Circulating bone morphogenetic protein-7 and transforming growth factor-β1 are better predictors of renal end points in patients with type 2 diabetes mellitus
2012; Elsevier BV; Volume: 83; Issue: 2 Linguagem: Inglês
10.1038/ki.2012.383
ISSN1523-1755
AutoresMuh Geot Wong, Vlado Perkovic, Mark Woodward, John Chalmers, Qiang Li, Graham S. Hillis, Dania Yaghobian Azari, Min Jun, Neil R Poulter, Pavel Hamet, Bryan Williams, Bruce Neal, Giuseppe Mancia, Mark E. Cooper, Carol A. Pollock,
Tópico(s)Dialysis and Renal Disease Management
ResumoAlbuminuria and a reduced glomerular filtration rate are conventional predictors of a future decline in kidney function in patients with type 2 diabetes mellitus. Using a nested case–control study we assessed whether circulating transforming growth factor-β1 (TGF-β1) and bone morphogenetic protein-7 (BMP-7) levels more accurately predict renal end points than the conventional markers. Cases were defined as those who developed a renal end point (doubling of serum creatinine to at least 200μmol/l, the need for renal replacement therapy, or death due to renal disease) during the study. Using propensity scoring, two controls were selected for each of 281 cases. Participants who developed renal end points had significantly higher total TGF-β1, lower BMP-7 levels, and a higher total TGF-β1 to BMP-7 ratio at baseline. A graded increase in risk was found in individuals with lower BMP-7 levels (odds ratio 24.07, for the lowest to the highest tertile), or significantly higher TGF-β1 levels (odds ratio for the highest to the lowest tertile, 8.43). The area under the receiver operating characteristic curve (c-statistic) for the conventional predictors was 0.73. Using BMP-7 and total and active TGF-β1, the c-statistic was 0.94 (significantly higher to conventional predictors). Thus, our results suggest these novel kidney markers are better predictors of renal progression than the conventional predictors in patients with type 2 diabetes mellitus. Albuminuria and a reduced glomerular filtration rate are conventional predictors of a future decline in kidney function in patients with type 2 diabetes mellitus. Using a nested case–control study we assessed whether circulating transforming growth factor-β1 (TGF-β1) and bone morphogenetic protein-7 (BMP-7) levels more accurately predict renal end points than the conventional markers. Cases were defined as those who developed a renal end point (doubling of serum creatinine to at least 200μmol/l, the need for renal replacement therapy, or death due to renal disease) during the study. Using propensity scoring, two controls were selected for each of 281 cases. Participants who developed renal end points had significantly higher total TGF-β1, lower BMP-7 levels, and a higher total TGF-β1 to BMP-7 ratio at baseline. A graded increase in risk was found in individuals with lower BMP-7 levels (odds ratio 24.07, for the lowest to the highest tertile), or significantly higher TGF-β1 levels (odds ratio for the highest to the lowest tertile, 8.43). The area under the receiver operating characteristic curve (c-statistic) for the conventional predictors was 0.73. Using BMP-7 and total and active TGF-β1, the c-statistic was 0.94 (significantly higher to conventional predictors). Thus, our results suggest these novel kidney markers are better predictors of renal progression than the conventional predictors in patients with type 2 diabetes mellitus. Diabetic nephropathy is the most common cause of end-stage kidney disease (ESKD) in most of the developed and many developing countries.1.Wild S. Roglic G. Green A. et al.Global prevalence of diabetes: estimates for the year 2000 and projections for 2030.Diabetes Care. 2004; 27: 1047-1053Crossref PubMed Scopus (11954) Google Scholar At present, albuminuria and reduced glomerular filtration rate (GFR) are the best available markers, and predict the risk of future decline in kidney function in patients with type 2 diabetes mellitus. However, albuminuria does not uniformly detect early disease,2.Ruggenenti P. Fassi A. Ilieva A.P. et al.Preventing microalbuminuria in type 2 diabetes.N Engl J Med. 2004; 351: 1941-1951Crossref PubMed Scopus (905) Google Scholar and recent reports have described a significant proportion of patients with diabetes mellitus who progress to stage 5 chronic kidney disease (CKD) with normal or low levels of albuminuria.3.Caramori M.L. Fioretto P. Mauer M. Low glomerular filtration rate in normoalbuminuric type 1 diabetic patients: an indicator of more advanced glomerular lesions.Diabetes. 2003; 52: 1036-1040Crossref PubMed Scopus (225) Google Scholar,4.Fioretto P. Stehouwer C.D. Mauer M. et al.Heterogeneous nature of microalbuminuria in NIDDM: studies of endothelial function and renal structure.Diabetologia. 