 |
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2007 JASN IMPACT FACTOR 7.111 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Frontiers in Nephrology: Diabetic Nephropathy: Understanding Mechanism and Defining Risk |
* Medicine; Pediatrics, University of Minnesota, Minneapolis, Minnesota; and Department of Medical and Surgical Sciences, University of Padova, Padova, Italy
Address correspondence to: Dr. Michael Mauer, Department of Pediatrics, MMC 491, 420 Delaware Street S.E., Minneapolis, MN 55455. Phone: 612-626-2922; Fax: 612-626-2791; E-mail: mauer002{at}umn.edu
 |
Abstract |
|---|
|
Introduction |
|---|
In the early 1980s, independent investigators reported that some patients with diabetes had increased urinary albumin excretion rates (AER) that were not detectable by standard laboratory methods and termed this microalbuminuria (MA). Initial retrospective studies reported that approximately 80% of patients with type 1 diabetes would progress from MA to proteinuria (P) over 6 to 14 yr (10–12). Because the post hoc values for MA that are predictive of progression to P differed among these initial investigations, a conference was convened to achieve consensus on the definition of MA (13), and these values have been used up to the present time.
|
DN Risk in Patients with Type 1 Diabetes |
|---|
More recent data from the Steno Diabetes Center showed higher progression rates in an inception cohort of patients who had type 1 diabetes with a much older age of onset (approximately 27 yr) than is typical for type 1 diabetes (18). In this study (18), 79 (28.5%) of 277 patients progressed from NA to MA and 27 (9.7%) progressed to P over a median follow-up period from type 1 diabetes onset of 18 yr (1.0 to 21.5). However, it is important to consider that >30% of the patients who progressed to MA later reverted back to NA, either transiently (15 patients) or permanently (13 patients). Thus, at the end of the study, progression from NA to MA was approximately 24%.
Although MA remains the best available marker for DN risk, recent studies suggested that the percentage of patients who have MA and progress to P over approximately 10 yr is approximately 30% (14), much less than the initial reports of approximately 80% (10–12). Forsblom et al. (19) were the first to suggest that patients with MA might revert to NA. He evaluated 20 patients with type 1 diabetes and MA and 16 to 36 yr of type 1 diabetes duration and found progression to P 10 yr later in only 25%, whereas 35% had NA on follow-up and 40% still had MA. Rossing et al. (20) indicated that the risk for progression to P in patients with MA and <15 yr of diabetes duration was 45 versus 26% progression rate in patients with >15 yr duration. Drs. Peter Rossing and Hans-Henrik Parving performed analyses that we had requested (14) and found that only 30% of 132 patients with MA and with diabetes duration of 20.3 ± 8.7 yr (7 to 40) developed P after a mean follow-up period of 9.1 yr. Moreover, 20% of these patients returned to NA, and 50% still had MA on follow-up. Duration of diabetes was shorter at baseline in patients whose MA progressed to P (17 ± 8 yr) than in those who still had MA (22 ± 9 yr; P < 0.005) but not different from those who reverted to NA (20 ± 9 yr). It is interesting that data that were extracted from conventionally treated patients who were in the Diabetes Control and Complications Trial (DCCT) and had MA after 7 to 15 yr of diabetes duration (14) revealed a 23% progression rate from MA to P over the 7 yr of this study. This was less than the 45% progression rate in the similar but much larger cohort of Rossing et al. (20). Rudberg et al. (15) found an even lower progression rate (18%) in children and adolescents who had MA and were followed for 8 yr. Since we published a review of the predictive value of AER for DN risk (14), several additional papers on his topic have been published. Perkins et al. (21) followed for 6 yr 220 patients who had type 1 diabetes and MA and found that 40% of them had reverted to NA and only 15% had progressed to P. Similarly, 35% of patients with MA in the Steno Diabetes Center inception cohort reverted to NA (half of them permanently) and 34% progressed to P, after a median follow-up of 7.5 yr (0.1 to 15.8) (18). In the large EURODIAB study, 50.6% of the patients who had MA reverted to NA whereas only 13.9% progressed to P after 7 yr of follow-up (22).
