Evolution of Glomerular Filtration Rate in Proteinuric NIDDM Patients
Objective: To evaluate, by means of a precise method, the rate of decline of glomerular filtration rate in proteinuric non-insulin-dependent diabetic (NIDDM) patients. Research Design and Methods: The study was comprised of seven NIDDM patients who visited an outpatient clinic and had a 24-h urinary protein excretion rate ^500 mg in the absence of heart failure, urinary tract infection, or other nephropathies. Results: Glomerular filtration rate (51Cr-labeled EDTA, singleinjection protocol) and 24-h proteinuria (turbidimetric method) were assessed at periodic intervals (2-6 mo). Correlation of the measurements with time (Pearson's r, with Student's t test used to assess the significance, a = 0.05) was used to evaluate the trend of evolution of glomerular filtration rate. Renal biopsies were performed in four patients. In three of four patients, renal histopathology was consistent with the diagnosis of diabetic nephropathy (in the 4th patient measurements were not satisfactory). Neither glomerular filtration rate nor proteinuria correlated significantly with time, except in one patient who had multiple myeloma. Conclusions: The decline of glomerular filtration rate in proteinuric NIDDM patients is different from that observed in insulin-dependent diabetic patients, which is probably much slower. Diabetes Care 14:355-59, 1991
Rogerio Friedman, MD, MSc Jorge L. Gross, MD, PhD
duction of GFR. In contrast, it is well known that insulindependent diabetic subjects present an inexorable decrease of GFR of —10-12 ml/yr after the appearance of clinically detectable proteinuria (2,3). However, there is no evidence to suggest that the nature of glomerular disease is different in the two types of diabetes; it may be impossible to distinguish in histopathological terms the glomerular lesions found in the two groups. The aim of this study was to prospectively follow the GFR of a cohort of proteinuric NIDDM patients without interfering with the therapeutic regimen.
RESEARCH DESIGN AND METHODS
This noncontrolled cohort study was conducted exclusively with outpatients. Every NIDDM patient who visited the outpatient clinic was considered for the study. We based our definition of NIDDM on World Health Organization criteria. The following were used as inclusion criteria: /) no episodes of ketoacidosis, 2) absence of ketonuria, 3) use of insulin because of secondary failure of oral agents or to improve control, 4) insulin introduced after at least 5 yr of disease, 5) diagnosis of diabetes after 30 yr of age, and 6) a positive proteinuria with a reactive tape test (N-Multistix, Ames, Elkhart, IN). Patients were then evaluated for persistent macroprohe natural history of diabetic nephropathy in teinuria in the absence of urinary tract infection, heart non-insulin-dependent diabetic (NIDDM) patients is still essentially unknown. In a cross-sectional study of 510 NIDDM patients, Fabre et al. (1) From the Endocrine Unit, Hospital de Clfnicas de Porto Alegre, and School of Universidade Federal do Rio Grande do Sul, Rio Grande do Sul, suggested that the decline in glomerular filtration rate Medicine, Brazil. (GFR) was of a lesser magnitude than that observed in Address correspondence and reprint requests to Jorge L. Gross, MD, Servigo insulin-dependent diabetic patients. This suggestion was de Endocrinologia, Hospital de Clfnicas de Porto Alegre, Rua Ramiro Barcellos 2350/635 (90210), Porto Alegre, RS, Brazil. based on the observation that only 13% of patients with Received for publication 16 January 1990 and accepted in revised form 10 proteinuria >150 mg/24 h presented a progressive re- December 1990.