1998; 41: 233-236Crossref PubMed Scopus (88) Google Scholar Furthermore, the presence of albuminuria and impaired renal function usually indicate established nephropathy where current therapy is at best only partially effective.5.Lewis E.J. Hunsicker L.G. Bain R.P. et al.The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. The Collaborative Study Group.N Engl J Med. 1993; 329: 1456-1462Crossref PubMed Scopus (5024) Google Scholar,6.Brenner B.M. Cooper M.E. de Zeeuw D. et al.Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy.N Engl J Med. 2001; 345: 861-869Crossref PubMed Scopus (6163) Google Scholar In the RENAAL (Reduction of Endpoints in NIDDM with the Angiotensin II Antagonist Losartan) study in people with established proteinuric diabetic nephropathy, treatment with losartan reduced the risk of ESKD by 28%, but only delayed progression to ESKD by 4–6 months.7.Vilayur E. Harris D.C. Emerging therapies for chronic kidney disease: what is their role?.Nat Rev Nephrol. 2009; 5: 375-383Crossref PubMed Scopus (65) Google Scholar Therefore, there is a need for biomarkers that can predict renal outcomes earlier in patients with diabetes mellitus so as to allow better identification and deployment of preventative therapies for those at high risk. Transforming growth factor-β1 (TGF-β1) and bone morphogenetic protein-7 (BMP-7) are secretory cytokines belonging to the TGF-β superfamily.8.Allendorph G.P. Vale W.W. Choe S. Structure of the ternary signaling complex of a TGF-beta superfamily member.Proc Natl Acad Sci USA. 2006; 103: 7643-7648Crossref PubMed Scopus (207) Google Scholar TGF-β1 is well known for its pro-fibrotic role in many renal diseases including diabetic nephropathy.9.Sharma K. Jin Y. Guo J. et al.Neutralization of TGF-beta by anti-TGF-beta antibody attenuates kidney hypertrophy and the enhanced extracellular matrix gene expression in STZ-induced diabetic mice.Diabetes. 1996; 45: 522-530Crossref PubMed Scopus (0) Google Scholar,10.Ziyadeh F.N. Sharma K. Ericksen M. et al.Stimulation of collagen gene expression and protein synthesis in murine mesangial cells by high glucose is mediated by autocrine activation of transforming growth factor-beta.J Clin Invest. 1994; 93: 536-542Crossref PubMed Scopus (575) Google Scholar Increased expression of TGF-β1 has been demonstrated in renal biopsies from patients with established diabetic nephropathy.11.Sharma K. Ziyadeh F.N. Alzahabi B. et al.Increased renal production of transforming growth factor-beta1 in patients with type II diabetes.Diabetes. 1997; 46: 854-859Crossref PubMed Scopus (258) Google Scholar In patients with type 2 diabetes mellitus, circulating total and active TGF-β1 have been shown to be elevated when compared with the nondiabetic population.11.Sharma K. Ziyadeh F.N. Alzahabi B. et al.Increased renal production of transforming growth factor-beta1 in patients with type II diabetes.Diabetes. 1997; 46: 854-859Crossref PubMed Scopus (258) Google Scholar,12.Hellmich B. Schellner M. Schatz H. et al.Activation of transforming growth factor-beta1 in diabetic kidney disease.Metabolism. 2000; 49: 353-359Abstract Full Text PDF PubMed Scopus (44) Google Scholar However, these studies had small sample sizes and were unable to ascertain whether circulating active and/or total TGF-β1 could be used as biomarkers to predict patients at risk of developing diabetic nephropathy. BMP-7 is recognized as a natural antagonist to TGF-β1, with antifibrotic and anti-inflammatory properties.13.Zeisberg M. Bone morphogenic protein-7 and the kidney: current concepts and open questions.Nephrol Dial Transplant. 2006; 21: 568-573Crossref PubMed Scopus (71) Google Scholar, 14.Gould S.E. Day M. Jones S.S. et al.BMP-7 regulates chemokine, cytokine, and hemodynamic gene expression in proximal tubule cells.Kidney Int. 2002; 61: 51-60Abstract Full Text Full Text PDF PubMed Scopus (197) Google Scholar, 15.Wang S. Hirschberg R. BMP7 antagonizes TGF-beta-dependent fibrogenesis in mesangial cells.Am J Physiol Renal Physiol. 2003; 284: F1006-F1013Crossref PubMed Scopus (37) Google Scholar, 16.Zeisberg M. Hanai J. Sugimoto H. et al.BMP-7 counteracts TGF-beta1-induced epithelial-to-mesenchymal transition and reverses chronic renal injury.Nat Med. 2003; 9: 964-968Crossref PubMed Scopus (1184) Google Scholar, 17.Zeisberg M. Bottiglio C. Kumar N. et al.Bone morphogenic protein-7 inhibits progression of chronic renal fibrosis associated with two genetic mouse models.Am J Physiol Renal Physiol. 2003; 285: F1060-F1067Crossref PubMed Scopus (257) Google Scholar Loss of tubular BMP-7 is seen in progressive diabetic nephropathy,18.Wang S.N. Lapage J. Hirschberg R. Loss of tubular bone morphogenetic protein-7 in diabetic nephropathy.J Am Soc Nephrol. 