Chapter 54. Diabetic Nephropathy

• Afflicts 35–40% of type 1 and type 2 diabetics with an increasing incidence associated with type 2 diabetes mellitus.
• An initial period of glomerular hyperfiltration.
• Progressively increasing proteinuria.
• A gradual decline in the glomerular filtration rate (GFR).
• Results in renal failure.

Diabetic nephropathy is a serious public health concern because it has become the leading cause of end-stage renal disease (ESRD) in most developed countries and is associated with increased cardiovascular mortality. Tracking both the incidence and prevalence of ESRD attributed to diabetes underscores its annual growth rate over the past decade in excess of 9%. According to the 2004 report of the U.S. Renal Data System (USRDS), in 2002, of 419,263 patients in the United States receiving either dialytic therapy or a kidney transplant, 149,614 had diabetes, a prevalence rate of 35.6%. The incidence rate was 44.5% in 2002, with 42,665 of 149,614 new (incident) cases of ESRD attributed to diabetes (Figure 54–1). This increase is mainly attributed to an increase in the occurrence of diabetes, especially type 2 diabetes; the extended life span of diabetic patients due to improved management of comorbid conditions; and the acceptance of patients for replacement therapy who in the past were excluded.

Diabetic nephropathy is characterized by an initial period of glomerular hyperfiltration associated with progressively increasing proteinuria, followed by a gradual decline in the GFR, eventually resulting in renal failure. Diabetic nephropathy afflicts 35–40% of type 1 and type 2 diabetic patients. While the natural history of diabetic nephropathy is well studied in type 1 diabetic patients, recent studies have shown a similar course of diabetic nephropathy in type 2 diabetic patients as well. In the past 2 decades, much has been learned about the possible pathogenesis of diabetic nephropathy leading to the development and use of specific therapies that have been effective in slowing progression to renal failure.

A cumulative incidence of diabetic nephropathy has been documented after 20–25 years of diabetes in both type 1 and type 2 individuals. Recent studies demonstrated that present treatment strategies substantially reduce the progression and incidence of diabetic nephropathy in type 1 diabetes. For example, a study from Sweden showed a substantial decline in albuminuria after 25 years of diabetes from 30% in patients in whom diabetes developed from 1961 to 1965 to 8.5% in those with onset from 1966 to 1970 and 13% in those diagnosed from 1971 to 1975. Similarly, the Steno Diabetes Center reported that in the same cohorts, the cumulative incidence of diabetic nephropathy after 20 years fell from 31.3% to 13.7%. Improved glycemic control, better control of blood pressure, and reduced prevalence of smoking were associated with the lower incidence of nephropathy.

In contrast to the decreasing incidence of diabetic nephropathy in type 1 diabetes, the incidence of diabetic nephropathy associated with type 2 diabetes mellitus has been increasing over the past 50 years, so that in the United States about 44% of all patients beginning ESRD replacement therapy have diabetes compared to 25–50% in Europe and about 25% in Australia. The incidence of diabetic nephropathy is about 1–2% per year in patients with type 1 diabetes.

Among young nonwhite patients with type 2 diabetes, such as Pima Indians, Japanese, and African-Americans, the incidence of nephropathy is similar to that of type 1 diabetes. However, the incidence of diabetic nephropathy is much lower in elderly white type 2 diabetic patients than in nonwhite patients.

Several large population-based studies found substantial racial and ethnic differences in the incidence rate of ESRD attributed to type 2 diabetes. The highest incidence of ESRD secondary to diabetes has been reported in Native Americans followed by Hispanics and African-Americans. In Pima Indians the cumulative incidence of ESRD after the onset of clinically detectable proteinuria is 40% at 10 years and 61% at 15 years. By contrast, ESRD developed in only 11% of white patients after 10 years of proteinuria and in 17% after 15 years. These ethnic and racial differences in the incidence of diabetic nephropathy reflect a complex and still poorly understood interplay between genetic and environmental factors.

Among patients starting renal replacement therapy, the incidence of diabetic nephropathy doubled from the years 1991 to 2001. Fortunately, the rate of increase is slowing down, probably because of adoption in clinical practice of several measures that contribute to the early diagnosis and delay of diabetic nephropathy, thereby slowing the progression to clinically noted renal disease. However, implementation of effective preemptive therapy in diabetic patients falls far below desirable goals.

The key to the development of diabetic nephropathy is hyperglycemia, which is postulated to mediate its effects in several different ways. First, glucose in high concentrations may be directly toxic to cells, altering cell growth and gene and protein expression, thus increasing extracellular matrix (ECM) and growth factor production. Second, hyperglycemia may induce its adverse effect indirectly through the formation of metabolic end products such as oxidative and glycation products. In vitro, transforming growth factor (TGF)-β modulates ECM production in glomerular mesangial and epithelial cells. In addition, TGF-β inhibits the synthesis of collagenases and stimulates production of metalloproteinase inhibitors, an effect that could lead to reduced degradation of ECM and hence ECM accumulation. High glucose concentrations also increase TGF-β mRNA expression in renal cells, stimulating TGF-β mRNA expression and bioactivity, cellular hypertrophy, and collagen transcription in proximal tubules, providing in vitro evidence for a role of TGF-β in the development of diabetic nephropathy.

