Radionuclide Staging of Renal Function in Type 1 Diabetes Mellitus

Abstract

Aim. The aim of this study was to assess renal function in different stages of type 1 diabetes mellitus by radionuclide methods. Additionally, glomerular and tubular functions were correlated with urinary albumin (UAER) and N-acetyl-β-D-glucosaminidase (NAGA) excretion rates. Patients and methods. Fifty-three patients with diabetes mellitus were classified into four groups: normoalbuminuric (NA, 18 patients), microalbuminuric (MiA, 12 patients), macroalbuminuric (MaA, 13 patients), and chronic renal failure group (CRF, 10 patients). Glomerular filtration rate (GFR) was estimated by diethylenetriamine pentaacetic acid-technetium 99m (^99mTc-DTPA) clearance rate while tubular function was calculated as a percentage of net injected activity fixed in both kidneys, 4 h after intravenous injection of dimercaptosuccinate acid-technetium 99m (^99mTc-DMSA). Additionally, ^99mTc-DTPA clearance was correlated with estimated GFR (eGFR) by using modified Modification of Diet in Renal Disease (MDRD) Study Group formula. Results. ^99mTc-DTPA clearance and ^99mTc-DMSA fixation were found significantly higher in normoalbuminuric group (p < 0.05 and p < 0.02, respectively), unchanged in microalbuminuric group (p > 0.05, p > 0.05), and decreased in both macroalbuminuric (p < 0.0001, p < 0.00001) and chronic renal failure group (p < 0.0001, p < 0.00001). Renal function was denoted as normal, increased (hyperfunction), or decreased (hypofunction). It was found normal in a high percentage of patients with normalbuminuria (filtration 44.4%, fixation 72.2% pts) and microalbuminuria (66.7% and 66.7%). Renal hyperfunction was not only found frequent in normalbuminuric group (55.6% and 27.8%), but was also recorded in microalbuminuric group (8.3% and 8.3%). Renal hypofunction was present in all macroalbuminuric patients and in one-quarter of those with microalbuminuria as well. Such distribution of renal function conditions indicated normalbuminuric and microalbuminiric groups functionally heterogeneous. Regression analysis showed a significant correlation between ^99mTc-DTPA clearance and eGFR in MaA and CRF groups only. Although urinary NAGA excretion rate was shown as a less sensitive staging parameter, being significantly increased when compared to control group only in MaA and CRF groups (p < 0.05), it significantly correlated with ^99mTc-DTPA clearance rate (r = −0.485, p = 0.0004) and ^99mTc-DMSA tubular fixation (r = −0.526, p = 0.0002). Conclusions. The results of this study favor the performance of radionuclide studies together with the determination of urinary albumin excretion rate in patients with type 1 diabetes mellitus in order to achieve more reliable staging of diabetic kidney disease. The demonstration of glomerular hyperfiltration and tubular hyperfunction by radiopharmaceuticals contributes to the early detection of diabetic kidney disease, while the quantification of renal function enables the follow-up of the progressive function loss in the later course of the disease.

Keywords:

• type 1 diabetes mellitus
• renal function
• ^99mTc-DMSA
• ^99mTc-DTPA
• urinary albumin
• NAGA

INTRODUCTION

In patients with type 1 diabetes mellitus, none of the body organs is spared from structural damage and subsequent functional impairment. Among them, the kidney is very early affected by a variety of pathogenetic factors. Therefore, the quantification of renal function in these patients is of great importance, as it enables the application of the appropriate treatment for prevention of disease progression. Nowadays, the staging of diabetic kidney disease on the basis of urinary albumin excretion rate (UAER) is generally applied. Renal injury may be detected by biochemical markers such as urinary albumin, immunoglobulin G or transferrin for glomerular injury, and urinary beta-2 microglobulin, alpha-1 microglobulin, retinol binding protein and N-acetyl-β-D-glucosaminidase (NAGA) for tubular impairment. Glomerular filtration rate (GFR) may be calculated on the basis of endogenous creatinine in 24 h urine collection, serum cystatin C, cysteine proteinase, radionuclide methods or formulas, such as Modification of Diet in Renal Disease (MDRD) Study[[1,], [2]] and Cockcroft-Gault formula. The MDRD formula has proven to be more accurate than the Cockroft-Gault formula for prediction of GFR in diabetic patients[[3]] and in patients with low GFR.[[4]]

