Transcript Risk Management Strategies Litigation During Drug Development

Assessing Risk
in Drug Development
David M. Benjamin, Ph.D.
Clinical Pharmacologist & Toxicologist
Adjunct Assistant Professor - Tufts Medical School
Fellow, American Academy of Forensic Sciences (Toxicology)
Fellow, American Society for Healthcare Risk Management
Fellow, American College of Clinical Pharmacology
Fellow, American College of Legal Medicine
Goals of the Presentation
 Provide Criteria for Assessing Risk in Drug
Development
 Describe Theories of Drug Product Liability
 Examine Omniflox, Baycol and other Recent
Drug Withdrawals
 Identify Reasons for Drug Withdrawal
 Review Criteria for “Adequate” labeling
What Got Everything Started?
 The Institute of Medicine issued:
To Err is Human which reported that
44,000 - 98,000 patients died each year as
a result of “Medical Error” (e.g., medication
errors, surgical errors, missed diagnoses)
 At an estimated cost to the US
economy of $17-$29 billion
Major Findings of the IOM Study
 Based on two reports from three states:
New York (1984), Utah and Colorado (1992)
 NY study sampled 30,000 charts from 51 state
hospitals and found:
 3.7% of pts suffered injury severe enough to
disable them or prolong hospitalization
 58% of these injuries were due to error; 13.6%
were fatal
Major Findings of the IOM
Study-2
 The 98,000 number was extrapolated from the NY
study, based on the number of hospitalizations for
1997
 Based on reports from Utah and Colorado (1992)
 15,000 charts were sampled and a value of 44,000
deaths was extrapolated from these data
 The IOM report has been criticized by one of the
investigators, Troyen Brennan, MD, JD, MPH of the
Harvard School of Public Health
Drug Product Liability:
Murphy’s Law Rules
 “Often, however, the benefits to the
many come at a high cost to the few, for
there are increasingly more
opportunities for mishaps, not only in
the manufacturing process, but also in
the marketing and use of the finished
products.”
Page Keeton, Former Dean, U. of Texas
School of Law, (Michigan Law Review,
1966)
What About Investigational
Drugs During Development?




Chemical
Animal Pharmacology
Animal Toxicology
Clinical Phases I-III: 2,5005,000 patients only detects
ADRs of 1/1000 (0.1%)
 Post-Marketing: hundreds of
thousands to millions
Types of Adverse Drug Reactions
(ADRs) Frequently Encountered







Overmedication
Side Effect
Secondary Effect
Allergic
Idiosyncratic
Drug-Drug Interaction
Investigational Drugs - Fewer pts.
studied; less info available to IRBs;
e.g., Investigator’s Drug Brochure
Evaluating ADRs for Causation
 Does alleged ADR: (1) follow a
reasonable temporal sequence from drug
administration
 Does alleged ADR represent a known
effect of the drug?
 Could the alleged ADR be caused by: (1)
the patient’s pathophysiological (clinical)
condition or (2) an intercurrent illness?
 Could alleged ADR be caused by: (1) prior
exposure to another agent or (2)
concomitant medication?
Evaluating ADRs for Causation -2
 Is there a prior History of Sensitivity?
 What were the: Dosage, Frequency and
Route of Administration?
 What was the patient’s Cardiac, Pulmonary,
Hepatic and Renal Status?
 Was the patient receiving any Concomitant
Medications, OTC meds, herbals or
alternative medical products?
 Was suspect drug D/C’d? Was suspect
drug restarted? With what effects on ADR?
Theories of Drug &
Device Product Liability:




Defective Warning (Failure to Warn)
Inadequate Testing (Animal or Human)
Overpromotion (Lack of Fair Balance)
Defectively Designed Drug or Device
(Unfit for intended purpose)
 Adulterated or Contaminated Drug
Communication of Information
Pharmaceutical Company ---warns-->
Physician of known adverse effects
Physician serves as “Learned Intermediary”
for patient
Physician provides patient with Informed
Consent
Including reasonably foreseeable,
material risks
Once fully informed, the patient either:
accepts treatment or states informed refusal
Vicarious Defendants in Drug &
Device Product Liability Suits
 Pharmaceutical Manufacturer is
Primary, vicarious defendants include:
 Physician Prescribing Medication
 Distributor
 Pharmacist Dispensing Medication
 Nurse Administering Medication
 Hospital Employing MD, Reg.Ph., or RN
 Apparent or ostensible agents of entity or
medical school, e.g., IRB members
Defenses for Vicarious
Defendants
 Not negligent, or negligence was not a
“Proximate Cause” of injury to patient
(e.g., Bad Outcome)
 Cannot warn when not informed or misled
 Cross-Claim against manufacturer
 Subrogation claim by insurance company
 Claim for injury to MD’s reputation see:
Washington State Physicians Ins. Exch &
Ass’n v. Fisons Corp. 858 P.2d 1054
(Wash. 1993)
Legal Theories of Liability, aka
What the lawyer says to the . . .
Physician: “Shouldn’t have prescribed it”
Nurse: “Shouldn’t have administered it”
Pharmacist: “Shouldn’t have dispensed it”
Hospital: “Shouldn’t have had it
in your formulary ”
 IRB: “Should have monitored
more closely or not
approved protocol”




