NHS Galleri Clinical Trial Outcomes Data Request

GRAIL Bio UK Ltd and Queen Mary University London (QMUL), as joint Controllers, are requesting access to NHS England pseudonymised record-level Data in order to carry out follow-up analysis based on a cohort of patients who were recruited to a clinical trial called ‘NHS-Galleri’ following being invited to participate if they were found to meet the required eligibility criteria.

Note that members of the team running the trial are based at the Cancer Prevention Trials Unit (CPTU) at QMUL. For the sake of consistency, and as QMUL is listed as a Controller, when referring to QMUL throughout the application, this encompasses the team at CPTU from QMUL.

A new Multi-Cancer Early Detection (MCED) test has been developed that can detect many types of cancer from a single blood sample. This test is called Galleri (Registered Trademark). This trial aims to find out whether it is better at detecting cancer early, compared to other tests that the NHS currently uses. The purpose of NHS-Galleri is to demonstrate the clinical utility of the MCED blood test for individuals in a general screening population in a real world NHS setting. The rationale behind this trial is that MCED is a novel screening paradigm, and assessment of the use and impact of test results is necessary to enable integration into clinical practice. This will be the first randomised, double blind, controlled trial statistically powered to assess clinical utility of a MCED test.

Recruitment to the trial has been aided by NHS England's DigiTrials team via a sister Data Sharing Agreement (DARS-NIC-456778-J0G3H) and took place over 10 – 12 months from August 2021 with a recruitment target of 140,000 people. Recruitment successfully completed and the target was met. Different geographic areas were targeted for recruitment, beginning with North-West England. Recruitment in a particular area lasted for 4 – 6 weeks, before moving on to another area. Recruitment was managed through mobile units on scheduled routes around a particular area. Batches of invitations were constructed to target whatever area the mobile units were in at the time of writing out to potential participants. Potential participants were only contacted once and there were no follow-ups sent after the original invitation letter.

Requesting Not to Take Part in Trial:

In addition to the National Data Opt-out, members of the public were able to specifically request not to be contacted for the NHS-Galleri Trial. GRAIL Bio UK Ltd promoted this information via the dedicated trial website (nhs-galleri.org) and NHS England did the same via a dedicated page on its website (https://digital.nhs.uk/services/nhs-digitrials/nhs-galleri-trial). There was also an option for people to register their request not to take part in NHS-Galleri by telephone. NHS England records these requests not to take part and ensure that anyone who has registered for this will be excluded from the cohort selection and as a result their details will not be on the list for invitations to be sent.

• Primary Objective of the NHS-Galleri Trial:

- For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of stage III and IV cancers diagnosed in the intervention arm as compared to the control arm following three annual test and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a. For a prespecified group of cancer types, including lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver / bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder.

b. For all cancer types, excluding prostate cancer.

c. For all cancer types.

• Secondary Objectives of the NHS-Galleri Trial:

1) For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of advanced cancers (defined as a cancer diagnosed at stage III or IV or one that results in a cancer-specific death) in the intervention arm as compared to the control arm following three annual tests and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a) First, evaluate for a statistically significant difference in a prespecified group of primary cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

b) If a statistically significant reduction in absolute numbers is found, continue by evaluating for a difference in all cancer types excluding prostate cancer.

c) If the above evaluations are both significant, evaluate for a difference in all cancer types.

2) For cancers that are routinely staged, determine whether there is a difference in the proportion of stage I and II cancers between the two arms of the study in the third testing round (i.e. for cancers diagnosed with an average 16-18 months of follow-up after the third test).

3) Evaluate the performance (overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three intervals of testing.

4) Evaluate the safety, including harms, of the testing pathway in the intervention arm among individuals with positive MCED test results.

a) Assess participant-reported anxiety among those receiving a positive MCED test result following the initial blood draw at various time points.

b) Examine the radiation exposure by participants associated with follow-up diagnostic procedures following a positive MCED test result

c) Describe the number and type of invasive procedures performed, and number of complications and deaths associated with follow-up diagnostic procedures.

5) Determine whether there is a statistically significant reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following the initial screening round and 12 months of follow-up (excluding cancers identified by the test performed at the 12-month visit)

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer;

c) For all cancer types.

6) Determine whether there is a reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following three annual tests and an average of 16-18 months follow-up after the third test

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer

c) For all cancer types.

7) Determine whether there is a statistically significant difference in cancer detection rates in the two arms of the study with 12 months follow-up after the first test.

8) Model the expected differences in cancer-specific mortality at 7 years post-randomisation based on cancers diagnosed within an average of 40-42 months of randomisation.

9) Summarize stage distribution per cancer type for the two arms, and compare the absolute numbers of stage IV alone as well as stage III and IV cancers, following the third screening and an average 16-18* months of follow-up.

10) Determine whether there is a reduction in the absolute numbers of stage III and IV cancers, excluding breast, cervical, and colorectal (i.e., those without a standard of care screening modality) as compared to the control arm following the third annual screening and an average 16-18 months of follow-up.

11) Assess the impact on cancer-specific mortality:

a) (Endpoint powered) Assess the potential for avoidance/postponement of cancer death by comparing the rates of death from cancer with-specific mortality rates among individuals receiving a positive MCED test at any of the 3 screening rounds in the intervention arm with that among individuals retrospectively testing positive at any of the 3 screening rounds in the control arm, at 3 years after the last study visit (retrospectively testing all samples from cancer deaths in the control arm).

b) (Endpoint not powered) Compare cancer-specific mortality in the intervention and control arms for all cancer types at a) 3 years and b) 6 years after the last study visit.

12) Compare the absolute numbers of stage IV cancers following the second blood draw and 12 months of follow-up, with prevalent cases excluded.

13) Compare the absolute numbers of stage III and IV cancers following the third blood draw and an average 16-18 months of follow-up, with cases from the first and second screenings excluded.

14) Assess the impact of use of the MCED test across three annual timepoints on healthcare resource utilization for cancer diagnosis and treatment.

15) Assess the potential impact of overdiagnosis by comparing the cumulative incidence (Kaplan-Meier curves) of cancers diagnosed after a positive baseline MCED test up to 36-48 months after randomization between the intervention and control arms (retrospectively testing baseline samples from all participants diagnosed with cancer in the control arm).

NHS ENGLAND DATA SETS REQUESTED:

> NDRS Rapid Cancer Registrations

> NDRS Cancer Registry

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT)

> NDRS National Radiotherapy Dataset (RTDS)

> NDRS Linked Hospital Episode Statistics (HES) Admitted Patient Care (APC)

> NDRS Linked HES Outpatients (OP)

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals)

> NDRS Linked Diagnostic Imaging Dataset (DIDS) (provided Summer 2022 onward)

> Emergency Care Data Set (ECDS) (provided Summer 2022 onward)

FREQUENCY - every 6 months beginning March 2022 (Rapid Registrations every 2 months from March 2022).

FROM AUGUST 2022 (Version 2 of this Data Sharing Agreement)) - Frequency of drops

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT) - no change

> NDRS Linked HES APC - no change

> NDRS Linked HES Outpatients (OP) - no change

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals) - no change

> NDRS National Radiotherapy Dataset (RTDS) – No change

> NDRS Linked Diagnostic Imaging Dataset (DIDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Linked Emergency Care Data Set (ECDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Rapid Cancer Registrations - Rapid Registrations will be provided every month rather than every two months from August 2022 to enable frequent monitoring of cancer diagnoses in the study.

> NDRS Cancer Registry - will be provided every month rather than every other month from August 2022 to enable frequent reporting of cancer diagnoses in the study and use the data as a filter for participant invitations for annual follow up timepoints. Expand historic gold standard registration field selection for 1995-2021 and addition of field to allow the trial team to distinguish provisional from finalised gold standard tumours. These changes are required for primary and secondary analysis. Data will be used to characterize the cohort for risk factors (prior cancers), and also assess test performance in a subpopulation of cancer survivors. The Status of Registration is required to enable accurate reporting and filtering of provisional data vs finalized data to internal and external stakeholders, and the interim analysis in the study. Death fields added to dataset specification for analysis of key secondary endpoints in the study. Death Data will also be used for reporting subject disposition in the analysis. Death Data, while requested with the Cancer Registry Data, should include information on all participants in the cohort (including participants that have not been diagnosed with cancer).

JUSTIFICATION FOR DATA SETS REQUESTED AND DATA MINIMISATION

> Cancer Registry and Rapid Registrations - These central NHS England datasets will be the sole source for the data required for primary and secondary objectives analyses in the study. They will be used to identify Primary endpoints, which are incidence and stage at diagnosis for cancer types that are stage-able (e.g., with available staging systems). Cancer is defined as any of the following cancers:

- Invasive solid cancer, excluding non-melanoma skin cancer

- Haematologic malignancies, including lymphoma, lymphoid leukemia, myeloma/plasma cell neoplasm, myeloid neoplasms (including myelodysplastic and myeloproliferative neoplasms with behaviour code 3 based on ICD-O-3.2*).

[*The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) is the standard classification system for the registration of cancers in the United States (and most areas of the world) and contains widely-accepted histologic categories of brain tumors.]

The Cancer Registry will also be used to identify cohort characteristics/baseline risk - Personal history of cancer assessed using limited set of data fields. Historical cancer data dating back to 1995 will be available.

> Hospital Episode Statistics (HES) Outpatients (OP), Admitted Patient Care (APC) and Emergency Care Data Set (ECDS), Cancer Waiting Times (Treatments and Referrals) (CWT) and Diagnostic Imaging Dataset (DIDS) - Will be used to identify Secondary endpoints (including safety objectives). Among all test positive cases, number of follow-up procedures and number of invasive procedures (including all biopsies, surgical interventions, bronchoscopy, thoracoscopy and endoscopy) to achieve diagnostic resolution (i.e. cancer diagnosis, non-cancer diagnosis, or no diagnosis and discharge from the diagnostic follow-up). Will also identify:

- Number and type of invasive procedures performed in false positive cases;

- Number of complications and deaths resulting from diagnostic procedures;

- Radiation exposure measured in mSv per participant due to test result-directed evaluations

> Systemic Anti-Cancer Therapy Data Set (SACT) and Radiotherapy Data set (RTDS) will also be used (in addition to the above HES, CWT and DIDs data) for health care resource utilization objectives.

The NHS-Galleri trial has the following Secondary Objective: "Evaluate the performance (overall episode sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three rounds."

Participants who have had cancer prior to enrolment may be at risk of a recurrence. Previous research has demonstrated the Galleri MCED is capable of detecting cancer recurrence (1). As NDRS registers primary cancers (and not cancer recurrences), GRAIL Bio UK Ltd are keen to establish when there is evidence that a patient has had a recurrence of their cancer. This is to complete a sensitivity analysis for the Positive Predictive Value (PPV) of the test in which the detection of a recurrence will be considered a True Positive (this is in contrast to the main analysis in which, because there is no new cancer registration, it is considered a False Positive). GRAIL Bio UK Ltd believe this analysis will help provide a more complete description of test performance.

(1) Schrag D, et al. Presentation at European Society for Medical Oncology (ESMO) Congress; September 9-13, 2022.

For the purposes of this analysis, GRAIL Bio UK Ltd will share the Study IDs of a small number of consented NHS-Galleri participants with NHS England for review as to whether these participants have a cancer recurrence. NHS England will convert the study IDs into internal patient IDs and provide an expert clinical review of each case to determine whether a patient has a recurrence. NHS England will do this based on the NDRS Datasets already listed in this Data Sharing Agreement.

***************************

Patient and Public Involvement and Engagement (PPI&E)

Ongoing PPI&E is conducted throughout the trial. Historical PPI&E was conducted to develop the participant facing materials and on questionnaires, and including specific questions to PPI groups around the sharing of pseudonymised Data with GRAIL Inc (US). Ongoing PPI&E includes, for example, PPI&E on new participant facing materials, survey feedback on the participant experience, and questionnaires for follow-up of participants. An embedded behavioural scientist in QMUL's Cancer Prevention Trials Unit (previously located at King’s College London) has been involved in, and will continue to be involved in, analysis of participant communications, and ethical and wellbeing issues within the course of the trial, as well as periodically advise the study organizers and other stakeholder on emerging themes.

In the setting up of the trial, GRAIL has taken seriously strategies for equity of access and taken steps to ensure that under-represented minorities were actively recruited into the trial. Amongst this is the invitation strategy, which identified and actively sent invitations to a greater proportion of individuals from these minorities to take part in the trial. It is hoped that this will positively contribute towards equity of access and inform the provision of cancer services in these areas in the future.

UK GDPR Legal Basis for the Processing of Personal Data:

GRAIL Bio UK Ltd, as joint Controller, is using Article 6(1)(f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of Personal Data, in particular where the data subject is a child." Processing Personal Data is necessary for GRAIL Bio UK Ltd’s legitimate interests which are described in this application. The Data to which access is requested are proportionate and necessary to achieve those interests. GRAIL Bio UK Ltd has completed a legitimate interests assessment (LIA) and are satisfied that the interests of the data subjects do not override their legitimate interests; that they would reasonably expect the processing and it would not cause unjustified harm. The data subjects interests and fundamental rights are protected through appropriate minimisation of fields and patient records being processed; protection of the Data in a secure environment, and guaranteeing secure destruction at any stage at the request of NHS England and/or after a defined period on completion of the project.

