Rudy Bilous, MD, FRCP
Richard Donnelly, MD, PHD, FRCP, FRACP
Diabetic nephropathy is a clinical diagnosis based upon the detection of proteinuria in a patient with diabetes in the absence of another obvious cause such as infection. Many of these patients will also be hypertensive, have retinopathy and, in advanced stages, renal impairment.
The original definition was based upon relatively crude tests that detected protein concentrations of around 500 mg/L. With the development of accurate dipstick urine tests, this has fallen to 300 mg/L (largely albumin). It is now possible to detect much lower concentrations using laboratory and dipstick tests, and the concept of microalbuminuria or incipient nephropathy was developed in the 1980s. Persistent albuminuria in the microalbuminuric range is now widely accepted as a positive diagnosis for diabetic nephropathy (Table 16.1)….
National Institute of Health and Clinical Excellence (NICE) guidance suggests that positive tests for microalbuminuria should be confirmed within 3 – 4 months before making a diagnosis of nephropathy.
Classically, patients were considered to progress relentlessly from normoalbuminuria through microalbuminuria to clinical nephropathy. It is now recognized that patients with microalbuminuria may spontaneously revert to normoalbuminuria, up to 50% in some series of type 1 patients. Moreover, albuminuria can increase during periods of poor glycemic control (and then decrease with glycemic correction), and reduce with antihypertensive therapy (notably drugs which block the renin-angiotensin system (RAS)).
Microalbuminuria has also been found in patients with coronary artery disease and essential hypertension and who have either normal or only mildly impaired glucose tolerance. In the UKPDS, the majority of those with microalbuminuria did not progress to need renal replacement therapy and the median period spent in the microalbuminuric range was 11 years. Thus it is not inevitable that all microalbuminuric patients will progress to renal failure and it also means that the contribution from cardiovascular or renal disease or both will vary from patient to patient.
Glomerular filtration rate (GFR) is often abnormally high ( > 135 mL/min/1.73 m2) in patients with newly diagnosed type 1 and, to a lesser extent, type 2 diabetes. This is called hyperfiltration and its relationship to subsequent nephropathy risk is controversial. Improving glycemic control in hyperfiltering patients reduces GFR toward normal.
With increasing albuminuria, GFR declines and the rate of loss is greater in those with higher systemic blood pressure. Historically, patients with clinical nephropathy had a rate of loss of GFR of 10 – 12 mL/min/yr. More recent studies suggest rates of decline of < 4 mL/min/yr in patients with well-controlled blood pressure. Ultimately, GFR declines relentlessly in patients with clinical nephropathy towards end-stage renal disease (ESRD) < 15 mL/min/1.73 m2, albeit at very different rates in individual patients.
The other clinical concomitant of nephropathy is blood pressure. In type 1 diabetes this is normal until microalbuminuria starts to develop although some studies have shown that, on average, patients who go on to develop nephropathy have higher blood pressures when normoalbuminuric (although still well within the normal range). In type 2 diabetes, many patients have hypertension when diagnosed and these individuals are at higher risk of nephropathy. Thus, hypertension is a feature of developing nephropathy in type 1, whereas it may be an initiating factor in type 2 diabetes; it is an exacerbating factor in both.
Patients with nephropathy are at much greater risk of cardiovascular disease. In the UKPDS, annual mortality was over twice and three times higher for those with microalbuminuria or clinical nephropathy respectively compared to their normoalbuminuric comparators. For those with a plasma creatinine > 175 µmol/L and/or on renal replacement therapy, the mortality was 14 times greater. In type 1 diabetes, excess mortality in patients with clinical nephropathy is 10 – 20 times higher than for their age-matched non-proteinuric comparators (Figure 16.1).
This excess mortality explains why many patients with renal impairment do not survive to need dialysis, although with better treatment many more are now doing so.
Stages of nephropathy
An international classification of chronic kidney disease (CKD) has now been widely adopted, but unfortunately it does not map precisely to staging of diabetic nephropathy based upon albuminuria (Table 16.2). The new staging is based upon GFR as estimated (eGFR) from a plasma creatinine concentration using a formula derived from the Modification of Diet in Renal Disease (MDRD) Study.
eGFR (mL/min/1.73 m2) = 175x[serum creatinine (µmol/L) x 0.0113]-1.154 x age (years)-0.203
If female multiply by 0.742.
If African American multiply by 1.21.
This formula uses a serum creatinine assay aligned to an international reference method. It has not been specifically validated in people with diabetes, in other ethnic subgroups such as South Asians or Chinese/Japanese, in children or in pregnancy. Alternative formulae for estimating creatinine clearance (not GFR) exist, such as Cockroft – Gault, and for GFR using serum cystatin C. Both have their advocates but although cystatin C has advantages, it is expensive and not widely available in the UK.
The MDRD eGFR is much less accurate above 60 mL/ min/1.73 m2 and tends to underestimate true GFR below this value. Values below 60 mL/min/1.73 m2 are, however, associated with an increasing cardiovascular mortality in US and UK populations, and more so in those with diabetes. It is therefore now considered as an extra cardiovascular disease risk factor.
Despite its drawbacks as an accurate measure of kidney function, eGFR is useful in alerting the clinician (and patient) to renal impairment that would not be apparent from the serum creatinine concentration alone. This helps to guide therapy (therapeutic dose adjustment or drug avoidance) and also a rapidly declining eGFR should prompt referral for specialist assessment (see later).