T
DIABETES CARE, VOL. 14, NO. 5, MAY 1991
355
CLOMERUIAR FILTRATION RATE IN NIDDM
failure, or other glomerulopathies; body mass index was not >30 and 28.6 kg/m2 for men and women, respectively (4). Persistent macroproteinuria was defined as a 24-h urinary protein excretion rate ^0.5 g, taken at three separate occasions, with a 15-day interval between the measurements (5-8). Absence of urinary tract infection was characterized by a negative urine culture. Heart failure was ruled out by clinical examination, chest films, and measurement of the ejection fraction through resting radioisotopic ventriculography (with 99m Tc-labeled erythrocytes); we believe that an ejection fraction of ^0.56, in the absence of clinical and radiographic signs, excludes heart failture (9). Hemoglobinuria, erythrocyte casts, previous history of nephropathies, and serious systemic hypertension (beginning before the 4th decade and before the diagnosis of diabetes) were considered as evidence of other glomerulopathies. A normal body mass index was necessary for the method used in the measurement of GFR. Patients were then evaluated to determine other longterm complications of diabetes. Arterial pressure was measured in the sitting position in the left arm after a 5min rest with a standard 12.5-cm cuff mercury sphygmomanometer (phases 1-4 of the Korotkoff sounds). Hypertension was defined as an arterial blood pressure level >160/95 mmHg in untreated patients or any values in patients under antihypertensive therapy (10). Fundus examination was performed by the ophthalmologist after mydriasis. Patients were classified as having no signs of retinopathy, presenting with nonproliferative retinopathy, or being affected by proliferative retinopathy (11). Peripheral neuropathy was considered whenever vibratory threshold was impaired (tuning fork test), primarily if it was associated with a diminished Achilles tendon reflex and compatible history. All patients were given five autonomic cardiovascular tests, and autonomic neuropathy was diagnosed in the occurrence of at least one abnormal test (12,13). Macroangiopathy was diagnosed if history and/or evidence of peripheral arterial disease (history of intermittent claudication, amputation due to ischemic complications, diminished distal pulses), cerebrovascular disease (history of stroke, confirmed by case notes, compatible neurological examination), or coronary heart disease (case history and electrocardiographic findings) were present. After the evaluation and at periodic intervals (every 2-6 mo), patients had their GFR, 24-h creatininuria, and 24-h proteinuria assessed. Glycemic control was assessed through the measurement of HbA1c. No intervention in the therapeutic regimen was made throughout the study. At each visit, the patient would come in at 0800 after an overnight fast, bringing a 24-h urine collection. Blood pressure was measured as described, and blood was taken for measurement of HbA1c and glucose. A single bolus of 51Cr-labeled EDTA was then injected, and blood samples were taken after 2, 3, and 4 h for estimation of GFR (14). In our unit, this technique had a
356
coefficient of variation of 11.2%, and the normal values for men and women >46 yr of age ranged from 73.2 to 142.2 ml • min" 1 • 1.73 m~2. Urinary proteins were measured by the turbidimetric method of Denis and Ayer (15,16), and urinary creatinine was quantified by the method of Jaffe" in an autoanalyzer (17,18). Patients who gave their informed consent were submitted to needle renal biopsies under ultrasound monitoring. The tissue fragments were examined by a pathologist under light electron and immunofluorescent microscopy. The pathologist was blinded to the identity of clinical data of patients. Statistical analysis consisted of Pearson's r (GFR vs. time in months) and the determination of the slope (a). Student's t test was used to establish statistical significance (a = 0.05).
RESULTS Seven patients fulfilled the admission criteria. They were followed for a mean period of 17 mo and 20 days, ranging from 4 mo and 20 days (patient 7) to 25 mo and 11 days. If patient 7 is excluded (as explained below), the mean period of observation was 19 mo and 28 days. Table 1 gives the clinical characteristics of the patients; no patients smoked. The degree of metabolic control, although not satisfactory in the group (HbA1c 5.5-17.0%), did not show a significant correlation with GFR in the individual patients throughout the study. Blood pressure levels remained unchanged. Figure 1 shows the evolution of GFR during the observation period. The evolution of patient 7 was markedly different from the others. Table 2 shows the correlation between GFR and time in months. This correlation was significant only for patient 7. The quick evolution, proteinuria level (19-22 g/day), and rapid development of anemia led to a more extensive investigation and to the diagnosis of multiple myeloma. This patient was excluded from the rest of the statistical analysis. In the remaining six patients, the correlation was not significant and the slopes of the lines followed different patterns. The 24-h urinary protein excretion rate (and the protein-creatinine ratio) did not correlate with time in the group (Table 3), with the exception of patient 3, in whom the tendency was toward the reduction of protein excretion. Renal biopsies were performed in patients 1-4. The procedure was not performed in patient 5 because of her age. Patient 6 had a horseshoe kidney identified through ultrasound, and a biopsy was cancelled because of the risks involved. Patient 7 was not biopsied because of the clinical diagnosis of multiple myeloma. The material obtained from patient 1 was not satisfactory. The glomeruli of the remaining three patients showed morphological alterations of diabetic glomeru-
DIABETES CARE, VOL 14, NO. 5, MAY 1991
R. FRIEDMAN AND J.L GROSS
TABLE 1 Clinical characteristics of patients Known duration of disease (yr)
Body mass
Patient
Age (yr)
Sex
index (kg/m2)
1 2 3 4 5 6 7
66 57 71 47 75 52 60
M M F M F M F
26.2 26.8 24.1 26.2 26.0 22.8 28.5
12 5 20 7 11 20 ?
Treatment
Other complications
Insulin Oral agent Insulin Oral agent Insulin Oral agent Oral agent
PR PR, PN, AN BR, PN, AN BR, PN, AN PN, AN PR, AN, MA PN
Associated conditions
Other drugs
HBP, hypoth
Methyldopa, furosemide
HBP Glaucoma, hyperth HBP, glaucoma HBP, myeloma
Clonidine, thiazide Thiazide Thiazide
PR, proliferative retinopathy; BR, background retinopathy; PN, proliferative neuropathy; AN, autonomic neuropathy; MA, macroangiopathy; HBP, arterial hypertension; hypoth, treated primary hypothyroidism (patient euthyroid under L-thyroxine); hyperth, treated hyperthyroidism (patient consistently euthyroid after 131I treatment); ?, unknown.
losclerosis. Electron microscopy confirmed the findings. Furthermore, immunohistochemical staining did not show immunocomplexes suggestive of other glomerulopathies.