Three of several pathways by which hyperglycemia may induce diabetic nephropathy are actively being explored.

Advanced Glycation End-Products

In health, reducing sugars such as glucose react nonenzymatically and reversibly with free amino groups in proteins to form small amounts of stable Amadori products (eg, hemoglobin A1c) through Schiff base adducts. In normal aging, spontaneous further irreversible modification of proteins by glucose results in the formation of advanced glycation end-products (AGEs), a heterogeneous family of biologically and chemically reactive compounds with cross-linking properties. This process of protein modification is amplified by the high ambient glucose concentration present in diabetes. In cultured glomerular endothelial and mesangial cells in vitro, glycated albumin and AGE-rich proteins have been shown to enhance the expression of type IV collagen and TGF-β1 and increase protein kinase C (PKC) activity. When performed under physiological glucose conditions, in vitro studies provide evidence that early glycation products may contribute to the pathogenesis of diabetic glomerulopathy independently of glucose.

Aminoguanidine, a hydrazine-like compound, reacts with early glycation products inhibiting further AGE formation. Aminoguanidine retards the development of nephropathy and other complications of diabetes in long-term experimental rats.

Other AGE inhibitors and AGE cross-link breakers are able to ameliorate diabetic nephropathy in experimental animals.

Aldose Reductase Pathway

The enzyme aldose reductase (AR) converts a variety of toxic aldehyde derivatives from lipid peroxidation to inactive alcohol.

AR is the rate-limiting enzyme in the polyol pathway, and facilitates the reduction of glucose to sorbitol. Sorbitol dehydrogenase converts sorbitol to fructose using nicotinamide adenine dinucleotide (NAD). In hyperglycemia, when the pathway of glucose to glucose-6-phosphate is saturated, excess glucose enters the polyol pathway and aldose reductase is activated, resulted in an accumulation of sorbitol. In in vitro experiments in mesangial cells, increased expression of glucose transporter 1 leads to increased AR expression and activity along with sorbitol accumulation and increased PKC-α protein levels, promoting stimulation of matrix protein synthesis. Numerous experimental and clinical studies with different AR inhibitors (ARI) implicate the diabetes-induced increased flux of glucose through the polyol pathway in the development of diabetic retinopathy and neuropathy; however, only a few studies have investigated the influence of ARI in diabetic nephropathy.

Diacylglycerol-Protein Kinase C Activation

PKC is a family of serine-threonine kinases, consisting of at least 10 structurally related isoforms that regulate a variety of cell functions including proliferation, gene expression, cell differentiation, cell migration, and apoptosis. In vitro studies have shown that PKC is activated in vascular tissues and glomerular mesangial cells exposed to high glucose concentration. Activated PKC increases production of cytokines and ECM and vasoconstrictor endothelin-1. These changes contribute to basement membrane thickening, vascular occlusion, and increased permeability. Several PKC inhibitors used in experimental diabetes yielded promising results.

Ruboxistaurin (LY333531) mesylate, a bisindolymaleimide, shows a high degree of specificity within the protein kinase gene family for inhibiting PKC-β isoforms. In experimental rodent models of diabetes, ruboxistaurin normalized glomerular hyperfiltration, decreased urinary albumin excretion, and reduced glomerular TGF-β1 and ECM protein production, despite continued hypertension and hyperglycemia.

Symptoms and Signs

In health, daily urinary excretion of albumin is less than 25 mg. Diabetic nephropathy follows a characteristic course starting with microalbuminuria defined as albuminuria ranging from 30–299 mg/24 hours or 20–199 μg/minute to overt proteinuria defined as albuminuria ≥300 mg/24 hours or ≥200 μg/minute and worsening azotemia. There are several clinical similarities between diabetic nephropathy in patients with type 1 diabetes and in patients with type 2 diabetes; however, the clinical course differs in some respects between these two groups. In patients with type 1 diabetes in whom nephropathy develops the clinical course is relatively well defined. Nephropathy usually becomes clinically evident after 15–25 years of diabetes and almost always progresses to ESRD.

However, because of the frequently insidious onset of type 2 diabetes, the advanced age of many patients, and the common presence of coexisting vascular disease and hypertension, early renal involvement is frequently missed. In elderly patients with type 2 diabetes, it is not always clear whether renal failure is due solely to or even caused by diabetes. However, in young patients with type 2 diabetes, recent studies have shown a course similar to that in type 1 diabetes. Thus, the clinical course of diabetic nephropathy is best defined in type 1 diabetes and proceeds through several stages (Figure 54–2).

Laboratory Findings

Stage 1: Glomerular Hyperfiltration and Renomegally

At the onset of type 1 diabetes, the GFR is above normal up to 140% in the majority of individuals. No single pathogenesis fully explains both the nephromegaly and glomerular hyperfiltration characteristic of type 1 diabetes; a correlation between renal enlargement and glomerular hyperfiltration has been inferred from the correction of both perturbations after the establishment of euglycemia. Intensive insulin therapy normalizes hyperglycemia and corrects glomerular hyperfiltration; the GFR begins to decline within 8 days of initiation of insulin therapy and falls further during 3 months of insulin treatment. A substantial subset of individuals (˜25–40%) with type 1 diabetes achieving usual levels of plasma glucose under insulin therapy continues to manifest a persistently elevated GFR; it is within this subgroup of hyperfiltering diabetic patients that the initial reductions in the GFR are first noted, with progression to clinical nephropathy.

