Blog entry by Meguid El Nahas
Intensive glucose control improves kidney outcomes in patients with type 2 .
Kidney International (2013) 83, 346–348; doi:10.1038/ki.2012.431
Intensive glycemic control in type 2 diabetics at high cardiovascular risk: do the benefits justify the risks?
- 1Department of Medicine, University of Alberta, Edmonton, Canada
- 2Alberta Kidney Disease Network, Canada
Correspondence: Marcello Tonelli, Department of Medicine, University of Alberta, Alberta Kidney Disease Network, 7-129 Clinical Science Building, Edmonton, Alberta T6B 2G3, Canada. E-mail: email@example.com
Perkovic et al. use novel data from the ADVANCE study to report on the potential renal benefits of standard glycemic control, compared with intensive glycemic control (mean hemoglobin A1c 7.3 and 6.5%, respectively). Intensive glycemic control reduced the risk of new-onset microalbuminuria, new-onset macroalbuminuria, and progression of albuminuria. The risk of end-stage renal disease was also reduced in patients treated with intensive glycemic control, although the number of events was small.
Most guidelines, including those from the Kidney Disease Outcomes Quality Initiative (K/DOQI) (National Kidney Foundation), suggest that glycemic control is an important clinical objective for all diabetic patients with and without chronic kidney disease (CKD). These guidelines recommend a target hemoglobin A1c of approximately 7.0% ‘to prevent or delay complications of diabetes, including diabetic kidney disease,’ noting that more intensive treatment improves albuminuria, but evidence for any effect on loss of glomerular filtration rate (GFR) is sparse.1
Perkovic et al.2 (this issue) explore the potential benefits of intensive glycemic control for renal outcomes, using a post hoc analysis of the ADVANCE trial. ADVANCE3 randomly assigned 11,140 patients to standard glycemic control following local guidelines versus intensive glycemic control (target A1c6.5%). Included patients had type 2 diabetes (average duration 8 years) and were more than 55 years old (average age 66 years). Only patients at high risk were included, based on a history of major macrovascular disease, microvascular disease (overt nephropathy or retinopathy), or one major cardiovascular risk factor. After a median duration of 5 years, mean A1c was 7.3 vs. 6.5%, respectively, in the two groups. There was no difference in the risk of macrovascular events between groups (hazard ratio (HR) 0.94, 95% confidence interval (CI) 0.84–1.06, P=0.32). However, patients who were in the intensive glycemic control group had fewer microvascular events (HR 0.86, 95% CI 0.77–0.97, P=0.01), primarily due to a 21% reduction in ‘new or worsening nephropathy’ (HR 0.79, 95% CI 0.66–0.93, P=0.006); there was no effect on retinopathy.
The new analysis by Perkovic et al.2 begins by providing us with more insight into the prevalence of preexisting renal disease in the 11,140 ADVANCE participants. At baseline, approximately 27% of patients had microalbuminuria (an inclusion criterion) but only 3.6% had macroalbuminuria. Most patients had no CKD (55%), while CKD stages 2 and 3 was present in 15% and 19% of patients, respectively. Advanced CKD (stages 4 and 5) was present in 0.5% of all patients. As expected based on the main finding of ADVANCE, outcomes related to proteinuria appeared more favorable with intensive glycemic control: new-onset microalbuminuria (33.5 vs. 36.3%), new-onset macroalbuminuria (3.0 vs. 4.3%), and progression of albuminuria by 1 stage (23.3 vs. 25.3%) (allP<0.012). Patients treated with intensive glycemic control had more regression of albuminuria by 1 stage (61.2 vs. 56.3%) and more regression to normoalbuminuria (56.3 vs. 50.2%) (all P0.002).
The authors acknowledge that albuminuria is of questionable reliability as a surrogate marker for renal outcomes and appropriately focus on the risk of end-stage renal disease (ESRD). In these new analyses, the risk of ESRD was significantly reduced in patients treated with intensive glycemic control (vs. standard treatment) (HR 0.35, 95% CI 0.15–0.83, P=0.017). Furthermore, patients with preexisting renal disease seemed to derive more benefit from intensive glycemic control as reflected by a lower number needed to treat (NNT): the NNT was 152 for any albuminuria, 147 for estimated GFR <60 ml/min, 85 for estimated GFR <60 ml/min per1.73 m2 with any albuminuria, and 41 for macroalbuminuria (irrespective of GFR). These findings were consistent in various subgroups, including participants with baseline A1c above or below median (7.2%), with or without retinopathy, in both assigned blood pressure treatment groups (ADVANCE was a two-by-two factorial trial of glycemic and blood pressure control), both men and women, and with age above or below median. On the basis of these data, the authors suggest that intensive glycemic control (presumably A1c <6.5% as targeted in ADVANCE) may be a useful strategy to prevent the development of ESRD in patients with type 2 diabetes.