CONCLUSIONS There is agreement in the literature concerning the evolution of GFR in macroproteinuric insulin-dependent diabetic patients. GFR falls inexorably at a mean rate of 1 ml • min" 1 • mo" 1 (2,3). At the same time, proteinuria tends to increase (2,19). The findings in NIDDM are not so clear. Fabre et al. (1), studying 510 NIDDM patients, observed that GFR was stable, despite the fact that their patients were macroproteinuric for long periods. Nevertheless, in only 65% of patients, GFR was in the normal range. The remaining 35% had a lower than normal GFR. This study did not observe the patients serially; the evolution of GFR was inferred from the estimated duration of di-
abetes. In our unit, Stein (20) made similar observations; 65 NIDDM patients had a stable GFR (as estimated by creatinine clearance) in the presence of macroproteinuria for as long as 36 mo, but the GFR of these patients was lower than that of nonmacroproteinuric ones. The known limitations of creatinine clearance as a marker of GFR made us question whether this finding was due to the method used (21-25). We then decided to conduct another trial using a more precise technique, the measurement of GFR through the single injection of 51 Cr-EDTA. Of course, this decision brought up another limitation; due to the extreme unpredictability of the volume of body fat distribution in obese patients, this method can be used only in nonobese individuals. When studying NIDDM patients, this is a considerable limitation, for at least 80% of them are obese. Nevertheless, there is no evidence that the GFR of obese NIDDM patients is different from the GFR of nonobese NIDDM patients. The results of this study are similar to those mentioned above: with the use of a more precise method, GFR did not change during the period of observation, no matter what the initial value was. The finding of a reduced GFR in macroproteinuric TABLE 2 Evolution of glomerular filtration rate (GFR) and its relationship to time in 7 patients with non-insulin-dependent diabetes mellitus and macroproteinuria
II 1
Patient
(months)
1 2 3 4
FIG. 1. Evolution of glomerular filtration rate (GFR) (ml • min"1 • 1.73 m~2) in 7 patients with non-insulin-dependent diabetes mellitus and persistent macroproteinuria. GFR measured by single-injection method with 51Cr-labeled EDTA. Patient 7 had multiple myeloma.
DIABETES CARE, VOL. 14, NO. 5, MAY 1991
5 6 7*
Slope (ml • min" 1 • mo"1)
Pearson's r (GFR vs. time)
P
0.09 -0.72 0.37 -0.02 0.36 0.80 -5.50
0.22 -0.75 0.45 -0.03 0.36 0.76 -0.95
>0.20 >0.05 >0.20 >0.20 >0.20 >0.05 0.200 >0.200 0.050 >0.200 >0.200
-0.01 -0.09 -0.06 0.03 0.01 -0.05
-0.58 -0.42 -0.91 0.79 0.27 -0.43
patients, associated with a longer evolution, suggests that the GFR falls (1,20). This may occur so slowly that it cannot be detected in short periods with the methods used. Alternatively, it could be speculated that GFR could remain unaltered, although in a lower level than that of nonproteinuric patients. In this case, an explanation would be needed concerning how, when, and why this reduction occurs. Other nephropathies can occur in 1 5 - 3 1 % of NIDDM patients (1,26). Could this different behavior be attributed to the association of nondiabetic glomerulopathies in these patients? One patient with multiple myeloma was included in this study on the basis of classic criteria: diabetes mellitus and proteinuria. Nevertheless, in three patients, diabetic glomerulosclerosis was confirmed (and other glomerulopathies excluded) by renal biopsy. Similar considerations should be made for the lack of increase in proteinuria with time. This seems to occur, to a certain extent (not necessarily related to the deterioration of GFR), in insulin-dependent diabetes mellitus (2,19) but has not been confirmed in NIDDM. The GFR of macroproteinuric NIDDM patients has a different evolution from that described for insulin-dependent diabetic patients. The deterioration of kidney function is probably much slower and the decline of GFR so discrete from month to month that the variation is not of statistical significance. The establishment of a pattern of evolution in these patients will require longer studies with a greater number of individuals or the development of methods of measurement of kidney function whose variability will be smaller than that obtained with current methods. We suggest that any additional studies involving proteinuric NIDDM patients should include renal biopsies. Any generalization of our conclusions may be risky, considering the small number of patients and the fact that we could not study obese individuals.
10. 11. 12. 13.
14. 15.
16.
17. 18.
ACKNOWLEDGMENTS
19.
Parts of this study were presented in abstract form at the 49th annual meeting of the American Diabetes Association, Detroit, Michigan, 1-6 June 1989.
20.
