Original Investigation Predicting Kidney Transplantation Outcomes Using Proteinuria Ascertained From Spot Urine Samples Versus Timed Urine Collections Hari Talreja, MD,1,* Ayub Akbari, MD, MSc,1,2,3,* Christine A. White, MD, MSc,4 Tim O. Ramsay, PhD,3 Swapnil Hiremath, MD, MPH,1,2,3 and Greg Knoll, MD, MSc1,2,3 Background: Proteinuria has been associated with transplant loss and mortality in kidney transplant recipients. Both spot samples (albumin-creatinine ratio [ACR] and protein-creatinine ratio [PCR]) and 24-hour collections (albumin excretion rate [AER] and protein excretion rate [PER]) have been used to quantify protein excretion, but which measurement is a better predictor of outcomes in kidney transplantation remains uncertain. Study Design: Observational cohort study. Setting & Participants: Tertiary care center, 207 kidney transplant recipients who were enrolled in a prospective study to measure glomerular filtration rate. Consecutive patients who met inclusion criteria were approached. Predictors: ACR and PCR in spot urine samples, AER and PER in 24-hour urine collections. Outcomes: Primary outcome included transplant loss, doubling of serum creatinine level, or death. Measurements: Urine and serum creatinine were measured using a modified Jaffe´ reaction that had not been standardized by isotope-dilution mass spectrometry. Urine albumin was measured by immunoturbidimetry. Urine protein was measured by pyrogallol red molybdate complex formation using a timed end point method. Results: Mean follow-up was 6.4 years and 22% developed the primary end point. Multivariable-adjusted areas under the receiver operating characteristic curves were similar for the different protein measurements: ACR (0.85; 95% CI, 0.79-0.89), PCR (0.84; 95% CI, 0.79-0.89), PER (0.86; 95% CI, 0.80-0.90), and AER (0.83; 95% CI, 0.780.88). C Index values also were similar for the different proteinuria measurements: 0.87 (95% CI, 0.79-0.95), 0.86 (95% CI, 0.79-0.94), 0.88 (95% CI, 0.82-0.94), and 0.86 (95% CI, 0.77-0.95) for log(ACR), log(PCR), log(PER), and log(AER), respectively. Limitations: Single-center study. Measurement of proteinuria was at variable times posttransplantation. Conclusions: Spot and 24-hour measurements of albumin and protein excretion are similar predictors of doubling of serum creatinine level, transplant loss, and death. Thus, spot urine samples are a suitable alternative to 24-hour urine collection for measuring protein excretion in this population. Am J Kidney Dis. 64(6):962-968. ª 2014 by the National Kidney Foundation, Inc. INDEX WORDS: Proteinuria; albuminuria; kidney graft survival; outcomes; protein-creatinine ratio (PCR); albumin-creatinine ratio (ACR); albumin excretion rate (AER); protein excretion rate (PER); spot urine sample; 24-hour urine collection; end-stage renal disease (ESRD).

B

oth proteinuria and albuminuria are strong risk factors for the development of kidney failure in the general population.1-3 Similarly, in kidney transplant recipients, total urinary albumin and protein excretion have been associated with progressive

From the 1Division of Nephrology, Department of Medicine, University of Ottawa; 2Kidney Research Centre and 3Clinical Epidemiology Program, Ottawa Hospital Research Institute, Ottawa; and 4Division of Nephrology, Department of Medicine, Queen’s University, Kingston, Canada. * H.T. and A.A. contributed equally to this work. Received January 20, 2014. Accepted in revised form July 28, 2014. Originally published online October 7, 2014. Address correspondence to Greg Knoll, Division of Nephrology, The Ottawa Hospital, Riverside Campus, 1967 Riverside Drive, Ottawa, Ontario, Canada K1H 7W9. E-mail: [email protected]  2014 by the National Kidney Foundation, Inc. 0272-6386/$36.00 http://dx.doi.org/10.1053/j.ajkd.2014.07.027 962