Additionally (as health Data is a special category of Personal Data), GRAIL Bio UK Ltd is using Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

QMUL, as joint Controller, is using Article 6:1(e): Specific task in the ‘public interest’ or task that has a clear basis in law, and additionally (as health Data is a special category of Personal Data) Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

Organisation Roles and Responsibilities:

• GRAIL Bio UK Ltd are a joint Controller and Lead for this DSA, who also process the Data. It is responsible for sponsoring and funding the trial, as well as the building and maintenance of the Clinical Records Management System.

• QMUL are a joint Controller who also process the Data and are responsible for implementing the trial and for primary objective analysis. The Principal Investigator and the Cancer Prevention Trial Unit (CPTU) are based at QMUL.

• GRAIL Inc. is the manufacturer of the MCED test, Galleri (Registered Trademark), and has set up the above UK subsidiary, GRAIL Bio UK Ltd, which is the sponsor of the NHS-Galleri trial. GRAIL Inc. will be receiving pseudonymised record-level NHS England Data from GRAIL BIO UK Ltd, and are therefore listed as a Processor in this DSA.

KCL will be subcontracted by QMUL to perform certain work related to the trial. KCL are not a Controller in this Data Sharing Agreement as they do not make any decisions determining the purposes and means of the processing in the NHS Galleri study, but, rather, remain under the direction of QMUL. No NHS England Data, as described in this Data Sharing Agreement, will be stored at KCL but will be accessed by substantive employees of KCL under affiliate QMUL staff status and subject to a Data Processor Agreement between QMUL and KCL (see Section 5b Processing activities for more details).

NHS England are listed as a strategic partner in relation to NHS-Galleri, but have no input into how the trial is run or determining the outcome. NHS England therefore do not consider them to be a Controller. Cancer Research UK appear on trial-related documentation by virtue of the fact that QMUL's Cancer Prevention Trials Unit are a Cancer Research UK Trials Unit. Cancer Research UK as an organisation in their own right have no involvement in the running of the trial, nor are they funding any of the trial activities. NHS England therefore do not consider them to be a Controller.

In the future, GRAIL Bio UK Ltd may receive commercial benefit (including intangible or indirect commercial benefits such as positive publicity) from the successful outcomes of the trial. GRAIL Bio UK Ltd and the NHS have entered into a partnership whereby the Galleri test will be piloted in clinical trials within the NHS in England. If the test is shown to work as intended in these clinical trials, the NHS may purchase the test from GRAIL and make the test routinely available in the future to benefit patient.

All outputs will be aggregated with small number suppressed as per the HES analysis guide or according to the specific data set's suppression guidance.

The results of the NHS-Galleri trial outputs will be published in peer-reviewed publications, presented at conferences, and used by NHS England in their decisions around implementation of the test within the NHS. The interim analysis, which will include analysis of a key secondary objective, which was completed in Q1 2024. The primary objective readout is expected by the end of 2025 or beginning of 2026. Planned dissemination activities will aim to enable the engagement with the scientific and policy making communities to ensure that knowledge developed by this research can benefit these communities to help the NHS detect cancer early, when it is easier to treat.

The Processors continue to receive NHS England Data to undertake ongoing data analysis for the study, but at the present time there are no Yielded Benefits to report under this amendment. This is because, as stated above, the outputs are not due until the end of 2026.

GRAIL Bio UK Ltd and King’s College London (KCL), as joint Controllers, are requesting access to NHS England pseudonymised record-level Data in order to carry out follow-up analysis based on a cohort of patients who are being recruited to a clinical trial called ‘NHS-Galleri’ following being invited to participate if they are found to meet the required eligibility criteria.

Note that members of the team running the trial are based at the Cancer Prevention Trials Unit (CPTU) at King's College London. For the sake of consistency, and as King's College London is listed as a Controller, when referring to King's College London throughout the application, this encompasses the team at CPTU from King’s College London. *** New for Version 5 of this agreement: Some members of the trial team will also be based at Queen Mary University of London (QMUL) from December 2023 - please see further details in this section below***

A new Multi-Cancer Early Detection (MCED) test has been developed that can detect many types of cancer from a single blood sample. This test is called Galleri (Registered Trademark). This trial aims to find out whether it is better at detecting cancer early, compared to other tests that the NHS currently uses. The purpose of NHS-Galleri is to demonstrate the clinical utility of the MCED blood test for individuals in a general screening population in a real world NHS setting. The rationale behind this trial is that MCED is a novel screening paradigm, and assessment of the use and impact of test results is necessary to enable integration into clinical practice. This will be the first randomised, double blind, controlled trial statistically powered to assess clinical utility of a MCED test.

Recruitment to the trial has been aided by NHS England's DigiTrials team via a sister Data Sharing Agreement (DARS-NIC-456778-J0G3H) and took place over 10 – 12 months from August 2021 with a recruitment target of 140,000 people. Recruitment successfully completed and the target was met. Different geographic areas were targeted for recruitment, beginning with North-West England. Recruitment in a particular area lasted for 4 – 6 weeks, before moving on to another area. Recruitment was managed through mobile units on scheduled routes around a particular area. Batches of invitations were constructed to target whatever area the mobile units were in at the time of writing out to potential participants. Potential participants were only contacted once and there were no follow-ups sent after the original invitation letter.

Requesting Not to Take Part in Trial:

In addition to the National Data Opt-out, members of the public were able to specifically request not to be contacted for the NHS-Galleri Trial. GRAIL Bio UK Ltd promoted this information via the dedicated trial website (nhs-galleri.org) and NHS England did the same via a dedicated page on its website (https://digital.nhs.uk/services/nhs-digitrials/nhs-galleri-trial). There was also an option for people to register their request not to take part in NHS-Galleri by telephone. NHS England records these requests not to take part and ensure that anyone who has registered for this will be excluded from the cohort selection and as a result their details will not be on the list for invitations to be sent.

• Primary Objective of the NHS-Galleri Trial:

- For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of stage III and IV cancers diagnosed in the intervention arm as compared to the control arm following three annual test and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a. For a prespecified group of cancer types, including lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver / bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder.

b. For all cancer types, excluding prostate cancer.

c. For all cancer types.

• Secondary Objectives of the NHS-Galleri Trial:

1) For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of advanced cancers (defined as a cancer diagnosed at stage III or IV or one that results in a cancer-specific death) in the intervention arm as compared to the control arm following three annual tests and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a) First, evaluate for a statistically significant difference in a prespecified group of primary cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

b) If a statistically significant reduction in absolute numbers is found, continue by evaluating for a difference in all cancer types excluding prostate cancer.

c) If the above evaluations are both significant, evaluate for a difference in all cancer types.

2) For cancers that are routinely staged, determine whether there is a difference in the proportion of stage I and II cancers between the two arms of the study in the third testing round (i.e. for cancers diagnosed with an average 16-18 months of follow-up after the third test).

3) Evaluate the performance (overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three intervals of testing.

4) Evaluate the safety, including harms, of the testing pathway in the intervention arm among individuals with positive MCED test results.

a) Assess participant-reported anxiety among those receiving a positive MCED test result following the initial blood draw at various time points.

b) Examine the radiation exposure by participants associated with follow-up diagnostic procedures following a positive MCED test result

c) Describe the number and type of invasive procedures performed, and number of complications and deaths associated with follow-up diagnostic procedures.

5) Determine whether there is a statistically significant reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following the initial screening round and 12 months of follow-up (excluding cancers identified by the test performed at the 12-month visit)

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer;

c) For all cancer types.

6) Determine whether there is a reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following three annual tests and an average of 16-18 months follow-up after the third test

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer

c) For all cancer types.

7) Determine whether there is a statistically significant difference in cancer detection rates in the two arms of the study with 12 months follow-up after the first test.

8) Model the expected differences in cancer-specific mortality at 7 years post-randomisation based on cancers diagnosed within an average of 40-42 months of randomisation.

9) Summarize stage distribution per cancer type for the two arms, and compare the absolute numbers of stage IV alone as well as stage III and IV cancers, following the third screening and an average 16-18* months of follow-up.

10) Determine whether there is a reduction in the absolute numbers of stage III and IV cancers, excluding breast, cervical, and colorectal (i.e., those without a standard of care screening modality) as compared to the control arm following the third annual screening and an average 16-18 months of follow-up.

11) Compare cancer-specific mortality in the intervention and control arms following the third screening and an average 16-18 months of follow-up, for a pre-specified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

12) Compare cancer-specific mortality in the intervention and control arms for all cancer types at

a) 3 years and

b) 6 years after the last study visit.

13) Compare the absolute numbers of stage IV cancers following the second blood draw and 12 months of follow-up, with prevalent cases excluded.

14) Compare the absolute numbers of stage III and IV cancers following the third blood draw and an average 16-18 months of follow-up, with cases from the first and second screenings excluded.

15) Assess the impact of use of the MCED test across three annual timepoints on healthcare resource utilization for cancer diagnosis and treatment.

16) Assess the potential impact of overdiagnosis by comparing the cumulative incidence (Kaplan-Meier curves) of cancers diagnosed after a positive baseline MCED test up to 36-48 months after randomization between the intervention and control arms (retrospectively testing baseline samples from all participants diagnosed with cancer in the control arm).

NHS ENGLAND DATA SETS REQUESTED:

> NDRS Rapid Cancer Registrations

> NDRS Cancer Registry

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT)

> NDRS National Radiotherapy Dataset (RTDS)

> NDRS Linked Hospital Episode Statistics (HES) Admitted Patient Care (APC)

> NDRS Linked HES Outpatients (OP)

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals)

> NDRS Linked Diagnostic Imaging Dataset (DIDS) (provided Summer 2022 onward)

> Emergency Care Data Set (ECDS) (provided Summer 2022 onward)

FREQUENCY - every 6 months beginning March 2022 (Rapid Registrations every 2 months from March 2022).

FROM AUGUST 2022 (Version 2 of this Data Sharing Agreement)) - Frequency of drops

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT) - no change

> NDRS Linked HES APC - no change

> NDRS Linked HES Outpatients (OP) - no change

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals) - no change

> NDRS National Radiotherapy Dataset (RTDS) – No change

> NDRS Linked Diagnostic Imaging Dataset (DIDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Linked Emergency Care Data Set (ECDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Rapid Cancer Registrations - Rapid Registrations will be provided every month rather than every two months from August 2022 to enable frequent monitoring of cancer diagnoses in the study.

> NDRS Cancer Registry - will be provided every month rather than every other month from August 2022 to enable frequent reporting of cancer diagnoses in the study and use the data as a filter for participant invitations for annual follow up timepoints. Expand historic gold standard registration field selection for 1995-2021 and addition of field to allow the trial team to distinguish provisional from finalised gold standard tumours. These changes are required for primary and secondary analysis. Data will be used to characterize the cohort for risk factors (prior cancers), and also assess test performance in a subpopulation of cancer survivors. The Status of Registration is required to enable accurate reporting and filtering of provisional data vs finalized data to internal and external stakeholders, and the interim analysis in the study. Death fields added to dataset specification for analysis of key secondary endpoints in the study. Death Data will also be used for reporting subject disposition in the analysis. Death Data, while requested with the Cancer Registry Data, should include information on all participants in the cohort (including participants that have not been diagnosed with cancer).

JUSTIFICATION FOR DATA SETS REQUESTED AND DATA MINIMISATION

> Cancer Registry and Rapid Registrations - These central NHS England datasets will be the sole source for the data required for primary and secondary objectives analyses in the study. They will be used to identify Primary endpoints, which are incidence and stage at diagnosis for cancer types that are stage-able (e.g., with available staging systems). Cancer is defined as any of the following cancers:

- Invasive solid cancer, excluding non-melanoma skin cancer

- Haematologic malignancies, including lymphoma, lymphoid leukemia, myeloma/plasma cell neoplasm, myeloid neoplasms (including myelodysplastic and myeloproliferative neoplasms with behaviour code 3 based on ICD-O-3.2*).

[*The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) is the standard classification system for the registration of cancers in the United States (and most areas of the world) and contains widely-accepted histologic categories of brain tumors.]

The Cancer Registry will also be used to identify cohort characteristics/baseline risk - Personal history of cancer assessed using limited set of data fields. Historical cancer data dating back to 1995 will be available.