358
Fabre J, Balant LP, Dayer P, Fox HM, Vernet AT: The kidney in mature onset diabetes mellitus: a clinical study of 510 patients. Kidney Int 21:730-78, 1982 Parving H-H, Smidt UM, Frisberg B, Bonnevie-Nielsen V, Andersen AR: A prospective study of glomerular filtration rate and arterial blood pressure in insulin-dependent diabetics with diabetic nephropathy. Diabetologia 20:45761, 1981 Viberti GC, Bilous RW, Mackintosh D, Keen H: Monitoring glomerular function in diabetic nephropathy: a prospective study. Am J Med 74:256-64, 1983 Obesity: a report of The Royal College of Physicians. J Royal Coll Physicians Lond 17:5-65, 1983 Deckert T, Andersen AR, Christiansen JS, Andersen JK: Course of diabetic nephropathy: factors related to development. Ada Endocrinol 97 (Suppl. 242): 14-15, 1981 Mauer SM, Steffes MW, Goetz FC, Sutherland DER, Brown DM: Diabetic nephropathy: a perspective. Diabetes 32 (Suppl. 2):52-55, 1983 Mogensen CE: Renal function changes in diabetes. Diabetes 25:872-79, 1976 Mogensen CE, Christensen CK, Vittinghus E: The stages in diabetic renal disease with emphasis on the stage of incipient diabetic nephropathy. Diabetes 32 (Suppl. 2): 64-78, 1983 Fishman AP: Heart failure. In Cec/7 Textbook of Medicine. Wyngaarden JB, Smith LH Jr., Eds. Philadelphia, PA, Saunders, 1985, p. 189-210 World Health Organization: Arterial Hypertension. Report of a WHO Expert Committee. Geneva, World Health Org., 1978 (Tech. Rep. Sen, no. 628) Blankenship G, Skyler JS: Diabetic retinopathy: a general survey. Diabetes Care 1:127-37, 1978 Ewing DJ, Martyn CN, Young RJ, Clarke BF: The value of autonomic function tests: 10 yr experience in diabetes. Diabetes Care 8:491-89, 1985 Neumann C, Amon LC, Canani LH, Gross JL, Schmid H: Testes autonomicos no diabete me'lito: padronizacao dos valores normais. In Anais do Congr. Brasileiro de Diabetes, VI, Salvador, Brazil, Soc. Brasileira de Diabetes, 1987 ChantlerC, BarrattTM: Estimation of glomerular filtration rate from plasma clearance of 51-chromium edetic acid. Arch Dis Child 47:613-17, 1972 Mattock MB, Keen H, Viberti GC, El Gohary MR, Murrells TJ, Scott GS, WingJR, Jackson PG: Coronary heart disease and urinary albumin excretion in type 2 (non-insulindependent) diabetic patients. Diabetologia 31:82-87, 1988 Schriever H, Gambino SR: Protein turbidity produced by trichloroacetic acid and sulfosalicylic acid at varying temperatures and varying ratios of albumin and globulin. Am I Clin Pathol 44:667-72, 1965 Chasson A, Grady H, Stanley M: Determination of creatinine by means of automatic chemical analysis. Am I Clin Pathol 35:83-88, 1961 Fabiny DL, Ertingshausen G: Automated reaction-rate method for determination of serum creatinine with the CentrifiChem. Clin Chem 17:696-700, 1971 Viberti GC, Keen H: The patterns of proteinuria in diabetes mellitus: relevance to pathogenesis and prevention of diabetic nephropathy. Diabetes 33:686-92, 1984 Stein ACR: Estudo comparativo e evolutivo de pacientes diabe"ticos com diferentes graus de proteinuria. In Curso DIABETES CARE, VOL. 14, NO. 5, MAY 1991
R. FRIEDMAN AND J.L. GROSS
c/e Pos-Graduagao em Medicina: Nefrologia, Porto Alegre, Brasil, Universidade Federal do Rio Grande do Sul, 24. 1984, p. 85 21. Brochner-Mortensen J: Routine methods and their reliability for assessment of glomerular filtration rate in adults. Dan Med Bull 25:181-202, 1978 25. 22. Carrie BJ, Golbetz HV, Michael AS, Myers BD: Creatinine: an inadequate filtration marker in glomerular dis26. eases. Am} Med 69:177-82, 1980 23. Dodge WF, Travis LB, Daeschner CW: Comparison of endogenous creatinine clearance with insulin clearance.