decreases in kidney function, transplant loss, and mortality.4-6 At 1 year posttransplantation, proteinuria with protein excretion . 150 mg/d is present in w40% and therefore is a common complication of kidney transplantation. Studies to date have shown that proteinuric kidney transplant recipients have a 2- to 5-fold increased risk for developing transplant failure.6 Therefore, estimation of urinary albumin or protein loss remains an important tool for prognostication. Timed urine specimens are considered the gold standard for measuring proteinuria and albuminuria. Because of convenience and cost, spot urine samples such as albumin-creatinine ratio (ACR) and proteincreatinine ratio (PCR) increasingly are being used for quantification of albuminuria and proteinuria. A recent KDIGO (Kidney Disease: Improving Global Outcomes) guideline suggests that ACR and PCR may be reasonable alternatives to 24-hour urine collection in kidney transplant recipients.7 These spot measures of protein excretion have been reported to Am J Kidney Dis. 2014;64(6):962-968

Spot Urine Protein and Kidney Transplant Outcomes Table 1. Baseline Characteristics All Patients (N5207)a

ACR , 30 (n5104)

ACR 30-300 (n575)

ACR . 300 (n521)

P for Trend

Mean age (y)

59.9 6 12

60 6 12

60.2 6 12

59.1 6 14

0.9

Male

135 (65)

62 (59.6)

51 (68)

16 (76.2)

0.5

Race White Black Asian Other

190 (92) 5 (2) 7 (3) 5 (2)

92 (89) 5 (4.8) 4 (3.9) 3 (2.9)

71 0 2 2

(95) (0) (3) (3)

20 0 1 0

(95) (0) (5) (0)

0.8 0.5 0.9 0.5

Weight (kg)

79.9 6 17

79 6 19

82.5 6 16

78.3 6 17

Height (cm)

167 6 10.2

168 6 10.7

168 6 9.3

164 6 11

0.2

Body surface area (m2)

1.89 6 0.2

1.89 6 0.2

1.92 6 0.2

1.84 6 0.2

0.9

Living donor

79 (38)

36 (34.6)

30 (40)

8 (38)

0.7

Time from Tx to study entry (y)

5.3 [9.1]

4.6 [5.9]

5.3 [8.9]

7.8 [10.7]

0.05

0.6

F/U after study entry (y)

6.2 [1.6]

6.3 [1.3]

6.5 [0.88]

6.2 [0.82]

0.3

Primary Tx

181 (87)

88 (84.6)

66 (88)

20 (95)

0.7

Cause of kidney disease Diabetes Polycystic kidney disease Glomerulonephritis Hypertension Other

27 (13) 33 (16) 54 (26) 12 (6) 81 (39)

13 (12.5) 18 (17.3) 26 (25) 5 (4.8) 42 (40.4)

7 12 20 4 32

(9) (16) (27) (5) (43)

4 2 6 2 7

(19) (10) (29) (10) (33)

0.8 0.5 0.8 0.5 0.9

Medication ACEi or ARB Prednisone Tacrolimus Cyclosporine Azathioprine Mycophenolate mofetil Sirolimus

72 (35) 204 (99) 80 (39) 107 (52) 35 (17) 140 (68) 6 (3)

31 (30) 103 (99) 40 (39) 51 (49) 20 (19) 72 (69) 1 (1)

28 74 28 41 11 51 3

(37) (99) (37) (55) (15) (68) (4)

10 (48) 20 (95) 8 (38) 12 (57) 3 (14) 13 (62) 1 (5)

0.3 0.9 0.9 0.6 0.5 0.8 0.2

GFR category $90 mL/min/1.73 m2 60-89 mL/min/1.73 m2 30-59 mL/min/1.73 m2 15-29 mL/min/1.73 m2 ,15 mL/min/1.73 m2

16 (8) 81 (39) 88 (42) 18 (9) 4 (2)

9 48 45 2 0

7 28 29 10 1

(9) (37) (39) (13) (1)

0 (0) 4 (19) 10 (48) 4 (19) 3 (14)

0.4 0.2 0.9 0.002 ,0.001 ,0.001

(9) (46) (43) (2) (0)

PER (mg/d)

210 [370]

100 [170]

270 [280]

1,330 [1,330]

PCR (mg/g)

150 [216]

61 [82]

209 [179]

1,003 [901]

,0.001

AER (mg/d)

37 [124]

16 [20]

81 [165]

738 [968]

,0.001

Doubling of serum creatinine

14 (7)

2 (2)

6 (8)

5 (24)