2001; 12: 2392-2399Crossref PubMed Google Scholar and exogenous administration of recombinant human BMP-7 or induced overexpression of BMP-7 in transgenic animals with diabetes mellitus have been shown to retard the progression of renal disease.19.Wang S. Chen Q. Simon T.C. et al.Bone morphogenic protein-7 (BMP-7), a novel therapy for diabetic nephropathy.Kidney Int. 2003; 63: 2037-2049Abstract Full Text Full Text PDF PubMed Scopus (207) Google Scholar,20.Wang S. de Caestecker M. Kopp J. et al.Renal bone morphogenetic protein-7 protects against diabetic nephropathy.J Am Soc Nephrol. 2006; 17: 2504-2512Crossref PubMed Scopus (151) Google Scholar Despite the suggestion that BMP-7 may be useful as a treatment for CKD, convincing clinical data are lacking.21.Dudas P.L. Argentieri R.L. Farrell F.X. BMP-7 fails to attenuate TGF-beta1-induced epithelial-to-mesenchymal transition in human proximal tubule epithelial cells.Nephrol Dial Transplant. 2009; 24: 1406-1416Crossref PubMed Scopus (47) Google Scholar Conversely, it is not clear whether an endogenous reduction in BMP-7 may predispose to future progressive nephropathy. The ADVANCE (Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation) Collaborative Group study was a factorial, randomized, double-blinded, controlled trial involving 11,140 patients with type 2 diabetes mellitus conducted at 215 collaborating centers in 20 countries from Asia, Australasia, Europe, and North America.22.Patel A. MacMahon S. Chalmers J. et al.Effects of a fixed combination of perindopril and indapamide on macrovascular and microvascular outcomes in patients with type 2 diabetes mellitus (the ADVANCE trial): a randomised controlled trial.Lancet. 2007; 370: 829-840Abstract Full Text Full Text PDF PubMed Scopus (1805) Google Scholar,23.Patel A. MacMahon S. Chalmers J. et al.Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes.N Engl J Med. 2008; 358: 2560-2572Crossref PubMed Scopus (44) Google Scholar We hypothesized that high serum TGF-β1 levels and low BMP-7 levels would indicate a profibrotic propensity that would be associated with an increased risk of progressive kidney disease. This study therefore aimed to assess baseline circulating levels of TGF-β1 (total and active) and BMP-7 in a subgroup of participants in the ADVANCE study, and to assess their predictive value for major kidney outcomes. The demographic and clinical baseline characteristics of study participants are shown in Table 1. The mean age, blood pressure, hemoglobin A1c (HbA1c), body mass index, and gender distribution were similar in those who developed major renal end points and those who did not (Table 1A). Among the 102 cases with blood samples available for analysis, 77 had a doubling of serum creatinine to a level of at least 200μmol/l, 14 required chronic renal replacement therapy, and 21 died from renal disease (Supplementary Table S1). As expected, the cases had higher baseline serum creatinine levels and a correspondingly lower estimated GFR (eGFR), as compared with the controls (both P<0.0001, Table 1B). The baseline urinary albumin/creatinine ratio values were also higher in the cases than the controls, as were diabetes duration and the proportion with evidence of microvascular disease at enrollment. Three (1%) of TGF-β1 values and 24 (9%) of BMP-7 values were below the detectable limits of 32 and 0.79pg/ml, respectively. In addition, one value of BMP-7 was missing because of inadequate serum for analysis.Table 1Characteristics of major renal cases and controls at baselineVariableCases (n=102)Controls (n=179)P-valueA. Clinical characteristicsMean (s.d.) Age (years)69.2 (6.9)68.9 (6.7)0.75 Systolic blood pressure (mmHg)155.3 (25.7)155.2 (23.8)0.98 Hemoglobin A1c (%)7.6 (1.3)7.5 (1.8)0.61 Body mass index30.2 (5.9)30.1 (5.0)0.92 Diabetes duration (years)10.4 (7.1)7.9 (6.1)0.002N (%) Gender (female)31 (30.4)54 (30.2)0.97 Past history of macrovascular disease (yes)40 (39.2)53 (29.6)0.010 Past history of microvascular disease (yes)28 (27.5)27 (15.1)0.012B. Established kidney markersMean (s.d.) Creatinine (µmol/l)126.1 (77.9)90.9 (19.9)<0.0001 eGFR (MDRD; ml/min per 1.73m2)54.3 (19.8)69.4 (16.6)<0.0001 eGFR (CKD-EPI; ml/min per 1.73m2)55.1 (19.8)70.7 (15.8)<0.0001Median (Q1, Q3) UACR (µg/mg)43.3 (15.0, 225.0)13.3 (6.4, 8.1)<0.0001C. Novel kidney markersMedian (Q1, Q3) Total TGF-ß1 (pg/ml)13,587 (8066, 18,837)6858 (4218, 11,760)<0.0001 BMP-7 (pg/ml)7.5 (2.0, 11.4)19.3 (9.8, 69.9)<0.0001 Total TGF-ß1/BMP-7 ratio1744.3 (917.0, 2623.1)277.8 (85.1, 719.8)<0.0001 Active TGF-ß1 (pg/ml)55.9 (17.4, 120.0)0 (0, 29.3)<0.0001Abbreviations: BMP-7, bone morphogenetic protein-7; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; MDRD, Modification of Diet in Renal Disease; Q1, lowest quartile; Q3, highest quartile; TGF-β1, transforming growth factor-β1; UACR, urinary albumin/creatinine ratio. Open table in a new tab Download .doc (.07 MB) Help with doc files Supplementary Tables 1–3 Abbreviations: BMP-7, bone morphogenetic protein-7; CKD-EPI, Chronic Kidney Disease Epidemiology Collaboration; eGFR, estimated glomerular filtration rate; MDRD, Modification of Diet in Renal Disease; Q1, lowest quartile; Q3, highest