Thus, these more recent studies suggest that the rate of progression from MA to P over 5 to 10 yr is approximately 15 to 30%, although as high as 45% in patients with <15 yr of diabetes duration but considerably lower than originally estimated. As mentioned, the Steno cohort study did find that many of the patients who had MA that reverted to NA during follow-up had redeveloped MA by the end of the study (18). However, risk for progression of this subgroup to P remains unknown.
|
DN Risk in Patients with Type 2 Diabetes |
|---|
The regularity with which patients with type 2 diabetes and P progress to ESRD is less well known than for patients with type 1 diabetes. Nelson et al. (24) suggested that the rate of decline of GFR among Pima Indians with type 2 diabetes is similar to that of white individuals with type 1 diabetes, but whether Pima Indian patients with their younger age of onset and racial specificity are fully representative of the larger body of patients with type 2 diabetes is not yet established.
Patients with NA type 2 diabetes, as a group, have diabetic glomerular lesions, but their rate of development is less clear, because, with the exception of the Pima Indian studies, duration is usually not precisely established in type 2 diabetes. Whether patients with NA and type 2 diabetes with more advanced diabetic renal lesions are at greater risk for progression needs to be determined.
Mogensen (25) was the first to explore the prognostic value of AER in patients with type 2 diabetes. MA, in this 10-yr retrospective study, was defined as urinary albumin concentration of 30 to 140 µg/ml. The 76 patients with MA and type 2 diabetes in this study experienced a 77.6% 10-yr mortality mostly from cardiovascular disease, whereas 22% progressed to P (25). Subsequently, three prospective studies with much lower mortality rates among patients with MA and type 2 diabetes have been published. These studies included a total of 152 patients who had MA and were followed for 5 to 6 yr and had an average risk for progression to P of approximately 40%. In one of these studies (26), no patients had NA at follow-up, whereas the other two studies (27,28) did not provide these data. Other studies described progression rates from MA to P from 33 to 40% (29–31). In one report with 21 patients, none progressed to P, whereas 43% reverted to NA (32).
The United Kingdom Prospective Diabetes Study (UKPDS) (33), using data from >5000 patients who were followed from the diagnosis of type 2 diabetes, found that progression from NA to MA occurred at a rate of 2.0% per year, from MA to P at 2.8% per year, and from P to increased serum creatinine (>175 µmol/L) or RRT at 2.3% per year. Cardiovascular death increased from an annual rate of 0.7% in NA to 2.0% in MA, 3.5% in P, and 12.1% in patients with increased serum creatinine or RRT. This very important analysis unfortunately did not provide any information on the risk factors associated with DN progression.
The Steno-2 study (34) evaluated the effects of multifactorial intervention on micro- and macrovascular events in 151 patients with type 2 diabetes and MA during 7.8 yr of follow-up. When all 151 patients were pooled, 46 (31%) patients reverted to NA, 58 (38%) still had MA, and 47 (31%) progressed to P (34). Baseline predictors of reversion and progression were AER levels and retinopathy status; predictors during follow-up were glycated hemoglobin (HbA1c), start of antihypertensive therapy, and start of RAS blockade (34) (see below). It is interesting that patients who reverted to NA had a GFR decline of 2.3 ml/min per yr, those who still had MA lost 3.7 ml/min per yr, and those who progressed to P lost 5.4 ml/min per yr (34). Recently, Araki et al. (35) reported that return to NA was frequent among Japanese patients with MA and type 2 diabetes. Indeed, among 216 patients with MA, during a follow-up of up to 6 yr, the prevalence of NA was observed to be 22, 33, and 23% at each successive 2-yr follow-up period; the prevalence of P was 9, 13, and 19% at the same time intervals (35). The factors that were associated with reversal to NA were short duration of MA, use of RAS-blocking drugs, and lower HbA1c and systolic BP (SBP) values (see below). Thus, also in type 2 diabetes, a significant proportion of patients with MA may return to NA.
In summary, AER remains the strongest indicator of DN risk in patients with type 1 diabetes and type 2 diabetes, and patients with MA have 200 to 400% higher risk for progression to P than patients with NA. Nonetheless, MA is not as strong a predictor of P as initially considered and a high proportion of patients who are at risk for severe DN will have NA at initial evaluation. For these reasons, it is important to consider other variables that could augment the predictive power of AER.
