Predicting diabetic kidney disease
If you knew you were going to be on dialysis in four years time, would you do something about it? Researchers in Australia are using precision medicine to predict the onset of chronic kidney disease years into the future, giving patients the chance to change the outcome...
The genes we are born with are the genes that we die with. Specific mutations in our genomes are responsible for many diseases such as cancer, but that is far from the whole story. Precision medicine is now moving beyond the genome to explore more complex classes of molecules - proteins and metabolites - and how they contribute to the way organisms function and their risks of developing diseases years into the future.
Picture the caterpillar and the butterfly. Sames genes, vastly different outcome. Cells use the genome as a set of instructions, a recipe book if you like, and it is the products of those genes - the proteins they encode - that are the executive and operational molecules of life, and that control how an organism appears and functions.
It is the same in the human body. Different pathways are switched on and off continuously and affect how our bodies function. If the process goes wrong, we get sick. In all cases of disease, there has to be a starting point, some kind of initial molecular level signal to start the disease. Consequently if we can detect this early change, this new molecular "fingerprint" that indicates trouble is coming, then a physician may be able to intervene very early in the process of disease, possibly stopping it before clinical symptoms appear.
At Proteomics International, we use mass spectrometers, sensitive instruments designed to look at these molecular fingerprints or "biomarkers", to analyse protein patterns found in biological systems. Proteomics is the industrial scale study of the structure and function of proteins. In 2009 we teamed up with the University of Western Australia (UWA) Medical School to investigate diabetes and see if we could apply proteomics to find new biomarkers for the disease and some of its main health consequences, such as kidney failure.
The rise of diabetes and chronic kidney disease
Diabetes has become a pandemic. Type 2 diabetes, generally caused by poor diet and lack of exercise, was hardly known at the beginning of the 1900's. Now there are over 425 million adults worldwide with type 2 diabetes, and this number is forecast to exceed 600 million within the next 20 years.
Diabetes is a disease where the body cannot control the level of glucose in the blood. It will not kill you, at least initially, but its complications can be life threatening, with cardiovascular disease and chronic kidney disease being the most serious.
The mechanisms causing these complications are still to be fully understood, but, simply put, diabetes damages the body's blood vessels. Today, approximately one in three adults with diabetes already have chronic kidney disease (CKD), and over 50% of those with diabetes will eventually develop kidney disease.
Kidney disease is a silent killer because the kidneys are highly efficient organs and patients rarely know their kidney function is dropping. You can lose 50% of your kidney function and not know. That's why a healthy person can donate a kidney and remain well. But if kidney function drops below 15-20% of normal, then renal replacement therapy is required. That means dialysis, or a kidney transplant. Many patients with diabetes can experience a slow decline in kidney function, but a sub-group, about 10% of diabetics, will have a rapid decline in kidney function and this is the most dangerous. CKD already causes nearly 90,000 deaths per year in the USA, with over half directly linked to diabetes.
The existing tests for kidney function have been around for over 50 years. The "albumin creatinine ratio" (ACR) measures protein in the urine, and the estimated glomerular filtration rate (eGFR) measures a single biomarker (creatinine) found in the blood. Both tests are not very sensitive or accurate: a walk around the block could change the test result. More importantly, the ACR and eGFR tests can only diagnose the disease once it's already present. They cannot predict future kidney disease.
When taken together, the lack of symptoms, poor sensitivity of existing tests, and rapidly rising numbers of patients with diabetes, means that the world faces a potential tsunami of CKD in the foreseeable future.
Investigating the protein fingerprint of disease
The clinical studies between Proteomics International and the University of Western Australia focused on a community based group of patients with type 2 diabetes living in the Perth metropolitan area: the Fremantle Diabetes Study. For the exploratory work, small groups of patients were analysed and compared - ten individuals from each of three groups: diabetes with no CKD, diabetes with mild CKD, and diabetes with severe CKD.