DIABETES CARE, VOL. 14, NO. 5, MAY 1991
Am ) Dis Child 113:683-92, 1967 Friedman R, Azevedo MJ, Gross JL: Is endogenous creatinine clearance still a reliable index of glomerular filtration rate in diabetic patients? Braz J Med Biol Res 21: 941_44, 1988 Shemesh O, Golbetz H, Kriss JP, Myers B: Limitations of creatinine as a filtration marker in glomerulopathic patients. Kidney Int 28:830-88, 1985 Grenfell A, Watkins PJ: Clinical diabetic nephropathy: natural history and complications. Clin Endocrinol Metab 15:783-805, 1986
359
Objective: To evaluate, by means of a precise method, the rate of decline of glomerular filtration rate in proteinuric non-insulin-dependent diabetic (NIDDM) patients. Research Design and Methods: The study was comprised of seven NIDDM patients who visited an outpatient clinic and had a 24-h urinary protein excretion rate ^500 mg in the absence of heart failure, urinary tract infection, or other nephropathies. Results: Glomerular filtration rate (51Cr-labeled EDTA, singleinjection protocol) and 24-h proteinuria (turbidimetric method) were assessed at periodic intervals (2-6 mo). Correlation of the measurements with time (Pearson's r, with Student's t test used to assess the significance, a = 0.05) was used to evaluate the trend of evolution of glomerular filtration rate. Renal biopsies were performed in four patients. In three of four patients, renal histopathology was consistent with the diagnosis of diabetic nephropathy (in the 4th patient measurements were not satisfactory). Neither glomerular filtration rate nor proteinuria correlated significantly with time, except in one patient who had multiple myeloma. Conclusions: The decline of glomerular filtration rate in proteinuric NIDDM patients is different from that observed in insulin-dependent diabetic patients, which is probably much slower. Diabetes Care 14:355-59, 1991
Rogerio Friedman, MD, MSc Jorge L. Gross, MD, PhD
duction of GFR. In contrast, it is well known that insulindependent diabetic subjects present an inexorable decrease of GFR of —10-12 ml/yr after the appearance of clinically detectable proteinuria (2,3). However, there is no evidence to suggest that the nature of glomerular disease is different in the two types of diabetes; it may be impossible to distinguish in histopathological terms the glomerular lesions found in the two groups. The aim of this study was to prospectively follow the GFR of a cohort of proteinuric NIDDM patients without interfering with the therapeutic regimen.
RESEARCH DESIGN AND METHODS
This noncontrolled cohort study was conducted exclusively with outpatients. Every NIDDM patient who visited the outpatient clinic was considered for the study. We based our definition of NIDDM on World Health Organization criteria. The following were used as inclusion criteria: /) no episodes of ketoacidosis, 2) absence of ketonuria, 3) use of insulin because of secondary failure of oral agents or to improve control, 4) insulin introduced after at least 5 yr of disease, 5) diagnosis of diabetes after 30 yr of age, and 6) a positive proteinuria with a reactive tape test (N-Multistix, Ames, Elkhart, IN). Patients were then evaluated for persistent macroprohe natural history of diabetic nephropathy in teinuria in the absence of urinary tract infection, heart non-insulin-dependent diabetic (NIDDM) patients is still essentially unknown. In a cross-sectional study of 510 NIDDM patients, Fabre et al. (1) From the Endocrine Unit, Hospital de Clfnicas de Porto Alegre, and School of Universidade Federal do Rio Grande do Sul, Rio Grande do Sul, suggested that the decline in glomerular filtration rate Medicine, Brazil. (GFR) was of a lesser magnitude than that observed in Address correspondence and reprint requests to Jorge L. Gross, MD, Servigo insulin-dependent diabetic patients. This suggestion was de Endocrinologia, Hospital de Clfnicas de Porto Alegre, Rua Ramiro Barcellos 2350/635 (90210), Porto Alegre, RS, Brazil. based on the observation that only 13% of patients with Received for publication 16 January 1990 and accepted in revised form 10 proteinuria >150 mg/24 h presented a progressive re- December 1990.