,0.001

Transplant loss

14 (7)

2 (2)

4 (5)

7 (33)

,0.001

Death

17 (8)

5 (5)

8 (11)

3 (14)

0.1

Note: Continuous values given as mean 6 standard deviation or median [interquartile range]; categorical values expressed as number (percentage). ACR given in mg/g of creatinine. Abbreviations: ACEi, angiotensin-converting enzyme inhibitor; ACR, albumin-creatinine ratio; AER, albumin excretion rate; ARB, angiotensin receptor blocker; F/U, follow-up; GFR, glomerular filtration rate; PCR, protein-creatinine ratio; PER, protein excretion rate; Tx, transplantation. a Seven patients did not have ACR measured.

perform equally as predictors of both renal and patient outcomes in the nontransplantation population.8 However, in the kidney transplant recipient population, the ability of spot measures of protein excretion to predict patient and transplant outcomes remains unclear. For instance, most transplant recipients are treated with steroids, which decrease muscle mass and Am J Kidney Dis. 2014;64(6):962-968

thus would lead to lower creatinine excretion, affecting ACR and PCR. Accordingly, the objective of this study was to determine whether ACR and PCR were better than standard 24-hour measurements of urine protein or albumin (protein excretion rate [PER] and albumin excretion rate [AER], respectively) in predicting outcomes in kidney transplant recipients. 963

Talreja et al Table 2. Hazard Ratio and C Index Values for the Primary Outcome

Table 3. Area Under the ROC Curve for the Primary Outcome Univariate Analysis

Multivariable Analysisa

ACR

0.78 (0.72-0.84)

0.85 (0.79-0.89)

PCR

0.82 (0.76-0.87)

0.84 (0.79-0.89)

Log(ACR) 2.61 (1.94-3.50) 1.82 (1.31-2.53) 0.87 (0.79-0.95)

AER

0.74 (0.68-0.80)

0.83 (0.78-0.88)

Log(PCR) 2.40 (1.89-3.04) 1.63 (1.22-2.17) 0.86 (0.79-0.94)

PER

0.79 (0.73-0.84)

0.86 (0.80-0.90)

Log(AER) 1.65 (1.38-1.98) 1.30 (1.07-1.58) 0.86 (0.77-0.95)

Note: Values are given as area under the curve (95% confidence interval). The primary outcome was a composite measure that included doubling of serum creatinine level, transplant loss (need for dialysis or repeat transplantation), or death. Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; PCR, protein-creatinine ratio; PER, protein excretion rate; ROC, receiver operator characteristic. a Adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate.

Univariate Analysis HR (95% CI)

Multivariable Analysisa HR (95% CI)

C Indexb (95% CI)

Log(PER) 2.51 (1.91-3.31) 1.66 (1.22-2.27) 0.88 (0.82-0.94) Note: The primary outcome was a composite measure that included doubling of serum creatinine level, transplant loss (need for dialysis or repeat transplantation), or death. HRs are given in terms of a 1–standard deviation increase in the variable. Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; CI, confidence interval; HR, hazard ratio; PCR, protein-creatinine ratio; PER, protein excretion rate. a Adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate. b The probability that a participant from the event group has a higher predicted probability of having an event than a participant from the nonevent group.

METHODS Study Population The prospective study was approved by the Ottawa Hospital Research Ethics Board and was designed to measure different markers of glomerular filtration rate (GFR) in adult kidney transplant recipients. In this analysis, we included adult kidney transplant recipients who had GFR measured and a 24-hour urine collection performed to measure PER or AER between April 2004 and December 2005 (for each patient, GFR and PER/AER measurements were performed on the same day). A total of 334 patients met the study criteria and consent was obtained from 261; a total of 54 patients withdrew from the study after consent was obtained but before investigations were conducted, leaving 207 patients. Inclusion criteria at the time of enrollment included: (1) stable transplant function (,0.4-mg/dL [,35-mmol/L] difference in serum creatinine between the 2 most recent values), and (2) at least 6 months posttransplantation. Those excluded: (1) were pregnant or breastfeeding, (2) had acute rejection within the preceding 3 months, (3) likely required renal replacement therapy within the next 3 months, or (4) were likely to die within the next 3 months. Baseline characteristics including age, sex, race, cause of kidney failure, date of transplantation, donor type (living or deceased), transplantation number (first, second, etc), and immunosuppression were recorded. The primary outcome was a composite measure that included doubling of serum creatinine level, transplant loss (need for dialysis or repeat transplantation), or death. The secondary outcome was doubling of serum creatinine level or transplant loss. Serum creatinine was measured approximately every 3 months, as is our routine in the transplantation clinic.