Glomerular hyperfiltration has also been reported in patients with a recent diagnosis of type 2 diabetes mellitus and is positively correlated with the development of proteinuria.

Stage 2: Early Glomerular Lesion

Expansion of the glomerular mesangial matrix and thickening of the glomerular basement membrane (GBM) are subtle morphologic changes noted 2–5 years after the onset of type 1 diabetes and persists for many years. During this stage of morphologic change, transient and recurrent microalbuminuria may be the only clinical evidence of renal involvement. A supplemental perspective of the early glomerular changes in type 2 diabetes is provided by the study of kidneys from nondiabetic donors transplanted into diabetic recipients. In biopsies of these recipient kidneys, the same sequence of changes—mesangial matrix expansion and GBM thickening within 3–5 years after transplant—was observed. Reversal of these sequential changes in type 1 diabetes has been noted after 10 years of euglycemia afforded by a functioning pancreas transplant. Stages 1 and 2 of diabetic nephropathy are usually clinically silent since the estimation of the GFR and renal biopsy are not performed routinely. However, early mesangial and basement membrane abnormalities are often seen on renal biopsy in stages 1 and 2, if this is performed.

Stage 3: Microalbuminuric Stage

In the early 1980s, studies in type 1 diabetic patients linked the presence of increased amounts of albumin in the urine, measured by immunoassay, to the subsequent development of overt diabetic nephropathy 10–14 years later.

Microalbuminuria is defined as increased urinary albumin excretion (UAE) (30–300 mg/24 hours or 20–200 μg/minute) not detected by standard bedside tests (dipstick) commonly employed to detect proteinuria. Several methods for quantifying low concentrations of urinary albumin are available, including radioimmunoassay, enzyme-linked immunoassay, and nephelometric immunoassay; a semiquantitative dipstick test has also been introduced (Micro-Bumintest, Miles Laboratories, Elkhart, IN).

A timed urine collection, either 24 hours or overnight, is the reference standard for assessment of microalbuminuria. Because of high intraindividual variability, transient elevations of proteinuria in the microalbuminuric range occur frequently, and clinical assessment should therefore be based on at least three measurements taken over 3–6 months. Persistent microalbuminuria is confirmed when at least two out of three consecutive, timed urine collections are in the range of 20–200 μg/minute.

Several factors can confound the assessment of microalbuminuria, including urinary tract infection, heavy exercise, high dietary protein intake, congestive heart failure, and acute febrile illness. For accurate evaluation, testing should be postponed if these factors are present.

There is accumulating evidence suggesting that the risk of developing diabetic nephropathy and cardiovascular disease starts when UAE values are still within the normoalbuminuric range. After 10 years of follow-up, the risk of diabetic nephropathy was 29 times greater in patients with type 2 diabetes with UAE values >10 μg/minute. The same was true for patients with type 1 diabetes. This favors the concept that the risk associated with UAE is a continuum, as is the case with blood pressure levels.

Although microalbuminuria has been considered a risk factor for macroalbuminuria, not all patients progress to this stage and some may regress to normoalbuminuria. In earlier studies ˜80% of type 1 diabetic patients with microalbuminuria progressed to proteinuria over a period of 5–15 years. In recent studies, only 30–45% of microalbuminuric patients progressed to proteinuria over 10 years. This change might be the consequence of more intensive glycemic and blood pressure control strategies.

Without specific intervention, as detailed below, once microalbuminuria becomes constant, a progressive decline in renal functional reserve marking a downhill course toward clinical renal failure is usual. Typically, albumin excretion increases by about 25 μg/minute/year, while the GFR remains normal or elevated. The GFR begins to decline, at a variable although individually constant rate, once the amount of microalbuminuria exceeds 70 μg/minute. Blood pressure is higher in type 1 diabetic patients with microalbuminuria than in normoalbuminuric patients, although not necessarily higher than 140/90 mm Hg. The combination of microalbuminuria and high blood pressure leads to near-term (5–15 years) deterioration to clinical nephropathy.

Stage 4: Clinical Nephropathy

Following a variable interval, usually after several years of microalbuminuria, the GFR declines below normal for age and gender and proteinuria is detectable by dipstick. Proteinuria, defined as albumin excretion greater than 300 mg/day, is the universal hallmark noted in diabetic nephropathy. Once overt nephropathy develops, blood pressure is usually elevated, and there is a progressive decline in the GFR that can be assessed as an absolute decline in milliliters/minute per year.