quartile; TGF-β1, transforming growth factor-β1; UACR, urinary albumin/creatinine ratio. Individuals who developed renal end points had a higher median (Q1, Q3) circulating total TGF-β1 level of 13,587 (8066, 18,837) pg/ml when compared with those who did not reach the defined renal end points, 6858 (4218, 11,760) pg/ml, P<0.0001 (Table 1C). Conversely, the median circulating BMP-7 levels were significantly lower in those who reached renal end points at 7.5 (2.0, 11.4) pg/ml versus 19.3 (9.8, 69.9) pg/ml (P<0.0001). Hence, the ratio of total TGF-β1/BMP-7 was higher in individuals with type 2 diabetes mellitus who progressed to a renal end point (P<0.0001). As with most biologically active cytokines, active TGF-β1 is generally undetectable or very low in the absence of active kidney disease. The median active TGF-β1 levels in the cases was 55.9 (17.4, 120.0) pg/ml and 0 (0, 29.3) pg/ml in the controls. Over half of the controls had undetectable circulating active TGF-β1 (58%), whereas the majority of cases had detectable circulating active TGF-β1 (93%, P<0.0001). As the circulating levels of the defined biomarkers were not normally distributed, analyses were performed by grouping patients into thirds according to the baseline circulating levels of each kidney marker. Because of the low circulating active TGF-β1 levels, the population was divided into those with detectable versus undetectable active TGF-β1. We then compared the odds of cases and controls developing renal end points between each third. In the unadjusted model, individuals in the highest compared with the lowest third of total TGF-β1 levels had an odds ratio (OR) of 8.43 for developing major renal end points (95% confidence interval 4.03–17.67, P<0.0001), whereas the middle third had an intermediate risk (OR 2.28, 95% confidence interval 1.17–4.45, P=0.016). The middle and lowest thirds of BMP-7 also had a graded increase in the odds of renal end points compared with the highest third. A similar graded relationship was observed for the lowest and middle compared with the highest thirds of total TGF-β1/BMP-7 ratio (all P<0.01, Figure 1). Full adjustment for age, sex, eGFR, HbA1c, urinary albumin/creatinine ratio, body mass index, diabetes duration, randomized treatment interventions, history of macrovascular disease, or retinopathy did not significantly alter the results. A detectable level of circulating active TGF-β1 compared with undetectable circulating active TGF-β1 was associated with an OR of more than 16 for the development of a major renal outcome (OR=16.15, 95% confidence interval 6.46–40.41, P<0.0001). For every s.d., on the log scale, in total TGF-β1 or decrease in BMP-7, the OR for renal end points was increased by a factor of 3–6 (Table 2). For the TGF-β1/BMP-7 ratio, the OR for s.d. increase was over 12, after full adjustment.Table 2Odds ratios for major renal end points associated with a 1 s.d. difference in log-transformed values of the novel kidney markersUnadjusted modelAdjusted modelaAdjusted for sex, age, estimated glomerular filtration rate (eGFR; Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)), randomized treatment interventions, urinary albumin/creatinine ratio (UACR), hemoglobin A1c (HbA1c), body mass index (BMI), diabetes duration, and history of macrovascular or microvascular disease.EffectOdds ratio95% Confidence limitsOdds ratio95% Confidence limitsLog of total TGF-β1 (per s.d.=1.08 increase)3.402.165.353.892.077.32Log of BMP-7 (per s.d.=1.72 decrease)3.512.325.325.952.8212.66Log of total TGF-β1/BMP-7 ratio (per s.d.=2.08 increase)6.373.5011.5812.324.2535.69Abbreviations: BMP-7, bone morphogenetic protein-7; TGF-β1, transforming growth factor-β1.All P-values <0.0001.a Adjusted for sex, age, estimated glomerular filtration rate (eGFR; Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI)), randomized treatment interventions, urinary albumin/creatinine ratio (UACR), hemoglobin A1c (HbA1c), body mass index (BMI), diabetes duration, and history of macrovascular or microvascular disease. Open table in a new tab Abbreviations: BMP-7, bone morphogenetic protein-7; TGF-β1, transforming growth factor-β1. All P-values <0.0001. Adding total or active TGF-β1 or BMP-7 to the clinical parameters significantly improve the fit (Table 3, P<0.0001). Adding both BMP-7 and total TGF-β1 to the baseline model provided better fit than each kidney marker alone (all P<0.0001). The strongest discrimination was observed when circulating BMP-7 along with total and active TGF-β1 were incorporated into the baseline model (c-statistic=0.9566, P<0.0001 vs. the baseline model). A model based on the cytokines alone gives a c-statistic of 0.9396 (Figure 2).Table 3Areas under the receiver operating characteristic curves (c-statistics) and measures of model fit for the conventional predictorsa with individual or combined novel kidney markersModelsAICBICc-Statistic (95% confidence