> Hospital Episode Statistics (HES) Outpatients (OP), Admitted Patient Care (APC) and Emergency Care Data Set (ECDS), Cancer Waiting Times (Treatments and Referrals) (CWT) and Diagnostic Imaging Dataset (DIDS) - Will be used to identify Secondary endpoints (including safety objectives). Among all test positive cases, number of follow-up procedures and number of invasive procedures (including all biopsies, surgical interventions, bronchoscopy, thoracoscopy and endoscopy) to achieve diagnostic resolution (i.e. cancer diagnosis, non-cancer diagnosis, or no diagnosis and discharge from the diagnostic follow-up). Will also identify:

- Number and type of invasive procedures performed in false positive cases;

- Number of complications and deaths resulting from diagnostic procedures;

- Radiation exposure measured in mSv per participant due to test result-directed evaluations

> Systemic Anti-Cancer Therapy Data Set (SACT) and Radiotherapy Data set (RTDS) will also be used (in addition to the above HES, CWT and DIDs data) for health care resource utilization objectives.

**** New for Version 5 of this agreement - November 2023 ****

The NHS-Galleri trial has the following Secondary Objective: "Evaluate the performance (overall episode sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three rounds."

Participants who have had cancer prior to enrolment may be at risk of a recurrence. Previous research has demonstrated the Galleri MCED is capable of detecting cancer recurrence (1). As NDRS registers primary cancers (and not cancer recurrences), GRAIL Bio UK Ltd are keen to establish when there is evidence that a patient has had a recurrence of their cancer. This is to complete a sensitivity analysis for the Positive Predictive Value (PPV) of the test in which the detection of a recurrence will be considered a True Positive (this is in contrast to the main analysis in which, because there is no new cancer registration, it is considered a False Positive). GRAIL Bio UK Ltd believe this analysis will help provide a more complete description of test performance.

(1) Schrag D, et al. Presentation at European Society for Medical Oncology (ESMO) Congress; September 9-13, 2022.

For the purposes of this analysis, GRAIL Bio UK Ltd will share the Study IDs of a small number of consented NHS-Galleri participants with NHS England for review as to whether these participants have a cancer recurrence. NHS England will convert the study IDs into internal patient IDs and provide an expert clinical review of each case to determine whether a patient has a recurrence. NHS England will do this based on the NDRS Datasets already listed in this Data Sharing Agreement.

***************************

Patient and Public Involvement and Engagement (PPI&E)

Ongoing PPI&E is conducted throughout the trial. Historical PPI&E was conducted to develop the participant facing materials and on questionnaires, and including specific questions to PPI groups around the sharing of pseudonymised Data with GRAIL LLC (US). Ongoing PPI&E includes, for example, PPI&E on new participant facing materials, survey feedback on the participant experience, and questionnaires for follow-up of participants. An embedded behavioural scientist in the Kings College London Cancer

Prevention Trials Unit has been involved in, and will continue to be involved in, analysis of participant communications, and ethical and wellbeing issues within the course of the trial, as well as periodically advise the study organizers and other stakeholder on emerging themes.

In the setting up of the trial, GRAIL has taken seriously, strategies for equity of access and has several plans underway to ensure that under-represented minorities are actively recruited into the trial. Amongst this is the invitation strategy, that will identify and actively send invitations to a greater proportion of individuals from these minorities to take part in the trial. It is hoped that this will positively contribute towards equity of access and inform the provision of cancer services in these areas in the future.

UK GDPR Legal Basis for the Processing of Personal Data:

GRAIL Bio UK Ltd, as joint Controller, is using Article 6(1)(f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of Personal Data, in particular where the data subject is a child." Processing Personal Data is necessary for GRAIL Bio UK Ltd’s legitimate interests which are described in this application. The Data to which access is requested are proportionate and necessary to achieve those interests. GRAIL Bio UK Ltd has completed a legitimate interests assessment (LIA) and are satisfied that the interests of the data subjects do not override their legitimate interests; that they would reasonably expect the processing and it would not cause unjustified harm. The data subjects interests and fundamental rights are protected through appropriate minimisation of fields and patient records being processed; protection of the Data in a secure environment, and guaranteeing secure destruction at any stage at the request of NHS England and/or after a defined period on completion of the project.

Additionally (as health Data is a special category of Personal Data), GRAIL Bio UK Ltd is using Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

King’s College London, as joint Controller, is using Article 6:1(e): Specific task in the ‘public interest’ or task that has a clear basis in law, and additionally (as health Data is a special category of Personal Data) Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

Organisation Roles and Responsibilities:

• GRAIL Bio UK Ltd are a jointController and Lead for this DSA, who also process the Data. It is responsible for sponsoring and funding the trial, as well as the building and maintenance of the Clinical Records Management System.

• King’s College London are a joint Controller who also process the Data and are responsible for implementing the trial and for primary objective analysis.

• GRAIL, Limited Liability Company (LLC). are the manufacturers of the MCED test, Galleri (Registered Trademark), and has set up the above UK subsidiary, GRAIL Bio UK Ltd, which is the sponsor of the NHS-Galleri trial. GRAIL, LLC will be receiving pseudonymised record-level NHS England Data from King’s College London, and are therefore listed as a Processor in this DSA.

****New for Version 5 of this Data Sharing Agreement - November 2023:****

• Queen Mary, University of London (QMUL) is added as a named processor to this Data Sharing Agreement. It is planned for the start of December 2023 that the clinical research agreement GRAIL Bio UK Ltd has with King’s College London's Cancer Prevention Trials Unit (CPTU) will novate to QMUL. This means that the current KCL CPTU trial staff working on NHS Galleri (including Patient Liaison, Clinical Project Managers and Clinical Trial Assistants) will cease employment at King’s College London and take up substantive employment with QMUL.

King’s College London will be subcontracted by QMUL to continue to work on NHS Galleri including allowing QMUL CPTU trial staff to access KCL facilities (software, equipment and data) under security and data control conditions that are appropriate for the data and facilities being accessed. Such subcontract will approximately last until August 2024, the end of the operational phase of the NHS Galleri study. Additionally, statisticians substantively employed by QMUL will access and analyse the NHS England Data.

To further clarify, pursuant to the Novation Agreement, QMUL shall be a Processor of GRAIL Bio UK Ltd. Further, QMUL has entered into a subcontract and a sub-data processing agreement with KCL in connection with the services KCL will provide to QMUL in relation to storing and managing the data for the NHS Galleri Trial. GRAIL Bio UK Ltd, as Controller, shall have responsibility for ensuring all processing requirements and obligations flow down to QMUL and to KCL respectively. These processing requirements and obligations shall include ensuring QMUL and KCL are fully aware of their processing obligations as set out in this Data Sharing Agreement, including NHS England auditing obligations.

QMUL are not considered a Controller in this Data Sharing Agreement as they do not make any decisions determining the purposes and means of the processing in the NHS Galleri study. NHS England Data, as described in this Data Sharing Agreement, will not leave KCL secure servers but will be accessed by substantive employees of QMUL under affiliate KCL staff status and subject to a Data Processor Agreement between King’s College London and QMUL (see Section 5b Processing activities for more details).

At the end of the Operational Phase of NHS Galleri (approximately the end of 2024), the subcontracting agreement between King’s College London and QMUL referenced above will end. To reflect this change an amendment to this Data Sharing Agreement will be submitted by the Controllers at such time.

***************************

NHS England are listed as a strategic partner in relation to NHS-Galleri, but have no input into how the trial is run or determining the outcome. NHS England therefore do not consider them to be a Controller. Cancer Research UK appear on trial-related documentation by virtue of the fact that King’s College London Cancer Prevention Trials Unit are a Cancer Research UK Trials Unit. Cancer Research UK as an organisation in their own right have no involvement in the running of the trial, nor are they funding any of the trial activities. NHS Englandl therefore do not consider them to be a Controller.

In the future, GRAIL Bio UK Ltd may receive commercial benefit (including intangible or indirect commercial benefits such as positive publicity) from the successful outcomes of the trial. GRAIL Bio UK Ltd and the NHS have entered into a partnership whereby the Galleri test will be piloted in clinical trials within the NHS in England. If the test is shown to work as intended in these clinical trials, the NHS may purchase the test from GRAIL and make the test routinely available in the future to benefit patient.

All outputs will be aggregated with small number suppressed as per the HES analysis guide or according to the specific data set's suppression guidance.

The results of the NHS-Galleri trial outputs will be published in peer-reviewed publications, presented at conferences, and used by NHS England in their decisions around implementation of the test within the NHS. The interim analysis, which will include analysis of a key secondary objective, is expected by the end of 2023 or beginning of 2024. The primary objective readout is expected by the end of 2025 or beginning of 2026. Planned dissemination activities will aim to enable the engagement with the scientific and policy making communities to ensure that knowledge developed by this research can benefit these communities to help the NHS detect cancer early, when it is easier to treat.

The Processors continue to receive NHS England Data to undertake ongoing data analysis for the study, but at the present time there are no Yielded Benefits to report under this amendment. This is because, as stated above, the outputs are not due until the end of 2026.

GRAIL Bio UK Ltd and King’s College London (KCL), as joint data controllers, are requesting access to pseudonymised record-level data in order to carry out follow-up analysis based on a cohort of patients who are being recruited to a clinical trial called ‘NHS-Galleri’ following being invited to participate if they are found to meet the required eligibility criteria.

Note that members of the team running the trial are based at the Cancer Prevention Trials Unit (CPTU) at King's College London. For the sake of consistency, and as King's College London is listed as a data controller, when referring to King's College London throughout the application, this encompasses the team at CPTU.

Background, Purpose and Rationale behind NHS-Galleri:

A new Multi-Cancer Early Detection (MCED) test has been developed that can detect many types of cancer from a single blood sample. This test is called Galleri (Registered Trademark). This trial aims to find out whether it is better at detecting cancer early, compared to other tests that the NHS currently uses. The purpose of NHS-Galleri is to demonstrate the clinical utility of the MCED blood test for individuals in a general screening population in a real world NHS setting. The rationale behind this trial is that MCED is a novel screening paradigm, and assessment of the use and impact of test results is necessary to enable integration into clinical practice. This will be the first randomised, double blind, controlled trial statistically powered to assess clinical utility of a MCED test.

Recruitment to the trial has been aided by NHS Digital's Digi-Trials team via a sister application, DARS-NIC-456778-J0G3H, and took place over 10 – 12 months from August 2021 with a recruitment target of 140,000 people. Recruitment successfully completed and the target was met. Different geographic areas were targeted for recruitment, beginning with North-west England. Recruitment in a particular area lasted for 4 – 6 weeks, before moving on to another area. Recruitment was managed through mobile units on scheduled routes around a particular area. Batches of invitations were constructed to target whatever area the mobile units were in at the time of writing out to potential participants. Potential participants were only contacted once and there were no follow-ups sent after the original invitation letter.

Requesting Not to Take Part in Trial:

In addition to the National Data Opt-out, members of the public were able to specifically request not to be contacted for the NHS-Galleri Trial. GRAIL Bio UK Ltd promoted this information via the dedicated trial website (nhs-galleri.org) and NHS Digital did the same via a dedicated page on its website (https://digital.nhs.uk/services/nhs-digitrials/nhs-galleri-trial). There was also an option for people to register their request not to take part in NHS-Galleri by telephone. NHS Digital records these requests not to take part and ensure that anyone who has registered for this will be excluded from the cohort selection and as a result their details will not be on the list for invitations to be sent.

Objectives of the Trial:

• Primary Objective:

- For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of stage III and IV cancers diagnosed in the intervention arm as compared to the control arm following three annual test and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a. For a prespecified group of cancer types, including lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver / bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder.

b. For all cancer types, excluding prostate cancer.

c. For all cancer types.

• Secondary Objectives:

1) For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of advanced cancers (defined as a cancer diagnosed at stage III or IV or one that results in a cancer-specific death) in the intervention arm as compared to the control arm following three annual tests and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a) First, evaluate for a statistically significant difference in a prespecified group of primary cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

b) If a statistically significant reduction in absolute numbers is found, continue by evaluating for a difference in all cancer types excluding prostate cancer.

c) If the above evaluations are both significant, evaluate for a difference in all cancer types.

2) For cancers that are routinely staged, determine whether there is a difference in the proportion of stage I and II cancers between the two arms of the study in the third testing round (i.e. for cancers diagnosed with an average 16-18 months of follow-up after the third test).

3) Evaluate the performance (overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three intervals of testing.

4) Evaluate the safety, including harms, of the testing pathway in the intervention arm among individuals with positive MCED test results.

a) Assess participant-reported anxiety among those receiving a positive MCED test result following the initial blood draw at various time points.

b) Examine the radiation exposure by participants associated with follow-up diagnostic procedures following a positive MCED test result

c) Describe the number and type of invasive procedures performed, and number of complications and deaths associated with follow-up diagnostic procedures.

5) Determine whether there is a statistically significant reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following the initial screening round and 12 months of follow-up (excluding cancers identified by the test performed at the 12-month visit)

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer;

c) For all cancer types.

6) Determine whether there is a reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following three annual tests and an average of 16-18 months follow-up after the third test

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer

c) For all cancer types.

7) Determine whether there is a statistically significant difference in cancer detection rates in the two arms of the study with 12 months follow-up after the first test.

8) Model the expected differences in cancer-specific mortality at 7 years post-randomisation based on cancers diagnosed within an average of 40-42 months of randomisation.