|
Family History |
|---|
|
Glycemia |
|---|
As discussed above, a significant proportion of patients with MA revert to NA. In the Joslin Clinic Study, patients who had type 1 diabetes and MA and reverted to NA had lower HbA1c levels (8.8 ± 1.5%) than the nonregressors (9.3 ± 1.6%; P < 0.001), and, compared with a HbA1c reference value of 
10.0%, the likelihood of regression increased with each 1% decrease in the HbA1c value (21). Thus, in patients with HbA1c <8.0%, the adjusted hazard ratio was reduced by 1.9-fold (1.2- to 2.9-fold) versus the reference group (21). A similar trend was seen in the smaller Steno cohort study (P = 0.09) (18). Although there was no HbA1c difference in patients who had type 1 diabetes and still had MA versus those who reverted to NA in the EURODIAB Study (22), patients who had MA that progressed to P had higher HbA1c levels. HbA1c values during follow-up in the Steno type 2 diabetes study were predictive of return to NA (34). Similarly, HbA1c values <6.95% were associated with more frequent reversal to NA in Japanese patients with type 2 diabetes (35). Taken together, these studies support the view that measures of glycemia provide added information to AER in the estimate of DN risk for development and progression and on the potential for reversal.
|
BP |
|---|
As already noted, reversion from MA to NA is much more common than initially anticipated. However, clinic BP values did not predict which patients with MA and type 1 diabetes would have NA on subsequent follow-up in the EURODIAB Study (22); SBP <115 mmHg was associated with a 1.4-fold (1.0- to 1.9-fold) likelihood of return to NA in another study of patients with type 1 diabetes (21), and this was strongly confirmed in the Steno type 1 diabetes cohort study (18). Moreover, return from MA to NA was more common in Japanese patients with type 2 diabetes and SBP <129 mmHg (35). In the Steno 2 study, initiation of antihypertensive therapy during follow-up was one of the significant predictors of reversion to NA in patients with type 2 diabetes (34). Taken together, these studies support the careful measure of systemic BP in patients with type 1 diabetes and type 2 diabetes, perhaps especially if normotensive by current standards. It is reasonable to recommend annual 24-h ABP measures for these "normotensive" patients to assist in the assessment to DN risk.
|
Albuminuria |
|---|
In a small study of 75 adolescents with type 1 diabetes, Lurbe et al. (58) found no predictive value of AER in the "normoalbuminuric" range. However, in a large study of 170 children, adolescents, and young adults with an average of approximately 9 yr of type 1 diabetes, we found that baseline AER within the NA range was higher in those who had MA versus those who had NA at the end of this 5-yr natural history study (64). Moreover, the group with persistent MA during the study (i.e., at least two of three consecutive values between 20 and 200 µg/min), whether they reverted to NA by the end of the study or not, had greater glomerular basement membrane (GBM) width at baseline. This indicated that higher AER within the NA range was associated with worse underlying DN lesions (64). Hovind et al. (18), in the Steno inception cohort study, found that AER was higher at the first assessment, 6 mo after onset of type 1 diabetes, in those who progressed to MA than in those who still had NA through a median of 18 yr (1.0 to 21.5 yr) of follow-up. The gap in AER values between the progressors and nonprogressors increased during follow-up studies before MA developed (18). In fact, those who progressed to MA had AER values that tended to increase, whereas those who still had NA had stable AER follow-up values (18). Larger epidemiologic studies in 1201 patients with type 1 diabetes also showed that baseline urinary albumin concentration above the normal range predicted the risk for progression to overt nephropathy in the subsequent 4 yr (2.5 to 5.5 yr) (65). Higher baseline AER in the NA range was also a predictor of progression to MA in patients with type 2 diabetes (66).
The absolute values of AER in the MA range were also evaluated as a risk predictor. These levels were nearly identical in patients who had type 1 diabetes and did or did not revert to NA in the study of Perkins et al. (21). Conversely, the cohort study of Hovind et al. (18) found lower AER values at final detection of MA in those who reverted to NA (29 µg/min [24 to 44 µg/min]) than in those who still had MA or progressed to P (41 µg/min [29 to 86 µg/min]; P = 0.01). Also the EURODIAB Progressive Complications Study found lower values for baseline AER in patients who had MA and type 1 diabetes and reverted to NA versus those who progressed to P, and this was an independent risk factor in multivariate analysis (22). Similarly, in patients with type 2 diabetes and MA, baseline AER values were predictive of the subsequent return to NA and progression to P (34).
These data are consonant with the idea that there is more to be gleaned from AER measurements than achieved by simple categorization and that absolute values and time trends within categories can provide additional prognostic precision. Given the approximately 40% day-to-day coefficient of variation in AER (67), these studies also suggest that this test should be done more frequently, and quarterly measures of AER are probably warranted.