We took blood plasma samples - less than a drop of blood is needed - and then used proteomics techniques to analyse the proteins present in the plasma, looking for changes in the protein fingerprints between the three groups. The aim was to find novel biomarkers that correlated with kidney disease.
The initial screen identified approximately 30 potential protein biomarkers. These we analysed in larger patient groups to confirm their statistical significance. These patients were also followed over time to see how the concentrations of the markers changed as their disease progressed. Ultimately, more than 1500 patients were followed over a four year period.
A new predictive test algorithm
After making allowances for statistical confounders, such as the patient's existing drug treatment, or the presence of other complications like heart disease, a panel of three protein biomarkers were identified as linked to diabetic kidney disease. When combined with simple clinical variables (such as the patient's age and cholesterol level), the result is a predictive algorithm that can diagnose existing kidney disease with greater accuracy than the existing tests, and, most exciting, can predict the future risk of CKD.
The new test is called PromarkerD and it was able to correctly predict 86% of patients who went on to have a rapid decline in kidney function within four years1. An alternative way of viewing the test result is that an adverse PromarkerD score means the patient has a much greater risk (between a 50% increase and a three times increase compared to normal) of developing CKD.
Now a patient and their doctor have much more precise information about the potential outcomes of their diabetes and can personalise the current medical treatment. If the test shows a patient is at moderate risk, the simplest treatment is tighter glucose monitoring and control of blood glucose levels, coupled with better diet and more exercise. Both patient and doctor knew that before, but, with the risk of dialysis or kidney transplant, the motivation for making a change is greater. Life expectancy on dialysis averages 5-10 years.
Crucially, if at high risk of CKD, there are now new therapeutic choices. Seventeen drugs are in mid-late stage clinical trials for this disease, which is notoriously difficult to treat. One of these drugs has recently been shown to be effective - it's a drug from the gliflozin (SGLT-2 inhibitor) class and is the first medicine in nearly 20 years to demonstrate significant reduction in risk of renal failure, dialysis or kidney transplantation. Others are likely to follow. This means medication can now be precisely tailored to the patient's needs. It is conceivable that early intervention will considerably improve a patient's clinical outcomes, and may even prevent the onset of chronic kidney disease.
How is the test used?
Highly specialist laboratories can run the PromarkerD test using a mass spectrometer, and then submit the data to the web portal to generate a test report. To expand this user-base, Proteomics International has also been working to adapt the test onto a platform technology commonly found in clinical pathology laboratories. That process has now been completed. Known as an immunoassay, it is simpler to use than the mass spectrometry technology used to develop the test. The immunoassay reagents for PromarkerD can be configured into a Laboratory Developed Test (LDT) for use by certified laboratories. A kit version of the test is also in final preparation for larger scale adoption.
Having just presented PromarkerD at the American Diabetes Association 79th Scientific Sessions in San Francisco, and the BIO 2019 conference in Philadelphia, the test is due to launch in major markets around the world in the coming weeks, with Europe and Central America the starting points.
What about other diseases?
The approach can be used for any disease. Proteomics International is now applying the same technology to find new biomarkers, and potentially develop new diagnostic tests, for endometriosis and lung diseases such as asthma and chronic obstructive pulmonary disease (COPD). Precision medicine is still in its infancy, but there are exciting times ahead in the treatment of chronic diseases.
About Proteomics International
Proteomics International (Perth, Western Australia) was established in 2001 and is a wholly owned subsidiary and trading name of Proteomics International Laboratories Ltd (ASX: PIQ), a medical technology company at the forefront of predictive diagnostics and bio-analytical services. The company specialises in the area of proteomics – the industrial scale study of the structure and function of proteins. It received the world’s first ISO 17025 laboratory accreditation for proteomics services, and operates from state-of-the-art facilities located on Perth's QEII Medical Campus. Proteomics International's business model is centred on the commercialisation of the company's pioneering test for diabetic kidney disease, PromarkerD. The company offsets the cash burn from R&D and product development through provision of specialist analytical services, whilst using its proprietary Promarker technology platform to create a pipeline of novel diagnostic tests.