interval)P-valueConventional predictors (CP)a297.19339.640.7531 (0.6860–0.8202)CP+total TGF-β1267.04313.030.8218 (0.7671–0.8765)CP+BMP-7238.43284.360.8663 (0.8207–0.9118)CP+active TGF-β1227.73273.710.8838 (0.8400–0.9275)CP+total TGF-β1+active TGF-β1+BMP-7155.32208.320.9566 (0.9321–0.9812) Open table in a new tab This analysis from the ADVANCE trial clearly demonstrates that total and active TGF-β1 are substantially increased, and BMP-7 levels decreased, in people with type 2 diabetes mellitus who subsequently develop major renal end points compared with matched controls. More importantly, these results suggest these novel kidney markers predict kidney end points much more strongly than the best currently available risk markers, and are additive such that combining them with conventional risk markers leads to further improvement in discrimination. Renal fibrosis is regarded as a central event in the progression of CKD regardless of the underlying cause of kidney injury. The role of TGF-β1 and BMP-7, and their pathways, in the pathogenesis of kidney fibrosis has been highlighted over recent years.24.Liu Y. Cellular and molecular mechanisms of renal fibrosis.Nat Rev Nephrol. 2011; 7: 684-696Crossref PubMed Scopus (931) Google Scholar,25.Boon M.R. van der Horst G. van der Pluijm G. et al.Bone morphogenetic protein 7: a broad-spectrum growth factor with multiple target therapeutic potency.Cytokine Growth Factor Rev. 2011; 22: 221-229Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar The histological hallmark of diabetic nephropathy is glomerulosclerosis (Kimmelstiel–Wilson nodules), manifesting clinically as albuminuria and eventually reduced GFR. More recently, the importance of vascular sclerosis and tubulointerstitial fibrosis in diabetic nephropathy has become apparent, with the latter consistently shown to be the best histological predictor of progression of chronic to ESKD.26.Bohle A. Mackensen-Haen S. von Gise H. Significance of tubulointerstitial changes in the renal cortex for the excretory function and concentration ability of the kidney: a morphometric contribution.Am J Nephrol. 1987; 7: 421-433Crossref PubMed Scopus (277) Google Scholar,27.Nath K.A. Tubulointerstitial changes as a major determinant in the progression of renal damage.Am J Kidney Dis. 1992; 20: 1-17Abstract Full Text PDF PubMed Scopus (864) Google Scholar In parallel, there has been growing recognition that many people with diabetes mellitus who progress to ESKD do not have high levels of proteinuria, consistent with the view that tubulointerstitial fibrosis as well as glomerular pathology play an important role in the progression of kidney disease in patients with diabetes mellitus, and may actually be more important than the glomerular abnormalities. Despite the extensive experimental in vitro and in vivo evidence, clinical observations relating circulating and urinary levels of TGF-β1 to CKD severity or progression are limited. In cross-sectional studies, Suthanthiran et al.28.Suthanthiran M. Li B. Song J.O. et al.Transforming growth factor-beta 1 hyperexpression in African-American hypertensives: a novel mediator of hypertension and/or target organ damage.Proc Natl Acad Sci USA. 2000; 97: 3479-3484Crossref PubMed Scopus (266) Google Scholar,29.Suthanthiran M. Gerber L.M. Schwartz J.E. et al.Circulating transforming growth factor-beta1 levels and the risk for kidney disease in African Americans.Kidney Int. 2009; 76: 72-80Abstract Full Text Full Text PDF PubMed Scopus (49) Google Scholar showed that serum levels of TGF-β1 and levels of TGF-β1 mRNA in peripheral blood mononuclear cells were higher in African-American subjects with ESKD compared with Caucasians, and also in hypertensive compared with normotensive individuals. They concluded that TGF-β1 may contribute to renal disease progression by modulating multiple risk factors more robustly in African-American subjects. Increasing evidence suggests that circulating TGF-β1 levels are linked to vascular target organ damage in both diabetic and nondiabetic populations.30.Zitouni K. Harry D.D. Nourooz-Zadeh J. et al.Circulating vitamin E, transforming growth factor beta1, and the association with renal disease susceptibility in two racial groups with type 2 diabetes.Kidney Int. 2005; 67: 1993-1998Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar,31.August P. Suthanthiran M. Transforming growth factor beta and progression of renal disease.Kidney Int Suppl. 