In an Editorial published in 2016, The Lancet Oncology commented on the plans of the company, GRAIL, to create a universal blood test that would allow screening of asymptomatic people for several types of cancer. We raised several questions, including the sensitivity and specificity of these tests, what tumours could be detected, the ethical implications for positive cases, and how to best inform clinical practice and guide the patient. Now, two new studies have been published that attempt to answer some of these questions: the CCGA study and the DETECT-A trial.
Initial results are encouraging. The CCGA study used a test to detect more than 50 types of cancer with high specificity (>99%). Sensitivity in 12 prespecified cancer types increased with tumour stage (from 39% sensitivity in stage I to 92% in stage IV disease). The DETECT-A trial combined the blood test with whole-body PET imaging. The specificity was also high at greater than 99%; 65% of cancers were detected at an early stage, and sensitivity varied with tumour type. These studies, however, reignite the debate about the ethical implications of general population testing and how to manage positive test results. They also raise the question of feasibility.
When technology evolves, price becomes an important lever in how widely it can be adopted. One clear lesson from the COVID-19 pandemic is that, even in high-income countries, mass testing can quickly exacerbate existing funding frailties in health systems. This consideration is particularly important with cancer, given numbers are increasing fast worldwide. The case for long-term savings made by detecting and treating cancers earlier must be made more clearly to governments and insurers alike.
4 years after GRAIL's ambition was announced, we now see encouraging scientific progress. However, several of our previous questions remain unanswered, and importantly, can these tests be refined to have a cost-benefit suitable for use worldwide?
© 2020 Fanatic Studio/ Gary Waters/Science Photo Library