1

Pharmacovigilance and the role

of Scientific Literature –

Challenges and Solutions

Presented by: Joyce de Langen, pharmacist

Sr Solution Manager Pharmacovigilance,

Elsevier Life Sciences

Date: 23rd of March 2016

2

Take Home Message

• Scientific and medical literature is a relevant source of adverse drug

reactions

• Literature screening for single adverse drug reactions does have

impact on patient safety and the life cycle of a drug

• MAHs are facing a number of challenges with regard to literature

management for pharmacovigilance

• Software solutions for literature management improve patient safety,

increase efficiency in literature management, and improve regulatory

compliance

3

Pharmacovigilance

• Science and activities relating to the detection, assessment,

understanding and prevention of adverse drug reactions (ADRs) or

any other medicine-related problem

• Pharmacovigilance is a science and activity that improves patients’

safety and public health

• Traditionally Pharmacovigilance is based on the gathering,

assessment and analysis of adverse drug reactions spontaneously

reported by Health Care Professionals to the regulatory authorities.

4

Preventing Adverse Drug Reactions Improves Public Health

• 5 % of all hospital admission in EU are related to adverse drug

reactions

• 28% of patients visit emergency department of hospital due to

adverse events

• Averse drug reactions are the 5th most common cause of death of

hospitalized patients.

• Nearly 197,000 death per year due to adverse drug reactions

• Costs 75 Billion US Dollars per year

- Lazarou J et al. Incidence of ADRs in hospitalized patients. JAMA 1998, 279 (15) 1200-1205

- Classen DC et al. ADRs in hospitalized patients: excess length of stay, extra costs, and attributable mortality.

Obstet. Gyncol Surv 1997, 52 (5):291-292

-Ahmad SR: Adverse drug event monitoring at FDA. J Gen Intern Med 2003, 18 (1): 57-60.

5

Literature is the 4th largest source of AE reporting

Adverse events reported in literature can have a high impact — major drug recalls

(e.g., Vioxx, Baycol) were initiated by published adverse reactions.

Often overlaps with reports from health professionals

FDA Adverse Event Reporting System (FAERS) Quarterly Data Files, Q2 and Q3 2013

Reported adverse events by report source,

Q2 & Q3 2013

N = 34,469 unique events

Overlap of adverse events sourced from

literature by reporting sources, Q2 & Q3 2013

N = 4,691 unique events

6

A. Good reporting Practice Spontaneous case reports of adverse events submitted to the sponsor and FDA, and reports from other sources, such as the medical literature or clinical studies, may generate signals of adverse effects of drugs. The quality of the reports is critical for appropriate evaluation of the relationship between the product and adverse events. FDA recommends that sponsors make a reasonable attempt to obtain complete information for case assessment during initial contacts and subsequent follow-up, especially for serious events, and encourages sponsors to used train…

VI.B.1.1.2. Literature reports The scientific and medical literature is a significant source of information for the monitoring of the safety profile and of the risk benefit balance of medicinal products, particularly in relation to the detection of new safety signals or emerging safety issues. Marketing authorization holders are therefore expected to maintain awareness of possible publications through a systematic literature review of widely used reference databases (e.g. Medline, Excerpta Medica or Embase) no less frequently than once a week. The marketing authorization holder should ensure that the literature review includes the use of reference databases that contain the largest reference of articles in relation to the medicinal product properties

Growing regulatory pressure and focus on literature

monitoring in EU & US

Marketing authorization holders are therefore

expected to maintain awareness of possible

publications through a systematic literature

review of widely used reference databases

(e.g. Medline, Excerpta Medica or Embase)

no less frequently than once a week.

The quality of the reports is critical for

appropriate evaluation of the relationship

between the product and adverse events.

7

Quality in literature review

• All MAHs need to implement a systematic approach to collect

information about suspected ADRs from literature sources

• This systematic approach should be:

– Documented in a Standard Operating Procedure

– Undergo Periodic Quality Control Checks >> to determine efficiency

• MAH should track/document:

– Search strategy

– Source Databases used

– Date of search

– Results of search >> specifically in case of ‘No results’

– Date of review & QC

– Names, details reviewer & QC

8

• Systematic review of widely used reference databases (eg. Medline,

Embase) and local journals in countries were medicinal product has

a marketing authorizaton

• Literature search should cover scientific & medical literature

including

– Full text or abstract publications

– Information presented at scientific meetings & conferences

– Systematic reviews & meta-analysis

– Data from competitor products

– Legal documents like Patents etc

– Dissertations & Thesis

– Lay press

• At least once weekly

Systematic review of literature for Safety Signals

9

Types of articles on ADRs between 1999-2003

10

Example 1: Thalidomide-induced Phocomelia (1961)

