1 of 18

5 Principles (5P’s) of Advancing Inclusive Research (AIR)

March 7, 2025

Shalini Hede, PharmD

1

2 of 18

The integration of inclusive research principles in clinical development is increasingly imperative. We propose to derive a comprehensive and unified set of principles that address inclusive research gaps and are adopted by all levels of the Roche organization across the clinical development timeline.

Problem statement

Despite significant pressure from Regulatory agencies to ensure the inclusion of representative patient cohorts in drug development initiatives, biotechnology companies continue to conduct clinical trials that do not fully represent diverse patient populations.

Medications are not thoroughly examined for safety, efficacy, and accessibility across all demographic groups.

Outcomes of clinical trials for marginalized populations remain uncertain and fail to provide valid real-world evidence.

Proposed solution

WHY: At Roche, we strive to ensure that

all populations are represented in biopharma research and development by Advancing Inclusive Research (AIR).

2

3 of 18

The 5 Principles (5Ps) Framework

The 5P framework was derived from simulating several studies in

a pharmaceutical late-stage portfolio where components formed the basis

for a data-informed framework to measure, standardize, and systematically define inclusive research.

5Ps can be considered a conceptual structure for a measurable,

transparent definition of clinical trial representativeness that

can function within organizational constraints.

3

4 of 18

The 5Ps of Advancing Inclusive Research (AIR)

eCRF=electronic case report forms with self-reported race / ethnicity categories applicable globally; SDOH=social determinants of health; SOGI=sexual orientation and gender identity

WHO?

Population Health�(Biological/�Genetic/ Demographic ) Considerations

1

WHERE?

Data-Informed Country & Site Selection to reach representative

patient cohorts

a. Geographical Proportionality Trial Enrollment (Region)

b. Accurate Demographic Representation (Country)

c. Representative real-world population (Site selection)

2

WHAT?

End-to-End

Inclusive

Trial Design (User-informed

and Data-Driven)

3

HOW?

Maximize Patient

Access to Trials

5

WHY ME?

Patient- Reported Data Collection Standards

(incl. eCRF,

SDOH, SOGI)

4

4

5 of 18

Population Health Considerations WHO are patients with the disease?

WHO

PRINCIPLE 1

WHERE

WHAT?

WHY ME?

H0W?

Population Science descriptors (via literature reviews) translated and captured in HE&PS DAPA reports (#20) across Oncology, Neurology, Neuroimmunology, Ophthalmology, Respiratory, and Cardio-metabolic-renal disease

Biological & Genetics Considerations:

  • Epidemiology
  • Ancestry
  • Risk factors
  • Others include:
    • Biomarkers
    • Disparities in diagnosis
    • Disease characteristics and progression
    • Treatment patterns
    • Morbidity and mortality
    • Quality of care

Demographics Considerations:

  • Race & Ethnicity
  • Age & Sex

5

6 of 18

Data-Informed Country & Site Selection

WHERE to reach representative patient cohorts

WHO

PRINCIPLE 2

WHERE

WHAT?

WHY ME?

H0W?

    • Global trial enrollment by region is recommended in ICH17 guidelines on multiple regional clinical trials and should be proportional to the geographical burden of disease, as defined by appropriate reference metrics
    • Data sources such as the US Cancer Statistics (USCS) for cancer in the US and the IHME GBD study globally, can be leveraged based on data quality, representativeness, and disease area coverage to provide epidemiology estimates at the indication or disease area level.

a.

Geographical Proportionality

Trial Enrollment

(Region)

    • Country-specific enrollment goals should adhere to standardized country-specific requirements and frameworks. Diversity outside the US cannot be captured or measured by US-specific demographic frameworks alone, such as the United States Office of Management and Budget’s (US OMB) standards on race and ethnicity

b.

Enable Access and Commercialization (Country)

    • Data-informed opportunities to maximize inclusion that leverage real-world evidence and ensure clinical trials are representative of real-world patient populations. Data sources such as the US Census, USCS, SEER, EHR-derived real-world data (e.g., TrinetX, IHME, Flatiron), and genomic databases are available to obtain epidemiological estimates for US and global populations.
    • Other HE&PS resources: AIR Calculator, R-index

c.

Representative real-world population

(Site)

Enrollment Goal Setting Guidance, DAP Template, AIR Calculator, R-Index (links)

6

7 of 18

End-to-End Inclusive Trial Design

WHAT considerations to include in protocol design

WHO

PRINCIPLE 3

WHERE

WHAT?

WHY ME?

H0W?

Artificial intelligence algorithms combined with real-world data can potentially improve several aspects of clinical trials including:

  • Patients screening
  • Predict likelihood of patients who will enroll
  • Extract features from electronic health records (EHRs)

Data-driven

Include input from providers, patient/patient advocacy groups, and investigators on the study trial design.