Laboratory Assessment Baseline urine and blood measurements were performed on the day of study entry. A 24-hour urine sample was collected for measurement of creatinine, PER, and AER. A morning urine sample was obtained for measurement of ACR and PCR. All measurements were performed on a Beckman Coulter LX20 Pro Clinical System using Beckman Coulter reagents. Urine and serum creatinine were measured using a modified Jaffé reaction that was not standardized to isotope-dilution mass spectrometry. Coefficients of variation for serum creatinine were 4.9% at 0.6 mg/dL 964

(55 mmol/L), 1.7% at 1.7 mg/dL (150 mmol/L), and 1.3% at 6.8 mg/dL (600 mmol/L). Coefficients of variation for urine creatinine were 1.7% at 69 mg/dL (6.1 mmol/L) and 1.5% at 167 mg/dL (14.8 mmol/L). Urine protein was measured by pyrogallol red molybdate complex formation using a timed end point method. Coefficients of variation for urine protein were 3.6% at 0.02 g/dL (0.2 g/L) and 2% at 0.077 g/dL (0.77 g/L). Urine albumin was measured by an immunoturbidimetric method. Coefficients of variation for urine albumin were 4.1% at 2.7 mg/dL (27 mg/L) and 3.1% at 12 mg/dL (120 mg/L). Patients were weighed and had their height measured on the day the 24-hour urine collection was completed. GFR was measured using plasma clearance of radiolabeled 99mtechnetium-diethylenetriaminepentaacetic acid (99mTc-DTPA).9

Statistical Analysis Continuous variables are expressed as mean 6 standard deviation, and categorical variables are expressed as number and percentage. Because proteinuria was not normally distributed, each measure of proteinuria (ACR, PCR, PER, and AER) was log transformed. P for trend was calculated by Cochran-Armitage trend test for categorical variables and linear regression for continuous variables. Receiver operating characteristic (ROC) curves were calculated for each measure of proteinuria. The area under the curve (AUC) was calculated from a model adjusted for age, sex, measured GFR (mGFR), and time from transplantation to study entry using Cox regression. The AUC for each protein measurement was compared to each other in a pair-wise manner. To account for multiple comparisons, P , 0.008 was considered statistically significant. Hazard ratio was calculated for each measure of proteinuria and adjusted for age, sex, mGFR, and time from transplantation to study entry. C Index10 (the probability that a participant from the event group has a higher predicted probability of having an event than a participant from the nonevent group) also was calculated. All analyses were performed using MedCalc, version 12 (MedCalc Software), and SAS, version 9.3 (SAS Institute Inc).

RESULTS A total of 207 patients were included in this study. Baseline characteristics of patients are presented in Table 1. Median duration of follow-up was 6.2 (interquartile range [IQR], 1.6) years. Five patients did not have PER measured, 6 patients did not have AER measured, and 7 patients did not have ACR measured. Most patients were men and of white race. Average age was 59.9 years and most had received Am J Kidney Dis. 2014;64(6):962-968

Spot Urine Protein and Kidney Transplant Outcomes Table 4. Pair-Wise Comparison of ROC Curves for Primary Outcome Univariate Difference (95% CI)

P

Multivariable Differencea (95% CI)

P

ACR vs AER PER PCR

0.021 (20.03 to 0.08) 0.020 (20.05 to 0.08) 0.032 (20.02 to 0.08)

0.4 0.6 0.2

0.005 (20.03 to 0.04) 0.004 (20.04 to 0.05) 0.004 (20.03 to 0.04)

0.8 0.8 0.8

PCR vs AER PER

0.053 (20.001 to 0.108) 0.051 (20.008 to 0.110)

0.06 0.09

0.001 (20.02 to 0.02) 0.008 (20.02 to 0.03)

0.9 0.5

AER vs PER

0.002 (20.05 to 0.05)

0.9

0.009 (20.02 to 0.03)

0.5

Note: The primary outcome was a composite measure that included doubling of serum creatinine level, transplant loss (need for dialysis or repeat transplantation), or death. Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; CI, confidence interval; PCR, protein-creatinine ratio; PER, protein excretion rate; ROC, receiver operator characteristic. a Adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate.

