Are We on the Cusp of a Diagnostic Revolution for Cancer?

Editor's note: The author's disclosure's have been updated

Every day in Britain, a thousand patients will be diagnosed with cancer. The UK falls behind most European countries in overall survival at 5 years – the key international metric of cancer outcomes. The reasons for this are complex but increasing evidence points to delay in the diagnostic process being the single most important factor.

Data looking at 1-year survival shows even greater disparity from Europe. The 1-year survival statistic is mainly determined by delays in diagnosis and not the quality of oncological care for the four common cancers - breast, lung, prostate, and colon.

Cancer survival data are driven by the stage of the disease at presentation. For all four of the major cancers, Stage I disease has an over 90% 5-year survival. Stage 2 disease represents extension of the cancer to neighbouring lymph nodes or surrounding tissues, whilst stage 3 and 4 have increasing spread of cancer around the body. Staging systems are slightly different but the principle is the same. The outcome drops precipitously as the stage increases. Any delay is likely to result in upstage migration and, therefore, result in poorer survival for that cancer. More advanced cancer is more expensive to treat as it may involve significantly more radiotherapy and chemotherapy, and so achieving an earlier diagnosis could well be cost effective.

The Problem

There are three causes for delay in cancer diagnostic pathways – the patient, the GP, and the hospital. The patient's decision to seek medical advice is the key entry point in any healthcare system. The threshold to seek help is complex and determined by multiple factors, including age, sex, ethnicity, socioeconomic status, and educational background. Educational campaigns have been successful in changing behaviours and are the responsibility of public health. But for many, access is becoming increasingly difficult and, therefore, the threshold to seeking help may be increased.

Traditionally in the UK, the GP has acted as the gatekeeper to referral of a patient to secondary care. This system has worked well in the past but the considerable changes in primary care over the last two decades have impinged adversely on getting an early cancer diagnosis. Many GPs are now part-time working defined sessions, so patient continuity is no longer a feature. The lack of 24-hour cover and the use of other agencies such as urgent care, walk-in centres, and NHS 111 mean that many patients are unable to name their own GP. This means that when new symptoms arise there is no knowledge of the patient, their past medical history, or their psychological background by the doctor.

But there are currently significant problems in GP access for patients and getting an appointment may take several weeks. The 2-week wait system was introduced as a pragmatic measure in 2000. The problem is that cancer symptoms are vague and over the last 20 years multiple studies have demonstrated that only 25-27% of cancer patients are diagnosed through this pathway. Creating a fast track in a resource-constrained diagnostic environment also reduces the speed of the routine pathway so may not necessarily enhance the overall group.

Although excellent NICE guidelines on recognition and referral were produced in 2015 and updated in 2021, even experienced GPs cannot distinguish the myriad of symptoms caused by cancer rather than less serious temporary problems.

System delay after referral is common in most hospitals. Investigations and specialist opinions are ordered sequentially often over several weeks. The capacity may not be adequate for imaging by CT, MRI, and endoscopy services are very stretched. The target from diagnosis to first definitive treatment is currently 62 days. This is being increasingly breached because of capacity problems.

Pan-Cancer Blood Tests

Recently, several pan-cancer blood tests are being trialled. Imagine a gadget in primary care that gives you an instant read out from a finger prick sample or a lateral flow test that we've all become familiar with over the last 3 years. You leave the surgery knowing if you have or don't have cancer, and ideally from which organ. A lot of hype is made of this concept through PR companies working for biotech, so positive articles repeatedly appear. But there now are some really promising approaches.

A variety of techniques are being used, including protein biomarkers, anti-tumour antibodies, patterns of DNA methylation, and measurements of changes in DNA and RNA sequences. It is difficult to know which will be the most effective, but all are being examined in the highly competitive environment of US biotechnology. Data mining using artificial intelligence system is being applied to seek our hitherto unseen correlations not obviously visible in the raw data.

The most well-known test is Galleri from a Californian company, Grail. Britain's NHS has partnered with them to study the early detection of lung cancer in London. It works by examining methylation patterns in circulating cell free DNA coming from cancer cells. As different tissues have different methylation signals it may also be able to identify the tissue of origin of the cancer, a useful feature in pinpointing the cancer's origin. Within trials, the test is free but it's on sale in the US for nearly $1000 USD.

The impeccably designed SUMMIT study aims to validate Galleri and to deliver low dose CT scan screening for lung cancer in smokers. Professor Sam Janes who set up the Lungs for Living Research Centre at University College Hospital in London, told me "The SUMMIT study has recruited over 13,000 people aged 55-77 at high risk of lung cancer. Its two primary aims are first to examine the feasibility of CT screening a large section of a major city; and second evaluates the performance of Galleri multi-cancer blood test. While the trial runs to 2023 enabling a ten-year follow up after three years of screening, the first results on the blood test performance should be ready this year".

He added "numerous sub-questions are being asked, including annual versus biennial screening, cost efficacy of screening, and the lessons learnt have contributed to the NHS Lung Health Check pilot studies. Importantly, the resources provided by the study funders (GRAIL Inc), including four CT scanners, have been gifted to the NHS to enable CT screening to continue in North London after screening closes in May 2023. An excellent example of mutually beneficial working between the NHS and industry."

There are several other methylation analytic approaches, including Singlera's PanSeer test and Adela's methylome enrichment technology. Methylation of DNA seems to be a major control mechanism for gene transcription, so it is perhaps not surprising that cancer cells streamlined for growth show different patterns.

Looking for cancer mutations in circulating tumour DNA (ctDNA) is another more obvious approach. The problem is the sensitivity of the test and the background noise produce false negatives and positives respectively. Exact Sciences with their CancerSEEK and StageZero with Aristotle which looks at changes in RNA are both in trial in US populations. In a study of 10,000 women, CancerSEEK detected 26 cancers not found by standard screening; 5 were Stage I, which is very encouraging. Aristotle had similar results, although the false-positive rate was between 0.3% and 7.0%. Yet another approach is to use patterns of DNA fragmentation, which become disturbed when shed into the bloodstream by malignant cells. The DELFI (DNA evaluation of fragments of early interception) test from Delfi Diagnostics is especially sensitive in detecting lung cancer. The key metrics of success are a test's Positive Predictive Value (PPV), its sensitivity and specificity measured at different stages of disease.

Downstream Investigations

Curiously, all these tests are being marketed direct to consumers even through the evidence of their efficacy is scanty at this stage. No drug would be licenced with such little evidence. This creates a problem for GPs and oncologists alike. In London, some of these tests are now advertised to the public by private clinics. Kits can even be ordered online to be sent back for analysis. What patients do with any abnormal results is not clear.

The educated 'worried well' who buy them outside of trials become even more worried with a positive result and naturally seek help. Further blood tests may clarify the situation but still leave the source of the tumour hazy for most. Total body scanning by MRI or PET-CT are possible but expensive especially if there is no information as to the primary tumour site. Oncologists have become very organ specialised so to whom does a patient with a positive test get referred? How do we know what to biopsy? And how do we deal with the psychological impact of having a positive test but feeling completely well? 

Only carefully controlled studies such as SUMMIT will yield the answers. But biotech investors want results tomorrow so they can get their products out there as rapidly as possible. The market is huge as the same tests have the potential to be used not just for diagnosis but for monitoring response to therapy and to detect recurrence before any symptoms arise. Ideally, they want to sell their created companies on to the big diagnostic players for billions of dollars as quickly as possible.

Pan-cancer tests are definitely the future, but we're just not there yet.