You are reading 1 of 2 free-access articles allowed for 30 days
In 1922 Leonard Thompson, a 14-year-old boy in Toronto was the first patient with diabetes to receive exogenous insulin. The boy who was near death, the expected outcome from diabetes, soon regained strength and appetite. This discovery, the product of work done by Frederick Banting, a surgeon, and his medical student assistant, Charles Best, would later transform type 1 diabetes from a death sentence characterised by a few years of prescribed starvation while on death row, to a chronic disease manageable by self-administered insulin treatment. The natural history of type 1 diabetes was thus transformed, however, this would lead to the recognition of the many long-term microvascular and macrovascular complications that could complicate the disease; nephropathy stands as one of the most devastating of these.
Diabetic kidney disease
Chronic kidney disease (CKD) exists when there is evidence of structural or functional abnormalities of the kidney for longer than three months. Such evidence of kidney damage may be supported by abnormalities of urinalysis, imaging, histology, or by a persistently lowered GFR (<60ml/min/1.732) for more than three months. In most patients with diabetes, CKD should be attributable to diabetes if there is macroalbuminuria present (Albumin creatinine ratio [ACR] > 300mg/g) or if there is microalbuminuria present (ACR 30-300mg/g) in association with diabetic retinopathy, or type 1 diabetes duration of at least 10 years. There is a strong relationship between the presence of diabetic retinopathy and diabetes being the cause of kidney disease in patients with type 1 diabetes, a relationship that is not as robust in the type 2 diabetic population.
A retrospective case series of 611 kidney biopsies in subjects with diabetes (all types) revealed that 36 per cent of cases had evidence of non-diabetic renal disease alone, with the remainder having diabetic nephropathy (37 per cent) or dual diabetic and non-diabetic pathology (27 per cent). However, this may be an over estimate of the prevalence of non-diabetic kidney pathology, reflecting selection bias: patients with diabetes who underwent biopsies tended to have atypical presentations. In an autopsy study of 168 diabetic subjects, changes related to diabetic nephropathy were present in the kidneys of 106, and in 20 of these no clinical manifestations associated with diabetic nephropathy had been evident. These two very different studies highlight both the important place that diabetes plays in causing kidney disease, but that vigilance is required in evaluating kidney disease in a diabetic patient. Diabetic nephropathy has an insidious onset, however, the abrupt onset of the nephrotic syndrome, associated haematuria, acute kidney injury, evidence of systemic complement system activation, paraproteinaemia or other systemic disease, which can cause kidney injury, should raise suspicion that kidney damage may not be caused by diabetes.
The clinical manifestations of diabetic kidney disease represent late changes within the kidneys of diabetic subjects. Pathological changes are present in the kidney at diagnosis of type 1 diabetes, characterised by early hyperfiltration and hypertrophy. These changes, which are present before insulin treatment commences, are at least partly reversible. Later, morphologic changes in the kidney can be present without signs of clinical disease. Even when clinical parameters are normal and urinary albumin excretion is normal in the resting situation, albuminuria can be demonstrated during exercise. Thereafter, persistent microalbuminuria develops at a rate of 1.8 cases per 100 person-years at risk with type 1 diabetes; often referred to as ‘incipient nephropathy’.
Natural history and epidemiology
Progression of microalbuminuria, from incipient nephropathy to overt diabetic nephropathy, is strongly influenced by degree of glycaemic control, with rates of progression as low as 1.3 per 100 person-years with HbA1c < 8 per cent (63.9mmol/mol) to 6.7 per 100 person-years with HbA1c of 10 per cent (85.8mmol/mol). Glomerular filtration rate (GFR) shows an inexorable decline over the ensuing years, the rate of which can be slowed, but not stopped. During these years, increasing proteinuria, often manifesting as the full nephrotic syndrome, hypertension and the myriad complications of chronic kidney disease present themselves as management challenges, as do the other microvascular and macrovascular complications of diabetes. The final destination of end-stage renal disease (ESRD), we know from large Finnish registry data, is reached by 7.8 per cent of all type 1 diabetics after 30 years.