In most patients with overt nephropathy and type 1 diabetes, the decrease in the GFR approaches linearity and is of the order of 11 mL/minute per year. In a prospective study, 227 white type 2 diabetic patients with nephropathy for 6.5 years (range 3–17 years) had a decline in the GFR of 5.2 mL/minute/year. In multivariate regression analysis, higher baseline albuminuria, systolic blood pressure, hemoglobin A1c (HbA1c), heavy smoking, and the presence of diabetic retinopathy were significantly associated with an increased rate of decline in the GFR. In the subset of diabetic patients who manifest nephropathy in both type 1 and type 2 diabetes, proteinuria is duration related, increasing in time to a florid nephrotic syndrome (proteinuria of more than 3.5 g/day coupled with hypoalbuminemia and hyperlipidemia). Anasarca develops in diabetic nephrotic patients at a higher serum albumin concentration than in nondiabetic patients, an observation probably explained by the fact that glycation of albumin leads to enhanced transcapillary permeability compared to normal albumin. Nearly 100% of diabetic patients who have reached the azotemic phase of diabetic nephropathy have coincident retinopathy when examined by fluorescein angiography; the absence of diabetic retinopathy in advanced renal disease is reason to doubt the diagnosis of diabetic nephropathy.

It should be kept in mind that because diabetic patients are at equal risk for unrelated renal disease, the quest for a renal diagnosis should be pursued, including renal biopsy, whenever the course does not fit the usual pattern of diabetic nephropathy, most often signaled by the absence of diabetic retinopathy and albuminuria, a rapid increase in serum creatinine, and the presence of active urinary sediment.

Stage 5: End-Stage Renal Disease

As noted above, after 20–30 years of type 1 diabetes, about 30–40% of patients manifest irreversibly failed kidneys. Prospective studies of populations with a high prevalence of type 2 diabetes such as blacks, Hispanics, and Native American tribes indicate that the interval between diagnosis of type 2 diabetes and onset of ESRD ranges from 5 years to 25 years. But because the diagnosis of type 2 diabetes may be delayed until a clinical complication or comorbidity is apparent, the timetable is imprecise. As the GFR decreases the uremic symptoms and signs become apparent and renal replacement therapy is needed.

Imaging Studies

The central pathologic change in the diabetic kidney is overproduction and impaired degradation of ECM components that lead to their accumulation in the basement membranes and mesangial region of the glomerulus.

Light microscopy findings show an increase in the solid spaces of the tuft, most frequently observed as coarse branching of solid material. An extensive accumulation of matrix is also seen as a nodule and the lesion is known as Kimmelstiel–Wilson nodules. Hyalin deposits may be seen in Bowman's capsule (“capsular drop”) and in the expanded mesangium. Diffuse global glomerular sclerosis, tubular atrophy, and interstitial fibrosis are late manifestations of the disease.

Immunofluorescence microscopy may show deposition of immunoglobulin G and albumin along the GBM in a linear pattern.

Electron microscopy shows thickening of the GBM in a very early stage of diabetic nephropathy, concomitant with the increase of mesangial matrix. In advanced disease, the mesangial regions occupy a large proportion of the tuft, with prominent matrix.

The severity of diabetic glomerulopathy has been gauged in some studies by the thickness of the basement membrane and mesangium and matrix expressed as a fraction of appropriate spaces (eg, volume fraction of mesangium/glomerulus, matrix/mesangium, or matrix/glomerulus).

Special Tests

Screening for diabetic nephropathy in type 1 diabetic patients is recommended 5 years after diagnosis; for type 2 diabetic patients screening should be started when diabetes is diagnosed since at least 7% of newly diagnosed patients have microalbuminuria. Furthermore, in type 1 diabetic patients with poor glycemic control, high blood pressure, and poor lipid regulation, the prevalence of microalbuminuria before 5 years may be as high as to 18%. It follows that in type 1 diabetes, screening for microalbuminuria is reasonable 1 year after diabetes is diagnosed. If microalbuminuria is absent, screening should be repeated annually in both type 1 and type 2 diabetic patients.

The first step in the screening and diagnosis of diabetic nephropathy is to measure albumin in a spot urine sample, collected either as the first urine in the morning or at random. Results of albumin measurements in spot urine collection may be expressed as a urinary albumin concentration (mg/L) or as a urinary albumin-to-creatinine ratio (mg/g or mg/mmol). The upper limit of 17 mg/L in a random urine specimen has a sensitivity of 100% and a specificity of 80% for the diagnosis of microalbuminuria using a 24-hour timed urine collection as the reference standard. All positive tests should be confirmed in two out of three samples collected over a 3- to 6-month period due to the known day-to-day variability in urinary albumin excretion.

Screening should not be performed in the presence of conditions that increase UAE such as urinary tract infection, hematuria, acute febrile illness, vigorous exercise, and heart failure. Although measurement of UAE is the cornerstone for diagnosing diabetic nephropathy, some patients with either type 1 or type 2 diabetes may have a decreased GFR with a normal UAE. In patients with type 1 diabetes, this finding is more frequent in women with long-standing diabetes, hypertension, and/or retinopathy. Therefore, the GFR and UAE should be routinely estimated for proper screening of diabetic nephropathy.

Early Treatment of Risk Factors

Early treatment of risk factors for diabetic nephropathy will slow and/or prevent its progression. These risk factors include hyperglycemia, hypertension, smoking, and dyslipidemia. These are also risk factors for cardiovascular disease that should be vigorously treated.