Federal Food, Drug, and Cosmetic
Act 1938
 Replaced US Pure Food and
Drug Act of 1906
 Enacted as a result of the 1937
sulfanilamide crisis
 Ensured that: (1) Foods were
pure and wholesome, and
produced under sanitary
conditions, and that (2) Drugs
were safe for their intended use
(i.e., not adulterated).
Chloramphenicol - Approved 1949
 June 1952 JAMA article describes
aplastic anemia - sometimes fatal
 Jan 1967 Cal Med Assoc & Dept
of Health reported an incidence
of aplastic anemia of: 1:25,000 in
chloramphenacol users vs
1:524,000 in the general
population (21X)
 Lesson to be learned: PostMarketing Surveillance is essential
to identify rare, previously
unrecognized ADRs
Thalidomide 1960
 FDA’s Dr. Frances Kelsey will not
approve thalidomide for US
market
 Mass Torts on teratogenic effects
postponed till Bendectin/Daubert
 1962 Harris-Kefauver Amendment
to FDCA required: (1) animal
testing prior to human
administration, (2) filing of an IND,
and (3) manufacturer must prove
efficacy (in addition to safety) NAS/NRC DESI Studies
Post-1960s Withdrawals
We Have Known and Loved






1970s
DES
IUDs
DPT/MMR
1980s
Oraflex, Zomax,
Suprofen
 Generic Drugs
 Bendectin







1990s
L-Tryptophan - EMS
Omniflox - 5-month life
Toradol - 5-day labeling
Imitrex - First-Year Effect
Silicone Breast Implants
Fen-Phen - “Off-Label”
use
 Duract - Where is Dr.
Kelsey when we still need
her?
Omniflox (temofloxacin) - PM
 Introduced: Jan 1992
 First Rx: February 24
 Voluntarily withdrawn: June
9th - 15 weeks
 US Clin Trials 4,600 pts
 Mkt’d in 8 countries Est.
300,000 pts rec’d drug
worldwide
Pre- vs Post- Marketing
Adverse Drug Reaction Reports
 Pre-Marketing
 4,261 pts
 Incidence > 1% N,V,D,
headache, rash,
itching
 Labs: Incr. BUN &
creatinine
 Renal Failure <0.1%
 Post-Marketing
 fewer than 300,000 pts
 1,700 non-fatal ADRs
reported to FDA
 54 cases of Acute
Renal Failure, 113
cases of hemolytic
anemia
 60 deaths, 25-50 may
be related; Globe 1/94
Recent Withdrawn Drugs





