Clinical studies for early cancer detection could be a game changer

A series of clinical trials examining biomarkers to reduce cancer deaths by around 20% could revolutionise cancer care. What does this mean for insurers?

The UK's NHS and biotech firm GRAIL announced that they are working together to make early cancer detection available for over 50 types of cancers. Whilst detection of cancer that can be treated earlier is a laudable aim, there are concerns from an insurance perspective, particularly for critical illness covers.  

What is it about?

In the UK, 43% of cancers are detected at stage 3 or 41. This means they have progressed sufficiently to either make treatment less successful or not curative, or that patients are less likely to survive than those who are treated earlier. In general, the UK cancer survival rate lags behind Europe, according to the International Cancer Benchmarking Project2.

As detecting cancer early on is linked to more favourable outcomes, GRAIL commenced a series of clinical studies (PATHFINDER3, STRIVE4, CCGA5 and SUMMIT6) with their Galleri blood test from 2017. This test examines biomarkers in the blood to indicate early stage cancer, often far earlier than other conventional tests. Their approach received FDA approval with breakthrough device designation in 2019. In the US, the SUMMIT and PATHFINDER trials are still recruiting, although on a much smaller scale than the recent UK studies. The UK NHS trial will have two cohorts of patients: a group with high pre-test probability, and one with no symptoms and no risk factors. Higher risk is defined as having smoked more than 100 cigarettes in a lifetime, having a family history of cancer, or having had cancer previously. Ongoing hormonal therapy for cancer is also a risk factor included in this group.

This study seeks to answer whether patients with no traditional cancer risk factors would benefit from early detection. This is a very exciting development but raises many questions.
Dr. Debbie Smith, Swiss Re's CMO for EMEA

What's the objective?

The goal is to improve detection of early cancers. The UK study aims to reduce cancer deaths by around 20%, achieved by detecting earlier cancers that are not subject to national surveillance programmes. Those include pancreatic, ovarian, head and neck and some blood cancers. Early detection of some cancers, such as pancreatic adenocarcinoma with a five-year survival rate, would likely revolutionise care.

What's planned?

In mid-2021, the NHS and GRAIL will recruit 165,000 participants, of which 140,000 will be 50 years or older and at low risk for cancer. A further 25,000 patients, aged 40 and over, will be the higher risk group as identified via their NHS medical health records. In the US, GRAIL is to complete the STRIVE trial in 2025. This trial is collecting blood samples from 100,000 women undergoing surveillance mammography. Its PATHFINDER trial is set to conclude in early 2022, with 6200 participants aged 50 or over being tested on the same criteria as the NHS trial.

How does it work?

A single blood sample is taken. The test is regarded as a liquid biopsy and identifies cfNA (circulating cell free nucleic acids) that indicates the presence of a potential malignancy. We released a briefing on liquid biopsies and their potential impacts on insurance.

What do the test results show?

If the test comes back positive for cancer, a "signal detected" result is given. The test identifies either the predicted organ where the cancer is lurking, or shows that the cancer's location cannot be identified – known as an indeterminate tissue of origin (TOO) result.

What happens next?

Clinical teams will decide the patient's treatment. In case of an identified organ, such as the ovaries or pancreas, specific imaging will likely be performed to confirm the test result. If this initial scan detects likely early cancer, then wider scanning would be expected. For example, a clinician could order a CT scan of the chest, abdomen and pelvis or an MRI. A blood panel would also be performed including any relevant tumour markers that are currently used for diagnosis. Treatment would carry on according to clinical guidelines.

What is less clear is what will be done when no specific organ target is noted – the TOO result. A CT chest, abdomen and pelvis scan or an MRI can detect lesions once they have grown to 3mm, PET scanning around 4mm. Cancers of unknown origin, so within this 3mm tolerance, accounted for 2% of all new UK cancer diagnoses in 2017, but caused 5% of cancer deaths, so this a clear area of interest.

What are potential pitfalls?

Clinicians have two main concerns. The first is about anxiety levels that patients will experience when they have a positive liquid biopsy test. This inevitably will be exacerbated if imaging doesn't reveal an identifiable cancer.  Who supports the patient through this journey once the trial is complete? The follow-up period of the latest GRAIL trial is 12 months – previous smaller trials extended to three years – and surveillance and support thereafter presumably falls to the NHS.

The second worry deals with resourcing. With a national shortage of radiologists7 available to report scans on existing cancer pathways, a parallel investment in AI-based cancer detection would be necessary to make a large-scale rollout workable. While AI-based approaches have been trialled before, they have not been tested at this level, and still require a qualified radiologist to review and confirm the results.

Dr. Debbie Smith, Swiss Re's CMO for EMEA, said: "This is a very exciting development, but one which raises many questions, particularly regarding differences in detection rates between the low- and high-risk groups. The focus of the smaller trials has been to identify cancers that present late, have poor survival and fall outside national screening programmes. This larger scale study seeks to also answer whether patients with no traditional risk factors would benefit. This widening of the net increases the potential for false positives, and the study objectives include measures of patient anxiety, thus confirming that this is a very genuine concern. The ability of the NHS to support significantly increased levels of diagnostic imaging is currently questionable, particularly in the wake of the COVID-19 pandemic."

When will the results be reported?

Currently, results in the UK are expected in 2023, with a further 1million patients planned for testing during 2024-2025 if the first trial is successful.

What is the impact on insurers?

The clinical trials may impact insurance as follows: 

  • Life: Improved efficacy and earlier diagnosis of cancer subtypes will result in earlier treatment, and therefore a likely improvement in mortality.
  • Critical illness (CI) & cancer only: Anti-selection risk around cancer already exists through early screening to detect initial stage cancers and may increase with biomarker testing. We confirm that an "invasive cancer" diagnosis is required to trigger a CI claim payment. This status is confirmed histopathologically by tissue biopsy. Liquid biopsy based detection would not meet this requirement. Any policy definitions that do not require a tissue diagnosis would be open to potential claims, regardless of whether the positive liquid biopsy test result materialises into a tumour which requires treatment.
  • Disability income (DI) & other living benefits: Given the proportionately low contribution of cancer to disability benefit claims, the impact here will likely be small. However, this area will certainly require ongoing monitoring as implementation of the practice continues.