Intensive Glycemic Control

The importance of strict glycemic control in preventing diabetic nephropathy has been confirmed by large clinical trials in both type 1 and type 2 diabetes. In the Diabetes Control and Complications Trial (DCCT), intensive treatment of diabetes reduced the incidence of microalbuminuria by 39%. Furthermore, patients randomized to strict glycemic control had a long-lasting reduction of 40% in the risk for development of microalbuminuria and hypertension 7–8 years after the end of the DCCT. Similarly, the UK Prospective Diabetes Study (UKPDS) of type 2 diabetes reported a 30% reduction in the risk of developing microalbuminuria in the intensively treated group as compared to the conventionally treated group.

The effect of strict glycemic control on the progression from microalbuminuria to macroalbuminuria and on the rate of renal function decline in macroalbuminuric patients is still controversial.

In the DCCT study, intensified glycemic control did not decrease the rate of progression to macroalbuminuria in patients with type 1 diabetes who were microalbuminuric at the beginning of the study. In another prospective study of 115 patients with diabetes and renal impairment, 50 with type 1 diabetes and 65 with type 2 diabetes, no relationship between HbA1c and fall in creatinine clearance was seen over a 7-year period. However, in two-combined analysis of the large Steno study better glycemic control was associated with a fall in the urinary albumin excretion rate and a decreased decline in the GFR. Similarly, in 18 patients with type 1 diabetes and diabetic nephropathy prospectively followed for 21 months, a direct relationship was noted between fall in the GFR and HbA1c, with the greatest fall in the GFR occurring in those with the highest HbA1c.

There are fewer details on the natural history of treated type 2 diabetes, although one study based in Japan randomized 110 patients with type 2 diabetes and 55 patients with retinopathy and microalbuminuria to receive either multiple injection therapy or conventional insulin therapy over a 6-year period. The cumulative percentage of progression of nephropathy in the multiple insulin therapy group was 11.5% as compared to 32.0% in the conventional insulin therapy group.

To date, no large trial of intensive therapy has been reported in overt nephropathy in either type 1 or type 2 diabetes. This lack of evidence of efficacy might be due to the complexity of delivering tight glycemic control and the increased risk of hypoglycemia in patients with impaired renal function. Therefore, intensive treatment of diabetes aiming at a HbA1c <7% (The American Diabetes Association standard) should be pursued as early as possible to prevent the development of microalbuminuria.

Intensive Blood Pressure Control

Multiple studies on the effect of pharmacologic induction of normotension in type 1 and type 2 diabetic patients with persistent microalbuminuria indicate that urinary albumin excretion may be reduced while clinically evident nephropathy is postponed and perhaps prevented. Hypertension is common in diabetic patients, even when renal involvement is not present. About 40% of type 1 and 70% of type 2 diabetic patients with normoalbuminuria have blood pressure levels <140/90 mm Hg.

In both type 1 and type 2 diabetic patients with overt diabetic nephropathy, blood pressure reduction, whether with angiotensin-converting enzyme (ACE) inhibitors or non-ACE inhibitors, reduces albuminuria, delays progression of nephropathy, postpones renal insufficiency, and improves survival. As a key example in UKPDS, a reduction in systolic blood pressure from 154 mm Hg to 144 mm Hg reduced the risk for the development of microalbuminuria by 29%.

Although retardation of the decline in the GFR has been shown with other antihypertensive medications, blockade of the renin–angiotensin system (RAS) with ACE inhibitors or angiotensin receptor blockers (ARBs) is thought to confer an additional “renoprotective” benefit on preserving renal function.

The sixth report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure first published the recommendation that the target level of blood pressure in diabetes be reduced to at or below 130/85 mm Hg. Similarly, the American Diabetes Association adapted the recommendation in 1999. Continuing this theme, a consensus report of the National Kidney Foundation (NKF) advises that the blood pressure goal should be less than 130/80 mm Hg for nonproteinuric patients and 125/70 mm Hg for those with proteinuria. To achieve this degree of blood pressure reduction usually necessitates more than one antihypertensive drug plus a willing patient. The ideal drug combination is far from clear. Few data compare combination drugs with single agents. One study examined the added antihypertensive effect of 12.5 mg of hydrochlorothiazide (HCTZ) in patients with diabetic nephropathy and GFR values within normal ranges. These patients were initially treated with an ACE inhibitor (cilazapril) or an α-blocker (doxazosin). A mean decline of 15 mm Hg in systolic blood pressure (SBP) and 8 mm Hg in diastolic blood pressure (DBP) was obtained. Addition of HCTZ induced a further decline in SBP of 8 mm Hg and in DBP of 5 mm Hg. The combination also gained a greater reduction in the urinary albumin excretion rate.