9) Summarize stage distribution per cancer type for the two arms, and compare the absolute numbers of stage IV alone as well as stage III and IV cancers, following the third screening and an average 16-18* months of follow-up.

10) Determine whether there is a reduction in the absolute numbers of stage III and IV cancers, excluding breast, cervical, and colorectal (i.e., those without a standard of care screening modality) as compared to the control arm following the third annual screening and an average 16-18 months of follow-up.

11) Compare cancer-specific mortality in the intervention and control arms following the third screening and an average 16-18 months of follow-up, for a pre-specified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

12) Compare cancer-specific mortality in the intervention and control arms for all cancer types at

a) 3 years and

b) 6 years after the last study visit.

13) Compare the absolute numbers of stage IV cancers following the second blood draw and 12 months of follow-up, with prevalent cases excluded.

14) Compare the absolute numbers of stage III and IV cancers following the third blood draw and an average 16-18 months of follow-up, with cases from the first and second screenings excluded.

15) Assess the impact of use of the MCED test across three annual timepoints on healthcare resource utilization for cancer diagnosis and treatment.

16) Assess the potential impact of overdiagnosis by comparing the cumulative incidence (Kaplan-Meier curves) of cancers diagnosed after a positive baseline MCED test up to 36-48 months after randomization between the intervention and control arms (retrospectively testing baseline samples from all participants diagnosed with cancer in the control arm).

DATA SETS REQUESTED:

> NDRS Rapid Cancer Registrations

> NDRS Cancer Registry

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT)

> NDRS National Radiotherapy Dataset (RTDS)

> NDRS Linked Hospital Episode Statistics (HES) Admitted Patient Care (APC)

> NDRS Linked HES Outpatients (OP)

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals)

> NDRS Linked Diagnostic Imaging Dataset (DIDS) (provided Summer 2022 onward)

> Emergency Care Data Set (ECDS) (provided Summer 2022 onward)

FREQUENCY - every 6 months beginning March 2022 (Rapid Registrations every 2 months from March 2022).

FROM AUGUST 2022 (Version 2 of this agreement) - Frequency of drops

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT) - no change

> NDRS Linked HES APC - no change

> NDRS Linked HES Outpatients (OP) - no change

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals) - no change

> NDRS National Radiotherapy Dataset (RTDS) – No change

> NDRS Linked Diagnostic Imaging Dataset (DIDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Linked Emergency Care Data Set (ECDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Rapid Cancer Registrations - Rapid Registrations will be provided every month rather than every two months from August 2022 to enable frequent monitoring of cancer diagnoses in the study.

> NDRS Cancer Registry - will be provided every month rather than every other month from August 2022 to enable frequent reporting of cancer diagnoses in the study and use the data as a filter for participant invitations for annual follow up timepoints. Expand historic gold standard registration field selection for 1995-2021 and addition of field to allow the trial team to distinguish provisional from finalised gold standard tumours. These changes are required for primary and secondary analysis. Data will be used to characterize the cohort for risk factors (prior cancers), and also assess test performance in a subpopulation of cancer survivors. The Status of Registration is required to enable accurate reporting and filtering of provisional data vs finalized data to internal and external stakeholders, and the interim analysis in the study. Death fields added to dataset specification for analysis of key secondary endpoints in the study. Death data will also be used for reporting subject disposition in the analysis. Death data, while requested with the cancer registry data, should include information on all participants in the cohort (including participants that have not been diagnosed with cancer).

JUSTIFICATION FOR DATA SETS REQUESTED AND DATA MINIMISATION

> Cancer Registry and Rapid Registrations - These central NHS datasets will be the sole source for the data required for primary and secondary objectives analyses in the Study. Will be used to identify Primary endpoints, which are incidence and stage at diagnosis for cancer types that are stage-able (e.g., with available staging systems). Cancer is defined as any of the following cancers:

- Invasive solid cancer, excluding non-melanoma skin cancer

- Haematologic malignancies, including lymphoma, lymphoid leukemia, myeloma/plasma cell neoplasm, myeloid neoplasms (including myelodysplastic and myeloproliferative neoplasms with behaviour code 3 based on ICD-O-3.2*).

[*The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) is the standard classification system for the registration of cancers in the United States (and most areas of the world) and contains widely-accepted histologic categories of brain tumors.]

The Cancer Registry will also be used to identify cohort characteristics/baseline risk - Personal history of cancer assessed using limited set of data fields. Historical cancer data dating back to 1995 will be available.

> Hospital Episode Statistics (HES) Outpatients (OP), Admitted Patient Care (APC) and Emergency Care Data Set (ECDS), Cancer Waiting Times (Treatments and Referrals) (CWT) and Diagnostic Imaging Dataset (DIDS) - Will be used to identify Secondary endpoints (including safety objectives). Among all test positive cases, number of follow-up procedures and number of invasive procedures (including all biopsies, surgical interventions, bronchoscopy, thoracoscopy and endoscopy) to achieve diagnostic resolution (i.e. cancer diagnosis, non-cancer diagnosis, or no diagnosis and discharge from the diagnostic follow-up). Will also identify:

- Number and type of invasive procedures performed in false positive cases;

- Number of complications and deaths resulting from diagnostic procedures;

- Radiation exposure measured in mSv per participant due to test result-directed evaluations

> Systemic Anti-Cancer Therapy Data Set (SACT) and Radiotherapy Data set (RTDS) will also be used (in addition to the above HES, CWT and DIDs data) for health care resource utilization objectives.

Patient and Public Involvement and Engagement (PPI&E)

Ongoing PPI&E is conducted throughout the trial. Historical PPI&E was conducted to develop the participant facing materials and on questionnaires, and including specific questions to PPI groups around the sharing of pseudonymised data with GRAIL LLC (US). Ongoing PPI&E includes, for example, PPI&E on new participant facing materials, survey feedback on the participant experience, and questionnaires for follow-up of participants. An embedded behavioural scientist in the Kings College London Cancer

Prevention Trials Unit has been involved in, and will continue to be involved in, analysis of participant communications, and ethical and wellbeing issues within the course of the trail, as well as periodically advise the study organizers and other stakeholder on emerging themes.

In the setting up of the trial, GRAIL has taken seriously, strategies for equity of access and has several plans underway to ensure that under-represented minorities are actively recruited into the trial. Amongst this is the invitation strategy, that will identify and actively send invitations to a greater proportion of individuals from these minorities to take part in the trial. It is hoped that this will positively contribute towards equity of access and inform the provision of cancer services in these areas in the future.

GDPR Legal Basis for the Processing of Personal Data:

GRAIL Bio UK Ltd, as joint Data Controller, is using Article 6(1)(f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child." Processing personal data is necessary for GRAIL Bio UK Ltd’s legitimate interests which are described in this application. The data to which access is requested are proportionate and necessary to achieve those interests. GRAIL Bio UK Ltd has completed a legitimate interests assessment (LIA) and NHS Digital are satisfied that the interests of the data subjects do not override their legitimate interests; that they would reasonably expect the processing and it would not cause unjustified harm. The data subjects interests and fundamental rights are protected through appropriate minimisation of fields and patient records being processed; protection of the data in a secure environment, and guaranteeing secure destruction at any stage at the request of NHS Digital or after a defined period on completion of the project.

Additionally (as health data is a special category of Personal Data), GRAIL Bio UK Ltd is using Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

King’s College London, as joint Data Controller, is using Article 6:1(e): Specific task in the ‘public interest’ or task that has a clear basis in law, and additionally (as health data is a special category of Personal Data) Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

Organisation Roles and Responsibilities:

• GRAIL Bio UK Ltd are a joint Data Controller and Lead for this agreement, who also process the data. It is responsible for sponsoring and funding the trial, as well as the building and maintenance of the Clinical Records Management System.

• King’s College London are a joint Data Controller who also process the data and are responsible for implementing the trial and for primary objective analysis.

• GRAIL, Limited Liability Company (LLC). are the manufacturers of the MCED test, Galleri (Registered Trademark), and has set up the above UK subsidiary, GRAIL Bio UK Ltd, which is the sponsor of the NHS-Galleri trial. GRAIL, LLC will be receiving pseudonymised record-level NHS Digital data from King’s College London, and are therefore listed as a Data Processor in this agreement.

NHS England are listed as a strategic partner in relation to NHS-Galleri, but have no input into how the trial is run or determining the outcome. NHS Digital therefore do not consider them to be a Data Controller. Cancer Research UK appear on trial-related documentation by virtue of the fact that King’s College London Cancer Prevention Trials Unit are a Cancer Research UK Trials Unit. Cancer Research UK as an organisation in their own right have no involvement in the running of the trial, nor are they funding any of the trial activities. NHS Digital therefore do not consider them to be a Data Controller.

In the future, GRAIL Bio UK Ltd may receive commercial benefit (including intangible or indirect commercial benefits such as positive publicity) from the successful outcomes of the trial. GRAIL Bio UK Ltd and the NHS have entered into a partnership whereby the Galleri test will be piloted in clinical trials within the NHS in England. If the test is shown to work as intended in these clinical trials, the NHS may purchase the test from GRAIL and make the test routinely available in the future to benefit patient.

All outputs will be aggregated with small number suppressed as per the HES analysis guide or according to the specific data set's suppression guidance.

The results of the NHS-Galleri trial outputs will be published in peer-reviewed publications, presented at conferences, and used by NHS England in their decisions around implementation of the test within the NHS. The interim analysis, which will include analysis of a key secondary objective, is expected by the end of 2023 or beginning of 2024. The primary objective readout is expected by the end of 2025 or beginning of 2026. Planned dissemination activities will aim to enable the engagement with the scientific and policy making communities to ensure that knowledge developed by this research can benefit these communities to help the NHS detect cancer early, when it is easier to treat.

Version 4: The Data Processors continue to receive NHS Digital data to undertake ongoing data analysis for the study, but at the present time there are no Yielded Benefits to report under this amendment.

GRAIL Bio UK Ltd and King’s College London (KCL), as joint data controllers, are requesting access to pseudonymised record-level data in order to carry out follow-up analysis based on a cohort of patients who are being recruited to a clinical trial called ‘NHS-Galleri’ following being invited to participate if they are found to meet the required eligibility criteria.

Note that members of the team running the trial are based at the Cancer Prevention Trials Unit (CPTU) at King's College London. For the sake of consistency, and as King's College London is listed as a data controller, when referring to King's College London throughout the application, this encompasses the team at CPTU.

Background, Purpose and Rationale behind NHS-Galleri:

A new Multi-Cancer Early Detection (MCED) test has been developed that can detect many types of cancer from a single blood sample. This test is called Galleri (Registered Trademark). This trial aims to find out whether it is better at detecting cancer early, compared to other tests that the NHS currently uses. The purpose of NHS-Galleri is to demonstrate the clinical utility of the MCED blood test for individuals in a general screening population in a real world NHS setting. The rationale behind this trial is that MCED is a novel screening paradigm, and assessment of the use and impact of test results is necessary to enable integration into clinical practice. This will be the first randomised, double blind, controlled trial statistically powered to assess clinical utility of a MCED test.

Recruitment to the trial has been aided by NHS Digital's Digi-Trials team via a sister application, DARS-NIC-456778-J0G3H, and took place over 10 – 12 months from August 2021 with a recruitment target of 140,000 people. Recruitment successfully completed and the target was met. Different geographic areas were targeted for recruitment, beginning with North-west England. Recruitment in a particular area lasted for 4 – 6 weeks, before moving on to another area. Recruitment was managed through mobile units on scheduled routes around a particular area. Batches of invitations were constructed to target whatever area the mobile units were in at the time of writing out to potential participants. Potential participants were only contacted once and there were no follow-ups sent after the original invitation letter.

Requesting Not to Take Part in Trial:

In addition to the National Data Opt-out, members of the public were able to specifically request not to be contacted for the NHS-Galleri Trial. GRAIL Bio UK Ltd promoted this information via the dedicated trial website (nhs-galleri.org) and NHS Digital did the same via a dedicated page on its website (https://digital.nhs.uk/services/nhs-digitrials/nhs-galleri-trial). There was also an option for people to register their request not to take part in NHS-Galleri by telephone. NHS Digital records these requests not to take part and ensure that anyone who has registered for this will be excluded from the cohort selection and as a result their details will not be on the list for invitations to be sent.

Objectives of the Trial:

• Primary Objective:

- For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of stage III and IV cancers diagnosed in the intervention arm as compared to the control arm following three annual test and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a. For a prespecified group of cancer types, including lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver / bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder.

b. For all cancer types, excluding prostate cancer.

c. For all cancer types.

• Secondary Objectives:

1) For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of advanced cancers (defined as a cancer diagnosed at stage III or IV or one that results in a cancer-specific death) in the intervention arm as compared to the control arm following three annual tests and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a) First, evaluate for a statistically significant difference in a prespecified group of primary cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

b) If a statistically significant reduction in absolute numbers is found, continue by evaluating for a difference in all cancer types excluding prostate cancer.

c) If the above evaluations are both significant, evaluate for a difference in all cancer types.