T
DIABETES CARE, VOL. 14, NO. 5, MAY 1991
355
CLOMERUIAR FILTRATION RATE IN NIDDM
failure, or other glomerulopathies; body mass index was not >30 and 28.6 kg/m2 for men and women, respectively (4). Persistent macroproteinuria was defined as a 24-h urinary protein excretion rate ^0.5 g, taken at three separate occasions, with a 15-day interval between the measurements (5-8). Absence of urinary tract infection was characterized by a negative urine culture. Heart failure was ruled out by clinical examination, chest films, and measurement of the ejection fraction through resting radioisotopic ventriculography (with 99m Tc-labeled erythrocytes); we believe that an ejection fraction of ^0.56, in the absence of clinical and radiographic signs, excludes heart failture (9). Hemoglobinuria, erythrocyte casts, previous history of nephropathies, and serious systemic hypertension (beginning before the 4th decade and before the diagnosis of diabetes) were considered as evidence of other glomerulopathies. A normal body mass index was necessary for the method used in the measurement of GFR. Patients were then evaluated to determine other longterm complications of diabetes. Arterial pressure was measured in the sitting position in the left arm after a 5min rest with a standard 12.5-cm cuff mercury sphygmomanometer (phases 1-4 of the Korotkoff sounds). Hypertension was defined as an arterial blood pressure level >160/95 mmHg in untreated patients or any values in patients under antihypertensive therapy (10). Fundus examination was performed by the ophthalmologist after mydriasis. Patients were classified as having no signs of retinopathy, presenting with nonproliferative retinopathy, or being affected by proliferative retinopathy (11). Peripheral neuropathy was considered whenever vibratory threshold was impaired (tuning fork test), primarily if it was associated with a diminished Achilles tendon reflex and compatible history. All patients were given five autonomic cardiovascular tests, and autonomic neuropathy was diagnosed in the occurrence of at least one abnormal test (12,13). Macroangiopathy was diagnosed if history and/or evidence of peripheral arterial disease (history of intermittent claudication, amputation due to ischemic complications, diminished distal pulses), cerebrovascular disease (history of stroke, confirmed by case notes, compatible neurological examination), or coronary heart disease (case history and electrocardiographic findings) were present. After the evaluation and at periodic intervals (every 2-6 mo), patients had their GFR, 24-h creatininuria, and 24-h proteinuria assessed. Glycemic control was assessed through the measurement of HbA1c. No intervention in the therapeutic regimen was made throughout the study. At each visit, the patient would come in at 0800 after an overnight fast, bringing a 24-h urine collection. Blood pressure was measured as described, and blood was taken for measurement of HbA1c and glucose. A single bolus of 51Cr-labeled EDTA was then injected, and blood samples were taken after 2, 3, and 4 h for estimation of GFR (14). In our unit, this technique had a
356
coefficient of variation of 11.2%, and the normal values for men and women >46 yr of age ranged from 73.2 to 142.2 ml • min" 1 • 1.73 m~2. Urinary proteins were measured by the turbidimetric method of Denis and Ayer (15,16), and urinary creatinine was quantified by the method of Jaffe" in an autoanalyzer (17,18). Patients who gave their informed consent were submitted to needle renal biopsies under ultrasound monitoring. The tissue fragments were examined by a pathologist under light electron and immunofluorescent microscopy. The pathologist was blinded to the identity of clinical data of patients. Statistical analysis consisted of Pearson's r (GFR vs. time in months) and the determination of the slope (a). Student's t test was used to establish statistical significance (a = 0.05).
RESULTS Seven patients fulfilled the admission criteria. They were followed for a mean period of 17 mo and 20 days, ranging from 4 mo and 20 days (patient 7) to 25 mo and 11 days. If patient 7 is excluded (as explained below), the mean period of observation was 19 mo and 28 days. Table 1 gives the clinical characteristics of the patients; no patients smoked. The degree of metabolic control, although not satisfactory in the group (HbA1c 5.5-17.0%), did not show a significant correlation with GFR in the individual patients throughout the study. Blood pressure levels remained unchanged. Figure 1 shows the evolution of GFR during the observation period. The evolution of patient 7 was markedly different from the others. Table 2 shows the correlation between GFR and time in months. This correlation was significant only for patient 7. The quick evolution, proteinuria level (19-22 g/day), and rapid development of anemia led to a more extensive investigation and to the diagnosis of multiple myeloma. This patient was excluded from the rest of the statistical analysis. In the remaining six patients, the correlation was not significant and the slopes of the lines followed different patterns. The 24-h urinary protein excretion rate (and the protein-creatinine ratio) did not correlate with time in the group (Table 3), with the exception of patient 3, in whom the tendency was toward the reduction of protein excretion. Renal biopsies were performed in patients 1-4. The procedure was not performed in patient 5 because of her age. Patient 6 had a horseshoe kidney identified through ultrasound, and a biopsy was cancelled because of the risks involved. Patient 7 was not biopsied because of the clinical diagnosis of multiple myeloma. The material obtained from patient 1 was not satisfactory. The glomeruli of the remaining three patients showed morphological alterations of diabetic glomeru-
DIABETES CARE, VOL 14, NO. 5, MAY 1991
R. FRIEDMAN AND J.L GROSS
TABLE 1 Clinical characteristics of patients Known duration of disease (yr)
Body mass
Patient
Age (yr)
Sex
index (kg/m2)
1 2 3 4 5 6 7
66 57 71 47 75 52 60
M M F M F M F
26.2 26.8 24.1 26.2 26.0 22.8 28.5
12 5 20 7 11 20 ?
Treatment
Other complications
Insulin Oral agent Insulin Oral agent Insulin Oral agent Oral agent
PR PR, PN, AN BR, PN, AN BR, PN, AN PN, AN PR, AN, MA PN
Associated conditions
Other drugs
HBP, hypoth
Methyldopa, furosemide
HBP Glaucoma, hyperth HBP, glaucoma HBP, myeloma
Clonidine, thiazide Thiazide Thiazide
PR, proliferative retinopathy; BR, background retinopathy; PN, proliferative neuropathy; AN, autonomic neuropathy; MA, macroangiopathy; HBP, arterial hypertension; hypoth, treated primary hypothyroidism (patient euthyroid under L-thyroxine); hyperth, treated hyperthyroidism (patient consistently euthyroid after 131I treatment); ?, unknown.
losclerosis. Electron microscopy confirmed the findings. Furthermore, immunohistochemical staining did not show immunocomplexes suggestive of other glomerulopathies.