Am J Kidney Dis. 2014;64(6):962-968

level or transplant loss. Hazard ratios for each proteinuria measurement are shown in Table 5. As with the primary outcome, C index values were nearly identical, again indicating that no one model had higher predictive ability than another (Table 5). In univariate analysis, AUCs were similar for the 4 proteinuria measures (Table 6). After adjustment for age, sex, mGFR, and time from transplantation to study entry, all AUC values showed some improvement (Table 6). None of the comparisons between proteinuria measures was statistically significant at the prespecified P , 0.008 level (Table 7; Fig 2). 100 80

Sensitivity

a deceased donor transplant. Mean mGFR was 58.9 6 23.1 mL/min/1.73 m2. Glomerulonephritis (26%) was the most common cause of kidney failure, followed by diabetes in 13%. Almost all patients (99%) were on immunosuppression protocols that included prednisone. Angiotensin-converting enzyme inhibitors or angiotensin receptor blockers were used in 35% of patients at baseline. Median PER was 210 (IQR, 370) mg/d, whereas median AER was 37 (IQR, 124) mg/d. There were 117 (58%) patients with PER . 150 mg/d and 45 (22%) with PER . 500 mg/d. There were 33 (16%) patients with AER . 300 mg/d. Median PCR was 150 (IQR, 216) mg/g, and median ACR was 27.9 (IQR, 93.4) mg/g. There was a wide range of transplant function in all stages of chronic kidney disease as assessed by mGFR. There were 45 (22%) patients who reached the primary outcome of doubling of serum creatinine level, transplant loss, or death. Hazard ratios for each proteinuria measurement are shown in Table 2. The multivariable model did not improve after the addition of a second measurement of proteinuria (data not shown). C Index values were nearly identical for each measure of proteinuria, indicating that no one model had higher predictive ability than another (Table 2). In univariate analysis, AUCs were similar for all proteinuria measures (Table 3). After adjustment for age, sex, mGFR, and time from transplantation to study entry, all AUC values showed some improvement (Table 3). The range between the lowest AUC (AER, 0.83; 95% confidence interval [CI] 0.78-0.88) and highest AUC (PER, 0.86; 95% CI, 0.80-0.90) was minor (Table 3). None of the comparisons between proteinuria measures was statistically significant at the prespecified P , 0.008 level (Table 4; Fig 1). There were 28 (13.5%) patients who reached the secondary outcome of doubling of serum creatinine

60 ACR

40

AER PCR

20

PER 0 0

20

40

60

100-Specificity

80

100

Figure 1. Receiver operator characteristic curve for multivariable analysis of the primary outcome (a composite measure that included doubling of serum creatinine level, transplant loss [need for dialysis or repeat transplantation], or death. Analysis adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate. Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; PCR, protein-creatinine ratio; PER, protein excretion rate. 965

Talreja et al Table 5. Hazard Ratio and C Index Values for the Secondary Outcome Univariate Analysis HR (95% CI)

Multivariable Analysisa HR (95% CI)

C Indexb (95% CI)

Log(ACR) 2.97 (2.08-4.24) 2.13 (1.38-3.29) 0.90 (0.84-0.96) Log(PCR) 2.68 (2.00-3.57) 1.86 (1.26-2.73) 0.89 (0.82-0.96) Log(AER) 1.91 (1.51-2.40) 1.46 (1.12-1.90) 0.89 (0.81-0.96) Log(PER) 2.89 (2.04-4.11) 1.80 (1.20-2.70) 0.89 (0.82-0.96) Note: HRs are given in terms of a 1–standard deviation increase in the variable. The secondary outcome was a composite measure that included doubling of serum creatinine level or transplant loss (need for dialysis or repeat transplantation). Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; CI, confidence interval; HR, hazard ratio; PCR, protein-creatinine ratio; PER, protein excretion rate. a Adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate. b The probability that a participant from the event group has a higher predicted probability of having an event than a participant from the nonevent group.