2) For cancers that are routinely staged, determine whether there is a difference in the proportion of stage I and II cancers between the two arms of the study in the third testing round (i.e. for cancers diagnosed with an average 16-18 months of follow-up after the third test).

3) Evaluate the performance (overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three intervals of testing.

4) Evaluate the safety, including harms, of the testing pathway in the intervention arm among individuals with positive MCED test results.

a) Assess participant-reported anxiety among those receiving a positive MCED test result following the initial blood draw at various time points.

b) Examine the radiation exposure by participants associated with follow-up diagnostic procedures following a positive MCED test result

c) Describe the number and type of invasive procedures performed, and number of complications and deaths associated with follow-up diagnostic procedures.

5) Determine whether there is a statistically significant reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following the initial screening round and 12 months of follow-up (excluding cancers identified by the test performed at the 12-month visit)

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer;

c) For all cancer types.

6) Determine whether there is a reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following three annual tests and an average of 16-18 months follow-up after the third test

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer

c) For all cancer types.

7) Determine whether there is a statistically significant difference in cancer detection rates in the two arms of the study with 12 months follow-up after the first test.

8) Model the expected differences in cancer-specific mortality at 7 years post-randomisation based on cancers diagnosed within an average of 40-42 months of randomisation.

9) Summarize stage distribution per cancer type for the two arms, and compare the absolute numbers of stage IV alone as well as stage III and IV cancers, following the third screening and an average 16-18* months of follow-up.

10) Determine whether there is a reduction in the absolute numbers of stage III and IV cancers, excluding breast, cervical, and colorectal (i.e., those without a standard of care screening modality) as compared to the control arm following the third annual screening and an average 16-18 months of follow-up.

11) Compare cancer-specific mortality in the intervention and control arms following the third screening and an average 16-18 months of follow-up, for a pre-specified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

12) Compare cancer-specific mortality in the intervention and control arms for all cancer types at

a) 3 years and

b) 6 years after the last study visit.

13) Compare the absolute numbers of stage IV cancers following the second blood draw and 12 months of follow-up, with prevalent cases excluded.

14) Compare the absolute numbers of stage III and IV cancers following the third blood draw and an average 16-18 months of follow-up, with cases from the first and second screenings excluded.

15) Assess the impact of use of the MCED test across three annual timepoints on healthcare resource utilization for cancer diagnosis and treatment.

16) Assess the potential impact of overdiagnosis by comparing the cumulative incidence (Kaplan-Meier curves) of cancers diagnosed after a positive baseline MCED test up to 36-48 months after randomization between the intervention and control arms (retrospectively testing baseline samples from all participants diagnosed with cancer in the control arm).

DATA SETS REQUESTED:

> NDRS Rapid Cancer Registrations

> NDRS Cancer Registry

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT)

> NDRS National Radiotherapy Dataset (RTDS)

> NDRS Linked Hospital Episode Statistics (HES) Admitted Patient Care (APC)

> NDRS Linked HES Outpatients (OP)

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals)

> NDRS Linked Diagnostic Imaging Dataset (DIDS) (provided Summer 2022 onward)

> Emergency Care Data Set (ECDS) (provided Summer 2022 onward)

FREQUENCY - every 6 months beginning March 2022 (Rapid Registrations every 2 months from March 2022).

FROM AUGUST 2022 (Version 2 of this agreement) - Frequency of drops

> NDRS Systemic Anti-Cancer Therapy Data Set (SACT) - no change

> NDRS Linked HES APC - no change

> NDRS Linked HES Outpatients (OP) - no change

> NDRS Linked Cancer Waiting Times (CWT) (Treatments and Referrals) - no change

> NDRS National Radiotherapy Dataset (RTDS) – No change

> NDRS Linked Diagnostic Imaging Dataset (DIDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Linked Emergency Care Data Set (ECDS) - Released August 2022 on bi-annual basis as planned.

> NDRS Rapid Cancer Registrations - Rapid Registrations will be provided every month rather than every two months from August 2022 to enable frequent monitoring of cancer diagnoses in the study.

> NDRS Cancer Registry - will be provided every month rather than every other month from August 2022 to enable frequent reporting of cancer diagnoses in the study and use the data as a filter for participant invitations for annual follow up timepoints. Expand historic gold standard registration field selection for 1995-2021 and addition of field to allow the trial team to distinguish provisional from finalised gold standard tumours. These changes are required for primary and secondary analysis. Data will be used to characterize the cohort for risk factors (prior cancers), and also assess test performance in a subpopulation of cancer survivors. The Status of Registration is required to enable accurate reporting and filtering of provisional data vs finalized data to internal and external stakeholders, and the interim analysis in the study. Death fields added to dataset specification for analysis of key secondary endpoints in the study. Death data will also be used for reporting subject disposition in the analysis. Death data, while requested with the cancer registry data, should include information on all participants in the cohort (including participants that have not been diagnosed with cancer).

JUSTIFICATION FOR DATA SETS REQUESTED AND DATA MINIMISATION

> Cancer Registry and Rapid Registrations - These central NHS datasets will be the sole source for the data required for primary and secondary objectives analyses in the Study. Will be used to identify Primary endpoints, which are incidence and stage at diagnosis for cancer types that are stage-able (e.g., with available staging systems). Cancer is defined as any of the following cancers:

- Invasive solid cancer, excluding non-melanoma skin cancer

- Haematologic malignancies, including lymphoma, lymphoid leukemia, myeloma/plasma cell neoplasm, myeloid neoplasms (including myelodysplastic and myeloproliferative neoplasms with behaviour code 3 based on ICD-O-3.2*).

[*The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) is the standard classification system for the registration of cancers in the United States (and most areas of the world) and contains widely-accepted histologic categories of brain tumors.]

The Cancer Registry will also be used to identify cohort characteristics/baseline risk - Personal history of cancer assessed using limited set of data fields. Historical cancer data dating back to 1995 will be available.

> Hospital Episode Statistics (HES) Outpatients (OP), Admitted Patient Care (APC) and Emergency Care Data Set (ECDS), Cancer Waiting Times (Treatments and Referrals) (CWT) and Diagnostic Imaging Dataset (DIDS) - Will be used to identify Secondary endpoints (including safety objectives). Among all test positive cases, number of follow-up procedures and number of invasive procedures (including all biopsies, surgical interventions, bronchoscopy, thoracoscopy and endoscopy) to achieve diagnostic resolution (i.e. cancer diagnosis, non-cancer diagnosis, or no diagnosis and discharge from the diagnostic follow-up). Will also identify:

- Number and type of invasive procedures performed in false positive cases;

- Number of complications and deaths resulting from diagnostic procedures;

- Radiation exposure measured in mSv per participant due to test result-directed evaluations

> Systemic Anti-Cancer Therapy Data Set (SACT) and Radiotherapy Data set (RTDS) will also be used (in addition to the above HES, CWT and DIDs data) for health care resource utilization objectives.

Patient and Public Involvement and Engagement (PPI&E)

Ongoing PPI&E is conducted throughout the trial. Historical PPI&E was conducted to develop the participant facing materials and on questionnaires, and including specific questions to PPI groups around the sharing of pseudonymised data with GRAIL LLC (US). Ongoing PPI&E includes, for example, PPI&E on new participant facing materials, survey feedback on the participant experience, and questionnaires for follow-up of participants. An embedded behavioural scientist in the Kings College London Cancer

Prevention Trials Unit has been involved in, and will continue to be involved in, analysis of participant communications, and ethical and wellbeing issues within the course of the trail, as well as periodically advise the study organizers and other stakeholder on emerging themes.

In the setting up of the trial, GRAIL has taken seriously, strategies for equity of access and has several plans underway to ensure that under-represented minorities are actively recruited into the trial. Amongst this is the invitation strategy, that will identify and actively send invitations to a greater proportion of individuals from these minorities to take part in the trial. It is hoped that this will positively contribute towards equity of access and inform the provision of cancer services in these areas in the future.

GDPR Legal Basis for the Processing of Personal Data:

GRAIL Bio UK Ltd, as joint Data Controller, is using Article 6(1)(f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child." Processing personal data is necessary for GRAIL Bio UK Ltd’s legitimate interests which are described in this application. The data to which access is requested are proportionate and necessary to achieve those interests. GRAIL Bio UK Ltd has completed a legitimate interests assessment (LIA) and NHS Digital are satisfied that the interests of the data subjects do not override their legitimate interests; that they would reasonably expect the processing and it would not cause unjustified harm. The data subjects interests and fundamental rights are protected through appropriate minimisation of fields and patient records being processed; protection of the data in a secure environment, and guaranteeing secure destruction at any stage at the request of NHS Digital or after a defined period on completion of the project.

Additionally (as health data is a special category of Personal Data), GRAIL Bio UK Ltd is using Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

King’s College London, as joint Data Controller, is using Article 6:1(e): Specific task in the ‘public interest’ or task that has a clear basis in law, and additionally (as health data is a special category of Personal Data) Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

Organisation Roles and Responsibilities:

• GRAIL Bio UK Ltd are a joint Data Controller and Lead for this agreement, who also process the data. It is responsible for sponsoring and funding the trial, as well as the building and maintenance of the Clinical Records Management System.

• King’s College London are a joint Data Controller who also process the data and are responsible for implementing the trial and for primary objective analysis.

• GRAIL, Limited Liability Company (LLC). are the manufacturers of the MCED test, Galleri (Registered Trademark), and has set up the above UK subsidiary, GRAIL Bio UK Ltd, which is the sponsor of the NHS-Galleri trial. GRAIL, LLC will be receiving pseudonymised record-level NHS Digital data from King’s College London, and are therefore listed as a Data Processor in this agreement.

NHS England are listed as a strategic partner in relation to NHS-Galleri, but have no input into how the trial is run or determining the outcome. NHS Digital therefore do not consider them to be a Data Controller. Cancer Research UK appear on trial-related documentation by virtue of the fact that King’s College London Cancer Prevention Trials Unit are a Cancer Research UK Trials Unit. Cancer Research UK as an organisation in their own right have no involvement in the running of the trial, nor are they funding any of the trial activities. NHS Digital therefore do not consider them to be a Data Controller.

In the future, GRAIL Bio UK Ltd may receive commercial benefit (including intangible or indirect commercial benefits such as positive publicity) from the successful outcomes of the trial. GRAIL Bio UK Ltd and the NHS have entered into a partnership whereby the Galleri test will be piloted in clinical trials within the NHS in England. If the test is shown to work as intended in these clinical trials, the NHS may purchase the test from GRAIL and make the test routinely available in the future to benefit patient.

All outputs will be aggregated with small number suppressed as per the HES analysis guide or according to the specific data set's suppression guidance.

The results of the NHS-Galleri trial outputs will be published in peer-reviewed publications, presented at conferences, and used by NHS England in their decisions around implementation of the test within the NHS. The interim analysis, which will include analysis of a key secondary objective, is expected by the end of 2023 or beginning of 2024. The primary objective readout is expected by the end of 2025 or beginning of 2026. Planned dissemination activities will aim to enable the engagement with the scientific and policy making communities to ensure that knowledge developed by this research can benefit these communities to help the NHS detect cancer early, when it is easier to treat.

Version 3: The Data Processors continue to receive NHS Digital data to undertake ongoing data analysis for the study, but at the present time there are no Yielded Benefits to report under this amendment.

GRAIL Bio UK Ltd and King’s College London (KCL), as joint data controllers, are requesting access to pseudonymised record-level data in order to carry out follow-up analysis based on a cohort of patients who are being recruited to a clinical trial called ‘NHS-Galleri’ following being invited to participate if they are found to meet the required eligibility criteria.

Note that members of the team running the trial are based at the Cancer Prevention Trials Unit (CPTU) at King's College London. For the sake of consistency, and as King's College London is listed as a data controller, when referring to King's College London throughout the application, this encompasses the team at CPTU.

Background, Purpose and Rationale behind NHS-Galleri:

A new Multi-Cancer Early Detection (MCED) test has been developed that can detect many types of cancer from a single blood sample. This test is called Galleri (Registered Trademark). This trial aims to find out whether it is better at detecting cancer early, compared to other tests that the NHS currently uses. The purpose of NHS-Galleri is to demonstrate the clinical utility of the MCED blood test for individuals in a general screening population in a real world NHS setting. The rationale behind this trial is that MCED is a novel screening paradigm, and assessment of the use and impact of test results is necessary to enable integration into clinical practice. This will be the first randomised, double blind, controlled trial statistically powered to assess clinical utility of a MCED test.

Recruitment to the trial has been aided by NHS Digital's Digi-Trials team via a sister application, DARS-NIC-456778-J0G3H, and took place over 10 – 12 months from August 2021 with a recruitment target of 140,000 people. Recruitment successfully completed and the target was met. Different geographic areas were targeted for recruitment, beginning with North-west England. Recruitment in a particular area lasted for 4 – 6 weeks, before moving on to another area. Recruitment was managed through mobile units on scheduled routes around a particular area. Batches of invitations were constructed to target whatever area the mobile units were in at the time of writing out to potential participants. Potential participants were only contacted once and there were no follow-ups sent after the original invitation letter.