Radionuclide methods, depending on specific kidney pathways of radiopharmaceuticals, can provide separate information on glomerular and tubular function. Diethylenetriamine pentaacetic acid-technetium 99m (^99mTc-DTPA) clearance is a well-established quantitative marker of glomerular function, eliminated from the blood mostly by glomerular filtration rate.[[5]] Predominant fixation of dimercaptosuccinate acid-technetium 99m (^99mTc-DMSA) in the renal tubular cells offers estimation of tubular function.[[6,], [7]]

The aim of this study was to define renal function stages by radionuclide methods in patients with type 1 diabetes mellitus. In addition, glomerular filtration rate measured by radionuclide method was compared with MDRD-GFR, and glomerular and tubular function was correlated with urinary albumin and NAGA excretion rates.

PATIENTS AND METHODS

Patients

The study was carried out on 53 patients (22 males, 31 females) with type 1 diabetes mellitus of both sexes who had suffered from the disease for over five years. All of them had sterile urine and had no renal disease unrelated to diabetes. Patients were divided into four groups according to urinary albumin excretion rate and serum creatinine levels:

1. normoalbuminuric (NA), UAER< 20 μg/min;

2. microalbuminuric (MiA), UAER 20–200 μg/min;

3. macroalbminuric (MaA), UAER > 200 μg/min; and

4. chronic renal failure group (CRF), with serum creatinine above 130 μmol/L.

Control group comprised 20 clinically healthy subjects (see Table 1). The study protocol was approved by the Medical Ethics Committee of the Clinical Center, University of Niš, Serbia, and all patients personally signed a written consent to take part in the study.

Table 1 Patients' data

Groups	n	Age (years)	Disease duration (years)	s-Glucose (mmol/L)	s-Creatinine (μmol/L.)
1. Control	20	41.5 ± 3.2	—	4.34 ± 0.94	69.3 ± 6.9
2. NA	18	39.2 ± 7.5	12.4 ± 6.6	13.0 ± 5.7^b	71.8 ± 13.6
3. MiA	12	37.2 ± 12.2	14.5 ± 6.7	16.1 ± 5.3^b	79.9 ± 12.2^c
4. MaA	13	48.8 ± 12.7^e	18.4 ± 4.5^e	13.1 ± 4.1^b	97.9 ± 15.8^a,^d,^f
5. CRF	10	47.0 ± 14.1^e,^g	20.8 ± 6.6^g	14.9 ± 5.9^b	191 ± 45^a,^d,^f,^h

Abbreviations: NA = normoalbuminuria, MiA = microalbuminuria, MaA = macroalbuminuria, CRF = chronic renal failure.

Vs. control: ^ap < 0.0001, ^bp < 0.001, ^cp < 0.05; vs. NA: ^dp < 0.0001, ^ep < 0.05; vs. MiA: ^fp < 0.005, ^gp < 0.05; vs. MaA: ^hp < 0.001.

Methods

Radionuclide Assessment of Renal Function

Renal function was assessed by ^99mTc-DTPA clearance and ^99mTc-DMSA tubular uptake (fixation). Both radiopharmaceuticals (Institute of Nuclear Sciences “Vinča”, Serbia) were injected intravenously in the activity of 1.85 MBq/kg/BW.

^99mTc-DTPA clearance was determined from one compartment system using a single sample method. Blood samples were taken on 3 h because at that time, the distribution volume of ^99mTc-DTPA correlates very closely with GFR.[[8]] Individual clearance rates were normalized on 1.73m² of the body surface.

Tubular fixation of ^99mTc-DMSA was measured four hours after the injection of radiopharmaceutical. Renal counts were acquired within one minute on 128×128 computer matrix in posterior patient projection using computerized gamma camera (Pho V, Siemens, Germany). Counts accumulated in each kidney were corrected by perirenal background counts, tissue absorption, and radionuclide decay. Tubular uptake of ^99mTc-DMSA was calculated as a percent of net injected activity that was fixed in both kidneys.[[9]]

Estimated GFR (eGFR) Using MDRD Formula

Glomerular filtration rate based on serum creatinine, age and sex was estimated using a modified Modification of Diet in Renal Disease Study equation, which gives normalized eGFR on 1.73m² of body surface area.[[1,], [2]]

Urinary NAGA Measurement

Urinary NAGA activity was determined in potassium acetate buffer with p-nitrophenol-N-acetyl-β-glucosaminide as a substrate. The amount of p-nitrophenol formed was measured by reading absorbance at 405 nm.[[10]] Obtained values were normalized to urinary creatinine.