Diabetes represents the single biggest cause of ESRD. It was the reported cause of kidney disease in 44 per cent of patients reaching ESRD in 2016. Mortality rates on dialysis have been falling year on year, however, they continue to be double that of congestive heart failure and more than double that of stroke and cancer. One-third of Finnish type 1 diabetic patients who started dialysis between 2000-2005 had died within five years; the median age at the start of dialysis in this cohort was 44.6 years, with 18 per cent being aged ≥55 years. This contrasts with a 23 per cent risk of death on dialysis over five years in patients who reached ESRD as a result of glomerulonephritis between 2000-2005, despite the glomerulonephritis patients being an older cohort with a median age of 57.2 and 55 per cent of whom were 55 years of age or older.
The rising incidence of type 1 diabetes, from 21.6 per 100,000 children aged 0-14 years in the time period 1978-1980 to 43.9 cases per 100,000 in the years 2005-2007 paves the way for a vastly increased number of adults living with diabetes and who are at risk of diabetic kidney disease. There is, therefore, an urgent need to develop effective therapies that may reverse diabetic kidney disease. Additionally, sensitive, dynamic biomarkers are needed that can accurately monitor therapeutic response and disease susceptibility.
The pathogenesis of diabetic nephropathy is contributed to by hyperglycaemia, glomerular hypertension, advanced glycation end-products (AGEs), angiotensin II, profibrotic growth factors such as connective tissue growth factor (CTGF) and transforming growth factor-β (TGF-β). Expansion of the mesangial cell matrix, glomerular basement membrane thickening, loss of podocytes, apoptosis and dedifferentiation of tubular epithelium, recruitment of macrophages and subsequent fibroblast activation are key processes in its development.
GFR, albuminuria and hypertension are typically considered to correlate strongly with mesangial expansion. However, in many patients the classic pattern of hyperfiltration progressing to persistent albuminuria is not observed and albuminuria is a dynamic parameter, which may regress.
Critical metabolic changes in early diabetic kidney disease include hyperglycaemia altering kidney haemodynamics promoting inflammation and fibrosis. Glomerular hyperfiltration is an early response in diabetic kidney disease. It is considered to reflect increased sodium-glucose transporter 2 [SGLT2] activity in the proximal tubule. This in turn reduces sodium delivery to the macula densa and decreases tubuloglomerular feedback. The proposed result is dilatation of the afferent arteriole, increasing glomerular perfusion, whereas production of Angiotensin II at the efferent arteriole produces vasoconstriction. The net effect is increased intraglomerular pressure and glomerular hyperfiltration. The reno-protective effects of glucose lowering by SGLT-2 inhibition as discussed below may reflect a restoration of tubuloglomerular feedback.
Later stages of kidney disease include, among other complications, the development of anaemia and mineral bone disease; the anaemia often appears earlier in diabetic kidney disease and the mineral bone pathology is more likely to be of an adynamic bone disease morphology in diabetes than in non-diabetic kidney disease.
The problem of diabetic kidney disease is huge, however, its development is not easily predicted based on glycaemic control alone, a traditional measure of the quality of diabetes self-care. Some patients with long-standing poor glycaemic control will not develop overt nephropathy, while others who have better glycaemic control will progress through the stages of nephropathy towards ESRD. Concordance rates for diabetic nephropathy in two sets of families, in which both probands and siblings had diabetes mellitus, provided evidence that heredity helped to determine susceptibility to diabetic nephropathy. Intensive glycaemic control in early type 1 diabetes exerts a long lasting benefit on the risk of developing diabetic kidney disease. Intensive glucose intervention targeting HbA1c <7 per cent reduced the nine-year risk of microalbuminuria by 34 per cent and macroalbuminuria by 56 per cent relative to conventional (standard) glucose control. Twenty-two year follow-up revealed a slower rate of decline in GFR in the intensive treatment group and a 50 per cent reduction in risk of eGFR <60ml/min/1.73m2. Decreases in retinopathy, neuropathy and cardiovascular events were also observed. Intriguingly, intensive glucose control after the onset of complications does not reduce the risk of diabetic kidney disease progression. This evidence of ‘metabolic memory’ suggests epigenetic modification of genes, which may contribute to the pathogenesis of diabetic complications.
