70- kg man)

• skeletal muscle - 43 about 30 kg
• skin, s.c. tissues - 26 about 18 kg
• bony skeleton - 17 about 12 kg
• liver - 2.1 about 1.5
  kg
• brain - 2.0 about 1.3 kg
• intestines - 2.0 about 1.3 kg
• kidneys - 0.5 about 0.3 kg
• heart - 0.5 about 0.3 kg

36
Non-Liver Transaminasemia

• acute muscle breakdown - rhabdomyolysis (both
  ALT, AST and bilirubin elevations)
• various muscular dystrophies, myopathies
• muscular exertion anorexia nervosa
• acute myocardial infarction
• intestinal celiac disease, untreated
  (becomes normal on gluten-free diet)

37
Can Muscle Injury Be Confused with Hepatotoxicity
?

• aspartate (AST) alanine aminotransferase
  (ALT), in addition to creatine phosphokinase
  (CPK) released
• release of muscle myoglobin into plasma -
  contains one molecule of heme that can become
  bilirubin
• renal failure (hepatorenal syndrome) also seen
  with acute liver failure . . . reversed by liver
  transplantation

38
But theyre still saying . .

• Whereas ALT is localized primarily to the liver,
  AST is present in a variety of tissues, including
  liver, heart, skeletal muscle, kidney, brain,
  pancreas, lungs, leukocytes, and erythrocytes.
• Zakim and Boyer. HEPATOLOGY, A Textbook
• of Liver Disease, 4th Edition, 2003. Friedman,
  Martin, Munoz page 662.

39
Functions of the Adult Liver

• extract and process nutrients from gut
• synthesize proteins, other molecules
• regulate intermediary metabolism
• metabolize steroid hormones, insulin
• extract bilirubin from plasma, excrete
• control cholesterol metabolism/bile acids
• handle xenobiotic substances, drugs
• but NOT to regulate serum enzyme levels !

40
Commonly Used Tests
injury
transaminases ALT (SGPT)
AST (SGOT)
hepatocellular obstructive
enzymes
alkaline phosphatase gamma-glutamyl transferase
function excretory synthetic
synthetic
bilirubin albumin prothrombin
substances
41
Is Serum ALT a Liver Function Test ?

• serum enzyme activity not just from liver but
  from skeletal and heart muscle, gut, etc.
• . . . so lets not say liver
• it is not a function or job of the liver to
  regulate the level of serum enzyme activity
• . . . so lets not say function
• elevated serum ALT activity MAY indicate
  hepatocellular injury

42
Maybe we should look closer . . .

• Note if serum transaminases elevated at the same
  time as serum CPK
• Work up immediately, with daily measures of CPK,
  AST, ALT, plus ALP, TBL and DBL, PT (INR), maybe
  GST, Cr
• Get full history of muscle exertion or injury and
  of liver diseases, alcohol, viruses A-C

43
Two questions

1) What is the source of the elevated serum
transaminase activities? 2) Does CPK gt10xULN
really indicate muscle disease (myopathy) or
rhabdomyolysis ?
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
Myopathy (muscle disease) ?

1) Unexplained muscle pain or weakness 2) CPK
gt10xULN
48
Rhabdomyolysis

1) Severe muscle breakdown 2)
Myoglobinuria 3) Renal insufficiency
49
rhabdo - myo - lysis (striped - muscle -
dissolution)
50
Heme-positive Urine

• Hemoglobinuria
• from red blood cells
• MW 64,500
• 4 hemes/molecule
• Cren slow, pink plasma
• methemalbuminemia
• HbO2 576-8 nm
• COHb 571 nm

• Myoglobinuria
• from muscle cells
• MW 17,500
• 1 heme/molecule
• Cren fast, clear plasma
• no methemalbuminemia
• MbO2 581-3 nm
• COMb 579 nm

51
Is it worthwhile ?

• statins becoming most used drugs in world
• widespread belief that the ALT, AST rises
  reflect liver injury
• hepatotoxicity probably vastly overstated
• mild muscle injury is not rhabdomyolysis, or even
  myopathy
• need data on closely time-related correlations
  of serum CPK, ALT, AST, other changes

52
More Conclusions

• serum transaminase elevations not all hepatic
• investigate AST, ALT elevations do CPK
• statin hepatotoxicity probably much overstated
• moderate exertional mild muscle injury is not
  rhabdomyolysis, or even myopathy
• need data on closely time-related correlations
  of serum CPK, ALT, AST, other changes
• serum T1/2 of CPK lt AST ltALT needs proof

53
Rich Findings in Placebo Data

• I. Concurrent bilirubin rise adds specificity
  to ALT testing, without losing sensitivity
• II. Serum transaminase activities vary greatly,
  as do CPK, and ALP less so
• III. Some AST, a little ALT comes from muscle
• IV. Baseline better determined by gt1 point

54
It may be DILI if its nothing else

• 1. Diagnosis of exclusion no test FOR DILI
• 2. Must gather data to rule out other causes
• 3. Need to educate people to do it better
• Develop model for quantitative likelihood
• 5. Prospective large studies needed
• - for true incidence
• - for risk factors
• - for omic analyses (gen-, prote-, metabon-)
  specimens
• - for mechanism elucidation