CONCLUSIONS There is agreement in the literature concerning the evolution of GFR in macroproteinuric insulin-dependent diabetic patients. GFR falls inexorably at a mean rate of 1 ml • min" 1 • mo" 1 (2,3). At the same time, proteinuria tends to increase (2,19). The findings in NIDDM are not so clear. Fabre et al. (1), studying 510 NIDDM patients, observed that GFR was stable, despite the fact that their patients were macroproteinuric for long periods. Nevertheless, in only 65% of patients, GFR was in the normal range. The remaining 35% had a lower than normal GFR. This study did not observe the patients serially; the evolution of GFR was inferred from the estimated duration of di-
abetes. In our unit, Stein (20) made similar observations; 65 NIDDM patients had a stable GFR (as estimated by creatinine clearance) in the presence of macroproteinuria for as long as 36 mo, but the GFR of these patients was lower than that of nonmacroproteinuric ones. The known limitations of creatinine clearance as a marker of GFR made us question whether this finding was due to the method used (21-25). We then decided to conduct another trial using a more precise technique, the measurement of GFR through the single injection of 51 Cr-EDTA. Of course, this decision brought up another limitation; due to the extreme unpredictability of the volume of body fat distribution in obese patients, this method can be used only in nonobese individuals. When studying NIDDM patients, this is a considerable limitation, for at least 80% of them are obese. Nevertheless, there is no evidence that the GFR of obese NIDDM patients is different from the GFR of nonobese NIDDM patients. The results of this study are similar to those mentioned above: with the use of a more precise method, GFR did not change during the period of observation, no matter what the initial value was. The finding of a reduced GFR in macroproteinuric TABLE 2 Evolution of glomerular filtration rate (GFR) and its relationship to time in 7 patients with non-insulin-dependent diabetes mellitus and macroproteinuria
II 1
Patient
(months)
1 2 3 4
FIG. 1. Evolution of glomerular filtration rate (GFR) (ml • min"1 • 1.73 m~2) in 7 patients with non-insulin-dependent diabetes mellitus and persistent macroproteinuria. GFR measured by single-injection method with 51Cr-labeled EDTA. Patient 7 had multiple myeloma.
DIABETES CARE, VOL. 14, NO. 5, MAY 1991
5 6 7*
Slope (ml • min" 1 • mo"1)
Pearson's r (GFR vs. time)
P
0.09 -0.72 0.37 -0.02 0.36 0.80 -5.50
0.22 -0.75 0.45 -0.03 0.36 0.76 -0.95
>0.20 >0.05 >0.20 >0.20 >0.20 >0.05 0.200 >0.200 0.050 >0.200 >0.200
-0.01 -0.09 -0.06 0.03 0.01 -0.05
-0.58 -0.42 -0.91 0.79 0.27 -0.43
patients, associated with a longer evolution, suggests that the GFR falls (1,20). This may occur so slowly that it cannot be detected in short periods with the methods used. Alternatively, it could be speculated that GFR could remain unaltered, although in a lower level than that of nonproteinuric patients. In this case, an explanation would be needed concerning how, when, and why this reduction occurs. Other nephropathies can occur in 1 5 - 3 1 % of NIDDM patients (1,26). Could this different behavior be attributed to the association of nondiabetic glomerulopathies in these patients? One patient with multiple myeloma was included in this study on the basis of classic criteria: diabetes mellitus and proteinuria. Nevertheless, in three patients, diabetic glomerulosclerosis was confirmed (and other glomerulopathies excluded) by renal biopsy. Similar considerations should be made for the lack of increase in proteinuria with time. This seems to occur, to a certain extent (not necessarily related to the deterioration of GFR), in insulin-dependent diabetes mellitus (2,19) but has not been confirmed in NIDDM. The GFR of macroproteinuric NIDDM patients has a different evolution from that described for insulin-dependent diabetic patients. The deterioration of kidney function is probably much slower and the decline of GFR so discrete from month to month that the variation is not of statistical significance. The establishment of a pattern of evolution in these patients will require longer studies with a greater number of individuals or the development of methods of measurement of kidney function whose variability will be smaller than that obtained with current methods. We suggest that any additional studies involving proteinuric NIDDM patients should include renal biopsies. Any generalization of our conclusions may be risky, considering the small number of patients and the fact that we could not study obese individuals.
10. 11. 12. 13.
14. 15.
16.
17. 18.
ACKNOWLEDGMENTS
19.
Parts of this study were presented in abstract form at the 49th annual meeting of the American Diabetes Association, Detroit, Michigan, 1-6 June 1989.
20.