The predictive value of AER studies in patients who are taking drugs that block the RAS system is unclear. Patients with MA and type 1 diabetes after 8 yr of angiotensin-converting enzyme inhibitor treatment had less increase in AER and less progression to P than patients who were given placebo (68). However, AER increased rapidly to above baseline values and, in several patients, into the P range within 2 mo of cessation of treatment (69). This suggested that progression in AER might have been masked by angiotensin-converting enzyme inhibitor treatment. In patients who had type 2 diabetes and MA and were treated with angiotensin receptor blocker (ARB) for 2 yr, a 1-mo hiatus in ARB therapy resulted in a disappearance in the AER-lowering effect of the lower dose of the ARB (150 mg/d irbesartan) and some loss of the effect of the 300 mg/d dose, albeit the significant difference from placebo remained at this higher dose (70). The BENEDICT trial (9) suggested a lower rate of MA development in patients who had NA and type 2 diabetes and were randomly assigned to ARB treatment, but the possibility of "masking" of AER increase by ARB therapy was not explored (9). RAS blockade did not increase the rate of remission from MA to NA in the type 1 diabetes study of Perkins et al. (21) but did so in the type 2 diabetes study of Araki et al. (35). In summary, it is possible that the prognostic value of AER levels may be influenced by treatment, but too little is known and much longer follow-up studies are critical. These factors should be kept in mind when interpreting AER values.
|
Total Intact and Immuno-Unreactive Albumin |
|---|
|
Retinopathy |
|---|
Given this background, it is not surprising that much stronger relationships between DN and DR than previously appreciated were found when patients with a broad spectrum of DN and DR were compared directly as to severity of respective eye and kidney lesions (60). Although serious DR lesions could occur in a subset of patients without clear DN lesions, this and the converse were, in fact, distinctly unusual (60). More recently, this research approach was applied to a cohort of 252 normotensive and normo- or hyperfiltering patients with NA and 12 ± 5 yr of type 1 diabetes duration (76). Eighty-nine patients had no DR, 136 had early nonproliferative DR, and 27 had moderate to severe nonproliferative DR or proliferative DR (76). The presence and the severity of DR lesions were strongly associated with DN glomerular lesions in these patients who had no renal functional abnormalities (76). Of interest, ABP parameters, which, as already noted, are related to underlying renal structure, were also related to DR in these patients with type 1 diabetes (Klein R, Mauer M, unpublished results).
The Steno cohort study found that the presence of simplex (77%) or proliferative (18%) DR antedated MA, whereas only 5% of the 79 patients who had NA and type 1 diabetes and progressed to MA had no DR (18). Conversely, of 198 patients who still had NA, 40% had no DR, 58% had simplex DR, and only 2% had proliferative DR (18). Patients with NA and proliferative DR had a nine-fold risk for MA; this risk was 1.3-fold in patients who had type 1 diabetes, NA, and simplex DR (18). The risk for MA was eight-fold less in those without DR (18). Whether DR status predicted subsequent reversion from MA to NA was not commented on in this paper (18) or in that of Perkins et al. (21). These studies in patients with type 1 diabetes suggest some common pathogenetic pathways for DR and DN and support the idea that the eye is a window on the kidney in type 1 diabetes.