Urinary Albumin Measurement

Urinary protein was measured by biuret method, while urinary albumin was assessed by immunoturbidimetric method using Ames “DCA 2000” analyzer (Germany), both in 24 h collections.

Statistics

Group data were expressed as mean±SD and median (NAGA), while the incidence rates where shown in percents. Student's t-test was applied to examine group value differences for all parameters in question, except urinary NAGA excretion values, where Mann-Whitney's U test was used. Criteria for glomerular hyper and hypofiltration as well as tubular hyper and hypofixation were ^99mTc-DTPA clearance and ^99mTc-DMSA uptake values out of control mean ± 2SDs. Pearson's coefficient was calculated to evaluate the correlation between different parameters in the same group. p values lower than 0.05 indicated statistical significance.

RESULTS

Mean group value of ^99mTc-DTPA clearance and ^99mTc-DMSA tubular uptake significantly increased in normoalbuminuric, remained unchanged in microalbuminuric, but decreased in macroalbuminuric and in patients with chronic renal failure in comparison to control group. As it is seen in Table 2 and Figure 1, when the obtained results in neighboring diabetic groups are compared (MiA vs. NA, MaA vs. MiA, and CRF vs. MaA), a progressive and significant decrease is detected.

Table 2 Mean ^99mTc-DTPA clearance and eGFR values and correlation between those values in type 1 diabetes mellitus patients

Groups	n	DTPA clearance (mL/min/1.73m^2)	eGFR (mL/min/1.73m^2)	eGFR/DTPA clearancer	p
1. Control	20	123 ± 7.0	91.7 ± 21.4	0.265	0.288
2. NA	18	131 ± 14.0^d	97.1 ± 29.4	0.212	0.431
3. MiA	12	116 ± 19.0^h	86.0 ± 24.3	0.317	0.342
4. MaA	13	83.4 ± 11.7^b,^f,^j	69.8 ± 21.4^c,^g	0.651	0.016
5. CRF	10	27.8 ± 9.3^b,^f,^j,^l	24.3 ± 8.0^a,^e,^i,^k	0.754	0.003

Vs. control: ^ap < 0.00001, ^bp < 0.0001, ^c p < 0.01, ^dp < 0.05; vs. NA: ^ep < 0.00001, ^fp < 0.0001, ^gp < 0.01, ^hp < 0.05; vs. MiA: ⁱp < 0.00001, ^jp < 0.0001; vs. MaA: ^kp < 0.0001, ^lp < 0.001.

Figure 1. ^99mTc-DMSA renal uptake values in type 1 diabetes mellitus patients. vs. control: ^ap < 0.00001, ^bp < 0.02; vs. NA: ^cp < 0.00001, ^dp < 0.01; vs. MiA: ^ep < 0.00001, ^fp < 0.0005; vs. MaA: ^gp < 0.001.

GFR determined by MDRD formula showed lower values than ^99mTc-DTPA clearance in patients from the control (p < 0.0001), NA (p < 0.0001), and MiA (p < 0.001) groups, while values in both groups with reduced GFR (MaA and CRF) were similar to those obtained by ^99mTc-DTPA clearance and correlated well with them (Table 2). Estimated GFR showed higher value in the NaA group than in both control and MiA groups, but the differences were insignificant. It also failed to show a considerable difference between MaA and MiA groups.

The incidence rates of normo, hyper-, and hypofiltraton as well as of normo, hyper-, and hypofixation are displayed in Figures 2 and 3, respectively. Normalbuminuric group involved patients with normal and increased renal function, microalbuminuric group comprised patients with various function states (normal function, hyperfunction and hypofunction), while both macroalbuminuric and CRF groups comprised only patients with renal hypofunction.

Figure 2. The incidence of hyper, normo, and hypofiltration of ^99mTc-DTPA in type 1 diabetes mellitus patients.

Figure 3. The incidence of hyper, normo, and hypofixation of ^99mTc-DMSA in type 1 diabetes mellitus patients.

Urinary albumin and NAGA excretion rates are presented in Table 3. UAER was higher in the MaA group than in the MiA group, but there was no further increase in CRF group related to the MaA group. As far as NAGA urinary activity is concerned, it was significantly increased only in macroalbuminuric and chronic renal failure patients.

