Close X

Personalising medicine in type 2 diabetes

1st December 2021

There is a wealth of new findings on how to optimise treatment for kidney and heart complications in type 2 diabetes and many sessions at the 57th Annual EASD Meeting were devoted to personalised medicine. One of the highlights was the Camillo Golgi Prize Lecture, with its presentation of the breadth of recent research – from biomarkers to combination therapy – and how this can be applied in diabetes care right now. Dr Susan Aldridge reports.

Hiddo Lambers Heerspink, Professor of Clinical Trials and Personalised Medicine at the University Medical Centre, Groningen, gave the Novo Nordisk/EASD Camillo Golgi Prize Lecture entitled ‘Personalising the treatment for type 2 diabetes: The mean is meaningless’.  

Heart failure is often one of the first complications seen in type 2 diabetes and many will also go on to develop chronic kidney disease. “Heart failure and chronic kidney disease are related too – having one causes or worsens the other. So, you get a vicious circle of renal, cardiac and metabolic risk. We have to interrupt this vicious circle in order to improve outcomes for our patients,” he said. 

The focus of his talk, accordingly, was how to personalise treatment of kidney and vascular complications in type 2 diabetes. In 2019, the SGLT-2 inhibitors and the endothelin receptor antagonist (ERA) atrasentan were found to be kidney protective. And at the end of last year, another trial showed that the mineralocorticoid receptor antagonist (MRA), finerenone, also protects the kidney and it’s now also known to protect against cardiovascular complications. “So, after many years of stagnation, we finally have new drugs for the treatment of diabetic kidney disease,” Hiddo said. Data from the DAPA-CKD and CREDENCE trials showed that dapagliflozin and canagliflozin protect the kidney in patients with and without diabetes. “Based on these results, some people now believe that we have cured diabetic kidney disease. I don’t believe that is true because there were still a number of kidney-related events in a relatively short follow-up of around two years in both the dapagliflozin and canagliflozin arms. So, although we have improved the outcome of kidney failure, we certainly have not cured it.” Why do these patients still develop kidney disease, despite the fact they have received an SGLT-2 inhibitor? Some can’t tolerate these drugs, but it’s also clear that not every patient responds beneficially. The SGLT-2 inhibitors also reduce risk of heart failure and cardiovascular death in patients with kidney disease, according to the CREDENCE trial, and across the entire eGFR range. But some of those on canagliflozin still do develop heart failure, so SGLT-2 inhibitors don’t cure this either.   

Hiddo then pointed to atrasentan and the SONAR trial, which showed that it does reduce  kidney failure – in carefully selected diabetes patients with kidney disease – but there are still patients who do not receive any benefit. Atrasentan can also cause fluid retention which, in itself, can lead to heart failure. Furthermore, the FIDELO-DKD showed that finerenone does protect the kidney, but a residual risk of kidney damage remains. 

“The remaining high risk of both kidney and heart failure, despite being on these recommended novel treatments is explained by suboptimal response. But it is difficult to see how the patient is responding because the progress of kidney and heart complications is so slow and it takes too long to monitor drug response.” One way his team is countering this is to use biomarkers to obtain an earlier assessment as to whether the patient is responding to the drug. “We already do this, in antihypertensive therapy, by measuring blood pressure and, if they’re not responding, up-titrate or change to another drug.” In kidney disease, measure albuminuria with the urine albumin:creatinine ratio (UACR). Hiddo’s team has done this with dapagliflozin and found that UACR did not decrease in 20% of patients and there was a large variation in individual response to the drug. But what are the factors that underlie these individual variations – and can we use them to individualise treatment?


Lack of response to a drug, as Hiddo describes above, could be due to its pharmacokinetics in a particular individual. This refers to aspects such as how much of the drug actually reaches its target and how long it remains in the body. It’s usual to assess this by measuring the concentration of the drug in blood or urine. But Hiddo’s team have gone beyond this by using telmisartan (an angiotensin-II receptor blocker or ARB) tagged with a radioactive label to track the drug’s journey through the body with positron emission tomography (PET). These experiments revealed that in one patient there was hardly any drug in the kidney, where you’d expect it to bind to the angiotensin receptor, or even in plasma. It could be that this patient is metabolising the drug and passing it out of the body very rapidly, before it has a chance to do anything. Another patient was the opposite. The PET study showed a “very nice” disposition throughout the kidney and in the blood, so this patient had a much lower first-pass metabolism. “If we can link these signals to clinical response, our hypothesis is that the first person will not respond to telmisartan and the second one will,” said Hiddo.  They have carried out a similar study with radiolabelled canagliflozin and were able to see active binding of the drug at the tissue level in the kidney. This approach will now be applied in future studies and will include drugs acting on the heart as well as the kidney. 

Severity of disease

One message Hiddo had during discussion at the end of his talk was that it is so important to catch kidney disease early. He spoke of patients’ regrets, once they faced dialysis, that they – and presumably their doctors – had not acted earlier on. But there may be some new hope for those whose kidney disease has progressed, for it seems they may have more to gain from treatment.

A study looking at variation in response to atrasentan found that, on average, the drug did slow the decline in eGFR compared with placebo. And it was those patients who went into the study with worse kidney function, shown by more rapid decline in eGFR (more than 5 ml per year) before the study, who benefited the most. The drug could not reverse the decline in kidney function, but it did slow it down – and the worse it was, the more the progression was slowed. “So, this shows that severity of disease may determine the response to atrasentan,” Hiddo remarked.  