2003; : S99-104Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar Although urinary TGF-β1 levels are believed to reflect intrarenal production of TGF-β1, circulating levels may be both a risk factor and risk marker for more severe or progressive CKD. In a small substudy including 63 patients from the Collaborative Study Group Captopril Trial in patients with type 1 diabetes mellitus and overt nephropathy, participants with an increased urinary TGF-β1 level showed the greatest GFR decline over the ensuing 2 years.32.Sharma K. Eltayeb B.O. McGowan T.A. et al.Captopril-induced reduction of serum levels of transforming growth factor-beta1 correlates with long-term renoprotection in insulin-dependent diabetic patients.Am J Kidney Dis. 1999; 34: 818-823Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar Lund et al.,33.Lund R.J. Davies M.R. Brown A.J. et al.Successful treatment of an adynamic bone disorder with bone morphogenetic protein-7 in a renal ablation model.J Am Soc Nephrol. 2004; 15: 359-369Crossref PubMed Scopus (89) Google Scholar using an animal model, suggested that decreased BMP-7 levels directly correlate with a loss of viable renal mass. Hence, the possibility of using BMP-7 as a marker to predict renal outcomes was raised and uniquely assessed in this study. The current evidence for clinical benefits from the administration of BMP-7 in diabetic nephropathy remains limited, and mainly derived from animal models. Wang et al.19.Wang S. Chen Q. Simon T.C. et al.Bone morphogenic protein-7 (BMP-7), a novel therapy for diabetic nephropathy.Kidney Int. 2003; 63: 2037-2049Abstract Full Text Full Text PDF PubMed Scopus (207) Google Scholar,20.Wang S. de Caestecker M. Kopp J. et al.Renal bone morphogenetic protein-7 protects against diabetic nephropathy.J Am Soc Nephrol. 2006; 17: 2504-2512Crossref PubMed Scopus (151) Google Scholar have shown in experimental models of diabetic nephropathy that exogenous administration of BMP-7 can reverse kidney hypertrophy, improve glomerular lesions, and preserve podocyte number. In addition, BMP-7 suppresses plasminogen activation inhibitor-1, reduces albuminuria, and improves GFR. These data, along with the compelling evidence provided by this study, further suggest the role of pro-fibrotic and antifibrotic cytokines in maintaining the delicate balance between tissue repair and pathological fibrosis. Although the present study specifically highlights the value of using these cytokines as biomarkers to predict poor renal outcomes, it also adds weight to the thesis that strategies to modify TGF-β1 and/or BMP-7 levels may be protective in people at risk of diabetic kidney disease. In contrast to BMP-7, accumulating evidence supports the possibility that lowering TGF-β1 may protect the kidney in people with diabetes. Studies of agents that lower TGF-β1, such as Pirfenidone,34.Sharma K. Ix J.H. Mathew A.V. et al.Pirfenidone for diabetic nephropathy.J Am Soc Nephrol. 2011; 22: 1144-1151Crossref PubMed Scopus (222) Google Scholar Tranilast,35.Soma J. Sato K. Saito H. et al.Effect of tranilast in early-stage diabetic nephropathy.Nephrol Dial Transplant. 2006; 21: 2795-2799Crossref PubMed Scopus (42) Google Scholar,36.Soma J. Sugawara T. Huang Y.D. et al.Tranilast slows the progression of advanced diabetic nephropathy.Nephron. 2002; 92: 693-698Crossref PubMed Scopus (48) Google Scholar aliskiren,37.Lizakowski S. Tylicki L. Renke M. et al.Aliskiren and perindopril reduce the levels of transforming growth factor-beta in patients with non-diabetic kidney disease.Am J Hypertens. 2012; 25: 636-639Crossref PubMed Scopus (18) Google Scholar and monoclonal anti-TGF-β1 antibody38.Benigni A. Zoja C. Corna D. et al.Add-on anti-TGF-beta antibody to ACE inhibitor arrests progressive diabetic nephropathy in the rat.J Am Soc Nephrol. 2003; 14: 1816-1824Crossref PubMed Scopus (169) Google Scholar,39.Collaborative Study Group (CSG) ELaC Multicenter, phase 2 study to evaluate the safety and renal efficacy of LY2382770 in patients with diabetic kidney disease due to type 1 or type 2 diabetes.ClinicalTrials.govDate: 2010Google Scholar (generally used in addition to inhibitors of renin–angiotensin–aldosterone system) in diabetic kidney disease, have shown promising results. The present study suggests that further trials of strategies to reduce TGF-β1 in clinical practice are warranted. It is unclear whether the circulating active form of TGF-β1 is locally activated in the kidney or is a result of peripheral activation. This remains difficult to study in vivo given the complex mechanisms for activation of TGF-β1. Active TGF-β1 was largely undetectable in the control group, whereas it was observed in 93% of the cases who progressed to renal end points, which is consistent with prior studies.11.Sharma K. Ziyadeh F.N. Alzahabi B. et al.Increased renal production of transforming growth factor-beta1 in patients with type II diabetes.Diabetes. 1997; 46: 854-859Crossref PubMed Scopus (258) Google Scholar,12.Hellmich B. Schellner M. Schatz H. et al.Activation of transforming growth factor-beta1 in diabetic kidney disease.Metabolism. 