DISCUSSION This study demonstrates that both spot proteinuria measurements (ACR and PCR) and 24-hour timed proteinuria measurements (AER and PER) have similar abilities to predict important clinical outcomes such as doubling of serum creatinine level, transplant loss, or death in kidney transplant recipients. After adjustment, the AUC for each proteinuria measurement was greater than 0.8, a value that is considered to show very good discrimination.11 These results are analogous to findings in patients with chronic kidney disease who have not received a kidney transplant, for whom spot and timed collections have been reported to perform equally for predicting all-cause mortality and renal outcomes.8 The novel aspect of our study is the direct comparison of spot urine protein measurements with the reference standard 24-hour collection in a kidney transplant population. Panek et al12 examined the performance of ACR and PCR in 500 kidney transplant recipients; in a mean follow-up of 2.8 years, ROC analysis found nearly identical results. The C statistic for the outcome of doubling of creatinine level, transplant loss, or death was 0.78 for ACR and 0.79 for PCR.12 These findings were very similar to our current study, but Panek et al12 did not report data for PER or AER. Nauta et al13 examined the performance of AER and PER in predicting death-censored transplant loss in 606 kidney transplant recipients. They found that the AUC for AER (0.78; 95% CI, 0.70-0.87) was significantly higher than that for PER (0.67; 95% CI, 0.59-0.76; P 5 0.001).13 These AUC values were lower than those we found for PER and AER, but the CIs were wide. The difference might be 966

Table 6. Area Under the ROC Curve for the Secondary Outcome Univariate Analysis

Multivariable Analysisa

ACR

0.83 (0.77-0.88)

0.91 (0.86-0.94)

PCR AER

0.87 (0.82-0.91) 0.82 (0.76-0.87)

0.90 (0.85-0.94) 0.90 (0.85-0.94)

PER

0.84 (0.78-0.89)

0.90 (0.85-0.94)

Note: Values are given as area under the curve (95% confidence interval). The secondary outcome was a composite measure that included doubling of serum creatinine level or transplant loss (need for dialysis or repeat transplantation). Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; PCR, protein-creatinine ratio; PER, protein excretion rate; ROC, receiver operator characteristic. a Adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate.

due to a relatively low number of events (deathcensored transplant loss occurred in 7% in Nauta et al13 and 13.5% in our study). Unfortunately, Nauta et al13 did not report the ROC analysis for ACR and PCR in addition to the AER and PER data. PCR and ACR had similar performance in our current analysis and in the study by Panek et al.12 However, National Kidney Foundation–Kidney Disease Outcomes Quality Initiative (NKF-KDOQI) guidelines suggest using ACR over PCR for screening and monitoring in the nontransplantation setting and recommend using PCR only if ACR is elevated.14 It has been suggested that this NKFKDOQI recommendation stems from the fact that the PCR cutoff is set too high and use of a lower cutoff of 150 mg/g may be more sensitive.15 Based on limited data in the transplant recipient population, the KDIGO kidney transplant guideline recommends using similar cutoffs to define albuminuria and proteinuria as are used in the general population.16 The type of protein excreted in urine varies with the type of kidney disease. For example, in glomerular diseases such as diabetic nephropathy, the predominant protein in urine is albumin, whereas in interstitial diseases, the predominant proteins are low-molecularweight proteins such as b2-microglobulin. If albumin was the only protein measured, this would lead to underestimation of total proteinuria. Our data suggest that the type of protein measured does not change the discriminatory ability to detect adverse patient outcomes. There are limited data for the use of spot urine measurements versus 24-hour urine collections in the routine management of kidney transplant recipients. Anecdotally, we are aware that some transplantation programs continue to measure PER or AER at routine intervals posttransplantation, whereas others use spot measurements such as ACR, PCR, or even the dipstick method. Spot samples certainly are more convenient Am J Kidney Dis. 2014;64(6):962-968

Spot Urine Protein and Kidney Transplant Outcomes Table 7. Pair-Wise Comparison of ROC Curves for Secondary Outcome Univariate Difference (95% CI)

P

Multivariable Differencea (95% CI)

