POST-MARKETING SURVEILLANCE: ONE DIVISION DIRECTOR’S VIEW

POST-MARKETING SURVEILLANCE: ONE DIVISION

DIRECTOR’S VIEW

RUSSELL KATZ, M.D.

DIRECTOR

DIVISION OF NEUROLOGY

PRODUCTS/CDER

POST-MARKETING SURVEILLANCE IS:

• Detecting a signal• Determining causality/frequency• Doing something about it • Informing relevant parties about the problem

TYPES OF SAFETY SIGNALS WE SEE

• Adverse Events• Medication Errors• Product Failures• Labeling Failures

WAYS THAT WE BECOME AWARE OF A SAFETY SIGNAL

• Literature• Spontaneous Reports (AERS)• Manufacturer• Private citizen• Trials• Registries

ADVERSE EVENTS

• Tysabri and Liver Injury• Botox and “Botulism”• Dopamine Agonists and Cardiac Valvulopathy• Pathological Gambling• Tysabri and Progressive Multifocal

Leukoencephalopathy (PML)• Anticonvulsants and Suicidality

ADVERSE EVENTS-SPONTANEOUS REPORTS (AERS)

• Tysabri and Significant Liver Injury– Several individual cases of elevated LFTs and

Bilirubin– Examination of individual cases for alternative

causes; two positive re-challenges• Botox and “Botulism”

– Several cases consistent with known pharmacology

ADVERSE EVENTS-SPONTANEOUS REPORTS (AERS)

• Adverse Events with “known” background rates

– Tolcapone and liver failure– Felbamate and aplastic anemia

HOW WE EVALUATE A SAFETY SIGNAL FROM SPONTANEOUS

REPORTS• Someone (clinical reviewer, safety reviewer, etc.)

detects a serious (possibly unlabeled) adverse event

• Task is to assess causality (also examine RCTs)• Review of individual case report usually not

definitive– Incomplete information– Logically flawed

HOW WE EVALUATE A SAFETY SIGNAL FROM SPONTANEOUS

REPORTS• EB 05

– Lower limit of the 90% Confidence Interval around the following metric:

– # of cases of event of interest/ # of total ADRs for the drug DIVIDED BY

– # of cases of event of interest of all drugs/ # of total ADRs for all drugs

– If >2, considered a possible signal

HOW WE EVALUATE A SAFETY SIGNAL FROM SPONTANEOUS

REPORTS• Calculation of a Reporting Rate (Not an

incidence)• RR = #cases reported/patient-years of exposure• Patient-years = #prescriptions/12 (assume each

prescription for 30 days)• Compare RR to background rate and other drug(s)

– Assume background rate is relevant– May not be (e.g., nefazadone)

HOW WE EVALUATE A SAFETY SIGNAL FROM SPONTANEOUS

REPORTS

• Any RR approaching the background rate is presumptive evidence that the drug caused the adverse event due to underreporting (many factors influence reporting or lack thereof)

• In many cases, the spontaneous reports serve as the basis for a definitive decision: i.e., spontaneous reports are not always just hypothesis generating

ADVERSE EVENTS-SPONTANEOUS REPORTS-LITERATURE

• Dopamine Agonists and Valvulopathy– One case control study, one echocardiographic

prevalence study in NEJM, 2007– Previous studies and AERS search in 2004 led

to conclusion about ergot-derived agonists– Recent AERS search for other agonists, 5HT2B

agonists

ADVERSE EVENTS-LITERATURE

• Dopamine agonists and Pathologic Gambling– Case series in the literature; mostly Mirapex

(some Requip)– AERS searched for all dopaminergic drugs for

impulse control disorders– Adequate background rates not available

VARIANT OF ADVERSE EVENT ASSESSMENT-LITERATURE

• Stevens-Johnson Syndrome known to occur with carbamazepine

• Literature reports of a marked increased incidence in Han Chinese patients

• Additional corroboration from international spontaneous reports

• Data “established” HLA B-1502 as an important risk factor for SJS

ADVERSE EVENTS-CLINICAL TRIALS

• Tysabri and PML– Three cases observed in open-label extensions

of controlled trials being done in Phase 4– Marketing discontinued and entire database

examined for any additional cases-none found – Causality assumed from mechanism

ADVERSE EVENTS-CLINICAL TRIALS

• AEDs and Suicidality– Sponsor notified us of a signal from their

controlled trials– Submission of a citizen petition raising

questions about another AED– All controlled trials from 11 recently approved

AEDs examined

ADVERSE EVENTS-CLINICAL TRIALS

• Hypnotics and cancer– Individual academic performed analyses of data

for recently approved hypnotics on the basis of which he concluded that they are associated with an increased risk for cancer

– Resulted in detailed review of numerous databases that did not confirm his conclusions

ADVERSE EVENTS-REGISTRIES

• Possible teratogenicity (cleft lip/palate) with Lamictal– Sponsor informed division of data from two

pregnancy registries– Analyses of other registries did not reveal

confirmatory findings

ADVERSE EVENTS-REGISTRIES

• Amendments to monitoring requirements for agranulocytosis for Clozapine– Requested by an interested party based on

family member’s experience– Resulted in detailed review of accumulated data

to date and ultimately a change to the monitoring regimen

MEDICATION ERRORS

• Lamictal/Lamisil• Reminyl/Amaryl• Fosphenytoin

HOW WE EVALUATE MEDICATION ERRORS

• Here, detailed review of cases can help pinpoint how the problem arises, and point the way to possible solutions

