5 March 2018
Scandinavian researchers have reported findings from a study involving nearly 9,000 newly diagnosed patients with Type 2 diabetes, which they identify as five distinct sub-groups based on disease characteristics.
This is strong evidence both that personalised medicine may be feasible outside the traditional areas of rare disease and cancer, and have some effect in combating one of the major public health problems of the modern age.
Diabetes has two major sub-groups: Type 1 diabetes is an auto-immune disease affecting around 10% of patients and typically requiring control with injections of artificial insulin via pen or pump devices, as well as other medication.
The majority of diabetes patients (around 90%) have Type 2 diabetes (T2D), which typically develops later in life; risk of this form of the disease has a major lifestyle component, and there is also scope to reduce risk, control and even reverse early stage disease through lifestyle modifications, particularly weight loss in the overweight and obese as well as maintaining a healthy diet and exercise regime. Many patients also need medication, and some may require insulin treatment. There are also some much rarer forms of diabetes, such as the inherited diseases Maturity Onset Diabetes of the Young (MODY) and neonatal diabetes, as set out in the 2014 PHG Foundation briefing The genomic contribution to diabetes, but T2D – incidence of which is currently soaring in the UK (where approaching five million people are thought to be affected) as in other countries around the world – poses the greatest challenge to healthcare systems.
Currently, T2D is diagnosed on the basis of blood glucose levels, but this is a relatively crude tool, and patients differ markedly in how they both present at diagnosis and how their disease progresses. The new research from Lund University Diabetes Centre in Sweden and the Institute for Molecular Medicine in Finland examined T2D patients, characterising them by features such as age and body mass index (BMI), as well as specific features of their disease and genetic markers from blood analysis.
Mathematical analysis showed that the patients could be separated into five different groups, associations that were maintained when results were tested and compared with healthy group controls. These five proposed T2D sub-groups were:
These sub-groups or clusters were verified by testing in other patient groups, and were also analysed for their relationship to clinical disease progression and complications. Interestingly, all except the SIRD group were associated with known genetic risk factors for T2D.
Generally speaking, it is thought that the capacity to sub-divide or stratify patient populations will increase capacity to offer more personalised care and treatment. Overall, whilst the first potential new T2D sub-group group (severe autoimmune diabetes) appears to represent a form of Type 1 diabetes, the second and third groups - severe insulin-deficient and severe insulin resistant diabetes – appear to be new and more serious forms of disease normally hidden amid the milder obesity and age-related cases. Taking steps to identify these groups and prioritising them for intensive interventions to prevent serious complications such as diabetic retinopathy, non-alcoholic fatty liver disease and kidney disease, as advocated by the authors of the study, would therefore make a lot of sense, assuming that the results of this research are found to apply to other patient populations too.
Using limited healthcare resources more efficiently by targeting the most intensive interventions towards those in greatest need is an important goal in personalised medicine, so it is heartening to see that there may well be scope to do this in diabetes.
Admittedly, if borne out across all or most diabetes populations, these data suggest that around 40% of patients will need potentially expensive interventions (and this would equate to a lot of people in total) – however, if by providing this it proves possible to prevent the most severe complications and keep those patients healthy for longer (and perhaps much longer, given their relatively low age of disease onset), then it will be a sensible investment. 60% of patients, a modest majority, would arguably need treatment primarily focused on healthy eating and exercise, which could be administered by community-based schemes such as clinics to provide dietary advice and exercise sessions.
However, whilst first steps have already been taken towards community services such as these in the UK, as a bid to create a more sustainable and patient-centred health service that is less focused on expensive hospital care, the reconfiguration needed to current services could take some doing. In policy terms, therefore, this finding in diabetes warrants close attention, and ideally consideration of expected changes in demand for current NHS diabetes services should proceed alongside with clinical verification of the research.
If the science of precision medicine diagnostics can combine with the technology for both prevention and self-management for mild disease plus remote monitoring, control and treatment of severe forms in a sustainable, affordable service, then perhaps diabetes need not be such a massive threat as at present.
In addition to this latest development, science and technology for better diabetes care and control are rapidly advancing. This ranges from novel biosensors for precise regulation of glucose levels and insulin requirement, and pioneering pancreatic stem cell transplants to help control disease more effectively in selected Type 1 diabetes patients, through to the more mundane (but potentially high impact) apps for personalised management of diet and exercise in patients with or at risk of T2D. If the science of precision medicine diagnostics can combine with the technology for both prevention and self-management for mild disease plus remote monitoring, control and treatment of severe forms in a sustainable, affordable service, then perhaps diabetes need not be such a massive threat as at present.