Reliability of FEUA in AKI and its Combination with Other Renal Failure Indices

Transcription

1 Med. J. Cairo Univ., Vol. 80, No. 2, September: , Reliability of FEUA in AKI and its Combination with Other Renal Failure Indices ADEL TAIE, MBBCh.; MOHAMAD HOSNY, M.D.; ENAS ABDELLATEEF, M.D. and NAEL SAMIR, M.D. The Department of Critical Care Medicine, Faculty of Medicine, Cairo University Abstract Purpose: An early detection of adult patients with acute kidney injury may provide the opportunity to treat and prevent the extension of kidney injury. Fractional excretion of sodium (FENa) has been used in the differentiation of acute kidney injury (AKI) into traditional categories of prerenal azotemia (PR) and acute tubular necrosis (ATN). However, many patients with PR have already received diuretics or saline at the time of diagnosis, which increase FENa. In contrast, the fractional excretion of uric acid (FEUA) and urea (FEUN) is less influenced by diuretics. We investigated the diagnostic significance of the FEUA and FEUN in differentiating between PR and ATN. Methods: The FENa, FEUA, and FEUN were calculated in 20 patients with PR and 20 patients with ATN at day 0 (D0), day 1 (D1) and day 2 (D2), sequentially. Results: FEUA (PR14.49±6.23% vs. ATN 47.09±23.35, p<0.001) and FEUN (PR ±9.54% vs. ATN 54.97± 17.14%, p<0.001) were lower in PR than in ATN patients. At the cutoff value of 1.04% FENa, sensitivity and specificity for the detection of PR was 75% and 85%, respectively. At the cutoff value of 39.4% FEUN, sensitivity and specificity for the detection of PR was 75% and 80%, respectively. At the cutoff value of 19.83% FEUA, sensitivity and specificity for the detection of PR was 85% and 90%, respectively. When FENa, FEUA and FEUN were combined, sensitivity and specificity was 84% and100%, respectively. Conclusion: FEUA and FEUN may be useful in differentiating between PR and ATN. The combination of FENa, FEUA and FEUN might increase diagnostic sensitivity and specificity in the differential diagnosis of AKI. Key Words: Acute kidney failure Fraction excretion of sodium Urea Uric acid Introduction ACUTE kidney injury (AKI) is a common complication of critical illness, seven to ten percent of intensive care unit (ICU) patients present with AKI during their ICU stay, [1] 45-60% of them is associated with high mortality [2,3]. An early detection of adult patients with kidney injury may provide Correspondence to: Dr. Mohamad Hosny, mhosny7474@hotmail.com the opportunity to treat and prevent the extension of kidney injury [4]. Several urinary indices have been proposed to improve the early diagnosis of AKI. But; the urinary indices currently available have several limitations. In current clinical practice, AKI is diagnosed by measuring serum creatinine concentration, which is an unreliable and delayed marker of the deterioration of kidney function during acute kidney injury and does not directly reflect cell injury, but rather the delayed functional consequences of the damage [4,5]. The fractional excretion of sodium (FENa) has been shown to discriminate reliably between prerenal azotemia and acute tubular necrosis (ATN) [6]. However, there are several reports of low FE Na (less than 1%) in conditions associated with intrinsic acute renal failure, such as sepsis, contrast nephropathy, myoglobulinuria and nonoliguric ATN [7,8]. Furthermore, a common major limitation in the use of FEN a stems from the fact that diuretic agents are employedtubular necrosis (ATN) frequently in the treatment of prerenal conditions such as congestive heart failure, liver failure with ascites, or to enhance urine output in oliguric patients. In addition, the excessive use of diuretics in a formerly euvolemic subject may lead to a prerenal state, often with increased urinary Na and thus increased FEN a. Other situations also exist where prerenal azotemia is associated with increased urinary sodium and increased FE N a. Most prominent among them is volume depletion and prerenal azotema due to vomiting or nasogastric suction. In such cases the ensuing bicarbonaturia maintains urinary sodium and FENa at high levels [9]. The fractional excretion of urea (FEUN) may result in more reliable determination of renal func- 181