Before widespread adoption of such a strategy, the reader should consider some important aspects of this post hoc analysis. First, despite an apparent reduction in the risk of ESRD with intensive glycemic control, there was no significant effect on serum creatinine over time—only a non-significant trend toward more frequent doubling of serum creatinine (to >200 μmol/l) in intensively treated patients (HR 1.15, P=0.42). The authors propose that doubling of serum creatinine may be an imprecise ‘surrogate’ for progression of diabetic nephropathy to ESRD, as it may capture patients suffering acute kidney injury due to sepsis, shock, and so on. In support of this hypothesis, there was a non-significant trend toward lower risk of sustained doubling of serum creatinine with intensive glycemic control (HR 0.83,P=0.38). Nonetheless, the possibility remains that other factors besides intensive glycemic control per se contributed to the apparent reduction of the risk of ESRD in the treatment group. For example, patients in the (unblinded) intensive treatment arm might have been observed more closely by their treating physician—which in turn could have reduced the risk of acute kidney injury or its consequences.
Second, the number of patients who developed ESRD during ADVANCE was exceedingly low (27 events in 11,140 patients=0.24%). This contributed to the high NNT (445 patients to prevent one case of ESRD over 5 years)—although the NNT was lower in patients with more advanced CKD at baseline, likely because of the higher absolute risk in this group. Despite the high quality of the analyses, the small number of events may reduce confidence in the findings.
The results of ADVANCE and the current analysis by Perkovic et al.2 must be interpreted in the context of other randomized controlled trials that have assessed the impact of intensive glycemic control on clinically relevant kidney outcomes. ACCORD was similar to ADVANCE and enrolled 10,251 patients with type 2 diabetes and high cardiovascular risk (>40 years old with known cardiovascular disease, or >55 years old with anatomical evidence of significant atherosclerosis, albuminuria, left ventricular hypertrophy, or two cardiovascular risk factors), with randomization to conventional glycemic control (A1c 7.0–7.9%) versus an even more intensive regimen targeting A1c <6%.4 This trial was terminated early (after 3.5 years) because of increased mortality in intensively treated patients. However, a subsequent analysis reported on renal end points at trial’s end.5 Intensive glycemic control resulted in lower A1c at one year (median 6.4 vs. 7.5%) and, similarly to ADVANCE, resulted in a 20–30% reduction in the risk of new-onset micro- and macroalbuminuria, but no reduction in the risk of doublings in serum creatinine (in fact, a significant increase: HR 1.07, P=0.016)—and, in contrast to the results reported by Perkovic et al.,2 no decrease in ESRD (HR 0.95, 95% CI 0.73–1.24, P=0.71). The most recent randomized trial of intensified glycemic control, VADT, was published in 2009 and enrolled 1791 military veterans with long-standing type 2 diabetes (mean duration 11.5 years), 40% of whom had known cardiovascular disease.6 Patients were randomized to standard therapy versus intensified glycemic control to decrease A1c by 1.5%; A1c between groups was 8.4 vs. 6.9%. After a median of 5.6 years, there was no difference in the risk of mortality or microvascular end points, other than a reduced risk of progression of albuminuria; the risks of doubling of serum creatinine and stage 5 CKD were similar between groups (P=0.99 and P=0.35, respectively).
What can we conclude about the effect of glycemic control on diabetic nephropathy in type 2 diabetes, and, more broadly, on patient survival and cardiovascular events? We believe that it is reasonable and generally safe to target an A1c of 7%. The UK Prospective Diabetes Study (UKPDS) showed that early, more intensive glycemic control (A1c of 7.0 vs. 7.9%) in patients with newly diagnosed type 2 diabetes safely reduced microalbuminuria and doubling of serum creatinine (as well as retinopathy).7
So, should clinicians routinely target an A1c of 6–7% or lower? As discussed, high-quality data from ACCORD, VADT, and ADVANCE all demonstrate that this strategy will improve proteinuria-based surrogate outcomes, but only the post hoc analysis of ADVANCE by Perkovic et al.2 suggests that such an intensive strategy may reduce the clinically relevant outcome of ESRD. However, adopting an intensive glycemic control strategy may also pose risks to patients. A1ctargets below 6.5% led to increased mortality (largely due to myocardial infarction) in ACCORD and had no significant effect on cardiovascular events or mortality in ADVANCE.