2000; 49: 353-359Abstract Full Text PDF PubMed Scopus (44) Google Scholar Therefore, having detectable circulating TGF-β1 is closely associated with poor renal outcomes in patients with type 2 diabetes mellitus, irrespective of the level. Importantly, the current data suggest that the effects of total TGF-β1, active TGF-β1, and low BMP-7 are independent and cumulative. Hence, the results of this study allow more precise identification of those individuals who are at high risk of major renal end points, thus allowing the targeting of established and experimental preventative strategies. This study was limited by the fact that the samples were collected from a clinical trial population that may not be representative of the general population with type 2 diabetes mellitus, although the participant characteristics were broadly representative of the overall population with diabetes in many countries.40.Chalmers J. Arima H. Importance of blood pressure lowering in type 2 diabetes: focus on ADVANCE.J Cardiovasc Pharmacol. 2010; 55: 340-347Crossref PubMed Scopus (23) Google Scholar In addition, the results of the study cannot necessarily be extrapolated to cohorts other than patients with diabetes mellitus. Despite the promising results demonstrated in the study, validation of these biomarkers in a separate population, with a cost-effectiveness assessment, is required. As our study had a nested case–control design, the absolute values of the c-statistic will be overestimates of what we would have expected to find in a corresponding cohort study. Notwithstanding this, the differences we found in various measures of predictive ability for different sets of variables in each model were consistent and are likely to be clinically important. In summary, these results suggest that the novel kidney markers TGF-β1 (total and active) and BMP-7, particularly when considered in combination, predict progression to a major renal end point, more reliably than the best currently utilized biomarkers of eGFR and urinary albumin excretion. Incorporating these biomarkers into clinical practice may identify those individuals with diabetes mellitus who are likely to benefit most from renoprotective therapies. The ADVANCE trial has been described in detail previously.22.Patel A. MacMahon S. Chalmers J. et al.Effects of a fixed combination of perindopril and indapamide on macrovascular and microvascular outcomes in patients with type 2 diabetes mellitus (the ADVANCE trial): a randomised controlled trial.Lancet. 2007; 370: 829-840Abstract Full Text Full Text PDF PubMed Scopus (1805) Google Scholar,23.Patel A. MacMahon S. Chalmers J. et al.Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes.N Engl J Med. 2008; 358: 2560-2572Crossref PubMed Scopus (44) Google Scholar Individuals with high risk of type 2 diabetes mellitus diagnosed at ≥30 years of age were eligible. High risk was as defined as an age of at least 55 years at the time of study entry, and a history of major macrovascular or microvascular disease or at least one other risk factor for vascular disease. Participants were randomized in a factorial manner to intensive glucose control, targeting a HbA1c of ≤6.5%, or to standard, guideline-based glucose control, as well as to combination perindopril–indapamide therapy or to matching placebo. Primary end points were composites of major macrovascular events (death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal stroke) and major microvascular events (new or worsening nephropathy or retinopathy), with a median follow-up period of 5 years. The results of the blood pressure22.Patel A. MacMahon S. Chalmers J. et al.Effects of a fixed combination of perindopril and indapamide on macrovascular and microvascular outcomes in patients with type 2 diabetes mellitus (the ADVANCE trial): a randomised controlled trial.Lancet. 2007; 370: 829-840Abstract Full Text Full Text PDF PubMed Scopus (1805) Google Scholar and glycemic23.Patel A. MacMahon S. Chalmers J. et al.Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes.N Engl J Med. 2008; 358: 2560-2572Crossref PubMed Scopus (44) Google Scholar arms of the study have been reported previously. Baseline blood samples collected from participants in the ADVANCE study were used in this study. At enrollment, participants provided informed consent for the future use of their samples for analyses relevant to the primary and secondary outcomes of the study. Ethical approval for this specific study was obtained from the Royal North Shore Hospital in Sydney, Australia. For the purpose of this study, cases were defined as individuals who developed a major renal end point during the 5-year follow-up period, including any of doubling of serum creatinine to at least 200μmol/l, need for renal replacement therapy, or death due to renal disease. The renal end points were predefined and verified by the independent end point adjudication committee with input from an independent academic nephrologist. Doubling of serum creatinine or requirement of renal replacement therapy in the context of an acute illness with evidence of recovery of renal function or in the context of the