358
Fabre J, Balant LP, Dayer P, Fox HM, Vernet AT: The kidney in mature onset diabetes mellitus: a clinical study of 510 patients. Kidney Int 21:730-78, 1982 Parving H-H, Smidt UM, Frisberg B, Bonnevie-Nielsen V, Andersen AR: A prospective study of glomerular filtration rate and arterial blood pressure in insulin-dependent diabetics with diabetic nephropathy. Diabetologia 20:45761, 1981 Viberti GC, Bilous RW, Mackintosh D, Keen H: Monitoring glomerular function in diabetic nephropathy: a prospective study. Am J Med 74:256-64, 1983 Obesity: a report of The Royal College of Physicians. J Royal Coll Physicians Lond 17:5-65, 1983 Deckert T, Andersen AR, Christiansen JS, Andersen JK: Course of diabetic nephropathy: factors related to development. Ada Endocrinol 97 (Suppl. 242): 14-15, 1981 Mauer SM, Steffes MW, Goetz FC, Sutherland DER, Brown DM: Diabetic nephropathy: a perspective. Diabetes 32 (Suppl. 2):52-55, 1983 Mogensen CE: Renal function changes in diabetes. Diabetes 25:872-79, 1976 Mogensen CE, Christensen CK, Vittinghus E: The stages in diabetic renal disease with emphasis on the stage of incipient diabetic nephropathy. Diabetes 32 (Suppl. 2): 64-78, 1983 Fishman AP: Heart failure. In Cec/7 Textbook of Medicine. Wyngaarden JB, Smith LH Jr., Eds. Philadelphia, PA, Saunders, 1985, p. 189-210 World Health Organization: Arterial Hypertension. Report of a WHO Expert Committee. Geneva, World Health Org., 1978 (Tech. Rep. Sen, no. 628) Blankenship G, Skyler JS: Diabetic retinopathy: a general survey. Diabetes Care 1:127-37, 1978 Ewing DJ, Martyn CN, Young RJ, Clarke BF: The value of autonomic function tests: 10 yr experience in diabetes. Diabetes Care 8:491-89, 1985 Neumann C, Amon LC, Canani LH, Gross JL, Schmid H: Testes autonomicos no diabete me'lito: padronizacao dos valores normais. In Anais do Congr. Brasileiro de Diabetes, VI, Salvador, Brazil, Soc. Brasileira de Diabetes, 1987 ChantlerC, BarrattTM: Estimation of glomerular filtration rate from plasma clearance of 51-chromium edetic acid. Arch Dis Child 47:613-17, 1972 Mattock MB, Keen H, Viberti GC, El Gohary MR, Murrells TJ, Scott GS, WingJR, Jackson PG: Coronary heart disease and urinary albumin excretion in type 2 (non-insulindependent) diabetic patients. Diabetologia 31:82-87, 1988 Schriever H, Gambino SR: Protein turbidity produced by trichloroacetic acid and sulfosalicylic acid at varying temperatures and varying ratios of albumin and globulin. Am I Clin Pathol 44:667-72, 1965 Chasson A, Grady H, Stanley M: Determination of creatinine by means of automatic chemical analysis. Am I Clin Pathol 35:83-88, 1961 Fabiny DL, Ertingshausen G: Automated reaction-rate method for determination of serum creatinine with the CentrifiChem. Clin Chem 17:696-700, 1971 Viberti GC, Keen H: The patterns of proteinuria in diabetes mellitus: relevance to pathogenesis and prevention of diabetic nephropathy. Diabetes 33:686-92, 1984 Stein ACR: Estudo comparativo e evolutivo de pacientes diabe"ticos com diferentes graus de proteinuria. In Curso DIABETES CARE, VOL. 14, NO. 5, MAY 1991
R. FRIEDMAN AND J.L. GROSS
c/e Pos-Graduagao em Medicina: Nefrologia, Porto Alegre, Brasil, Universidade Federal do Rio Grande do Sul, 24. 1984, p. 85 21. Brochner-Mortensen J: Routine methods and their reliability for assessment of glomerular filtration rate in adults. Dan Med Bull 25:181-202, 1978 25. 22. Carrie BJ, Golbetz HV, Michael AS, Myers BD: Creatinine: an inadequate filtration marker in glomerular dis26. eases. Am} Med 69:177-82, 1980 23. Dodge WF, Travis LB, Daeschner CW: Comparison of endogenous creatinine clearance with insulin clearance.
DIABETES CARE, VOL. 14, NO. 5, MAY 1991
Am ) Dis Child 113:683-92, 1967 Friedman R, Azevedo MJ, Gross JL: Is endogenous creatinine clearance still a reliable index of glomerular filtration rate in diabetic patients? Braz J Med Biol Res 21: 941_44, 1988 Shemesh O, Golbetz H, Kriss JP, Myers B: Limitations of creatinine as a filtration marker in glomerulopathic patients. Kidney Int 28:830-88, 1985 Grenfell A, Watkins PJ: Clinical diabetic nephropathy: natural history and complications. Clin Endocrinol Metab 15:783-805, 1986
359