Posicor - too many interactions
Rezulin - liver toxicity
Raxar - cardiac arrhythmias
Propulsid - too many interactions
Seldane/ Hismanyl - drugs inhibit
metabolism--> - long QT syndrome
 Rotashield - infant bowel obstruction
 Lotronex - off Nov 2000 GI ADRs &
deaths; Re-instituted 2002 w/pt FU
 Baycol (are other “statins” to follow?)
Why are Drugs Withdrawn
From the Market?
 Examples: Posicor,
Propulsid, Seldane, and
Hismanyl
 Removed because of too
many drug-drug interactions
 Manufacturers & FDA tried
to minimize interactions
by revising labeling - BUT
 How do physicians view
(and use) labeling?
FDA Studies on Practitioners’
Views About Drug Labeling
 1992 Focus Groups
 Oct 1993-March 1994: FDA-sponsored
phone interview surveys of office-based
MDs to determine:
 How physicians perceive & use drug
labeling, and
 How labeling could be made more useful
“Survey Said:”
 Office-based MDs used labeling primarily to
answer specific questions, rather than as a
general educational tool
 Labeling is typically consulted after a
prescribing decision has been made
 Physicians wanted: important information
summarized in the front of labeling, and
 More graphs, larger print, abstracting of
important data
Re: Clinical Pharmacology
Section Survey Said:
 Among the least useful section
 Information in this section is used less
often than other labeling information
 Physicians wanted it “moved toward the
end of labeling”, and
 Important data from Clinical
Pharmacology Section should be put
under other sections like “Special
Populations” or “Drug Interactions”
Revising Labeling Will Not
Reduce Drug-Drug Interactions
 Labeling is not read
 Dear Dr. Letters are
not read
 See Yarrow v. Sterling
Drug Inc, (on a
subsequent slide)
regarding Aralen
(chloroquine) and
labeled retinal damage
Where’s the Proof ?
 Cisapride (Propulsid) withdrawal
 Marketed August 1993
 By 1995, 5 million OutPt Rxs and
FDA had received reports of: 34
cases of torsade de pointes 23
cases of prolonged QT interval
including 4 deaths
 1995 “Black Box” warning added to
labeling contraindicating use in
pts also receiving drugs that block
the CYP 3A4 enzymes, e.g.
macrolides, “conazole” antifungals,
& indinavir/ritonavir for HIV
More Proof
 1998 “Black Box” warning expanded to
contraindicate concomitant use of agents
that caused prolongation of QT interval
due to reports of serious arrhythmias,
e.g., class IA or III antiarrhythmics,
tricyclic anti-depressants, and
antipsychotics, or in pts with “other
conditions” that predispose to cardiac
arrhythmias, e.g., vent. arrhythmia, CHF,
ASHD, electrolyte imbalance, renal
failure and respiratory failure.
Still More Proof
 June 26, 1998 - FDA circulated “Press Release”
and manufacturer sent “Dear Healthcare
Professional” letter to 800,000 MDs informing
them of the changes in the labeling
 July 2000 - Manufacturer terminated marketing
of cisapride in US
 December 20, 2000 JAMA 284:3036-3039:
Smalley et al reviewed regulatory history of
cisapride in Contraindicated Use of Cisapride
Impact of FDA Regulatory Action
Smalley, Shatin, Wysowski, et al
 Analyzed number of pts. at three centers
who received cisapride when it was
contraindicated
 Received cisapride July 1997 - June 1998
(pre-warning revision) vs.
 Received cisapride July 1998 - June 1999
(post-warning revision)
 Pre-Warning Year 7/97-6/98: 26%, 30% &
60%
 Post-Warning Year 7/98-6/99: 24%, 28% &
58%
Smalley, Shatin, Wysowski, et al
 Conclusions:
 “There was no material reduction in . . .
use at any of the study sites following FDA
regulatory action, which included a blackbox warning . . . and a Dear Health Care
Professional letter” page 3038
 “. . . no material change in contraindicated
use even in the group most likely to be
affected by the regulatory action” - pts.
beginning new cisapride use.
Smalley, Shatin, Wysowski, et al
 Conclusions:
 “ . . . despite the prominent publication of
case reports, label changes, and Dear Health
Care Professional letters . . .”
 “ . . . need to develop more effective methods
for modifying practice to reflect new
information about a drug’s risks and benefits.”
Ditto for Metformin
 Metformin marketed for Type 2 DM in 1995 with
“black box” warning regarding lactic acidosis
 Contraindications included: renal dysfunction
and CHF
 In 2000, 25 million Rxs for metformin
 Horlen, Malone, Dennis et al (letter in JAMA
287:2504-5) performed retrospective chart
review of pts. who received metformin at UNC
outpatient pharmacy between Jan 1, 2000 Sept 30, 2000.
Metformin (continued)
 Metformin was prescribed twice or more for 241
pts. - 100 charts randomly selected
 22% of pts. had either CHF or renal
dysfunction; 14 CHF, 5 renal dysfunction,
3
had both CHF and renal dysfunction
 Only 2 pts. had documentation in the medical
record considered metformin contraindications
 “. . . it is difficult to determine whether clinicians
are aware they are prescribing metformin
against a black-box warning.”
Yarrow v. Sterling re: Aralen
Where the doctor is inundated with the literature and
product cards of the various drug manufacturers . . . a
change in the literature or an additional letter intended to
present new information on drugs to the doctor is
insufficient. The most effective method employed by the
drug company in the promotion of new drugs is shown to
be the use of detail men; thus, the Court feels that this
would also present the most effective method of warning
the doctor about recent developments in drugs already
employed by the doctor, at no great additional expense.
Yarrow v. Sterling Drug, Inc. 263 F. Supp. 159, 1963
What ADRs will this Drug Have?
 Look to animal studies
 Examine other drugs of the same
pharmacologic or chemical class
 Review SBA of other similar drugs
 Toxicity is the interaction of frequency and
severity, e.g., Severe but infrequent = OK,
and Mild but common = OK
 True frequency cannot be determined
before marketing because “n” is too small
Assessing ADRs -1
Severity
Mild
Moderate
Severe
|<--------------------50%-------------->|100%
Nausea,
More Than
LifeRash?
Annoying
Threatening
Red & Itchy
vs.
SJS/TEN
Assessing ADRs - 2
Frequency
Rare: <1/10,000, e.g., chloramphenacol &
blood dyscrasias
Rare, but life-threatening
Frequent: >1/100, e.g., headache & dry mouth
Frequent, but not serious
Characteristics of Ideal Drug
 Novel pharmacologic
action or treats condition
for which no acceptable
alternative exists
 Higher Therapeutic Index
than other drugs in its class
 Lower frequency & severity
of reported reactions
 Manufactured inexpensively
 Secondary or multiple uses
What are the Most Common
Processes Associated with
Medication Errors?