McBride WG. Thalidomide and congenital abnormalities. Lancet. 1961;278:1358

11

To consider when reviewing literature articles for ICSRs

• Quality and completeness of report

• Report is not always proven, more opinions

• Under-reporting

• Exposure data are often missing >> frequency of ADR difficult to

determine

• Patient identifiers are missing or are presented as aggregated data

• Confounding by indication might play a role

ADRs may be the result of external/human factors (eg. Medication

errors)

• Biased reporting triggered by media attention

12

Literature and impact on benefit-risk balance of medicinal

products

• Literature monitoring for adverse drug reactions might reveal safety

signals that impact the benefit-risk balance of a medicinal product

• Several safety signals detected in literature reports had impact on

the life cycle of the drug

• Reporting rates can differ between reports from literature and other

spontaneous reported ADRs depending on the type of adverse

reaction reported

13

Central system (Eudravigilance) supports detection of

signals of new or changing safety issues

• In 2014, 2,030 potential signals were reviewed:

– 86.7% of potential signals originating from EudraVigilance.

– 8.6% from scientific literature (vs 5% in 2013)

– 3.2% from other regulatory authorities

– 1.5% from other sources.

• 90 signals confirmed

– 40% of the assessed

signals resulted in a

recommendation for an

update of the product

information

2014 annual report on EudraVigilance

14

Safety information from spontaneous and literature adverse

reactions reports differ

Klose J, Fröhling S, Kroth E, Dobmeyer T, Nolting A. Safety information from spontaneous and literature adverse reactions

reports: a comparison. Ther Innov Regul Sci. 2013;47:248–55.

Extracted from Table 4

Reporting rates can differ depending on the type of adverse reaction reported

Drug Substance System Organ Class Literature Cases (%) Spontaneous Cases (%) % Difference

Acetylsalicylic Acid

Nervous System

Disorders 25.6 8 17.6

Gastrointestinal

Disorders 8.4 25.4 17.0

Fentanyl Injury, poisoning,

complications 35.9 7.5 28.3

Alendronic acid

Gastrointestinal disorders 4.6 21.0 16.5

Injury, poisoning,

complications 28.3 5.4 22.9

Tamsulosin Injury, poisoning,

complications 50 4.1 45.9

Etoposide Congenital, familial,

genetic 0 24.3 24.3

15

Example 2: Tamsulosin and ‘Floppy Iris Syndrome’ (2005)

Chang DF, Campbell JR,. Intraoperative floppy iris syndrome associated with tamsulosin.

J Cataract Refract Surg. 2005;3: 664-73

• Intraoperative floppy iris syndrome

occurred in approximately 2% of a

cataract surgery population

• Appeared to be caused by

tamsulosin, a systemic sympathetic

alpha-1A antagonist

• Chang et al. mention 15 patients with

IFIS

• At the time of publication, none had

been reported to the Regulatory

Authorities!

16

Example 3: Off-label use of GM-CSF in HIV-patients

• A systematic qualitative review of the literature was performed to

assess the safety of granulocyte macrophage colony-stimulating

factor (GM-CSF)

• Off-label use in US for treatment of neutropenia in AIDS-patients

• Concerns about the safety as in vitro data showed HIV up-regulation

by GM-CSF

• Meta-analysis showed an increased risk for viral-replication caused

by GM-CSF in AIDS-patients who where not treated with anti-

retroviral products

• Signal was detected based on information from in vitro studies

published in scientific literature

Challenges in Literature Management

for Pharmacovigilance

18

Challenges in literature screening

• Lack of harmonization across Regulatory Authorities and Industry

thought leaders

• Building the perfect search strategy. Does it even exist?

• Meeting the authority requirements and increased attention for

literature management during inspections

19

Challenges in literature screening: Lack of Harmonization

across Regulatory Authorities & thought leaders

20

ICH E2D

Regularly screen worldwide scientific literature by assessing widely

used systematic literature reviews or reference databases

Frequency: according to local/national requirements or at least every

two weeks

Scientific and medical literature

including relevant published meeting & conference abstracts

Including draft manuscripts (Article in press & Article in Process)

Reporting Rules

One form per identifiable patient

Report Source: publication reference

Copy of article might be required (differs between local authorities)

21

ICH E2D

• Local language journals need to be monitored by local affiliates

• Day 0 = The moment the MAHs (or one of his third parties) has

received the minimum criteria for needed to meet the reporting rules.

• When the product source, brand or trade name is not mentioned, the

MAH should assume that it was its product, but the case report

should mention that the brand could not be identified.