ACE inhibitors are the drugs of choice for the treatment of hypertension and for retarding the decline in renal function in diabetic patients. Evidence that ACE inhibitors are indeed renoprotective against deterioration of renal failure was afforded by a study that demonstrated a significant decline in primary endpoints (doubling of serum creatinine and the development of ESRD) in the cohort treated with captopril. Seven years later, the EUCLID study, which was done in 18 European centers, randomized type 1 diabetic patients with normoalbuminuria or microalbuminuria to treatment with an ACE inhibitor (lisinopril) or a placebo. After 24 months, there was a significant difference favoring the lisinopril cohort both in terms of mean UAE and in the ratio of transition from normoalbuminuria to microalbuminuria. Similar benefits of ACE inhibitors have also been seen in patients with type 2 diabetes. The MICRO-HOPE study involved 1140 patients with type 2 diabetes and microalbuminuria randomized to ramipril 10 mg/day or a placebo. The blood pressure of all patients was to be kept at normal (target) values permitting the addition of other medications. The study aim was to determine whether ACE inhibition has organ protective effects independent of its antihypertensive action. Sustaining this thesis, at 4.5 years ramipril-treated patients had a combined primary outcome risk reduction of 25%, with myocardial infarction being decreased by 22%, stroke by 33%, cardiovascular death by 37%, and total mortality by 24%. There was a slower rise in UAE in ramipril-treated patients while fewer patients on ramipril progressed from microalbuminuria to macroalbuminuria.

ARBs are antihypertensive agents that inhibit the renin–angiotensin system by selectively blocking the AT1 subtype of A11 receptors. A renoprotective effect of ARBs was proven by the results of three large international prospective controlled studies published in one issue of the New England Journal of Medicine. These three studies enrolled type 2 diabetic patients with microalbuminuria or overt proteinuria and renal impairment. The renoprotective effect of irbesartan and losartan was established as these agents not only reduced albuminuria but also, over period of 2–4 years, significantly retarded the decline in the GFR and decreased the proportion of patients progressing to end-stage renal failure.

These three studies, in addition to the MICRO-HOPE and EUCLID studies, provide a strong rationale for the use of ACE inhibitors and/or ARBs in diabetic patients with nephropathy. The present standard of care for proteinuric diabetic patients is based on the inclusion of ACE inhibitors or ARBs. Recently, several studies have indicated benefits from dual blockade of the renin–angiotensin system by ACE inhibitors plus ARBs as noted by a greater reduction in albuminuria and blood pressure in combination-treated diabetic patients compared to the maximal dose of an ACE inhibitor alone.

Dietary Protein Restriction

In health and in diabetes, dietary protein intake modulates renal hemodynamics. Several reports suggest that in type 1 diabetes, ingestion of a high protein diet increases the risk of nephropathy. It is also known that protein restriction is effective in alleviating the symptoms of uremia and may delay the need for dialysis. Beyond symptomatic relief, a reduced protein intake is advocated to slow or/and prevent a decline in renal function. The proposed mechanism for halting progression is the reduction in hyperfiltration that occurs in the remaining nephrons after renal injury is established. A benefit from dietary protein restriction was noted in a small prospective randomized controlled study of 35 patients with type 1 diabetes and clinical nephropathy. The low-protein diet contained 0.6 mg/kg/day and patients were followed for a mean period of 35 months. A 4-fold decrease in the rate of fall of the GFR was found in the low-protein diet group compared to controls after 3 months. The mean urinary protein excretion fell by 24% in the study group but rose by 22% in the control group. At the end of the study, the reduction in proteinuria in the study population was only 6%, whereas the controls had a 24% increase. Adding to the case for dietary protein restriction, it was found that albumin excretion rates decreased when patients with microalbuminuria were fed a predominantly vegetarian diet in the absence of significant change in either blood glucose control or arterial blood pressure. Although a clear benefit of dietary restriction has not been shown in a large randomized prospective trial, based on these small positive clinical studies a protein intake of <0.8 g/kg is a reasonable regimen for patients with macroalbuminuria with further restriction as the GFR falls.

Lipid-Lowering Drugs

Hyperlipidemia is a risk factor for the development of vascular disease including nephropathy in both rat models of diabetes and human type 1 and type 2 diabetes. The effect of lipid reduction by antihyperlipidemic agents on the progression of diabetic nephropathy has not been established. Although large prospective trials of the effect of treatment of dyslipidemia on the progression of diabetic nephropathy have not been reported, some evidence indicates that lipid reduction by antihyperlipidemic agents preserves the GFR and decreases proteinuria in diabetic patients.

A recent randomized, double-blind placebo-controlled study compared simvastatin and diet versus placebo and diet on albuminuria in 39 patients with type 1 diabetes and nephropathy. Although not attaining significance, the rise in albuminuria was slower in the simvastatin-treated cohort compared to placebo over 2 years. A meta-analysis on lipid-lowering therapy on progression of renal disease assessed 13 prospective controlled trials, seven of which were exclusively in diabetic patients. Lipid lowering was associated with a lower rate of decline in renal function compared to controls (p = 0.008) inducing beneficial effects equivalent to ACE inhibition in the preservation of renal function. The discerned effect on the GFR did not correlate with either the type of lipid-lowering agent or the etiology of renal disease. As for other components of renoprotection, large prospective clinical trials with longer follow-up are needed to validate the value of lipid-lowering agents.

However, as cardiovascular disease is the number one cause of death in diabetic patients with nephropathy, optimizing lipid control is now a standard of care.