New Roche PDG mandate: New protocols due to be finalised Q1 ‘24 onwards are developed with

consideration for the patient.

    • Use Framework for patient input during protocol development
    • Include appropriate patient-centric tactics

User Informed

ASCO evidence-based recommendations that laboratory tests should be used as exclusion criteria only when necessary. (link)

FDA has also developed a series of guidelines for eligibility criteria, however these are not standards and have not been widely implemented by many companies.

7

8 of 18

Patient-Reported Data Collection Standards

(eg. eCRF, SOGI)

WHO

PRINCIPLE 4

WHERE

WHAT?

WHY ME?

H0W?

Current Roche eCRF Race or

Ethnicity Categories

Proposed Roche eCRF Race or Ethnicity Categories

Race or Ethnicity eCRF

Sub-categories

Directive No 15 (1997)

Directive No 15 (2024)

HISPANIC OR LATINO

HISPANIC OR LATINO

Mexican

Hispanic or Latino

Hispanic or Latino

Puerto Rican

Salvadoran

Cuban

Dominican

Guatemalan

Other Hispanic or Latino culture or origin

ALASKA NATIVE OR

AMERICAN INDIAN

ALASKA NATIVE OR

AMERICAN INDIAN

North American Indian Or Alaska Native

American Indian or

Alaska Native

American Indian or

Alaska Native

Central American Indian

South American Indian

Other Indigenous Not Listed

ASIAN

ASIAN

Chinese

Asian

Asian

Filipino

Japanese

Korean

South Asian

Vietnamese

Other Asian Not Listed

BLACK

BLACK

African American

Black or African American

Black or African American

African Caribbean

Black African

Other Black Identity Not Listed

INDIGENOUS AUSTRALIAN OR PACIFIC ISLANDER

INDIGENOUS AUSTRALIAN OR PACIFIC ISLANDER

Aboriginal Or Torres Strait Islander

Native Hawaiian or

Other Pacific Islander

Native Hawaiian or

Other Pacific Islander

Maori

Native Hawaiian

Other Pacific Islander Not Listed

MIDDLE EASTERN,

NORTH AFRICAN, OR WHITE

MIDDLE EASTERN OR

NORTH AFRICAN

Middle Eastern

White

Middle Eastern or

North African

North African

WHITE

White

White

White

1. What is your sexual orientation?

  1. Heterosexual
  2. Lesbian
  3. Gay
  4. Bisexual
  5. Queer
  6. Questioning
  7. Other, please specify
  8. Prefer not to answer

2. What is your current gender identity?

  1. Cisgender male
  2. Cisgender female
  3. Transgender male
  4. Transgender female
  5. Nonbinary
  6. Other, please specify
  7. Prefer not to answer

Optional SOGI Questionnaire for

Clinical Research Survey

Reconciliation of eCRF to FDA/OMB Race Categories

8

9 of 18

Patient-Reported Data Collection Standards

(cont. SDOH Questionnaire)

WHO

PRINCIPLE 4

WHERE

WHAT?

WHY ME?

H0W?

Education

What is your level of education/schooling?

[ ] Elementary School

[ ] Middle School

[ ] Some High School

[ ] Graduated from High School

[ ] Some College

[ ] Graduated from College

[ ] Post-Graduate

Employment Status

Before your diagnosis, what was your employment status?

[ ] Full-time Employed

[ ] Part-time Employed

[ ] Self-employed (either full-time or part-time)

[ ] Retired

[ ] Not employed

Health Insurance Status

What type of health insurance do you have, if any. Please select one:

[ ] Private/Commercial

[ ] Medicare

[ ] Medicaid

[ ] Veteran Affairs (VA) Health Plan

[ ] Other

[ ] I do not have a health insurance

Income Status

What is your annual household income?

[ ] E< $15,000

[ ] $15,000 - $34,999

[ ] $35,000 - $49,999

[ ] $50,000 - $74,999

[ ] $75,000- $99,999

[ ] $100,000 - $149,999

[ ] >$150,000

Marital Status

What is your marital status?

[ ] Single (never married)

[ ] Married (including common law)

[ ] Separated

[ ] Divorced

[ ] Widowed

[ ] Domestic Partnership [ ] (same sex, opposite sex or unregistered)

Zip Code of Home Residence

What is the zip code of your home residence?

9

10 of 18

Maximize Patient Access to trials

WHO

PRINCIPLE 5

WHERE

WHAT?

WHY ME?

HOW?

Partnerships to enable access during trials ie. transportation

A fundamental and ethical aspect of study placement and

access is Roche will only conduct trials in countries where

a sustainable path to regulatory and access approval exists.