Table 3 Urinary excretion rates of albumin and NAGA in type 1 diabetes mellitus patients

Groups	UAER (g/24 h) (mean±SD)	NAGA (U/mmolCreatinine) median (range)
1. Control	—	0.878 (0.664–2.444)
2. NA	—	0.878 (0.156–2.989)
3. MiA	0.150 ± 0.100	1.690 (0.685–3.200)
4. MaA	1.380 ± 0.920^a	1.720 (1.340–5.150)^b
5. CRF	1.300 ± 1.465^b	2.700 (1.250–8.460)^b

Vs. MiA: ^ap < 0.0001, ^bp < 0.05.

Using regression analysis, a significant correlation was found between ^99mTc-DTPA clearance and ^99mTc-DMSA uptake values (r = 0.872, p = 0.00001) in all diabetic patients. Urinary NAGA excretion rate significantly correlated with ^99mTc-DTPA clearance (r = −0.485, p = 0.0004) and ^99mTc-DMSA renal uptake (r = −0.526, p = 0.0002) in all examined patients. However, a correlation between ^99mTc-DTPA clearance and UAER was proved only in the MaA group (r = −0.580, p = 0.008).

DISCUSSION

In this study, glomerular function was assessed by measuring blood clearance rate of ^99mTc-DTPA. This radiopharmaceutical is a well-documented marker of glomerular filtration rate, as its clearance rate has proven to correlate well with that of inulin.[[11]] ^99mTc-DMSA renal uptake is a marker of global tubular mass function.[[7,], [9]] After active extraction from peritubular circulation it is bound in a high percentage to tubular cells,[[9]] whereas only its smaller amount, which is non-bound to plasma proteins, is actually filtered through glomeruli.[[12]] Therefore, renal uptake of this radiopharmaceutical expressed in percentage of the applied net activity can be considered a semiquantitative marker of tubular function. Both ^99mTc-DTPA clearance and ^99mTc-DMSA renal uptake were found reduced in some kidney diseases.[13–15]

In the current study, mean values of ^99mTc-DTPA clearance and ^99mTc-DMSA tubular fixation were increased in normoalbuminric, unchanged in microalbuminuric, and decreased in macroalbuminuric groups of patients. Positive correlation between ^99mTc-DTPA clearance and ^99mTc-DMSA uptake results from the fact that both glomerular and tubular damage occurs in diabetes mellitus.

Glomerular filtration rate was also estimated by MDRD equation.[[1]] This formula, based on the multicenter trial of the Modification of Diet in Renal Disease, is widely acknowledged for GRF calculation in adult patients. It embodies age, serum creatinine and sex variables, and is indexed for body surface area and race. It is recommended that MDRD formula may offer a more precise estimate of GFR than Cockcroft-Gault formula,[[3,], [4]] but when compared to radionuclide methods that give real GFR, it may underestimate GFR in patients with normal renal function or GFR ≥ 60 mL/min/1.73m².[[3]] In this study, generally lower GFR estimated by MDRD equation than that measured by ^99mTc-DTPA clearance in healthy persons as well as in patients with normo and microalbuminria was shown. In patients with reduced GFR, values of ^99mTc-DTPA clearance and eGFR were similar and significantly correlated. Lower values of GFR, calculated by MDRD formula, than that determined by radionuclide renal clearance were also found in microalbuminric patients with diabetes mellitus.[[16]] Correlation between eGFR and GFR measured by ⁵¹Cr-EDTA clearance in diabetic patients with moderately (GFR < 60 mL/ min/1.73 m²) or severely reduced renal function (GFR < 30 mL/min/1.73 m²) was found previously.[[17]] According to the data obtained in our study, it seems that MDRD formula underestimates GFR not only in patients with normal but also in patients with the increased GFR. This may be the result of serum creatinine concentration, which is affected by factors other than glomerular filtration. On the other hand, ^99mTc-DTPA, being neither reabsorbed nor secreted by renal tubuli and not dependent on muscle mass and food intake, enables the precise measurement of GFR.

Analysis of individual values showed increased glomerular filtration rate measured by ^99mTc-DTPA in more than one-half of normalbuminuric patients and also in a small number of patients with microalbuminuria. This finding is the consequence of various pathogenetic mechanisms involved in the development of hyperfiltration. According to some authors, hyperfiltration is induced by vasoactive substances, which lead to renal vasodilatation[[18]] associated with the increased blood flow, resulting in the augmentation of intraglomerular capillary pressure. Enlarged filtration surface also leads to the elevation of glomerular filtration rate,[[19]] while renal hypertrophy is being considered a prerequisite for hyperfiltration development.[[20]] The pathogenetic effect of supranormal filtration results in nephron injury followed by renal function decline.[[21]] Thus, normal glomerular function in currently studied normo and microalbuminuric patients may represent a decline of the previously increased glomerular filtration rate. Hyperfiltration incidence in normoalbuminiruc patients is higher in the present study than previously described.[[22,], [23]] This may result from poor glycemic control of our patients, whose serum glucose levels diverged from 7.6 to 19.0 mmol/L in the past two years. On the contrary, predominantly unchanged glomerular filtration rate and low incidence of hypofiltration in microalbuminuric patients is in line with the other studies.[[22,], [24]]