Investigations on the molecular mechanisms underlying the initiation and progression of diabetic kidney disease have implicated several key pathways in the pathophysiology of disease. From an experimental perspective these investigations have included analysis of differential gene expression in human renal biopsy tissue and in vitro and in vivo models of diabetic kidney disease and analysis of genome wide associations in diabetic kidney disease. Critical processes emerging from these investigations include inflammation, its defective resolution, oxidant stress and fibrosis. In this context several therapeutic targets have emerged, which may be amenable to intervention. Pre-eminent among the drivers of fibrosis is TGF-β. However, a recent trial targeting TGF-β with a monoclonal antibody in patients with diabetic kidney disease and type 1 or type 2 diabetes was halted due to lack of efficacy. Evidence of oxidant stress as a driver of kidney damage led to the development of bardoxolone methyl as a prototypic antioxidant immune modulator, however, phase 3 trials of bardoxolone methyl in type 2 diabetics with CKD Stage 4 were halted prematurely owing to increased adverse events in the bardoxolone-treated subjects versus controls.
Evidence from family studies displaying familial aggregation of diabetic nephrology in type 1 and type 2 diabetes and differential susceptibility to diabetic nephrology between ethnic groups supports a genetic basis for diabetic kidney disease. Genome wide association studies [GWAS] have been employed in an effort to understand the genetic basis of diabetic kidney disease. Several genetic loci have been implicated and significant effort is being made to understand the potential role of such loci in diabetic kidney disease. A large meta-analysis and GWAS of over 6,000 Europeans with type 1 diabetes, with and without nephropathy, has identified novel candidate genes; some of these genes have potentially important roles in maintaining kidney tubule integrity and regulating cell proliferation and inflammation. A single nucleotide polymorphism (SNP), in the promoter of FERM domain containing 3 (FRMD3), is associated with nephropathy. This gene (FRMD3) encodes a membrane protein that helps determine the interaction between kidney cells and the actin cytoskeleton, a critical process in filtration. FRMD3 is co-expressed with bone morphogenetic proteins (BMPs) in renal tissue. BMPs include a family of proteins which suppress fibrosis. These findings have led to the hypothesis that reduced FRMD3 expression is associated with diabetic nephropathy. The grem-1 gene which encodes a BMP antagonist has also been implicated by GWAS studies. Interestingly, genetic manipulation to reduce gremlin expression in mice renders them less vulnerable to diabetic kidney disease under experimental conditions. Levels of gremlin gene expression in the kidney correlate with disease severity in human diabetic nephropathy. Other GWAS candidate genes include AFF3. Insights into the potential functional relevance of such polymorphisms have been generated in experimental systems where manipulating [decreasing] AFF3 expression is associated with attenuating fibrotic responses to TGF-β1 in cultured renal epithelial cells. Ongoing research collaborations between our group and investigators at Beaumont Hospital, Dublin, led by Prof Peter Conlon, and at the Mater Misericordiae University Hospital, Dublin, led by Dr Denise Sadlier, have established a renal biopsy biobank, which is supporting our investigations of the pathogenesis of CKD in general and diabetic nephropathy in particular.
Integrating data that has emerged from GWAS, studies of transcription profiles, epigenetics, proteomics and metabolomics presents a huge challenge, which will allow an in-depth understanding of the cause and pathogenesis of the disease. To date, such analyses have identified the renin-angiotensin, complement cascade, coagulation pathways and peroxisome proliferator-activated receptor (PPAR) pathways as potential targets for therapeutic intervention.
The ability to delay the onset and slow the progression of diabetic nephropathy has been known about since the 1990s, following the randomisation of a large group of type 1 diabetics to intensive (target pre-meal glucose 3.9-6.7mmol/l, post-prandial glucose 10mmol/l, HbA1c < 6.05 per cent) glycaemic control compared to conventional control. Even though these difficult targets were not attained in full, those patients who achieved HbA1c < 7 per cent saw a 34 per cent reduction in the development of microalbuminuria and a 56 per cent reduction in development of overt nephropathy.