• When an article mentions multiple drugs, only a report should be

submitted for the product identified as suspect by the author of the

article

22

FDA

• Serious, unexpected adverse experiences reported in scientific

literature (or in any unpublished scientific paper) that are know to

the applicant must be submitted within 15 days.

• Reports of serious, unexpected adverse experiences described in

scientific literature should be submitted for products with the same

active substance as a product marketing in the US. This is true even

if the excipient, dosage forms, strengths, routes of administration

and indications vary.

• Serious unexpected adverse experiences based on a foreign

language article or manuscript should be translated by the applicant

into English promptly….

23

EMA

• Reports of suspected adverse reactions from scientific and medical

literature, including relevant published abstracts from meetings and

draft manuscripts, should be reviewed and assessed by MAHs to

identify and record ICSRs originating from spontaneous reports, or

non-interventional post-authorisation studies

• If multiple medicinal products are mentioned in the publication, only

those which are identified by the publication’s author as having at

least a possible causal relationship with the suspected adverse

reaction should be considered by the concerned MAHs’

24

EMA -Exclusions

• Exclusion based on primary source country or country of origin of

the ADR, if MAH can demonstrate that the suspect drug never has

been supplied or marketed in that country

• Literature ICSRs from local language journals from countries where

MAH never has supplied or marketed the drug

• Literature ICSRs from analysis from competent authority database

within EU.

• Literature with data analysis from public databases or literature

summarizing aggregated data

25

Challenges in literature screening: When to start?

• When do you start your literature surveillance of a new medicinal

product?

A. After the submission of the application for marketing authorization

B. After the regulatory approval of the marketing authorization

C. After the marketing of an authorized medicinal product

26

Challenges in literature screening: When to start?

• When do you start your literature surveillance of a new medicinal

product?

A. On submission of the application for marketing authorization

B. After the regulatory approval of the marketing authorization

C. After the marketing of an authorized medicinal product

It is expected that literature screening should start on submission of a

marketing authorization application and should continue while

authorization is active.

27

Challenges in literature screening: which sources to monitor?

• Search well recognized scientific and medical journals

• At least two well recognized medical databases Embase

– Medline

– Cochrane Library

– CINAHL

– Other relevant databases

• Conference abstracts, draft manuscripts

• Local Language Journals

28

Challenges in literature screening: handling large amount

of data Number of records added to Embase each year

Embase, July 2015 Millions of Records

The large amount of data increases workload and costs, and may even increase the

risk of losing oversight of the data. Quality of the reports may decrease. Potentially

relevant information might be missed and the risk for missing signals might be

increased. Growing amount of data to be processed requires a structured approach

to handle all data.

29

Searching for safety-relevant information

• Optimizing search strategy by

– Choosing the best source databases and search engines

– Creating the best possible search strategy and search terms

– Using the right search limits

– Using defined review criteria

30

Choice of source databases and Search Engines

• Databases differ in accessibility, coverage & overlap

• No single database has full sensitivity for literature screening

• Periodic literature search and review should be performed in

multiple, carefully selected databases

31

Choice of search strategy

• Balancing between recall and precision = balancing between

‘capturing all relevant information’ and ‘avoiding all noise’

• Ideally the goal would be to build a search strategy with low recall

with high precision

• However, usually when recall , precision ; and visa versa

• So in real-life the goal of building a search strategy is to keep

precision as high as possible while recall is optimized

32

Choice of search strategy

• In case of a very large number of results from the search strategy,

MAHs may prepare two different search strategies and run these

parallel on a weekly basis:

• One for ICSR detection

• One for signal detection

• Search strategy for ICSR detection can exclude records for

pharmaceutical forms or routes of administration that are not

approved for that MAH

33

Challenges in literature screening: ongoing regulatory focus

during inspections

https://www.gov.uk/government/statistics/pharmacovigilance-inspection-metrics-2009-to-present

Finding 2011-2012 2012-2013 2013-2014 2014-2015

Critical

Findings NA NA NA NA

Major Findings 2.3% 15% 11% 11%

Minor Findings 11.2% 8% 11% 5%

34

Challenges in literature screening: Integration EMA MLM

Services

• Since September 2015, EMA started insource literature monitoring

for 300 active chemical substance groups (generics) and 100

herbal substance groups in order to alleviate workload for MAHs

• MAHs need to integrate the dataflow resulting from the EMA MLM

Service into their existing workflow for literature management and

ICSR-management

35

Challenges in literature screening: Integration EMA MLM

Services

• Pharmaceutical manufacturers are responsible for monitoring

scientific and medical publications in local journals

• EMA does not aggregate data for PSURs. Literature searches for

PSURs should be wider than those for individual adverse reaction

cases as they should also include studies reporting safety outcomes

in groups of subjects and other products, including:

– pregnancy outcomes (including termination) with no adverse outcomes

– use in pediatric populations

– compassionate supply, named patient use

– lack of efficacy

– asymptomatic overdose, abuse or misuse

– medication error where no adverse events occurred

– important non-clinical safety results

36

Summary: Challenges in literature screening

• Lack of harmonization across Regulatory Authorities and Industry

thought leaders

• Building the perfect search strategy

• Increasing amount of scientific information needs to be processed

- Increase of costs, more resources needed

- Risk for missing safety information/safety signals

- Difficulties with management of work process/workflow

• Ongoing attention for literature searching during inspections

• Integration EMA MLM Services into company literature

management system

37

Distribution of keywords between abstract and full-text

37 Information extraction from full text scientific articles: where are the keywords?

PMID[12775220]

Conclusion:

Although the abstract contains the best ratio of keywords per total of words, other sections of the

article may be a better source of biologically relevant data

Keyword measure (K) – depends on number and strength of co-occurrence relations within the text

Corpus: 104 articles published in Nature Genetics from June

1998 (volume 19, issue 2) to June 2001 (volume 28, issue 2),

which comply with the AIMRD structure.

“Major” keywords are distributed evenly between

sections, the difference is more noticeable for

“minor keywords”

38

Full text contains more information

38

39

Take Home Message

• Scientific and medical literature is a relevant source of adverse drug

reactions

• Literature screening for single adverse drug reactions does have

impact on patient safety and the life cycle of a drug

• MAHs are facing a number of challenges with regard to literature

management for pharmacovigilance

• Software solutions for literature management improve patient safety,

increase efficiency in literature management, and improve regulatory

compliance

Solutions for literature

management challenges

41

Literature management solutions: decrease risk of missing

information without redundant reviewing

Make sure all relevant

articles are captured

• Capture data from

most comprehensive

source of journals/

conference abstracts

• Develop custom

search strategies to

find all relevant data

Avoid redundant reviews

of the same input

• Save time and avoid

redundant reviewing with

automatic deduplication

of articles

Save time and

stay current

• Stay current and

work more efficiently

with automated

article curation

process

42

Literature management solutions: improve workflow

management and regulatory compliance

Track review process

in case of audit

• Demonstrate procedures

done correctly and

on time with traceable

review process

Improve article pipeline

management

• Identify most relevant

articles with text mining

• Ensure appropriate

personnel receive/

review required data

with alert system

Capture metrics behind

article reviews

• Calculate efficiency

and ensure strict

quality control by

capturing metrics

43

Introducing QUOSA PV

Empowering rapid, transparent literature surveillance and case triage

Conference

Documents

Online literature

database

Journal RSS

Feeds

PSURs

ICSRs

OTHER

SAFETY &

COMPLIANCE

REPORTS

A browser-based tool that enables pharmacovigilance

teams to save time and money by centralizing and

automating the literature review and triage process

Receive alerts to

rapidly identify

adverse events

in literature

Review and

annotate articles

from a scalable

central library

Use pre-formatted

output to easily

create reports

44

Introducing QUOSA PV

Empowering rapid, transparent literature surveillance and case triage

Conference

Documents

Online literature

database

Journal RSS

Feeds

PSURs

ICSRs

OTHER

SAFETY &

COMPLIANCE

REPORTS A combination of software and services that allows customers to

reduce risk, remain compliant and ensure that workgroups have

the latest scientific literature

Supervisors

Track deadlines

And bottlenecks

• Automatic alerts import

• Article deduplication

• E2B case data export

• Document listing

exports in Vancouver

format for PSURs

Administrators

View all aspects of

the review process

45

Reviewer view — Decision tree supports rapid identification

and capturing of cases

View abstracts or full-text and classify relevance with a fully trackable process

• View abstracts

or full text

• Audit trail for

every action

• Supports

multi-drug

annotation

46

Reviewer view — Decision tree supports rapid identification

and capturing of cases

Collaborate with colleagues for additional input/specialist review

47

Take Home Message

• Scientific and medical literature is a relevant source of adverse drug

reactions

• Literature screening for single adverse drug reactions does have

impact on patient safety and the life cycle of a drug

• MAHs are facing a number of challenges with regard to literature

management for pharmacovigilance

• Software solutions for literature management improve patient safety,

increase efficiency in literature management, and improve regulatory

compliance

48

Our mission in pharmacovigilance

Elsevier provides the solutions necessary for

Pharmacovigilance and drug safety groups

to be more efficient, stay compliant and

mitigate risks.

Thank you! Any questions?