2 182 Reliability of FEUA in AKI tion than sodium in the presence of oliguricazotemia; however, its usefulness remains controversial, perhaps due to an evolving understanding of urea transport within the kidney [8]. The FEUN appears more accurate in patients receiving diuretics; however, the FENa may have an advantage in patients with infection [10]. In prerenal azotemia, decreased fractional excretion of uric acid (FEUA) may represent a reliable indicator of prerenal azotemia in the differential diagnosis of acute renal failure [11]. The combination of FENa, FEUA and FEUN might increase diagnostic sensitivity and specificity in the differential diagnosis of AKI. When FENa, FEUA and FEUN were combined, sensitivity and specificity was PR (84%) and (100%) in ATN [11]. Aim of the work: This study investigated the diagnostic significance of the FEUA in differentiating between PR and ATN and benefits of its combination with other renal failure indices. Patients and Methods This was a prospective; Unrandomized Uncontrolled Un limited study that took place at the Critical Care Medicine Department, Cairo University from May 2010 March Forty consecutive adult patients referred to the Intensive Care Units for evaluation of acute renal failure. The major reason was a rapidly increasing blood urea nitrogen (BUN) and creatinine (BUN 30mg/dL and creatinine 1.5mg/dL) with or without oliguria. Urine output measurements were available for all patients, all of whom had Foley catheters. In all cases, the diagnosis based on the medical history, physical examination, and evaluation of urine sediment morphology and composition [1] is shown in (Table 1). Table (1): Criteria for diagnosis of acute renal failure and how to differentiate prerenal azotemia and acute tubular necrosis (ATN). A- Criteria for diagnosing acute renal failure: 1-Azotemia-rapidly increasing BUN and creatinine (BUN >30mg/dL and creatinine >1.5mg/dL) with or without oliguria. 2- Serum creatinine increase in excess of 0.5mg/dL in the preceding 2 days. B- Criteria to differentiate ATN from prerenal azotemia: 1- History (volume depletion, decreased cardiac output or vasodilation related to sepsis, liver failure and anaphylaxis favor pre-renal azotemia, while exogenous toxins such as medications, or endogenous toxins as in the case of myoglobin, or even prolonged renal hypoperfusion that became unresponsive to appropriate corrective measures or to high dose of loop diuretics, all favor ATN). 2- Physical examination (blood pressure, heart rate, orthostatic changes, cardiac sounds, pulmonary findings, presence of ascites or pedal edema). 3- Findings of the urine analysis (urinary sediment non-revealing in pre-renal, presence of muddy brown granular casts in patients with ATN) as performed by a member of the renal service. Retrospective confirmation (response to therapy) also was recorded. 4- The urinary indices evaluated on ICU admission were: Urinary sodium (U): UN <15mEq/L favors prerenal, while a value higher than 20 is consistent with ATN. Urinary to plasma creatinine ratio (U/P). U/Pr>20 is consistent with prerenal while levels <15 suggest ATN. Fractional excretion of sodium (FENa ). FEN a <1% is suggestive of prerenal azotemia, while levels >1% indicate the presence of ATN. Nine patients had received diuretic therapy (furosemide or thiazides). Patients with urinary tract obstruction (completely anuric), patients with acute glumerionephritis or patients with chronic renal failure on regular dialysis were excluded from this study. Prerenal azotemia was the result of circulatory failure secondaryto various causes such as sepsis, gastrointestinal bleeding, hepatic, respiratory or cardiac failure. The diagnosis of prerenal azotemia was established in circulatory failure when improvement of heart function, ces- sation of diuretic therapy, or treatment of shock affected a prompt increase in urinary output and creatinine clearance. Of the 20 cases of prerenal azotemia, 4 had sepsis, 5 congestive heart failure, 5 hepatic failures and 1 patient with malignancy. The remaining patients had miscellaneous or less definite diagnoses. Ten were men and 10 were women, with an average age of ± years (mean±sd). As expected, the majority of the prerenal cases receiving diuretics (9 patients) were subjects with congestive heart failure. The demo-