Requesting Not to Take Part in Trial:

In addition to the National Data Opt-out, members of the public were able to specifically request not to be contacted for the NHS-Galleri Trial. GRAIL Bio UK Ltd promoted this information via the dedicated trial website (nhs-galleri.org) and NHS Digital did the same via a dedicated page on its website (https://digital.nhs.uk/services/nhs-digitrials/nhs-galleri-trial). There was also an option for people to register their request not to take part in NHS-Galleri by telephone. NHS Digital records these requests not to take part and ensure that anyone who has registered for this will be excluded from the cohort selection and as a result their details will not be on the list for invitations to be sent.

Objectives of the Trial:

• Primary Objective:

- For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of stage III and IV cancers diagnosed in the intervention arm as compared to the control arm following three annual test and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a. For a prespecified group of cancer types, including lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver / bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder.

b. For all cancer types, excluding prostate cancer.

c. For all cancer types.

• Secondary Objectives:

1) For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of advanced cancers (defined as a cancer diagnosed at stage III or IV or one that results in a cancer-specific death) in the intervention arm as compared to the control arm following three annual tests and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a) First, evaluate for a statistically significant difference in a prespecified group of primary cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

b) If a statistically significant reduction in absolute numbers is found, continue by evaluating for a difference in all cancer types excluding prostate cancer.

c) If the above evaluations are both significant, evaluate for a difference in all cancer types.

2) For cancers that are routinely staged, determine whether there is a difference in the proportion of stage I and II cancers between the two arms of the study in the third testing round (i.e. for cancers diagnosed with an average 16-18 months of follow-up after the third test).

3) Evaluate the performance (overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three intervals of testing.

4) Evaluate the safety, including harms, of the testing pathway in the intervention arm among individuals with positive MCED test results.

a) Assess participant-reported anxiety among those receiving a positive MCED test result following the initial blood draw at various time points.

b) Examine the radiation exposure by participants associated with follow-up diagnostic procedures following a positive MCED test result

c) Describe the number and type of invasive procedures performed, and number of complications and deaths associated with follow-up diagnostic procedures.

5) Determine whether there is a statistically significant reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following the initial screening round and 12 months of follow-up (excluding cancers identified by the test performed at the 12-month visit)

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer;

c) For all cancer types.

6) Determine whether there is a reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following three annual tests and an average of 16-18 months follow-up after the third test

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer

c) For all cancer types.

7) Determine whether there is a statistically significant difference in cancer detection rates in the two arms of the study with 12 months follow-up after the first test.

8) Model the expected differences in cancer-specific mortality at 7 years post-randomisation based on cancers diagnosed within an average of 40-42 months of randomisation.

9) Summarize stage distribution per cancer type for the two arms, and compare the absolute numbers of stage IV alone as well as stage III and IV cancers, following the third screening and an average 16-18* months of follow-up.

10) Determine whether there is a reduction in the absolute numbers of stage III and IV cancers, excluding breast, cervical, and colorectal (i.e., those without a standard of care screening modality) as compared to the control arm following the third annual screening and an average 16-18 months of follow-up.

11) Compare cancer-specific mortality in the intervention and control arms following the third screening and an average 16-18 months of follow-up, for a pre-specified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

12) Compare cancer-specific mortality in the intervention and control arms for all cancer types at

a) 3 years and

b) 6 years after the last study visit.

13) Compare the absolute numbers of stage IV cancers following the second blood draw and 12 months of follow-up, with prevalent cases excluded.

14) Compare the absolute numbers of stage III and IV cancers following the third blood draw and an average 16-18 months of follow-up, with cases from the first and second screenings excluded.

15) Assess the impact of use of the MCED test across three annual timepoints on healthcare resource utilization for cancer diagnosis and treatment.

16) Assess the potential impact of overdiagnosis by comparing the cumulative incidence (Kaplan-Meier curves) of cancers diagnosed after a positive baseline MCED test up to 36-48 months after randomization between the intervention and control arms (retrospectively testing baseline samples from all participants diagnosed with cancer in the control arm).

DATA SETS REQUESTED:

> Rapid Registrations Cancer Data Set

> Cancer Registry

> Systemic Anti-Cancer Therapy Data Set (SACT)

> Radiotherapy Data set (RTDS)

> Hospital Episode Statistics (HES) Admitted Patient Care (APC)

> HES Outpatients (OP)

> Cancer Waiting Times (Treatments and Referrals) (CWT)

> Diagnostic Imaging Dataset (DIDS) (To be released Summer 2022 onward)

> Emergency Care Data Set (ECDS) (To be released Summer 2022 onward)

FREQUENCY - every 6 months beginning March 2022 (Rapid Registrations every 2 months from March 2022).

************** CHANGES FROM AUGUST 2022 (Version 2 of this agreement) **************

> Systemic Anti-Cancer Therapy Data Set (SACT) - no change

> Hospital Episode Statistics (HES) Admitted Patient Care (APC) - no change

> HES Outpatients (OP) - no change

> Cancer Waiting Times (Treatments and Referrals) (CWT) - no change

> Radiotherapy Data Set – No change

> Diagnostic Imaging Dataset (DIDS) - Released August 2022 on bi-annual basis as planned.

> Emergency Care Data Set (ECDS) - Released August 2022 on bi-annual basis as planned.

> Rapid Registrations Cancer Data Set - Rapid Registrations will be provided every month rather than every two months from August 2022 to enable frequent monitoring of cancer diagnoses in the study.

> Cancer Registry - will be provided every month rather than every other month from August 2022 to enable frequent reporting of cancer diagnoses in the study and use the data as a filter for participant invitations for annual follow up timepoints. Expand historic gold standard registration field selection for 1995-2021 and addition of field to allow the trial team to distinguish provisional from finalised gold standard tumours. These changes are required for primary and secondary analysis. Data will be used to characterize the cohort for risk factors (prior cancers), and also assess test performance in a subpopulation of cancer survivors. The Status of Registration is required to enable accurate reporting and filtering of provisional data vs finalized data to internal and external stakeholders, and the interim analysis in the study. Death fields added to dataset specification for analysis of key secondary endpoints in the study. Death data will also be used for reporting subject disposition in the analysis. Death data, while requested with the cancer registry data, should include information on all participants in the cohort (including participants that have not been diagnosed with cancer).

**************

JUSTIFICATION FOR DATA SETS REQUESTED AND DATA MINIMISATION

> Cancer Registry and Rapid Registrations - These central NHS datasets will be the sole source for the data required for primary and secondary objectives analyses in the Study. Will be used to identify Primary endpoints, which are incidence and stage at diagnosis for cancer types that are stage-able (e.g., with available staging systems). Cancer is defined as any of the following cancers:

- Invasive solid cancer, excluding non-melanoma skin cancer

- Haematologic malignancies, including lymphoma, lymphoid leukemia, myeloma/plasma cell neoplasm, myeloid neoplasms (including myelodysplastic and myeloproliferative neoplasms with behaviour code 3 based on ICD-O-3.2*).

[*The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) is the standard classification system for the registration of cancers in the United States (and most areas of the world) and contains widely-accepted histologic categories of brain tumors.]

The Cancer Registry will also be used to identify cohort characteristics/baseline risk - Personal history of cancer assessed using limited set of data fields. Historical cancer data dating back to 1995 will be available.

> Hospital Episode Statistics (HES) Outpatients (OP), Admitted Patient Care (APC) and Emergency Care Data Set (ECDS), Cancer Waiting Times (Treatments and Referrals) (CWT) and Diagnostic Imaging Dataset (DIDS) - Will be used to identify Secondary endpoints (including safety objectives). Among all test positive cases, number of follow-up procedures and number of invasive procedures (including all biopsies, surgical interventions, bronchoscopy, thoracoscopy and endoscopy) to achieve diagnostic resolution (i.e. cancer diagnosis, non-cancer diagnosis, or no diagnosis and discharge from the diagnostic follow-up). Will also identify:

- Number and type of invasive procedures performed in false positive cases;

- Number of complications and deaths resulting from diagnostic procedures;

- Radiation exposure measured in mSv per participant due to test result-directed evaluations

> Systemic Anti-Cancer Therapy Data Set (SACT) and Radiotherapy Data set (RTDS) will also be used (in addition to the above HES, CWT and DIDs data) for health care resource utilization objectives.

Patient and Public Involvement and Engagement (PPI&E)

Ongoing PPI&E is conducted throughout the trial. Historical PPI&E was conducted to develop the participant facing materials and on questionnaires, and including specific questions to PPI groups around the sharing of pseudonymised data with GRAIL LLC (US). Ongoing PPI&E includes, for example, PPI&E on new participant facing materials, survey feedback on the participant experience, and questionnaires for follow-up of participants. An embedded behavioural scientist in the Kings College London Cancer

Prevention Trials Unit has been involved in, and will continue to be involved in, analysis of participant communications, and ethical and wellbeing issues within the course of the trail, as well as periodically advise the study organizers and other stakeholder on emerging themes.

In the setting up of the trial, GRAIL has taken seriously, strategies for equity of access and has several plans underway to ensure that under-represented minorities are actively recruited into the trial. Amongst this is the invitation strategy, that will identify and actively send invitations to a greater proportion of individuals from these minorities to take part in the trial. It is hoped that this will positively contribute towards equity of access and inform the provision of cancer services in these areas in the future.

GDPR Legal Basis for the Processing of Personal Data:

GRAIL Bio UK Ltd, as joint Data Controller, is using Article 6(1)(f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child." Processing personal data is necessary for GRAIL Bio UK Ltd’s legitimate interests which are described in this application. The data to which access is requested are proportionate and necessary to achieve those interests. GRAIL Bio UK Ltd has completed a legitimate interests assessment (LIA) and NHS Digital are satisfied that the interests of the data subjects do not override their legitimate interests; that they would reasonably expect the processing and it would not cause unjustified harm. The data subjects interests and fundamental rights are protected through appropriate minimisation of fields and patient records being processed; protection of the data in a secure environment, and guaranteeing secure destruction at any stage at the request of NHS Digital or after a defined period on completion of the project.

Additionally (as health data is a special category of Personal Data), GRAIL Bio UK Ltd is using Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

King’s College London, as joint Data Controller, is using Article 6:1(e): Specific task in the ‘public interest’ or task that has a clear basis in law, and additionally (as health data is a special category of Personal Data) Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

Organisation Roles and Responsibilities:

• GRAIL Bio UK Ltd are a joint Data Controller and Lead for this agreement, who also process the data. It is responsible for sponsoring and funding the trial, as well as the building and maintenance of the Clinical Records Management System.

• King’s College London are a joint Data Controller who also process the data and are responsible for implementing the trial and for primary objective analysis.

• GRAIL, Limited Liability Company (LLC). are the manufacturers of the MCED test, Galleri (Registered Trademark), and has set up the above UK subsidiary, GRAIL Bio UK Ltd, which is the sponsor of the NHS-Galleri trial. GRAIL, LLC will be receiving pseudonymised record-level NHS Digital data from King’s College London, and are therefore listed as a Data Processor in this agreement.

NHS England are listed as a strategic partner in relation to NHS-Galleri, but have no input into how the trial is run or determining the outcome. NHS Digital therefore do not consider them to be a Data Controller. Cancer Research UK appear on trial-related documentation by virtue of the fact that King’s College London Cancer Prevention Trials Unit are a Cancer Research UK Trials Unit. Cancer Research UK as an organisation in their own right have no involvement in the running of the trial, nor are they funding any of the trial activities. NHS Digital therefore do not consider them to be a Data Controller.

In the future, GRAIL Bio UK Ltd may receive commercial benefit (including intangible or indirect commercial benefits such as positive publicity) from the successful outcomes of the trial. GRAIL Bio UK Ltd and the NHS have entered into a partnership whereby the Galleri test will be piloted in clinical trials within the NHS in England. If the test is shown to work as intended in these clinical trials, the NHS may purchase the test from GRAIL and make the test routinely available in the future to benefit patient.

All outputs will be aggregated with small number suppressed as per the HES analysis guide or according to the specific data set's suppression guidance.

The results of the NHS-Galleri trial outputs will be published in peer-reviewed publications, presented at conferences, and used by NHS England in their decisions around implementation of the test within the NHS. The interim analysis, which will include analysis of a key secondary objective, is expected by the end of 2023 or beginning of 2024. The primary objective readout is expected by the end of 2025 or beginning of 2026. Planned dissemination activities will aim to enable the engagement with the scientific and policy making communities to ensure that knowledge developed by this research can benefit these communities to help the NHS detect cancer early, when it is easier to treat.

-

GRAIL Bio UK Ltd and King’s College London (KCL), as joint data controllers, are requesting access to pseudonymised record-level data in order to carry out follow-up analysis based on a cohort of patients who are being recruited to a clinical trial called ‘NHS-Galleri’ following being invited to participate if they are found to meet the required eligibility criteria.