Early studies in small numbers of type 1 diabetics suggested that the mean yearly increase in albumin clearance could fall and the rate of renal function decline could be slowed with good blood pressure control. Captopril demonstrated its ability to protect against deterioration in renal function more than blood pressure control alone in 1993, thus laying the foundation for the angiotensin converting enzyme (ACE) inhibitor as the cornerstone of our treatment of diabetic kidney disease. Targets for blood pressure control remain a topic of debate, as is the case in every area of hypertension care, but current recommendations suggest targeting <130/80mmHg, with the preferential use of an ACE inhibitor. The combination usage of ACE inhibitors and angiotensin-receptor blocking (ARB) agents is no longer recommended, for blood pressure control or reducing proteinuria, owing to the increased risk of hyperkalaemia and acute kidney injury.
The sodium-glucose co-transporter protein in the kidney proximal tubule epithelium has been identified as a useful target for glucose lowering in type 2 diabetes, and drugs have now come to market inhibiting the transporter in this context. These agents have also been studied in type 1 diabetic subjects and a recent meta-analysis suggests that these may be an effective adjunct to insulin, providing improved glycaemic control, reduced body weight and total daily insulin dose, without an increase in total adverse events. The slowed progression of kidney disease with empagliflozin in type 2 diabetes represents a potentially promising avenue for new therapy in type 1 diabetic kidney disease, should it be replicated here. The ability of these drugs to inhibit glucose, and sodium, absorption in the proximal tubule may restore tubuloglomerular feedback and thereby help in slowing hyperfiltration damage.
Incretin mimetics, such as liraglutide and semaglutide, can lower albuminuria in early stage kidney disease in type 2 diabetes, which suggests these drugs may be having direct effects on the kidney. Further elucidation of their mechanisms and study of these drugs in type 1 diabetes may create a place for them in the treatment of diabetic kidney disease in this population.
Amylin is co-secreted with insulin from the pancreatic β-cell; it slows gastric emptying, suppresses the post-prandial rise in glucose and promotes satiety. An analogue, pramlintide, is licensed for use in the US for type 1 and type 2 diabetes. It can reduce HbA1c and body weight when compared to placebo and carries minimal risk for hypoglycaemia. There is no data supporting its use in diabetic kidney disease, per se, but it can be used as an adjunct to insulin to achieve better glycaemic control.
The ability of persistent normoglycaemia, as achieved after pancreas transplantation, to reverse the lesions of diabetic nephropathy was identified in 1998. Eight patients who had received pancreas transplants were successfully followed up for 10 years from the time of engraftment. All had lesions of diabetic nephropathy at the time of pancreas transplant, but did not have uraemia.
Biopsy studies at five years on these subjects did not reveal reversal of the lesions, but by 10 years these lesions were found to have reversed. Glomerular basement and tubular basement membrane thickness, mesangial volume and mesangial matrix had all reduced.
This work tells us that, although it takes a long time, the changes in the diabetic kidney are not permanent, at least in their early stages and is encouraging in supporting the value of basic research in elucidating the pathogenic mechanisms of disease and potential therapeutic modification of implicated pathways. Indeed, using analysis of transcriptomic networks, upregulation of the JAK-STAT pathway has been identified in early and progressive diabetic nephropathy in type 2 diabetes and inhibition with baricitinib has been demonstrated to reduce albuminuria.
Kidney disease in type 1 diabetes remains a challenging area of practice. Whereas preventive care for adults with diabetes has improved substantially in recent decades an analysis of diabetes-related complications in the US 1990-2010 has revealed that ESRD shows the least improvement. Rates of MI and stroke declined by over 50 per cent, whereas the decline in ESRD was approximately 25 per cent. As we continue to understand more of the genetic and biochemical pathways of the disease, we must learn to translate this knowledge into clinically applicable diagnostic and therapeutic tools.
Unfortunately, there is kidney damage taking place at the tissue level in the early years of diabetes, which we cannot detect based on current clinical tools. The kidney biopsy studies performed in recipients of pancreas transplants should provide hope to us that these changes are reversible and encourage endeavours to identify disease biomarkers and novel therapeutic targets.
As we approach the centenary of Banting and Best’s pioneering work, the landscape to diabetes has changed, but the struggle of people living with type 1 diabetes and those that care for them remains.
References on request