P

ACR vs AER PER PCR

0.010 (20.05 to 0.07) 0.018 (20.06 to 0.09) 0.037 (20.02 to 0.09)

0.7 0.6 0.2

0.006 (20.02 to 0.03) 0.010 (20.02 to 0.04) 0.008 (20.02 to 0.04)

0.7 0.5 0.6

PCR vs AER PER

0.047 (20.01 to 0.10) 0.055 (20.00 to 0.11)

0.1 0.05

0.002 (20.02 to 0.02) 0.002 (20.01 to 0.02)

0.8 0.8

AER vs PER

0.007 (20.04 to 0.06)

0.8

0.004 (20.01 to 0.02)

0.6

Note: The secondary outcome was a composite measure that included doubling of serum creatinine level or transplant loss (need for dialysis or repeat transplantation). Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; CI, confidence interval; PCR, protein-creatinine ratio; PER, protein excretion rate; ROC, receiver operator characteristic. a Adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate.

Am J Kidney Dis. 2014;64(6):962-968

rejection. Fifth, only 13% of patients had diabetes as the cause of kidney failure and our data may not be generalizable to population with a high incidence of diabetic nephropathy. Sixth, we used a single measurement of serum creatinine performed in an outpatient setting for the outcome of doubling of serum creatinine level; however, it is unlikely that a patient in this setting would have acute kidney injury. In conclusion, independent of GFR, commonly used measures of urine protein quantification (ACR, PCR, PER, and AER) are very good at discriminating

100

80

Sensitivity

for patients and currently are less costly. Although there is good correlation between 24-hour and spot urine samples, the accuracy of ACR and PCR in transplant recipients recently has been questioned.17 The simultaneous assessment of both spot and timed urine protein and albumin measurements in this analysis suggests that for prognostication, spot measures provide similar information to the more cumbersome timed collections. However, further research is required on the diagnostic accuracy of ACR and PCR to determine whether current cutoffs are sufficiently accurate to guide major changes in management, such as need for a biopsy or change in immunosuppression. This study has several strengths. We accounted for the major influence of kidney function on kidney transplant outcomes by adjusting for mGFR rather than using a GFR estimate based on serum creatinine level, which has many limitations in the transplant recipient population.9 Urine samples were obtained in the morning rather than at random, which has been shown to improve accuracy.18 Timed urine collections were returned to the laboratory at the same time the spot urine sample was obtained, ensuring that testing was carried out under similar clinical circumstances. Limitations to our study also should be noted. First, the study was a single-center design, which may limit generalizability. Second, we selected patients with stable transplant function and thus our results do not apply to patients with proteinuria and acute deterioration in kidney function. Third, we measured proteinuria at variable times posttransplantation rather than a fixed time point such as 3 months after transplantation surgery. Although time posttransplantation will have an effect on outcomes such as transplant loss and death, we adjusted for this potential confounding factor in the ROC analysis. Fourth, we had only 45 outcomes, so we could not control for many factors such as different immunosuppressive medications and episodes of acute

60

40

ACR AER

20

PCR PER

0 0

20

40

60

80

100

100-Specificity Figure 2. Receiver operator characteristic curve for multivariable analysis of the secondary outcome (a composite measure that included doubling of serum creatinine level or transplant loss [need for dialysis or repeat transplantation]). Analysis adjusted for time from transplantation to study entry, age, sex, and measured glomerular filtration rate. Abbreviations: ACR, albumin-creatinine ratio; AER, albumin excretion rate; PCR, protein-creatinine ratio; PER, protein excretion rate. 967

Talreja et al

who will and will not develop doubling of serum creatinine level, transplant loss, or death. In addition, each measurement performs equally well, with none being a significant improvement over another. Because PCR and ACR are more convenient and less costly than timed urine collections, we recommend use of either method to predict clinical outcomes in kidney transplant recipients. Further research is needed to determine whether routine monitoring with ACR or PCR followed by interventions to reduce the ACR or PCR will lead to improvements in patient and transplant survival.