Note that all members of the team running the trial are based at the Cancer Prevention Trials Unit (CPTU) at King's College London. For the sake of consistency, and as King's College London is listed as a data controller, when referring to King's College London throughout the application, this encompasses the team at CPTU.

Background, Purpose and Rationale behind NHS-Galleri:

A new Multi-Cancer Early Detection (MCED) test has been developed that can detect many types of cancer from a single blood sample. This test is called Galleri (Registered Trademark). This trial aims to find out whether it is better at discovering cancer early, compared to other tests that the NHS currently uses. The purpose of NHS-Galleri is to demonstrate the clinical utility of the MCED blood test for individuals in a general screening population in a real world NHS setting. The rationale behind this trial is that MCED is a novel screening paradigm, and assessment of the use and impact of test results is necessary to enable integration into clinical practice. This will be the first randomised, double blind, controlled trial statistically powered to assess clinical utility of a MCED test.

Recruitment to the trial is being aided by NHS Digital's Digi-Trials team via a sister application, DARS-NIC-456778-J0G3H, and is taking place over 10 – 12 months from August 2021 with a recruitment target of 140,000 people. Different areas are being targeted for recruitment, beginning with North-west England. Recruitment in a particular area will last for 4 – 6 weeks, before moving on to another area. Recruitment is being managed through mobile units on scheduled routes around a particular area. Batches of invitations are being constructed to target whatever area the mobile units will be in at the time of writing out to potential participants. Potential participants will only be contacted once and there will be no follow-ups sent after the original invitation letter.

Requesting Not to Take Part in Trial:

In addition to the National Data Opt-out, members of the public are able to specifically request not to be contacted for the NHS-Galleri Trial. GRAIL Bio UK Ltd have promoted this information via the dedicated trial website (nhs-galleri.org) and NHS Digital have done the same via a dedicated page on its website (https://digital.nhs.uk/services/nhs-digitrials/nhs-galleri-trial). There is also an option for people to register their request not to take part in NHS-Galleri by telephone. NHS Digital is recording these requests not to take part and ensuring that anyone who has registered for this will be excluded from the cohort selection and as a result their details will not be on the list for invitations to be sent.

Objectives of the Trial:

• Primary Objective:

- For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of stage III and IV cancers diagnosed in the intervention arm as compared to the control arm following three annual test and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a. For a prespecified group of cancer types, including lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver / bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder.

b. For all cancer types, excluding prostate cancer.

c. For all cancer types.

• Secondary Objectives:

1) For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of advanced cancers (defined as a cancer diagnosed at stage III or IV or one that results in a cancer-specific death) in the intervention arm as compared to the control arm following three annual tests and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a) First, evaluate for a statistically significant difference in a prespecified group of primary cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

b) If a statistically significant reduction in absolute numbers is found, continue by evaluating for a difference in all cancer types excluding prostate cancer.

c) If the above evaluations are both significant, evaluate for a difference in all cancer types.

2) For cancers that are routinely staged, determine whether there is a difference in the proportion of stage I and II cancers between the two arms of the study in the third testing round (i.e. for cancers diagnosed with an average 16-18 months of follow-up after the third test).

3) Evaluate the performance (overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three intervals of testing.

4) Evaluate the safety, including harms, of the testing pathway in the intervention arm among individuals with positive MCED test results.

a) Assess participant-reported anxiety among those receiving a positive MCED test result following the initial blood draw at various time points.

b) Examine the radiation exposure by participants associated with follow-up diagnostic procedures following a positive MCED test result

c) Describe the number and type of invasive procedures performed, and number of complications and deaths associated with follow-up diagnostic procedures.

5) Determine whether there is a statistically significant reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following the initial screening round and 12 months of follow-up (excluding cancers identified by the test performed at the 12-month visit)

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer;

c) For all cancer types.

6) Determine whether there is a reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following three annual tests and an average of 16-18 months follow-up after the third test

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer

c) For all cancer types.

7) Determine whether there is a statistically significant difference in cancer detection rates in the two arms of the study with 12 months follow-up after the first test.

8) Model the expected differences in cancer-specific mortality at 7 years post-randomisation based on cancers diagnosed within an average of 40-42 months of randomisation.

9) Summarize stage distribution per cancer type for the two arms, and compare the absolute numbers of stage IV alone as well as stage III and IV cancers, following the third screening and an average 16-18* months of follow-up.

10) Determine whether there is a reduction in the absolute numbers of stage III and IV cancers, excluding breast, cervical, and colorectal (i.e., those without a standard of care screening modality) as compared to the control arm following the third annual screening and an average 16-18 months of follow-up.

11) Compare cancer-specific mortality in the intervention and control arms following the third screening and an average 16-18 months of follow-up, for a pre-specified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

12) Compare cancer-specific mortality in the intervention and control arms for all cancer types at

a) 3 years and

b) 6 years after the last study visit.

13) Compare the absolute numbers of stage IV cancers following the second blood draw and 12 months of follow-up, with prevalent cases excluded.

14) Compare the absolute numbers of stage III and IV cancers following the third blood draw and an average 16-18 months of follow-up, with cases from the first and second screenings excluded.

15) Assess the impact of use of the MCED test across three annual timepoints on healthcare resource utilization for cancer diagnosis and treatment.

16) Assess the potential impact of overdiagnosis by comparing the cumulative incidence (Kaplan-Meier curves) of cancers diagnosed after a positive baseline MCED test up to 36-48 months after randomization between the intervention and control arms (retrospectively testing baseline samples from all participants diagnosed with cancer in the control arm).

DATA SETS REQUESTED:

> Rapid Registrations Cancer Data Set

> Cancer Registry

> Systemic Anti-Cancer Therapy Data Set (SACT)

> Radiotherapy Data set (RTDS)

> Hospital Episode Statistics (HES) Admitted Patient Care (APC)

> Emergency Care Data Set (ECDS) (released Summer 2022 onwards)

> HES Outpatients (OP)

> Cancer Waiting Times (Treatments and Referrals) (CWT)

> Diagnostic Imaging Dataset (DIDS) (released Summer 2022 onwards)

Patient and Public Involvement and Engagement (PPI&E)

Ongoing PPI&E is conducted throughout the trial. Historical PPI&E was conducted to develop the participant facing materials and on questionnaires, and including specific questions to PPI groups around the sharing of pseudonymised data with GRAIL LLC (US). Ongoing PPI&E includes, for example, PPI&E on new participant facing materials, survey feedback on the participant experience, and questionnaires for follow-up of participants. An embedded behavioural scientist in the Kings College London Cancer

Prevention Trials Unit has been involved in, and will continue to be involved in, analysis of participant communications, and ethical and wellbeing issues within the course of the trail, as well as periodically advise the study organizers and other stakeholder on emerging themes.

In the setting up of the trial, GRAIL has taken seriously, strategies for equity of access and has several plans underway to ensure that under-represented minorities are actively recruited into the trial. Amongst this is the invitation strategy, that will identify and actively send invitations to a greater proportion of individuals from these minorities to take part in the trial. It is hoped that this will positively contribute towards equity of access and inform the provision of cancer services in these areas in the future.

JUSTIFICATION FOR DATA SETS REQUESTED AND DATA MINIMISATION

> Cancer Registry and Rapid Registrations - These central NHS datasets will be the sole source for the data required for primary and secondary objectives analyses in the Study. Will be used to identify Primary endpoints, which are incidence and stage at diagnosis for cancer types that are stage-able (e.g., with available staging systems). Cancer is defined as any of the following cancers:

- Invasive solid cancer, excluding non-melanoma skin cancer

- Haematologic malignancies, including lymphoma, lymphoid leukemia, myeloma/plasma cell neoplasm, myeloid neoplasms (including myelodysplastic and myeloproliferative neoplasms with behaviour code 3 based on ICD-O-3.2*).

[*The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) is the standard classification system for the registration of cancers in the United States (and most areas of the world) and contains widely-accepted histologic categories of brain tumors.]

The Cancer Registry will also be used to identify cohort characteristics/baseline risk - Personal history of cancer assessed using limited set of data fields. Historical cancer data dating back to 1995 will be available.

> Hospital Episode Statistics (HES) Outpatients (OP), Admitted Patient Care (APC) and Emergency Care Data Set (ECDS), Cancer Waiting Times (Treatments and Referrals) (CWT) and Diagnostic Imaging Dataset (DIDS) - Will be used to identify Secondary endpoints (including safety objectives). Among all test positive cases, number of follow-up procedures and number of invasive procedures (including all biopsies, surgical interventions, bronchoscopy, thoracoscopy and endoscopy) to achieve diagnostic resolution (i.e. cancer diagnosis, non-cancer diagnosis, or no diagnosis and discharge from the diagnostic follow-up). Will also identify:

- Number and type of invasive procedures performed in false positive cases;

- Number of complications and deaths resulting from diagnostic procedures;

- Radiation exposure measured in mSv per participant due to test result-directed evaluations

> Systemic Anti-Cancer Therapy Data Set (SACT) and Radiotherapy Data set (RTDS) will also be used (in addition to the above HES, CWT and DIDs data) for health care resource utilization objectives.

GDPR Legal Basis for the Processing of Personal Data:

GRAIL Bio UK Ltd, as joint Data Controller, is using Article 6(1)(f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child." Processing personal data is necessary for GRAIL Bio UK Ltd’s legitimate interests which are described in this application. The data to which access is requested are proportionate and necessary to achieve those interests. GRAIL Bio UK Ltd has completed a legitimate interests assessment (LIA) and NHS Digital are satisfied that the interests of the data subjects do not override their legitimate interests; that they would reasonably expect the processing and it would not cause unjustified harm. The data subjects interests and fundamental rights are protected through appropriate minimisation of fields and patient records being processed; protection of the data in a secure environment, and guaranteeing secure destruction at any stage at the request of NHS Digital or after a defined period on completion of the project.

Additionally (as health data is a special category of Personal Data), GRAIL Bio UK Ltd is using Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

King’s College London, as joint Data Controller, is using Article 6:1(e): Specific task in the ‘public interest’ or task that has a clear basis in law, and additionally (as health data is a special category of Personal Data) Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

Organisation Roles and Responsibilities:

• GRAIL Bio UK Ltd are a joint Data Controller and Lead for this agreement, who also process the data. It is responsible for sponsoring and funding the trial, as well as the building and maintenance of the Clinical Records Management System.

• King’s College London are a joint Data Controller who also process the data and are responsible for implementing the trial and for primary objective analysis.

• GRAIL, Limited Liability Company (LLC). are the manufacturers of the MCED test, Galleri (Registered Trademark), and has set up the above UK subsidiary, GRAIL Bio UK Ltd, which is the sponsor of the NHS-Galleri trial. GRAIL, LLC will be receiving pseudonymised record-level NHS Digital data from King’s College London, and are therefore listed as a Data Processor in this agreement.

NHS England are listed as a strategic partner in relation to NHS-Galleri, but have no input into how the trial is run or determining the outcome. NHS Digital therefore do not consider them to be a Data Controller. Cancer Research UK appear on trial-related documentation by virtue of the fact that King’s College London Cancer Prevention Trials Unit are a Cancer Research UK Trials Unit. Cancer Research UK as an organisation in their own right have no involvement in the running of the trial, nor are they funding any of the trial activities. NHS Digital therefore do not consider them to be a Data Controller.

In the future, GRAIL Bio UK Ltd may receive commercial benefit (including intangible or indirect commercial benefits such as positive publicity) from the successful outcomes of the trial. GRAIL Bio UK Ltd and the NHS have entered into a partnership whereby the Galleri test will be piloted in clinical trials within the NHS in England. If the test is shown to work as intended in these clinical trials, the NHS may purchase the test from GRAIL and make the test routinely available in the future to benefit patient.

All outputs will be aggregated with small number suppressed as per the HES analysis guide or according to the specific data set's suppression guidance.

The results of the NHS-Galleri trial outputs will be published in peer-reviewed publications, presented at conferences, and used by NHS England in their decisions around implementation of the test within the NHS. The interim analysis, which will include analysis of a key secondary objective, is expected by the end of 2023 or beginning of 2024. The primary objective readout is expected by the end of 2025 or beginning of 2026. Planned dissemination activities will aim to enable the engagement with the scientific and policy making communities to ensure that knowledge developed by this research can benefit these communities to help the NHS detect cancer early, when it is easier to treat.

-

GRAIL Bio UK Ltd and King’s College London (KCL), as joint data controllers, are requesting access to pseudonymised record-level data in order to carry out follow-up analysis based on a cohort of patients who are being recruited to a clinical trial called ‘NHS-Galleri’ following being invited to participate if they are found to meet the required eligibility criteria.

Note that all members of the team running the trial are based at the Cancer Prevention Trials Unit (CPTU) at King's College London. For the sake of consistency, and as King's College London is listed as a data controller, when referring to King's College London throughout the application, this encompasses the team at CPTU.