Cessation of Cigarette Smoking

Convincingly, a linkage between cigarette smoking and progression of diabetic nephropathy has been established. The effect of cigarette smoking on diabetic renal and retinal complications in 359 type 1 diabetic patients was assessed. The prevalence of an increased rate of albumin excretion was 2.8 times higher in smokers than in those who do not smoke. Even after correction for glycohemoglobin level and duration of diabetes, smoking was a significant factor in their logistic regression model for albuminuria. Significant improvement in urinary albumin excretion was noted when subjects ceased smoking. Similar findings were reported in patients with type 2 diabetes and nephropathy. As is true for preventing pulmonary and cardiovascular disease, reducing or quitting smoking should be part of all regimes for renoprotection in diabetic patients.

Uremia Therapy

Although normalizing blood pressure, optimizing glycemic control, and adhering to a low-protein diet may retard the development and progression of diabetic nephropathy, many diabetic patients still progress to ESRD. Continuing empathetic interactions as renal function deteriorates builds confidence and minimizes panic, despair, and frantic behavior as the need for ESRD therapy becomes pressing. Patients with diabetic nephropathy should be referred to a nephrologist early in the course of their disease in order to optimize their pre-ESRD care. Reports from both Europe and the United States found that a high proportion of patients with diabetic nephropathy are referred to a nephrologist late in the course of their disease, resulting in suboptimal treatment.

Management of diabetic patients with progressive renal insufficiency is a challenge because of comorbid conditions that accompany the nephropathy. These preexisting comorbid conditions (cardiovascular disease, retinopathy, cerebrovascular and peripheral vascular disease) play a major role in the decreased survival of diabetic patients on renal replacement therapy. In practice, a team of collaborating specialists optimize the pre-ESRD diabetic care. This team should include a nephrologist, diabetologist, nutritionist, cardiologist, ophthalmologist, podiatrist, and other specialists when necessary. Pre-ESRD management of diabetic patients with advanced renal failure includes sustaining a hemoglobin level above 11 g/dL by administering erythropoietin and supplemental iron, minimizing metabolic bone disease due to secondary hyperparathyroidism by use of phosphate binders, along with use of synthetic vitamin D and/or calcimimetics. Potential familial kidney donors should be interviewed and tissue typed; when hemodialysis appears to be likely, preserving forearm cutaneous veins by avoiding venous punctures and intravenous catheters and maintaining good nutritional status are essential.

Renal Replacement Therapy

Diabetic patients with ESRD have options for renal replacement therapy similar to nondiabetic patients. For diabetic ESRD patients those options include hemodialysis, peritoneal dialysis, kidney transplant alone, and a combined pancreas and kidney transplant, which is unique to diabetic patients. Before a patient is assigned a specific modality of renal replacement therapy, it is important that patients and their families be properly informed about the advantages and disadvantages of each treatment option.

Selecting the treatment that is best for a particular patient is made by considering the patient's age, level of education, severity of comorbid conditions, social and family support, and geographic location. Once the decision for a preferred option is made, preparation for renal replacement therapy should be started. For instance, an arteriovenous fistula should be established in patients who will be undergoing hemodialysis, and a peritoneal catheter should be inserted in those who will be treated with peritoneal dialysis. As a general guide, renal replacement therapy is initiated when the GFR of a diabetic patient falls to 10–15 mL/minute.

Maintenance Hemodialysis

As reported in the USRDS 2004 registry, 75% of all diabetic patients with ESRD are treated with hemodialysis (center or home), 7.4% with peritoneal dialysis [continuous ambulatory peritoneal dialysis (CAPD) or continuous cyclic peritoneal dialysis (CCPD)], and 17% receive a functioning kidney transplant. Hemodialysis treatment for diabetic patients is similar to that for nondiabetic patients. An ideal hemodialysis regimen consists of three weekly dialysis sessions, each lasting 3.5–4.5 hours, as determined by individual blood chemistry and clinical response during which extracorporeal blood flow is maintained at 300–500 mL/minute.

The survival and rehabilitation of diabetic patients on maintenance hemodialysis are distinctly inferior to those of nondiabetic patients, mainly because of preexisting severe vascular disease. Peripheral vascular calcification in middle-sized arteries plus atherosclerosis of small vessels may pose a challenge for the vascular surgeon attempting to construct a vascular access in a diabetic patient. Although the preferred vascular access is an arteriovenous fistula, preexisting vascular disease limits its utility in diabetic patient, who have a primary fistula failure rate of 30–40%. A less desirable, but necessary alternative vascular access can be constructed using a polytetrafluoroethylene graft, which has a half-life in excess of 1 year.

The proportion of arteriovenous fistulas compared to grafts can be increased in diabetic patients by careful preoperative investigation to select an adequate location for initial placement to the fistula. Large-diameter arteries and veins, frequently requiring use of the elbow region, are employed, thereby avoiding an initial polytetrafluoroethylene graft. Continuous surveillance by a nephrologist and the dialysis staff concerning early elective access revision and avoidance of thrombosis improves the viability of the primary fistulas. Complications of vascular access are the leading cause of hospitalization in diabetic patients with ESRD undergoing hemodialysis.

Glycemic control in diabetic patients undergoing dialysis is difficult. Insulin dosage is more complex because of unrecognized gastroparesis, which disconnects absorption of ingested food from timed insulin administration and because of reduced renal insulin catabolism, which results in the prolonged action of exogenous insulin. This combination causes erratic glucose regulation complicated by frequent hypoglycemic episodes, a potentially serious complication. Glycemic control should remain a priority in the dialyzed diabetic patient as it may retard further complications of microvascular disease. Survival during the long-term management of ESRD in diabetes has been linked to the quality of glycemic control achieved.