The relationships between DR and DN are less well understood in patients with type 2 diabetes. Østerby et al. (77) found a correlation between diabetic glomerulopathy and DR lesions in patients with P and type 2 diabetes. However, discordance between DR and renal structure has been described in Japanese patients with type 2 diabetes (78). Mathian et al. (79) also showed that several patients with P and type 2 diabetes had normal fundi despite DN that was proved by kidney biopsy and that, when present, DR was predictive of nodular glomerulosclerosis, whereas the absence of DR was predictive of diffuse mesangial expansion lesions. We and others (80,81) reported that approximately 40% of patients with type 2 diabetes and MA or P had typical diabetic glomerulopathy, whereas the remaining patients had minimal structural changes or tubular, interstitial, vascular, and/or global glomerulosclerosis lesions that are advanced relative to the severity of DN glomerular lesions. When only those with diabetic glomerulopathy were considered, almost all patients in our cohort had DR. All patients with proliferative DR had typical diabetic glomerulopathy. The proportion of patients with type 2 diabetes and without DR was 44% among patients with MA and 25% among patients with P (Fioretto P, unpublished data). Thus, there is considerable discordance between DR and DN in type 2 diabetes when retinal structural abnormalities are compared with renal functional parameters. However, when key structural parameters of diabetic glomerulopathy (mesangial expansion and GBM thickening) were considered, we found a strong concordance between DR and DN lesions in patients with type 2 diabetes. The proportion of patients with DR was 81% in the group of patients with mesangial expansion, whereas it was only 38% in patients without mesangial expansion (Fioretto P, unpublished data). Longitudinal studies on the prognostic impact of DR on risk for progression of DN in patients with NA and MA and type 2 diabetes are lacking; it is known, however, that the presence of DR is associated with an increased risk of overt nephropathy in patients with type 2 diabetes and MA (34), and that the presence of DR predicts a faster loss of renal function in patients with type 2 diabetes and P (82). It therefore is likely that patients with DR are at higher risk to progress to overt nephropathy and that, as for type 1 diabetes, the eye is a window on the kidney in patients with type 2 diabetes.
|
Cigarette Smoking |
|---|
|
GFR |
|---|
Cross-sectional studies of Danish patients with type 2 diabetes found higher mean GFR values and frequency of hyperfiltration in patients with MA (92). However, in a longitudinal study of black patients with type 2 diabetes, in which the baseline prevalence of hyperfiltration was 15%, hyperfiltration did not predict deterioration of renal function (93). Silveiro et al. (66) found that increased GFR did not predict the subsequent development of MA in patients with type 2 diabetes, but patients with hyperfiltration had an increased GFR loss over the 5 yr of the study. This could have represented regression to the mean. In summary, the literature to date is mixed, and hyperfiltration does not consistently emerge as a strong predictor of progression from NA to MA or P, independent of glycemia and BP.
Reduced GFR may be a useful predictor of DN risk. The typical sequence for functional consequences of DN in type 1 diabetes is increasing AER and ABP values within the normal range followed by MA and overt hypertension. Proteinuria typically ushers in acceleration of GFR loss, culminating in uremia. The sequence in type 2 diabetes is more complex, with many patients already hypertensive at type 2 diabetes diagnosis and with a subset manifesting increases in AER without significant DN lesions (81). However, not all patients follow these patterns. We initially described a subset of patients, all women, who had type 1 diabetes patients and low GFR (<90 ml/min per 1.73 m2) despite NA (94). These patients had significantly worse DN lesions than women with NA and similar type 1 diabetes duration and normal to high GFR. It is interesting that most of these patients with low GFR and type 1 diabetes had self-selected a relatively low-protein diet (94). More recently, we confirmed these observations in a larger cohort of 105 patients with NA and type 1 diabetes, 23 (19 women, four men) of whom had GFR <90 ml/min per 1.73 m2 (95). These patients with low GFR much more often were female and had a higher incidence of hypertension and more severe DR than the normal or high GFR group and had significantly worse diabetic glomerulopathy lesions (95). These findings were independent of the use of RAS-blocking drugs.
Reduced GFR has also been described in patients with NA and type 2 diabetes (96). Using a GFR cutoff of 60 ml/min per 1.73 m2 in an older cohort of patients with type 2 diabetes (mean age 73 yr), the prevalence of low GFR among patients with NA was 23% (96). Also, the prevalence of NA was high (39%) among these patients with type 2 diabetes and reduced GFR (96). Compared with patients with P and similarly reduced GFR, patients with NA were older and more frequently female (58% in NA versus 18% in P) (96). These group differences for age and female gender remained significant after patients who did not have NA before RAS inhibitor therapy were excluded. So configured, 23% of patients with NA and long-term type 2 diabetes had low GFR, 65% of whom were female (96). This compared with 22% of patients with long-standing NA and type 1 diabetes, 83% of whom were female (95). In one report, 24% of the patients with NA, low GFR, and type 2 diabetes had AER values >20 µg/min when both immunoreactive and immuno-unreactive albumin were measured (97).
