Transcript Isoniazid-Related Hepatitis among Pregnant and

Size of the Safety Database:
Population Considerations
Richard O’Brien
Foundation for Innovative New Diagnostics
TB Drug Forum
Arlington, VA
December 6-7. 2005
Safety Database
• Complication of information related to drug
safety collected during clinical development
• Major items
– Serious Adverse Event (AE): death, including lifethreatening event, hospitalization, disability, birth
defect
– Discontinuations due to adverse events
• Other: common AE*s, laboratory data, vital
signs, ECG, Phase 1 safety data, pregnancies,
overdose experience, drug-drug interaction,
dose dependency re AEs
*any adverse finding temporarily associate with drug use, Including signs,
symptoms, lab/VS/ECG, etc., or cluster of events
Sources of Safety Data
• Phase I trials
– Common AEs
– Lab, vital sign, and/or EKG data
– Metabolic profile, AEs in special populations
• Phase II/III randomized controlled trials
– Common AEs, serious less common AEs (maybe)
– Discontinuations due to AEs
– Lab, vital sign, and EKG data
• Phase II/III Open Label trials
– Serious AEs
NDA Safety Review
• Required by FDC Act, section505
– Includes “all tests reasonable applicable to
show…drug is safe…under…proposed labeling”
– Results show “drug is safe under such conditions”
• To critically examine sponsor’s claim of safety
– To assess adequacy of testing for safety: numbers
of patients, duration of exposure, quality of data
– To identify safety issues that could affect approval
• To identify safety issues to be described in
product label
Risk Factors to Explore
• Patient factors
–
–
–
–
–
Age
Gender
Race
Genetic vulnerability
Comorbid illness
• Drug factors
–
–
–
–
Dose
Plasma level
Duration
Concomitant
medications
Safety Database: Numbers of
Patients
• Generally accepted that premarketing safety
database is not able to detect rare SAEs
• ICH guidelines for chronically administered
drug:
– 300-600 patients for 6 months
– 100 patients for 1 year
– 1500 total patients
• If labeling will recommend dose range:
– How much exposure was observed at highest
dose?
• Subgroup analyses problematic
Probability of Observing AEs
Event
Rate
Sample Size Probability of observing
> 1 event > 2 events
1%
500
0.99
0.96
0.5%
500
1000
0.92
0.99
0.71
0.96
0.1%
1500
3000
0.78
0.95
0.44
0.80
0.01%
6000
10,000
20,000
0.45
0.63
0.86
0.12
0.26
0.59
Subpopulations of special interest
in TB trials
• Extremes of age, especially children
• Patients with comorbid conditiions
– hepatic or renal impairment
– HIV infection
•
•
•
•
•
Pregnant/lactating women
Different races/ethnicities
Different disease forms
Differing severity of disease
Different geographic regions
Hepatic Toxicity in South Indian TB
Patients Treated with H,R,Z*
• Data from studies of 1686 TB patients with
TB treated at TRC with SCC regimens
• Hepatitis usually accompanied by jaundice
• Rates varied by site of disease
– 16-39% in children with TBM
– 10% in spinal TB
– 2-8% in pulmonary TB
• Rates varied by acetylator phenotype
– 11% in 317 slow acetylators
– 1% in 244 rapid acetylators
*Parthasarathy eta al. Tubercle 1986;67:99-108.
Urinary Excretion of Toxic Metabolites
of INH by Acetylator Status*
Slow Acetylators (7) Rapid Acetylators (5)
% Dose in 24 hr
% Dose in 24 hr
Isoniazid
32.4 (+ 4.0)
Acetylisoniaizid
25.9 (+ 3.4)
Acetlyhydrazine
3.1 (+ 1.0)
Diacetylhydrazine 5.1 (+ 2.4)
Hydrazine
1.0 (+ 0.3)
9.2 (+ 2.9)
42.9 (+ 7.1)
1.6 (+ 0.4)
22.6 (+ 3.9)
0.4 (+ 0.1)
All differences statistically significant, p < 0.01-0.05
*Peretti et al. Eur J Clin Pharmacol 1987;33:283-6.
Hepatic Toxicity Children in U.S.
Treated with Short-Course Therapy*
• Survey of state TB programs following
ATS/CDC recommendation on SCC
• 874 reports from children treated 1977-79
• 16 hepatotoxic reactions (15/430, 3.3% in
430 children on SCC)
• Half in the first month, all in the first 10 wks
• Higher in those on high HR doses and on
those with disseminated TB
• Recommendations on limiting doses and
monitoring those with serious disease
*O’Brien el al. Pediatrics 1983;72:491-9.
TBTC Study 22: Relationship of
Acetylator Status to Outcome in
HIV-Negative Patients*
Failure/Relapse
Cured
Fast
13 (59%)
9 (41%)
Medium
13 (35%)
24 (65%)
Slow
8 (25%)
24 (75%)
P=0.04, chi square
*Weiner et al. AJRCCM 2003;167:1341-7.
Isoniazid-Related Hepatitis in
USPHS Cooperative Surveillance Study*
• 18,838 patients in IPT in 21 HDs in U.S.
• Age most significant hepatitis risk factor
– 0.1% among persons < 20 yr
– 2.3% among persons 50-64 yr
• Related to alcohol consumption
– 0.6 % in non-drinkers vs. 2.7% in daily drinkers
• Eight hepatitis deaths
– 5 in black women
– 7 in one geographic region
*Kopanoff et al. ARRD 1978;117:991-1001.
