Written by:
Global Clinical Head, Clinical Metabolism, Late-stage Development, Cardiovascular, Renal and Metabolism (CVRM), BioPharmaceuticals R&D, AstraZeneca
Executive Director, Early Clinical Development, Research and Early Development, Cardiovascular, Renal and Metabolism (CVRM), BioPharmaceuticals R&D, AstraZeneca
AI & Analytics, Director, Data Science & Artificial Intelligence, R&D, AstraZeneca
Elevated levels of protein in your urine – known as ‘albuminuria’ or ‘proteinuria’ – are associated with an increased risk of kidney function loss over time, leading to chronic kidney disease (CKD). Clinical research has demonstrated that the level of proteinuria reduction positively correlates with long-term renal protection; so the larger the initial reduction in proteinuria in the first few months of treatment, the lower the risk of end stage renal disease during treatment.1 However a number of patients carry residual proteinuria although on current standard of care, remaining at risk for renal disease progression.
The leakage of even small amounts of protein in the urine is one of the earliest signs of kidney damage and is a predictor for progressive organ function decline. There is also emerging evidence that proteinuria has a toxic effect on renal tissues, further driving progressive loss of kidney function.1
Proteinuria is measured via a urine ACR (albumin-to-creatinine ratio) test; a normal amount of urine albumin is less than 30 mg/g and anything above 30 mg/g may indicate you have kidney disease.2
Chronic kidney disease (CKD) affects one in 10 people worldwide,3 and people living with CKD progressively lose kidney function, which can lead to kidney failure. In reality, CKD is an umbrella term for many diseases and encompasses various primary disorders and stages of progression. Therefore, patient populations are highly heterogeneous, and treatment can be complex.
The kidneys play a critical role in human physiology by removing waste products, balancing the body’s fluids, salts and minerals, and ensuring optimal conditions for many of the body’s organ systems. The kidneys accomplish this through complex interplay with other organs like the heart. Not surprisingly, up to one in five patients diagnosed with kidney disease develop cardiovascular complications, and both share common risk factors like diabetes, obesity and hypertension.
The existing unmet need despite today’s standard of care
Current treatments for CKD include those that interact in the renin-angiotensin-aldosterone system (RAAS) to help regulate blood volume, electrolyte balance and systemic vascular resistance. Clinical studies have shown that with these treatments there is an initial reduction in proteinuria which positively correlates with the reduction in risk for renal disease progression. However these treatments can be underused because they risk increasing potassium in the blood, known as hyperkalaemia.4 In addition, research presented at European Renal Association (ERA) Annual Congress shows that people with more advanced kidney disease, as indicated by their proteinuria levels, are more susceptible to hyperkalaemia.5 As kidney function declines, the kidneys are less able to remove potassium from blood.6 Hyperkalaemia is a common complication in CKD, affecting up to 40-50% of patients.7 It is associated with an increased risk of cardiovascular events and death.8-10
These challenges may limit current treatment options for patients with poor kidney function and significant proteinuria. There is a need reduce for future treatments to reduce proteinuria whilst managing the risks of hyperkalaemia to slow disease progression.
Delivering a step change in identifying patients with proteinuria
Currently, despite compelling data, only a minority of patients are screened for high proteinuria. Proteinuria is assessed by measuring the albumin creatinine ratio (ACR) in a patient’s urine and, results may vary even within the same individual, making it harder to identify meaningful biologic changes and increasing complexity of clinical trials.
In research presented at ERA Annual Congress 2023 we presented a new diagnostic approach to identify patients with proteinuria using machine learning to predict ACR levels from Electronic Health Records.11 The model may in the future support identifying undiagnosed proteinuria and be applied in pre-screening for clinical trials. We intend to validate the model further in an upcoming CKD trial. Further development of the technique may lead to a future where we can predict progression of kidney disease without needing to take a urine test at all.
Harnessing the power of machine learning to improve screening of proteinuria in chronic kidney disease