Although the survival of diabetic patients on dialysis has been improving over the past decade, morbidity and mortality remain significantly higher than in those without diabetes. This grim reality is mostly attributed to the progression of comorbid conditions. Cardiovascular disease, infections, and withdrawal from dialysis are the leading cause of mortality in patients with diabetes and ESRD.

Peritoneal Dialysis

Peritoneal dialysis is a satisfactory alternative mode of dialytic therapy available for diabetic ESRD patients. In the United States, peritoneal dialysis is applied to only 7% of all diabetic patients on renal replacement therapy. As is true for hemodialysis, preparation of the patient for CAPD, the most frequently utilized form of peritoneal dialysis, necessitates education, repetitive explanation, and facilitating surgery to insert an intraperitoneal permanent catheter. CAPD can be mastered as a home regimen in about 4 weeks. An alternative to manual cycling of the dialysate is the use of a mechanical cycling device in a CCPD regimen, which can be performed during sleep.

CAPD offers some advantages over hemodialysis, such as freedom from a machine, performance at home, reduced cardiovascular stress, better preservation of renal function, avoidance of heparin, and less dietary restrictions. However, the disadvantages of peritoneal dialysis include the risk of peritonitis, the high rate of technical failure, and less adequate dialysis when residual renal function is low.

Some nephrologists consider peritoneal dialysis as the preferred choice of treatment for diabetic ESRD patients. Indeed, the CAPD or CCPD may be life sustaining when vascular access sites for hemodialysis have been exhausted, or in those with severe congestive heart failure, angina, or severe dialysis-related hypotension. Peritoneal dialysis involves less vascular stress because of its relatively slow ultrafiltration rate coupled with less rapid solute removal.

During the course of both CAPD and CCPD, there is a constant risk of peritonitis as well as a gradual decrease in peritoneal surface area, which may ultimately prove to be insufficient for adequate dialysis. Diabetic patients on CAPD experience twice as many hospitalization days as nondiabetic patients: peritonitis accounts for 30–50% of these hospital days.

Of the two major options in dialytic therapy for ESRD in diabetes, the USRDS consistently reports superior survival in those treated by hemodialysis, except in those under age 45 years, compared to peritoneal dialysis. However, there are few studies reporting equivalent patient survival for peritoneal dialysis versus hemodialysis in the first 2 years of treatment. The two leading causes of death in diabetic CAPD patients are cardiovascular events and infection.

Kidney Transplantation

In the 1970s and early 1980s many transplant programs excluded diabetic ESRD patients from consideration for renal transplantation. However, in centers that performed transplants during this period, the survival of diabetic patients receiving transplants exceeded that of diabetic patients remaining on dialysis. Today, the improved management of diabetic and uremic complications established kidney transplantation as the preferred form of treatment for diabetic ESRD patients.

Although the survival of diabetic ESRD patients after renal transplantation is continuously improving, over a 5-year period it is 10–20% below that of patients with other causes of renal disease. A further decrease in survival of diabetic renal transplant recipients after 5 or more years is the consequence of coronary, cerebral, and other macrovascular diseases. The most recent analysis of patient survival in diabetic kidney transplant recipients indicated 93.7% 1-year and 85.5% 3-year survival for recipients of grafts from deceased donors and 95.4% 1-year and 91.3% 3-year survival for those receiving living donor transplants. The annual death rates of transplant recipients are approximately one-third those of diabetic patients remaining on dialysis. In fairness, it must be noted that a strong selection bias extracts the fittest patients for kidney transplants leaving a residual pool of dialysis patients with extensive life-threatening comorbidities.

To respond to the real risk of cardiac events in diabetic renal transplant recipients and to assess the extent of cardiovascular disease present prior to transplantation, at least annual reassessment with noninvasive studies in asymptomatic high-risk patients should be performed.

Pancreas Transplantation

Over the past decade, highly successful results have been reported in type 1 diabetic patients for pancreatic transplants inserted concurrently with a renal allograft. Although combined pancreas and kidney transplants do not result in an increase in immediate perioperative mortality, perioperative morbidity is markedly increased over that of a kidney transplant alone. Almost 13,000 pancreas transplants were reported to the international Pancreas Transplant Registry from 1966 through September 1999, with 75% of them performed in the United States. The majority of pancreas transplants in the United States have been combined kidney–pancreas transplants (SPK). One year patient survival has improved from 90% for 1987–1988 cases to 95% for 1995–1996 cases. In addition, pancreas graft survival has improved from 74% to 85% at 1 year for the same time period; kidney graft survival improved from 83% to 91%. Solitary pancreas transplants, either pancreas alone (PTA) or pancreas after kidney transplant (PAK), constitute only a small proportion of the total pancreas transplants in the United States. Graft survival of PTA or PAK is worse compared to SPK. Early reports have surprisingly found a substantial survival benefit for pancreas plus kidney transplants in type 2 diabetic ESRD patients.