Given that GFR estimates from the application of serum creatinine formulas to patients with diabetes are, at best, imprecise and, at worst, misleading (98), other means of estimating GFR need to be considered. Serum cystatin C levels were found to be a better approximation of formal GFR measures than serum creatinine values in several studies (99–103) in patients with type 1 diabetes and type 2 diabetes. Alternatively, plasma clearance of 51Cr-EDTA or iohexol (104,105) may be more convenient as direct measures of GFR than methods that involve timed urine collections. The important point here is that there are many exceptions to the concept that clinical manifestations of DN are lock step, increased AER followed by hypertension, and then GFR decline. Given the magnitude of the clinical problem of DN and high risks that patients with DN face, we are no longer justified in using inadequate GFR estimates in these patients, especially in certain subsets, such as women with NA, long-standing diabetes, retinopathy, and/or hypertension. The time and money required for more precise GFR estimates, in our view, is warranted. Otherwise, to take the analogy to DR one step further, not only are we using renal functional tests to diagnose and follow DN while ophthalmologists follow retinal structure, but we also are using poor renal function tests, such as inaccurate GFR estimates, limited value clinic BP, and AER with all of its inherent problems as discussed above.
|
Renal Biopsy |
|---|
No studies have identified structural predictors of progression from NA to MA in patients with type 2 diabetes. However, studies in Pima Indians with MA and type 2 diabetes found that reduced baseline podocyte number predicted greater increase in AER after 4 yr of follow-up (r = 0.57, P = 0.02) (107). Baseline Vv(Mes/glom) had a similar predictive value (r = 0.48, P = 0.06) for the AER increase over this time (107). The odds ratio for having progressive GFR loss in white patients with type 2 diabetes and abnormal AER increased across the quartiles of electron microscopy (EM) morphometrically measured GBM width and Vv(Mes/glom) (108). Baseline AER levels did not predict subsequent changes in GFR in this study (108), perhaps because several patients had MA without significant DN glomerular lesions.
Light Microscopy
Although the above discussion focuses on quantitative EM data, it should be noted that the experienced reader of light microscopy (LM) can develop semiquantitative estimates of mesangial (109) and interstitial (110) expansion, which closely approach unbiased morphometric measurements of these parameters. Also, severity of arteriolar hyalinosis lesions and global glomerulosclerosis are closely related, and the latter is associated with GFR loss (111). Given the greater heterogeneity of lesions in type 2 diabetes versus type 1 diabetes (see the Retinopathy section), LM evaluations in type 2 diabetes can be very helpful in predicting outcomes. Thus, patients with typical diabetic nephropathology (balanced severity of glomerular, tubular, interstitial, global glomerulosclerosis, and vascular lesions) have more rapid GFR decline than the patients with less advanced diabetic glomerulopathy relative to other renal lesions (Fioretto P, unpublished observations). LM can also be helpful in diagnosing diseases other than or in addition to DN. However, at the earlier stages of DN, when LM may look normal, there are, unfortunately, no substitutes for unbiased EM morphometric studies.
In summary, renal biopsies that are done strictly for research purposes, given adequate diabetes duration, are renal functional outcome predictors in both type 1 diabetes and type 2 diabetes. It should be emphasized, however, that studies of patients who undergo biopsy for clinical indications (i.e., because these patients were following what was considered to be an atypical clinical course) have not been included here because of the strong possibility of selection bias.
|
Lipids |
|---|
In summary, the measurement of plasma lipid levels adds to the prognostic value of AER measures, and this association is consistent with a role for lipid abnormalities in DN pathogenesis.
|
Conclusion |
|---|
|
|
|
|
|
Although AER is important and should be measured more frequently, there are gains to be made in predictive precision in considering family history, smoking habits, glycemia, retinopathy status, GFR, BP including ABP recordings, lipid levels, and the possible influences of treatment on AER. We also recognize that current information is inadequate for the appropriate weighting of these additional DN parameters for risk assessment algorithms to be developed. It is hoped, in fact, that this discussion will stimulate cooperative merging of appropriate longitudinal databases to facilitate the emergence of such clinical tools that, we predict, would improve patient outcomes and clinical trial design. In the mean time, the clinician will need to apply a more qualitative thought process to integrate these predictive variables.
We also recognize that we need better biomarkers and risk predictors. Cystatin C as a surrogate measure of GFR and total intact urinary albumin as an early marker of altered permselectivity and increased risk deserve further exploration. Also promising are inflammatory markers; urinary proteomics; and the genetic, molecular, and cellular approaches outlined elsewhere in this issue (112). Although these newer approaches may improve DN risk prediction, the application of the integrated approach outlined here will, we believe, remain valuable.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
Top
Abstract
Introduction