ACKNOWLEDGEMENTS Support: This study was funded by The Physicians Services Inc Foundation (grant R03-59) and Astellas Pharma Canada. Financial Disclosure: The authors declare that they have no relevant financial interests. Contributions: Research idea and study design: AA, HT, GK, CAW; data acquisition: HT, CAW; data analysis/interpretation: AA, GK, HT, CAW; statistical analysis: AA, TOR, SH; supervision or mentorship: AA, GK. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. HT, GK, and AA take responsibility that this study has been reported honestly, accurately, and transparently; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

REFERENCES 1. Viberti GC, Hill RD, Jarrett RJ, Argyropoulos A, Mahmud U, Keen H. Microalbuminuria as a predictor of clinical nephropathy in insulin-dependent diabetes mellitus. Lancet. 1982;1(8287):1430-1432. 2. Bigazzi R, Bianchi S, Baldari D, Campese VM. Microalbuminuria predicts cardiovascular events and renal insufficiency in patients with essential hypertension. J Hypertens. 1998;16(9): 1325-1333. 3. Gerstein HC, Mann JF, Yi Q, et al. Albuminuria and risk of cardiovascular events, death, and heart failure in diabetic and nondiabetic individuals. JAMA. 2001;286(4):421-426. 4. Fernandez-Fresnedo G, Escallada R, Rodrigo E, et al. The risk of cardiovascular disease associated with proteinuria in renal transplant patients. Transplantation. 2002;73(8):1345-1348.

968

5. Fernandez-Fresnedo G, Escallada R, Rodrigo E, et al. Proteinuria is an independent risk factor of cardiovascular disease in renal transplant patient. Transplant Proc. 2002;34(1):367. 6. Knoll GA. Proteinuria in kidney transplant recipients: prevalence, prognosis, and evidence-based management. Am J Kidney Dis. 2009;54(6):1131-1144. 7. Kidney Disease: Improving Global Outcomes Transplant Work Group. KDIGO clinical practice guideline for the care of kidney transplant recipients. Am J Kidney Dis. 2009;9(suppl 3):S1-S155. 8. Methven S, MacGregor MS, Traynor JP, Hair M, O’Reilly DS, Deighan CJ. Comparison of urinary albumin and urinary total protein as predictors of patient outcomes in CKD. Am J Kidney Dis. 2011;57(1):21-28. 9. White C, Akbari A, Hussain N, et al. Estimating glomerular filtration rate in kidney transplantation: a comparison between serum creatinine and cystatin C-based methods. J Am Soc Nephrol. 2005;16(12):3763-3770. 10. Pencina MJ, D’Agostino RB. Overall C as a measure of discrimination in survival analysis: model specific population value and confidence interval estimation. Stat Med. 2004;23(13): 2109-2123. 11. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology. 1982;143(1):29-36. 12. Panek R, Lawen T, Kiberd BA. Screening for proteinuria in kidney transplant recipients. Nephrol Dial Transplant. 2011;26(4): 1385-1387. 13. Nauta FL, Bakker SJ, van Oeveren W, et al. Albuminuria, proteinuria, and novel urine biomarkers as predictors of long-term allograft outcomes in kidney transplant recipients. Am J Kidney Dis. 2011;57(5):733-743. 14. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39(2)(suppl 1):S1-S266. 15. Panek R, Lawen T, Kiberd BA. Screening for proteinuria in kidney transplant recipients. Nephrol Dial Transplant. 2011;26(4): 1385-1387. 16. Kidney Disease: Improving Global Outcomes (KDIGO) Transplant Work Group. KDIGO clinical practice guideline for the care of kidney transplant recipients. Am J Transplant. 2009;9(suppl 3):S1-S155. 17. Akbari A, White CA, Shahbazi N, Booth RA, Hiremath S, Knoll GA. Spot urine protein measurements: are these accurate in kidney transplant recipients? Transplantation. 2012;94(4):389-395. 18. Witte EC, Lambers Heerspink HJ, de Zeeuw D, Bakker SJ, de Jong PE, Gansevoort R. First morning voids are more reliable than spot urine samples to assess microalbuminuria. J Am Soc Nephrol. 2009;20(2):436-443.

Am J Kidney Dis. 2014;64(6):962-968

Predicting kidney transplantation outcomes using proteinuria ascertained from spot urine samples versus timed urine collections.

Proteinuria has been associated with transplant loss and mortality in kidney transplant recipients. Both spot samples (albumin-creatinine ratio [ACR] ...
287KB Sizes 1 Downloads 9 Views