One may speculate that patients enrolled in the latter two trials, most of whom had established type 2 diabetes and major cardiovascular risk factors, were more susceptible to the adverse consequences of hypoglycemia, as opposed to the younger, healthier UKPDS participants. Similarly, observational data suggest that achieved A1c<6.5% is associated with excess mortality in patients with diabetes and established CKD.8
Thus, intensive glycemic control appears to have both risks and benefits—and despite the important findings of Perkovic et al.,2 this strategy cannot be broadly recommended at present. Current data do not allow clinicians to confidently identify patients in whom the risk-to-benefit ratio of tighter glycemic control is especially favorable. Until such data are available, we suggest that an A1c target <6.5% for type 2 diabetes should be used cautiously, if at all—perhaps only in well-informed patients who are younger, at lower risk for hypoglycemia, and free of symptomatic cardiovascular disease.
This is an excellent and balanced review of the recent publication on a posthoc analysis of ADVANCE putting in the context of other intensive v conventional glycemia control stuides in T2DM.
It highlights the facts that in high risk T2DM patients:
1. Intensive glycemia control with HbA1c <7% is either associated with no CVD benefit or increase mortality (ACCORD)
2. Intensive glycemia control has NO effect on renal HARD ENDPOINTS such as decline of GFR or incidence of ESRD
3. Observational studeis also suggest increased mortality with HbA1c<6.5%
The study commenetd upon by Perkovic et al also highlights a number of issues:
1. The concern about endless mining of data to find positive results in posthoc analyses. These are at best hypothesis generating (difficult when considerable data suggest the opposite...) or at worst futile and misleading excercises.
2. The distinction between statistical analysis and the true clinical value of such observations; p vale <0.05 but number needed to treat to prevent 1 ESRD 445 patients over 5 years!!!!!
3. The use of serum creatinine as a marker of progressive CKD/DN, in elderly patients with CVD and a tendency to sarcopenia; thus dissociating further changes in sCr = eGFR and true measured GFR and true progression of Diabetic nephropathy. Intensive glycemia control with its induced side effecst and increased morbidity may be associated with a fall in sCr hence the apparently stable sCR in the Perkovic study in spite of possible worsening of true GFR/kidney function....?!
4. The use of ESRD in the absence of measured GFR as a hard endpoint; this can be misleading and observer baised as the decision to start RRT often has subjective elements to it; a good example of such dissociation was seen in REIN studies where in patients with reduced GFR<45; Ramipril had NO effect on measured GFR decline but decreased the number of those reaching ESRD....!!
5. Glycemia control may improve/lower sCr through improved on the impaired tubular secretion of Cr observed and associated with DM. http://www.ncbi.nlm.nih.gov/pubmed/15882297
6. The disconnect between reduction of albuminuria and progression of diabetic nephropathy/CKD; most of the studies showing that intensive glycemia control benefit albuminuria but NOT progression of kidney function decline. This may reflect that glycemia control can affect albuminuria in many ways unconnected to slowing the decline in GFR.
7. Lowering glycemia can improve urianry albumin excretion in the following way:
a. Increasing CVD morbidity and decreasing protein intake due to poor health; this would in turn reduce albuminuria that is often proportional to the protein intake.
b. Affecting glycation and charge of albumin which in turn decrease its filtration and reabsortion rates. http://www.ncbi.nlm.nih.gov/pubmed/9187409
c. Improving peritubular circulation an dimproving proximal tubular reabsorption of albumin; many beleive that microalbuminuria in DM is a reflection of vascular and fall in peritubular capillary perfusion impacting/decreasing proximal tubular reabsorption of albumin.
IT IS MISLEADING TO CLAIM PROTECTION FROM PROGRESSIVE DIABETIC NEPHROPATHY THROUGH THE REDUCTION IN ALBUMINURIA AND UNRELATED BIOMARKER.
IT IS MISLEADING TO EQUATE CHANGES IN SCR INDEPENDENTLY OF THE HARD ENDPOINT OF MEASURING GFR.
START OF RRT/ESRD IS DIFFICULT TO INTERPRET IN THE ABSENCE OF HARD DATA RELATING TO TEH RATE OF DECLINE OF KIDNEY FUNCTION.