Isoniazid-Related Hepatitis among
Pregnant and Postpartum Hispanics*
• CDC investigation following 2 INH-related
hepatitis deaths in post-partum women
• 3,681 women on IPT in pre-natal clinic
• Control group: 3948 women in earlier study
• Rate of hepatitis increased 2.5-fold among
prenatal/postpartum women
– For 15-34 yr: 0.3 % vs. 0.06%
– For 35-44 yr: 2.3% vs. 0.9%
*Franks et al. Pub Health Rpt 1989;104:151-5.
2RZ-Related Hepatitis
• Addition of PZA to INH/RIF in SCC did not
increase rate of hepatitis in CDC Study 21
– Enrolled 1451 newly diagnosed PTB patients
– 2.4% in 2HRZ/4HR vs 3.6% in 9HR*
• No increased hepatotoxicity in studies of
2RZ LTBI Rx in HIV+ patients
– International study enrolled 1583 patients**
– Randomized to 2RZ and 9H
– Abnormal LFTs in 1.4% on 2RZ and 3.3% on 9H
*Combs et al. Ann Int Med 1990;112:397-406.
**Gordin et al. JAMA 2000;283:1445-50 and
Multicenter CT of 2RZ in HIVNegative Patients*
• 589 HIV- adults eligible for LTBI treatment in
3 centers
• Randomized to 2RZ and 6H
• Hepatoxicity in persons completing treatment
– 16/207 (7.7%) for 2RZ
– 2/204 (1%) for 6H
• No difference in rate of completion of
treatment
*Jasmer et al. Ann Int Med 2002;137:640-7.
Toxic Hepatitis Among HIV+ and
HIV- Patients in Barcelona*
Drug Regimen
Toxic Hepatitis
HIV+ (%), RR (CI)
HIV- (%) RR (CI)
2RZ (n=281)
4(3.5),1.5(0.2-17)
12(11),4.4(1.3-19)
6-9H (n=320)
2(2)
3(3)
P = 0.7
P = 0.015
*P Sanchez. Personal communication.
Initial Report of Fatal Hepatitis with 2RZ*
* MMWR Weekly Report. 2001;50;289-91.
Cases of Serious 2RZ-Related
Hepatitis, 1999-2003*
10
Known case: Pre or Post-survey period
Known case: Captured by survey
8
Known case: NOT captured by survey
Cases
Additional case: Detected by survey
6
Additional case: Verispan® search
4
MMWR8
CDC/ATS
endorsement4
MMWR9
x
2
MMWR10
MMWR11
x
x x
x
x
x
x
x
x
x
x
1999
2000
2001
Survey Periods
*McElroy et al. Clin Inf Dis 2005;41:1125-33.
2002
2003
Oct
Jul
Apr
Jan
Oct
Jul
Apr
Jan
Oct
Jul
Apr
Jan
Oct
Jul
Apr
Jan
Oct
Jul
Apr
Jan
0
CDC Survey of 2RZ Use
• Survey of state/city TB controllers on 2RZ use
between Jan 2000 and June 2002
• 110/139 responded (79%), 87 (79%) used 2RZ
in 8087 patients
• Rates per 1000 initiations
–
–
–
–
Asymptomatic AST > 5 times ULN: 25.6
Symptomatic hepatitis: 18.7
Hospitalization for hepatitis: 2.8
Death from hepatitis: 0.9 (vs. 0.0-0.3 for INH)
• Completion rate: 64%
Acquired RIF-Resistance in HIV+ Patients*
TBTC Study 22 included
an arm which evaluated
the once-weekly
INH-Rifapentine
regimen in HIV+ persons
*Vernon et al. Lancet 1999;353:1843-7.
Acquired RIF-Resistance in Study 22
Item
Treatment Arm
INH+RPT INH+RIF
Both
Enrolled
36
35
71
Completed Rx
30
31
61
Relapsed1
5
3
8
RMR relapse2
4
0
4
1p=0.47,
Fisher’s Exact Test
2p=0.05, Fisher’s Exact Test
Acquired RIF-Resistance in Study 22
HIV+ Patients in Rifapentine Arm
Group
RIF-R
No RIF-R
P value
Number
4
25
Age (mean/sd)
31+10
43+14
0.10
CD4 (mean/sd)
57+88
227+189
0.03
EPTB
75%
12%
0.02
Use of azoles
75%
8%
0.01
Acquired RIF-Resistance in TBTC Study 23*
• Single arm study of largely intermittent
rifabutine in HIV+ persons
• 169 eligible patients enrolled and treated
• 9 patients had culture+ relapse (6) or failure (3)
– Overall rate of relapse (5%) not higher than
expected
– 12% in pts with CD4<100 vs 0% in pts > 100
• 8/9 had acquired rifampin resistance
• Resulted in change in ATS/CDC/IDSA
treatment recommendations for HIV+ TB Pts
*Burman et al. AJRCCM. E-Pub Aug 18, 2005
Summary
• Only more common adverse events will be detected
during clinical development
• Safety data from populations of great interest difficult
to obtain
• Cannot predict toxicity indirectly, e.g., HRZ vs. RZ
hepatoxicity
• Discussions about acceptable rates of serious AEs
for new TB drugs are needed (> 1/50,000 not
acceptable for less serious condition, those with other
therapeutic options
• Post-approval/marketing (Phase IV) studies are
needed. May be difficult to estimate true rate of
uncommon AEs, but planning should begin now