Physician ordering: 39-49%
Nursing administration: 26-38%
Transcription error: 11-12%
Pharmacy dispensing error: 11-14%
Source: Bates et al JAMA
1995;274(1):29-34 & Leape et al JAMA
1995;274(1):35-43.
Portrait of a Poor
Prescription
What is the name of the first drug written on
this prescription? See next slide for discussion.
The Correct Drug was . . .
 Final Answer: Isordil.
Not Plendil, which was
dispensed and caused
fatal hypotension for
which both the MD and
RPh were held jointly
liable for almost
$500,000.
Why Computerize?
 MD computerized order
entry decreased serious
medication errors 55%
and
 Potential undetected
Adverse Drug
Experiences (ADE)
declined 84%
 Bates et al JAMA
1998;280:1311-1316
Transcription Errors
 Transcription Error
 Data Entry Error
 Verification Error
Vigilance
Transcription Errors - 2
Confusion Over:
Drug Names or
Handwriting
 Larocin 250 vs.
Lanoxin 0.250
(Larocin changed to
Larotid after mix-up)
 enalapril vs.
Elderpryl
 Lamisil vs. Lamictal
The “Name Game”:
Risk Management Strategies
 Select drug names that are not easily
confused with other marketed products:
e.g., Losec 50 mg vs. Lasix 50 mg,
ultimately Losec changed to Prilosec
 Packaging should distinguish between
different dosage strengths of the same drug
 Size, shape and color of different dosage
strengths of the same product should make
it easy to tell the difference
Writing Adequate Labeling
 Clear Directions for Use
 Use adjectives and adverbs
carefully, e.g.,
 Rarely, occasionally vs.
 Frequently, commonly
 Mild, moderate, severe
 Frequencies: A few mild rashes
vs. a few cases of aplastic
anemia. Can you determine the
true incidence?
Use Imprecise “Stock”
Statements Sparingly
 The relationship to the drug
has not been proven
 What if you have a positive
dechallenge/rechallenge?
 Use only when the expected
benefit outweighs the potential
risk
 Caution should be exercised
 Long-term & short-term use
Graham, G. Labeling: What Should it Say, and How Should it
Say it? Drug Information Journal 1991;25:211-216
Risk Management Strategies to
Minimize the Risk of Litigation
 The Merck Paradigm:
 Patient took Flexeril
(cyclobenzaprine) with an
MAOI and developed NMS
 Patient sued for Failure to
Warn
 Labeling stated,
“cyclobenzaprine is closely
related to the TCAs, e.g.,
amitriptyline and
imipramine”
Risk Management Strategies to
Minimize the Risk of Litigation
Your Package
Insert
 Labeling also stated, “Flexeril may
interact with MAOIs.”*
 “Hyperpyretic crisis … and deaths
have occurred in patients
receiving TCAs and MAOIs”*
 Generalizing: Although this ADR
has not been reported with this
drug, it has been reported to
occur with other drugs of the
same (pharm/chemical) class.
*Benjamin, DM. Case Studies in Pharmaceutical Risk
Management, Medical Sciences Bulletin, November 1995.
Summary
 True safety of a drug usually cannot be
determined until after marketing
 Stay away from “me-toos” with poorer safety
profiles than already marketed drugs, e.g.,
bromfenac
 Doctors do not read the package insert or
Dear Doctor Letters, find another way to warn
them (Detailpersons?) to avoid withdrawal
from the market
 Look for new uses of old drugs, e.g.,
spironolactone