Genetic polymorphisms

Some of the drugs used in treating kidney and heart complications will bind to molecules in the liver called organic anion transporters (OATs). Hiddo’s team has carried out genetic analysis on OATs and found a number of polymorphisms that affect this binding. These can distinguish slow metabolisers and fast metabolisers of drugs like atrasentan. “The slow metabolisers have a higher exposure to atrasentan, which should help with kidney failure but, on other hand, it causes fluid retention.” In fact, analysis of clinical data shows that these slow metabolisers do not even benefit in terms of kidney function, despite their higher exposure to the drug, which is surprising. “Patients with this polymorphism should not, therefore, receive atrasentan, because it is of no benefit to the kidney and increases the risk of heart failure,” Hiddo concluded. Clearly, in the future, we might see people with type 2 diabetes being screened for a whole range of drug-metabolising polymorphisms to help personalise their treatment. The above study alone, if applied in clinical practice, could steer many patients towards a treatment that is beneficial rather than one that is potentially harmful. 

Switching treatment

There are many new treatments for kidney and heart complications, so one way forward, if a patient doesn’t respond to one treatment, is to try another. Hiddo described the ROTATE study in which this option was explored.  In an open label trial, participants were placed on four treatments, with four different modes of action, for four weeks in turn and the impact on the kidney measured by UACR. The four drugs were telmisartan, linagliptin, empagliflozin and the anti-inflammatory drug baricitinib. They then noted the best drug of the four for each participant and put them back on it for another four weeks for confirmation. 

On average, telmisartan was the best of the four, producing around a 30% reduction in UACR, followed by linagliptin and empagliflozin. Baricitinib had little effect. However, analysis showed a large amount of individual variation in response within these averages, even for telmisartan. “So, for some patients telmisartan is best, for others linagliptin or empagliflozin – or even baricitinib,”said Hiddo.   


Discovering biomarkers for drug response is a hot research area in personalised medicine. Hiddo’s team has been involved in the European Innovative Medicines Initiative BEAt-CKD study. This is a multiomics biomarker discovery approach for response to ERAs and SGLT-2 inhibitors. It has involved transcriptomics, proteomics and metabolomic studies within clinical trials of these two classes of drug, as well as cell and animal transcriptomic studies. “We are integrating all this information and relating it back to the gene level,” said Hiddo. “We have also constructed a pathway to see which bits of it are targeted by these two drug classes and also a pathway of diabetic kidney disease. We’ve overlaid the two to find the parts of the kidney disease pathway that are targeted by the drug of interest. We use this to develop and discover novel biomarkers.” They have already discovered four biomarkers that predict the effect of ERAs and are currently validating these. 

Personalising treatment – do it now

While some of the above findings may take time to be applied in clinical practice, there is plenty that can already be done. Hiddo said “Yes, we can start individualising treatment with currently available drugs,” and he gave a couple of practical examples.

First, the EMPA-Reg, CANVAS, CREDENCE and DAPA-HF trials all showed that the SGLT-2 inhibitors reduce the risk of hyperkalaemia. On the other hand, MRAs increase the risk of hyperkalaemia. So how about combining the two to give kidney protection, while also reducing the risk of hyperkalemia? That’s exactly what is being done in the ROTATE-3 trial, where participants are receiving an SGLT-2 inhibitor, then an MRA and then a combination in rotation. Results are due very soon. “This could provide a compelling argument to improve efficacy and safety of treatment,” Hiddo said.  

The second example involves a combination of an SGLT-2 inhibitor with an ERA, to try to offset the fluid retention risk of the latter. Hiddo’s group used weight gain as a measure of fluid retention in a trial of atrasentan alone compared with the combination of atrasentan with an SGLT-2 inhibitor. The results were very clear. Atrasentan alone increased body weight, indicating fluid retention, and UACR was down only 27.6% on average – whereas the combination reduced weight, so no fluid retention, and brought down UACR by 55%. 

One combination that is often discussed in this context is GLP-1 receptor agonists and SGLT-2 inhibitors. And there has been a trial that shows that the two together do, indeed, reduce albuminuria more than either drug alone. “This suggest that in future we should combine these two to enhance kidney protection,” Hiddo said. 

And if the patient in front of you has a simple phenotype, you really don’t need to wait for biomarker validation or other results. A simple algorithm will suffice to personalise treatment. “Suppose the patient has obesity, a high HbA1cand diabetic kidney disease Then a combination of a GLP-1 receptor agonist and an SGLT-2 inhibitor could be the best combination for them, because it reduces weight as well as protecting the kidney.” But if your patient has heart failure – so common in type 2 diabetes – as well as high albuminuria, but a low BMI, then Hiddo suggests an SGLT-2 inhibitor plus and ERA. “So you already have a relatively simple algorithm for finding the best treatment for each individual.” 

In conclusion then, there is a large variation in people’s response to different drugs for kidney and heart complications in type 2 diabetes. Many new drugs for diabetic kidney disease are now available and the challenge for healthcare professionals is to optimise the choice and combination of these drugs for each patient. Hiddo’s take-home message is: “It is now time to move from one-size-fits-all to one-fit-for everyone approach.”

Webcasts from the EASD’s 2021 virtual meeting are now available free to all on the EASD website.

View Professor Heerspink’s Camillo Golgi Prize Lecture

For more on the topics raised in this lecture, why not enrol on the following EASD e-Learning courses:

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.