Post-trial access/reimbursement

programs

Demonstrating

Trustworthiness

10

11 of 18

Maximize Patient Access to trials

  • A well defined catalogue of invoiceable services

  • Built into study budget for consistency and sustainability

  • Sites determine who is best suited to provide each service

  • Rates vary depending on seniority of role performing the service (~2,500 - ~2,900 USD per patient)

1.

P5

Enhanced Patient Support Services (EPSS)

11

12 of 18

Where are the patients most affected by multiple myeloma?

121.9-194.0

60.4-121.9

36.5-60.4

22.5-36.5

4.3-22.5

Crude rate per 100,000

Principle 2

Data-driven country and site selection

Incidence per 100,000 people

Who are the patients most affected by multiple myeloma?

Principle 1

Biological/genetic and population health considerations

Patient demographics

Patient risk factors

Region

Number of cases

%

Africa

6882

4

Americas

52,190

28

United States

4779

3

East Mediterranean

7309

4

Europe

53,889

29

Southeast Asia

23,117

12

Western Pacific

44,541

24

Race

White

74%

Black

21%

Asian/Pacific Islander

3%

American Indian/Alaska Native

0.7%

Ethnicity

Non-Hispanic

91%

Hispanic

9%

Age

<40

1%

40-64

34%

≥65

65%

Sex at birth

Male

55%

Female

45%

Race

2-3x higher incidence in African Americas than other race/ethnic origins

Gender

1.5x higher in men than women

Age

Older age of diagnosis, median age of diagnosis 69 years old

Younger median age of diagnosis among Black patients compared to White patients

Family history

2-4x risk 1st degree relative with MM

Environmental exposures

Post 9/11 workers, exposure to agent orange in Vietnam War, pesticides and chemicals

What will the multiple myeloma trial design look like?

  • Broaden inclusion and exclusion criteria for late-stage studies
  • Optimize assessments for patients and healthcare system (eg, blood volumes, visits)
  • Co-creation with the Clinical Trials Expert Council

Principle 3

User informed end-to-end inclusive trial design

Why are we focusing on these patients?

  • Standardized eCRF, SDOH, SOGI (optional, self-reported)
  • Race/ethnicity collection with county and region categorizations; subject to regulation limitations
  • Report on OMB categories at analytic stage

Principle 4

Patient-reported demographic identification

How will patients access trials and treatment?

  • Patient-centric clinical trial elements
  • Standardized Enhanced Patient Support Services
  • Conduct trials in countries where there is a sustainable path to regulatory and access approval

Principle 5

Enabling access to trials

12

13 of 18

Where are the patients most affected by breast cancer?

Principle 2

Data-driven country and site selection

Overall incidence per 100,000 people

71.1-105.4

52.2-71.1

39.2-52.2

27.8-39.2

4.6-27.8

Crude rate per 100,000

Prevalence rate, %

Prevalence of HR+ HER2- breast cancer

Clinical trial site locations

Who are the patients most affected by breast cancer?

Principle 1

Biological/genetic and population health considerations

Patient demographics

What will the breast cancer trial design look like?

  • Broadened ANC Reference Range
  • Potential inclusion of patients infected with HIV, HBV, or HCV
  • Inclusion of patients with stable
  • brain metastases
  • Inclusion of male patients in breast cancer trial
  • Resolution of acute toxicity

Principle 3

User informed end-to-end inclusive trial design

Why are we focusing on these patients?

  • Standardized eCRF, SDOH, SOGI (optional, self-reported)
  • Race/ethnicity collection with county and region categorizations; subject to regulation limitations
  • Report on OMB categories at analytic stage

Principle 4

Patient-reported demographic identification

How will patients access trials and treatment?

  • Patient-centric clinical trial elements
  • Standardized Enhanced Patient Support Services
  • Conduct trials in countries where there is a sustainable path to regulator and access approval

Principle 5

Enabling access to trials

US Target

Number on study

Percent on study

Race

American Indian/Alaska Native

0.4%

2

1.4%

Black or African American

11.1%

15

10.4%

Asian

5.7%

12

8.3%

Native Hawaiian/Pacific Islander

0

0

White

82.9%

101

70.1%

Multiple (eg, Black and Asian)

1

0.7%

Not reported/unknown

13

9.0%

Ethnicity

Non Hispanic White

91.1%

115

79.9%

Hispanic any race

8.9%

20

13.9%

Not reported/unknown

9

6.3%

North America

United States

Europe

Germany

Greece

Italy

Spain

Turkey

United Kingdom

Asia

Japan

Republic of Korea

Singapore

Taiwan

South America

Argentina

Africa

South Africa

13

14 of 18

Where are the patients most affected by COPD?

Principle 2

Data-driven country and site selection

COPD prevalence, 2021

Age standardized incidence per 100,000 people

South Asia

North America

East Asia

Latin America

Sub-Saharan Africa

Clinical site locations

What will the COPD trial design look like?