• Name confusion– Lamictal/Lamisil

• Written/verbal prescription

• Carton appearance; overlapping strengths

• Drug locations on pharmacy shelf

HOW WE EVALUATE MEDICATION ERRORS

• Here, detailed review of cases can help pinpoint how the problem arises, and point the way to possible solutions

• Name confusion– Reminyl/Amaryl

• Similar written appearance

• Several deaths due to hypoglycemia

HOW WE EVALUATE MEDICATION ERRORS

• Here, detailed review of cases can help pinpoint how the problem arises, and point the way to possible solutions

• Product Label– Fosphenytoin

• Total units in vial (100 mg/2 mL) vs concentration (50mg/mL)

• Automated displays still use old description

PRODUCT FAILURES

• Sometimes, the causes (or the errors) are obvious• Example:

– Diastat Accudial (prefilled syringes of diazepam for rectal administration)

• Leakage from cracked tips

• Inappropriate dialing of doses

PRODUCT FAILURES

• Potential “failures” of generic AEDs to perform adequately (possible ADRs and/or breakthrough seizures) reported to Agency by expert community

• Personal experience• Surveys• Literature

LABELING FAILURE-PHASE 4 STUDIES

• Novantrone (mitoxantrone): approved for progressive multiple sclerosis

• Two required Phase 4 studies– A study to evaluate incidence of

cardiomyopathy in patients prospectively monitored

– A “real-life” study to see if patients monitored according to labeling-insurance records for 400+ patients

LABELING FAILURE-PHASE 4 STUDIES

• Novantrone (mitoxantrone): approved for progressive multiple sclerosis– Data show:

• Cardiomyopathy occurs at much lower cumulative doses than previously believed

• Very few patients are monitored appropriately

AVAILABLE ACTIONS

• Early communications– Press Release, Public Health Announcement

• Pergolide and valvulopathy• Diastat failures

– Physician information sheets• AEDs and suicidality• Teratogenicity and Lamictal

– Early communications• Botox and botulism

AVAILABLE ACTIONS

• Labeling changes– Tysabri and PML, Liver injury– Dopamine agonists and impulsive behaviors– Novantrone– Carbamazepine

AVAILABLE ACTIONS

• Dear Health Care Practitioner Letters• Name Change

– Reminyl to Razadyne• Removal from Market

– Tysabri– Pergolide (“compassionate IND” instituted)

AVAILABLE ACTIONS

• Require studies– AEDs and generic “failures”

• Restricted distribution/observational studies– Tysabri

• Extensive education campaigns and other changes– Lamictal/Lamisil errors– Change carton appearance

FOOD AND DRUG ADMINISTRATION AMENDMENTS ACT

• Title IX-Drug Safety• Sec. 901

– FDA may require studies or clinical trials at the time of approval (or after approval if new safety information)

FOOD AND DRUG ADMINISTRATION AMENDMENTS ACT

• May require studies or clinical trials to:– Assess known serious risk related to drug use– Assess signals of serious risk related to drug

use– Identify an unexpected serious risk when

available data indicates the potential for a serious risk

FOOD AND DRUG ADMINISTRATION AMENDMENTS ACT

• Before a study/trial may be required, FDA must find:– AE reporting and “active postmarketing risk

identification and analysis system” are not sufficient to meet the purposes;

– A post-approval study is not sufficient before requiring a clinical trial

FOOD AND DRUG ADMINISTRATION AMENDMENTS ACT

• Required labeling changes– Agency must promptly inform sponsor of new

safety information– Within 30 days, the sponsor must submit a

labeling supplement or tell us why not– Rapid turnaround by Agency

FOOD AND DRUG ADMINISTRATION AMENDMENTS ACT

• Required labeling changes– Within 15 days of conclusion of discussions,

Agency can issue an order for a labeling change– Within 15 days of an order, sponsor must

submit a labeling supplement, or within 5 days, sponsor may appeal the order

FOOD AND DRUG ADMINISTRATION AMENDMENTS ACT

• Risk Evaluation and Mitigation Strategies (REMS)-Timetables Required– Pre-approval

• FDA may determine REMS needed to ensure benefits outweigh the risks

– Post-approval (if new safety information)• If needed to ensure benefits outweigh risks

NEW AGENCY SAFETY INITIATIVES

• Safety First– Associate Directors for Safety– Dedicated Project Manager for Safety– Standardized Record Keeping for Safety Issues– Deadlines to be applied for action on safety

issues

WHERE AGENCY CAN IMPROVE

• No systematic surveillance outside AERS• Most assessments take too long• Have not consistently followed up requests to

industry adequately (either for review of the data or requested labeling changes)

• Coordination with consulting divisions can be more efficient

WHERE AGENCY CAN IMPROVE

• May propose action with little prior notification to industry

• May confuse, mislead, (anger?) relevant constituencies with early communications, public announcements

WHERE INDUSTRY CAN IMPROVE

• Slow to respond to Agency requests• Unnecessary negotiations over labeling

– Class labeling• Often don’t inform Agency of a problem (even if

considerable work has been done)• May meet technical requirements for informing

Agency (e.g., PSUR) but problems not flagged

WHERE INDUSTRY CAN IMPROVE

• Even if Agency informed, may not have a proposal for a comprehensive plan

• On the other hand, some proposals appear to be overly “negative”– Proposed labeling may describe adverse event

which we all agree is not causally related to drug

THE FUTURE

CHANGE WE CAN BELIEVE IN

OR

SOLUTIONS FOR AMERICA