terminal stage of illness was not adjudicated as a renal end point. Death because of renal disease was defined as noncardiovascular death with the underlying cause being renal disease, supported by laboratory reports, clinical notes, death certificate, or autopsy findings. Using propensity score methods, two controls were matched to every case for a range of parameters recorded at baseline, including age, sex, race, eGFR, blood pressure, HbA1c, known macrovascular disease (cerebrovascular accident, myocardial infarct, peripheral vascular disease), history of retinopathy, and treatment allocation (blood pressure lowering, intensive glucose lowering, both, or neither). Only cases with at least one matched control were included in the final analysis. In all, 25 cases had one matched control whereas 77 had two matched controls, providing a total study population for analysis of 102 cases and 179 controls (Figure 3). The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula41.Levey A.S. Stevens L.A. Schmid C.H. et al.A new equation to estimate glomerular filtration rate.Ann Intern Med. 2009; 150: 604-612Crossref PubMed Scopus (16115) Google Scholar was used to calculate the eGFR. The Modification of Diet in Renal Disease (MDRD) study formula42.Levey A.S. Bosch J.P. Lewis J.B. et al.A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group.Ann Intern Med. 1999; 130: 461-470Crossref PubMed Scopus (13034) Google Scholar was used in alternative sensitivity analyses. Estimating GFR using either method yielded highly congruent results (Supplementary Tables S2 and S3 online). Lithium heparin anticoagulated blood samples were collected at randomization. The blood samples were immediately subjected to centrifugation at 2000g for 10min at 4°C. The supernatant plasma was aliquoted and stored at -80°C for future biomarker assay. Blood glucose, serum creatinine, and urinary albumin/creatinine ratio were performed in local laboratories at individual study sites. Each glycated hemoglobin measurement was standardized. Total and active plasma TGF-β1 protein concentration was measured using an isoform-specific TGF-β1 enzyme-linked immunosorbent assay according to the manufacturer’s protocol (TGF-β1 Emax ImmunoAssay System, Madison, WI). Total TGF-β1 was obtained by acidification with 1N HCl for 15min at room temperature, followed by neutralization with 1N NaOH. This assay has reported low crossreactivity with TGF-β2 and TGF-β3. Circulating BMP-7 levels were determined using an immunoassay kit assay (R&D Systems, Minneapolis, MN) with minimal modification. Heparinized neat plasma (150μl) without assay diluent was used. The absorbance values at 450nm, corrected by readings at 540nm, were obtained, and plotted against the standard to determine the circulating level of BMP-7. Descriptive results for the subjects’ baseline characteristics and biomarkers are expressed as either mean and s.d. or, when the variable is skewed, median and interquartile range. Values below detectable limits were given imputed values at the middle of the undetectable range in the statistical analyses. Statistical comparisons between groups were made by t-tests for data that were normally distributed, Wilcoxon tests for skewed continuous data, and χ2s for categorical data. Conditional logistic regression models were used to estimate the ORs for the novel biomarkers using unadjusted and adjusted models, both after dividing into thirds and in continuous form after log transformation to remove the effect of outliers. Discrimination was assessed through the area under the receiver operating characteristic curve, also known as the c-statistic. Akaike’s information criterion43.Akaike H. A new look at the statistical model identification.IEEE Trans Automat Contr. 1974; 6: 716-723Crossref Scopus (36083) Google Scholar and Schwartz’s Bayesian information criterion44.Schwarz G.E. Estimating the dimension of a model.Ann Statist. 1978; 2: 461-464Crossref Google Scholar were used to compare the fit of various models, with and without biomarkers and potential confounders. To allow for optimism according to self-testing, we bootstrapped the c-statistic. Using this method, the bias was estimated to be <2%. Analyses were performed using the software package SAS 9.2 (SAS Institute, Cary, NC). P-values of <0.05 were considered significant. We thank Miss Katherine Pegg, Miss Rachel Yong, and Miss Dijana Bosnjak for their technical assistance. Table S1. Numbers of study participants who developed renal endpoints. Table S2. Odds ratio for major renal endpoints associated with a 1 standard deviation difference in log transformed values of the novel kidney markers. Table S3. AUCs (C-statistics) and measures of model fit for the conventional predictors with individual or combined novel kidney markers. Supplementary material is linked to the online version of the paper at http://www.nature.com/ki
Referência(s)