  • Amendment of I/E criteria: inclusion of asthma and HIV/HBV/HCV
  • Integrating mobile nursing in trial offering for 18/27 hospital visits
  • Building Inclusive Respiratory Research Networks through enhanced site, patient community and sponsor connections
  • Co-created patient & caregiver study experience journey uncovered unmet needs which then were translated into action

Principle 3

User informed end-to-end inclusive trial design

Why are we focusing on these patients?

  • Standardized eCRF, SDOH, SOGI (optional, self-reported)
  • Race/ethnicity collection with county and region categorizations; subject to regulation limitations
  • Report on OMB categories at analytic stage

Principle 4

Patient-reported demographic identification

How will patients access trials and treatment?

  • Patient-centric clinical trial elements
  • Standardized Enhanced Patient Support Services
  • Conduct trials in countries where there is a sustainable path to regulator and access approval

Principle 5

Enabling access to trials

Who are the patients most affected by COPD?

Principle 1

Biological/genetic and population health considerations

Patient risk factors

Race

Latin American / Indigenous American ancestry had multiple protective factors (PRDM15, IL6R, CYP2F2P). Risk variants established in European-ancestry cohorts, such as KCNE2, GPR126

Age

NHBs are more likely to develop COPD at a younger age, with less cumulative smoking

Environmental exposures

Smoking, including secondhand smoke, global rural populations had higher COPD rates than urban populations, ambient particulate matter, household air pollution, occupational particulates, and ozone

North America

Canada

United States

South America

Argentina

Brazil

Chile

Africa

Kenya

South Africa

Europe

Austria

Denmark

Hungary

Netherlands

Sweden

Belgium

France

Israel

Poland

Switzerland

Bulgaria

Germany

Italy

Romania

Turkey

Czechia

Greece

Latvia

Spain

UK

Asia Pacific

Australia

Republic of Korea

China

New Zealand

Hong Kong

Philippines

India

Taiwan

Japan

Thailand

14

15 of 18

More variable; dependent on unique disease characteristics, and prioritization may differ across disease conditions

May be less dynamic; strategies can be less dependent on disease-specific factors

5P AIR Framework:

Interdependencies with Disease Condition

5

2

3

4

1

Population Health (Biological/Genetic/

Demographic) Considerations

Data-Driven Region, Country & Site

Selection

End-to-End Inclusive Trial Design (Data-driven & User-informed)

Patient-Reported data collection standards (eg. eCRF, SDOH, SOGI)

Enabling Access �to Trials

15

16 of 18

Applying 5Ps End-to-End

Population Health (Biological/Genetic/

Demographic) Considerations

Data-Driven Region, Country �& Site Selection

End-to-End Inclusive Trial Design (Data-driven and User-informed)

Patient-Reported data collection standards (eg. eCRF, SDOH, SOGI)

Enabling Access �to Trials

Pre Trial

Clinical Trial

Post Trial

Enrollment by global regions �should be proportional to the geographical burden of disease �both between and within countries

Enrollment by demographic �variables such as race, ethnicity, �age and sex should adhere to �country-specific standards

Understand country-specific operational constraints: capabilities, resource and regulatory guidelines

Guided by regulatory requirements, local demographics, disease burden for optimized representativeness �and efficiency

Established regional frameworks to benchmark to advance diverse trial enrolment on a global basis

If study targets are not met can activate resources or alternative sites for future trials

Decisions made at study level that are specific to each study

Collection of patient reported data to ensure equity and representativeness in trials (eCRF, SDOH, SOGI) to also reflect county appropriateness

Data collection should respect country and region categorizations

Modifications based on systematic vs ad hoc assessments of �available data

Modifications with the goal of removing barriers or ensuring appropriate considerations

Real world data can inform the 5Ps

Ensuring access to treatments and post-trial access programs

16

17 of 18

The 5Ps of Advancing Inclusive Research (AIR)

eCRF=electronic case report forms with self-reported race / ethnicity categories applicable globally; SDOH=social determinants of health; SOGI=sexual orientation and gender identity

WHO?

Population Health�(Biological/�Genetic/ Demographic ) Considerations

WHERE?

Data-Informed Country & Site Selection to reach representative

patient cohorts

a. Geographical Proportionality Trial Enrollment (Region)

b. Accurate Demographic Representation (Country)

c. Representative real-world population (Site selection)

WHAT?

End-to-End

Inclusive

Trial Design (User-informed

and Data-Driven)

WHY ME?

Patient- Reported Data Collection Standards

(incl. eCRF,

SDOH, SOGI)

HOW?

Maximize Patient

Access to Trials

1

2

3

4

5

17

18 of 18

Doing now what patients need next

18