Background, Purpose and Rationale behind NHS-Galleri:

A new Multi-Cancer Early Detection (MCED) test has been developed that can detect many types of cancer from a single blood sample. This test is called Galleri™ and this trial aims to find out whether it is better at discovering cancer early, compared to other tests that the NHS currently uses. The purpose of NHS-Galleri is to demonstrate the clinical utility of the MCED blood test for individuals in a general screening population in a real world NHS setting. The rationale behind this trial is that MCED is a novel screening paradigm, and assessment of the use and impact of test results is necessary to enable integration into clinical practice. This will be the first randomised, double blind, controlled trial statistically powered to assess clinical utility of a MCED test.

Recruitment to the trial is being aided by NHS Digital's Digi-Trials team via a sister application, DARS-NIC-456778-J0G3H, and is taking place over 10 – 12 months from August 2021 with a recruitment target of 140,000 people. Different areas are being targeted for recruitment, beginning with North-west England. Recruitment in a particular area will last for 4 – 6 weeks, before moving on to another area. Recruitment is being managed through mobile units on scheduled routes around a particular area. Batches of invitations are being constructed to target whatever area the mobile units will be in at the time of writing out to potential participants. Potential participants will only be contacted once and there will be no follow-ups sent after the original invitation letter.

Requesting Not to Take Part in Trial:

In addition to the National Data Opt-out, members of the public are able to specifically request not to be contacted for the NHS-Galleri Trial. GRAIL Bio UK Ltd have promoted this information via the dedicated trial website (nhs-galleri.org) and NHS Digital have done the same via a dedicated page on its website (https://digital.nhs.uk/services/nhs-digitrials/nhs-galleri-trial). There is also an option for people to register their request not to take part in NHS-Galleri by telephone. NHS Digital is recording these requests not to take part and ensuring that anyone who has registered for this will be excluded from the cohort selection and as a result their details will not be on the list for invitations to be sent.

Objectives of the Trial:

• Primary Objective:

- For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of stage III and IV cancers diagnosed in the intervention arm as compared to the control arm following three annual test and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a. For a prespecified group of cancer types, including lung, head and neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver / bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder.

b. For all cancer types, excluding prostate cancer.

c. For all cancer types.

• Secondary Objectives:

1) For cancers that are routinely staged, determine whether there is a statistically significant reduction in the absolute numbers of advanced cancers (defined as a cancer diagnosed at stage III or IV or one that results in a cancer-specific death) in the intervention arm as compared to the control arm following three annual tests and an average of 16-18 months follow-up after the third test, using a fixed-sequence strategy as below:

a) First, evaluate for a statistically significant difference in a prespecified group of primary cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

b) If a statistically significant reduction in absolute numbers is found, continue by evaluating for a difference in all cancer types excluding prostate cancer.

c) If the above evaluations are both significant, evaluate for a difference in all cancer types.

2) For cancers that are routinely staged, determine whether there is a difference in the proportion of stage I and II cancers between the two arms of the study in the third testing round (i.e. for cancers diagnosed with an average 16-18 months of follow-up after the third test).

3) Evaluate the performance (overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and cancer signal origin accuracy) of the MCED test in the intervention arm after each annual round of testing and in aggregate across the three intervals of testing.

4) Evaluate the safety, including harms, of the testing pathway in the intervention arm among individuals with positive MCED test results.

a) Assess participant-reported anxiety among those receiving a positive MCED test result following the initial blood draw at various time points.

b) Examine the radiation exposure by participants associated with follow-up diagnostic procedures following a positive MCED test result

c) Describe the number and type of invasive procedures performed, and number of complications and deaths associated with follow-up diagnostic procedures.

5) Determine whether there is a statistically significant reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following the initial screening round and 12 months of follow-up (excluding cancers identified by the test performed at the 12-month visit)

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer;

c) For all cancer types.

6) Determine whether there is a reduction in the absolute numbers of stage IV cancers diagnosed in the intervention arm compared to the control arm, following three annual tests and an average of 16-18 months follow-up after the third test

a) For a prespecified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder;

b) For all cancer types excluding prostate cancer

c) For all cancer types.

7) Determine whether there is a statistically significant difference in cancer detection rates in the two arms of the study with 12 months follow-up after the first test.

8) Model the expected differences in cancer-specific mortality at 7 years post-randomisation based on cancers diagnosed within an average of 40-42 months of randomisation.

9) Summarize stage distribution per cancer type for the two arms, and compare the absolute numbers of stage IV alone as well as stage III and IV cancers, following the third screening and an average 16-18* months of follow-up.

10) Determine whether there is a reduction in the absolute numbers of stage III and IV cancers, excluding breast, cervical, and colorectal (i.e., those without a standard of care screening modality) as compared to the control arm following the third annual screening and an average 16-18 months of follow-up.

11) Compare cancer-specific mortality in the intervention and control arms following the third screening and an average 16-18 months of follow-up, for a pre-specified group of cancer types: lung, head & neck, colorectal, pancreas, myeloma/plasma cell neoplasm, liver/bile duct, stomach, oesophagus, anus, lymphoma, ovary, and bladder.

12) Compare cancer-specific mortality in the intervention and control arms for all cancer types at

a) 3 years and

b) 6 years after the last study visit.

13) Compare the absolute numbers of stage IV cancers following the second blood draw and 12 months of follow-up, with prevalent cases excluded.

14) Compare the absolute numbers of stage III and IV cancers following the third blood draw and an average 16-18 months of follow-up, with cases from the first and second screenings excluded.

15) Assess the impact of use of the MCED test across three annual timepoints on healthcare resource utilization for cancer diagnosis and treatment.

16) Assess the potential impact of overdiagnosis by comparing the cumulative incidence (Kaplan-Meier curves) of cancers diagnosed after a positive baseline MCED test up to 36-48 months after randomization between the intervention and control arms (retrospectively testing baseline samples from all participants diagnosed with cancer in the control arm).

DATA SETS REQUESTED:

> Rapid Registrations Cancer Data Set

> Cancer Registry

> Systemic Anti-Cancer Therapy Data Set (SACT)

> Radiotherapy Data set (RTDS)

> Hospital Episode Statistics (HES) Admitted Patient Care (APC)

> Emergency Care Data Set (ECDS) (released Summer 2022 onwards)

> HES Outpatients (OP)

> Cancer Waiting Times (Treatments and Referrals) (CWT)

> Diagnostic Imaging Dataset (DIDS) (released Summer 2022 onwards)

Patient and Public Involvement and Engagement (PPI&E)

Ongoing PPI&E is conducted throughout the trial. Historical PPI&E was conducted to develop the participant facing materials and on questionnaires, and including specific questions to PPI groups around the sharing of pseudonymised data with GRAIL LLC (US). Ongoing PPI&E includes, for example, PPI&E on new participant facing materials, survey feedback on the participant experience, and questionnaires for follow-up of participants. An embedded behavioural scientist in the Kings College London Cancer

Prevention Trials Unit has been involved in, and will continue to be involved in, analysis of participant communications, and ethical and wellbeing issues within the course of the trail, as well as periodically advise the study organizers and other stakeholder on emerging themes.

In the setting up of the trial, GRAIL has taken seriously, strategies for equity of access and has several plans underway to ensure that under-represented minorities are actively recruited into the trial. Amongst this is the invitation strategy, that will identify and actively send invitations to a greater proportion of individuals from these minorities to take part in the trial. It is hoped that this will positively contribute towards equity of access and inform the provision of cancer services in these areas in the future.

JUSTIFICATION FOR DATA SETS REQUESTED AND DATA MINIMISATION

> Cancer Registry and Rapid Registrations - These central NHS datasets will be the sole source for the data required for primary and secondary objectives analyses in the Study. Will be used to identify Primary endpoints, which are incidence and stage at diagnosis for cancer types that are stage-able (e.g., with available staging systems). Cancer is defined as any of the following cancers:

- Invasive solid cancer, excluding non-melanoma skin cancer

- Haematologic malignancies, including lymphoma, lymphoid leukemia, myeloma/plasma cell neoplasm, myeloid neoplasms (including myelodysplastic and myeloproliferative neoplasms with behaviour code 3 based on ICD-O-3.2*).

[*The International Classification of Diseases for Oncology, Third Edition (ICD-O-3) is the standard classification system for the registration of cancers in the United States (and most areas of the world) and contains widely-accepted histologic categories of brain tumors.]

The Cancer Registry will also be used to identify cohort characteristics/baseline risk - Personal history of cancer assessed using limited set of data fields. Historical cancer data dating back to 1995 will be available.

> Hospital Episode Statistics (HES) Outpatients (OP), Admitted Patient Care (APC) and Emergency Care Data Set (ECDS), Cancer Waiting Times (Treatments and Referrals) (CWT) and Diagnostic Imaging Dataset (DIDS) - Will be used to identify Secondary endpoints (including safety objectives). Among all test positive cases, number of follow-up procedures and number of invasive procedures (including all biopsies, surgical interventions, bronchoscopy, thoracoscopy and endoscopy) to achieve diagnostic resolution (i.e. cancer diagnosis, non-cancer diagnosis, or no diagnosis and discharge from the diagnostic follow-up). Will also identify:

- Number and type of invasive procedures performed in false positive cases;

- Number of complications and deaths resulting from diagnostic procedures;

- Radiation exposure measured in mSv per participant due to test result-directed evaluations

> Systemic Anti-Cancer Therapy Data Set (SACT) and Radiotherapy Data set (RTDS) will also be used (in addition to the above HES, CWT and DIDs data) for health care resource utilization objectives.

GDPR Legal Basis for the Processing of Personal Data:

GRAIL Bio UK Ltd, as joint Data Controller, is using Article 6(1)(f) "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party except where such interests are overridden by the interests or fundamental rights and freedoms of the data subject which require protection of personal data, in particular where the data subject is a child." Processing personal data is necessary for GRAIL Bio UK Ltd’s legitimate interests which are described in this application. The data to which access is requested are proportionate and necessary to achieve those interests. GRAIL Bio UK Ltd has completed a legitimate interests assessment (LIA) and NHS Digital are satisfied that the interests of the data subjects do not override their legitimate interests; that they would reasonably expect the processing and it would not cause unjustified harm. The data subjects interests and fundamental rights are protected through appropriate minimisation of fields and patient records being processed; protection of the data in a secure environment, and guaranteeing secure destruction at any stage at the request of NHS Digital or after a defined period on completion of the project.

Additionally (as health data is a special category of Personal Data), GRAIL Bio UK Ltd is using Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

King’s College London, as joint Data Controller, is using Article 6:1(e): Specific task in the ‘public interest’ or task that has a clear basis in law, and additionally (as health data is a special category of Personal Data) Article 9:2(j): Special category data used for “Archiving in the public interest, scientific or historical research or statistical purposes”, with a basis in law.

Organisation Roles and Responsibilities:

• GRAIL Bio UK Ltd are a joint Data Controller and Lead for this agreement, who also process the data. It is responsible for sponsoring and funding the trial, as well as the building and maintenance of the Clinical Records Management System.

• King’s College London are a joint Data Controller who also process the data and are responsible for implementing the trial and for primary objective analysis.

• GRAIL, Limited Liability Company (LLC). are the manufacturers of the MCED test, Galleri™, and has set up the above UK subsidiary, GRAIL Bio UK Ltd, which is the sponsor of the NHS-Galleri trial. GRAIL, LLC will be receiving pseudonymised record-level NHS Digital data from King’s College London, and are therefore listed as a Data Processor in this agreement.

NHS England are listed as a strategic partner in relation to NHS-Galleri, but have no input into how the trial is run or determining the outcome. NHS Digital therefore do not consider them to be a Data Controller. Cancer Research UK appear on trial-related documentation by virtue of the fact that King’s College London Cancer Prevention Trials Unit are a Cancer Research UK Trials Unit. Cancer Research UK as an organisation in their own right have no involvement in the running of the trial, nor are they funding any of the trial activities. NHS Digital therefore do not consider them to be a Data Controller.

In the future, GRAIL Bio UK Ltd may receive commercial benefit (including intangible or indirect commercial benefits such as positive publicity) from the successful outcomes of the trial. GRAIL Bio UK Ltd and the NHS have entered into a partnership whereby the Galleri™ test will be piloted in clinical trials within the NHS in England. If the test is shown to work as intended in these clinical trials, the NHS may purchase the test from GRAIL and make the test routinely available in the future to benefit patient.

All outputs will be aggregated with small number suppressed as per the HES analysis guide or according to the specific data set's suppression guidance.

The results of the NHS-Galleri trial outputs will be published in peer-reviewed publications, presented at conferences, and used by NHS England in their decisions around implementation of the test within the NHS. The interim analysis, which will include analysis of a key secondary objective, is expected by the end of 2023 or beginning of 2024. The primary objective readout is expected by the end of 2025 or beginning of 2026. Planned dissemination activities will aim to enable the engagement with the scientific and policy making communities to ensure that knowledge developed by this research can benefit these communities to help the NHS detect cancer early, when it is easier to treat.

-