Considering ^99mTc-DMSA fixation values, tubular hypefuncton was found in more than one-quarter of normoalbuminuric patients and with the same frequency (8.3%) as glomerular hyperfiltration in microalbuminuric group. The appearance of tubular hyperfunction in the early phase of the disease is the consequence of kidney growth.[[25]] According to the “tubular hypothesis” of diabetic kidney disease, initially occurring tubular hyperfunction is followed by glomerular hyperfiltration through the action of tubuloglomerular feedback mechanism.[[25]]

Hypofunction was recorded by both ^99mTc-DTPA and ^99mTc-DMSA in 25.0% of patients with microalbuminuria. A similar incidence of decreased GFR measured by inulin clearance in patients with microalbuminuria was also previously reported.[[22]] In all patients with macroalbuminura, a considerable GFR and tubular function decrease was found, being more evident in renal failure. Renal function decrease in macroalbuminuric patients occurs during the disease evolution.[[15,], [26]] Pathogenetic mechanism of renal function injury in diabetes mellitus seems to be very complex. For instance, hyperfiltration itself injures the glomerulus because of capillary hypertension.[[18]] The main structural damages are the thickening of glomerular basement membrane, progressive expansion of mesangium, and arteriolar hyalinosis leading to filtration surface loss.[[19]] Tubular injury develops due to ischemia caused by blood vessels' obliteration or external compression.[[27]] Apart from this, tubuli are exposed to toxic effects of various substances, such as cytokines from the injured glomeruli, excessive amount of ammonia and reactive oxygen species generated due to hypermetabolism, and degradative products of filtered and reabsorbed proteins. Such alterations trigger the sequence of factors leading to tubular damage as the disease progresses.[[28]] Together with interstitial damage, tubular destruction is considered to be a prominent pattern in close connection with the disease progression leading to renal failure.

The results obtained by ^99mTc-DTPA and ^99mTc-DMSA in this study were compared with findings of other kidney damage markers, such as NAGA and UAER. Urinary excretion of NAGA is widely accepted as a marker of tubular cell damage.[[29]] This finding was observed even in the early stage of diabetic kidney disease.[[30,], [31]] However, reports on its predictive value for the development of diabetic nephropathy are controversial.[[32,], [33]] In this study, a significant inverse correlation of urinary NAGA excretion in relation to ^99mTc-DMSA renal uptake and ^99mTc-DTPA clearance was found. However, unchanged significantly in normoalbuminuric and microalbuminuric group, NAGA urinary excretion was found to be less sensitive in the early detection of the kidney function impairment.

Positive correlation between UAER and GFR was found in some studies[[34,], [35]] but not in others.[[24,], [36]] In this study, such a correlation was recorded only in macroalbuminuric and CRF groups. Its absence in other stages can have multifactorial origin. For instance, apart from glomerular vasodilatation that increases filter permeability and transglomerular passage of albumin,[[34]] decreased albumin endocytosis by tubular cells can account for the occurrence of this disorder. Additionally, there is no further increase of albuminuria rate in renal failure due to the loss of large number of glomeruli.

This study supports the capability of radionuclide methods to detect renal hyperfunction as an initial and transient disorder in type 1 diabetes. Glomerular and tubular hyperfunction is, with the help of these methods, found in the early stages of diabetic kidney disease. Such findings enable the application of an adequate treatment to prevent disease progression. After renal function decrease, its quantification by radionuclide methods gives an insight into residual function. The decrease of ^99mTc-DTPA clearance under 60 mL/1.73 m² represents the onset of renal failure occurrence, while clearance values remain a reliable functional indicator. It is demonstrated that^99mTc-DMSA renal uptake and ^99mTc-DTPA clearance studies could be used for an early detection and follow-up of kidney disease in patients with type 1 diabetes mellitus.

ACKNOWLEDGMENT

This work was supported by the Ministry of Science and Environmental Protection of Serbia, grant 145004.