3 Adel Taie, et al. 183 graphics were actually quite similar in both groups where the representation of men to women were 10 to 10 in the PR group and 8 to 12 in the ATN group. The twenty patients in the ATN group with an average age of 60.5 ± ATN resulted from nephrotoxic drugs (aminoglycoside antibiotics), myoglobinuria, sustained circulatory failure, or massive surgical volume losses unresponsive to replacement. Five patients had protracted sepsis, although the exact role of coexistent factors such as major surgical procedures or use of nephrotoxins (e.g; aminoglycoside therapy) does not allow for a clear-cut identification of the underlying etiology. The remainder of the cases had a mixed and often less well identified etiology. Table (2): Characteristics of patients with prerenal (azotemia) AKI and A.T.N. PR ATN p-value Diabetic 6 (30%) 7 (35%) 0.73 Hypertensive 2 (10%) 3 (15%) 1.00 Malignancy 1 (5%) 2 (10%) 1.00 Liver cirrhosis 5 (25%) 2 (10%) 0.40 Heart failure 5 (25%) 2 (10%) 0.40 Sepsis 4 (20%) 5 (25%) 1.00 CKD 4 (20%) 5 (25%) 1.00 Previous ICU admission 3 (15%) 6 (30%) 0.45 The diagnosis was made by the consulting attending nephrologists, based on the medical history, physical examination and renal investigations including routine renal functions test; serum BUN, CREAT, (COBAS INTEGRA 800) Na, K, Mg, Ca; blood gases (ABL800; RADIOMETER ), evaluation of urine sediment morphology and composition and renal imaging if available. The indices evaluated at the time of consultation were urinary sodium (UNa), urinary to plasma creatinine ratio (U/PCr), and fractional excretion of sodium (FENa). Fractional excretion of urea (FEUN) and the fractional excretion of uric acid (FEUA). The FEN a was calculated as: Urine sodium / plasma sodium Urine creatinine / plasma creatinine x 100 The FEUN was calculated as: Urine urea nitrogen / blood urea nitrogen x 100 Urine creatinine / plasma creatinine The FEUA was calculated as: Urine uric acid / blood uricd Urine creatinine / plasma creatinine x 100 Urine electrolytes were determined by flame photometry. Urine urea nitrogen and creatinine were determined spectrophotometrically on spot urine samples collected at the bedside [20,21]. Data were statistically described in terms of mean±standard deviation (±SD), median and range, or frequencies (number of cases) and percentages when appropriate. Comparison of numerical variables between the study groups was done using Student t-test for independent samples when comparing 2 groups and one way analysis of variance (ANOVA) test with posthoc 2-group comparisons when comparing more than 2 groups. For comparing categorical data, Chi square ( χ 2 ) test was performed. Exact test was used instead when the expected frequency is less than 5. Accuracy was represented using the terms sensitivity, specificity, +ve predictive value, -ve predictive value, over all accuracy. p-values less than 0.05 was considered statistically significant. All statistical calculations were done using computer programs SPSS (Statistical Package for the Social Science; SPSS Inc., Chicago, IL, USA) version 15 for Microsoft Windows. Results The patients were divided into two groups. The first group was composed of 20 patients with prerenal failure; the second group had 20 patients with established ATN. The indices evaluated in ARF are shown in Table (3). It is clear that FENa was characteristically low ( 1 %) in the prerenal group, it was substantially higher in the ATN group. In contrast, the increased BUN/creatinine ratio, the urine to plasma creatinine ratio (U/PCr), was consistently different in both prerenal groups from those in the ATN group (Table 3). Indeed, the FEUN in the prerenal groups (32.21 ±9.54) was significantly different from the FEUN found in ATN (54.97± 17.14, p ). The FEUA in the the prerenal groups (14.49 ±6.23) was also significantly different from the FEUA found in ATN group (47.09±23.35, p ). Table (3): Comparison of diagnostic indices on initial evaluation in patients with prerenal aki and acute tubular necrosis. Diagnostic Indices PR (No = 20) ATN (No = 20) p-value Serum creatinine (mg/dl) 3.99± ± * BUN (mg/dl) 81.85± ± * BUN/creatinine ratio 21.00± ± ** Urine sodium (meq/l) 37.77± ± U/P creatinine 32.17± ± ** U/P urea 9.93± ± * FENa (%) 1.14± ± ** FEUA (%) 14.49± ± ** FEUN (%) 32.21± ± ** * = Statistically significant. ** = Highly statistically significant.

4 184 Reliability of FEUA in AKI After we compared the 3 main indices (FEna, FEun, FEua) between the 2 groups (PR & ATN) we concentrated on the PR group (trying to diagnose the patients in this stage to avoid progression to ATN) so, we chosen the most sensitive and specific cut-off number for every index to reflect this pre-renal group and from our statistical data we chosen (1.04 for FEna, 39.4 for FEun & for FEua), see Table (4). Table (4) and Fig. (1) show the importance of the indices in patients in whom a given test was positive. It appears that FE N a, FEUN, and FEUA are the indices that discriminate best among the different types of acute renal failure. While 75% of patients with prerenal azotemia had a FENa 1 %, only 44% of patients who were prerenal and on diuretic therapy had a low FENa. By contrast, 55.5% of this latter group had a FEUN 39.44% and 77.7% had a FEUA If one combines all the prerenal cases, FEUA had the best sensitivity and specificity (85 and 90%, respectively), and the best positive and negative predictive value (89.5 and 85.7%, respectively) of all the indices examined in this study. Table (4): Number of pts with the most sensitive and specific cut-off number in each ARF group. Indices PR total PR Without- D WPh D ATN FENA (<1.04) 15/20(75%) 11/11 (100%) 4/9 (44%) 3/20 (15%) FEUN (<39.440) 15/20(75%) 10/11 (90%) 5/9 (55.5%) 4/20 (20%) FEUA (<19.83) 17/20(85%) 10/11 (90%) 7/9 (77.7%) 2/20 (10%) % FE na FE un FE ua PR total PR Without- D PR With-D ATN Fig. (1): of the three major indices for patients with prerenal situations with or without diuretic use, and patients with ATN. Table (5) and Fig. (2) evaluates the positivity of different levels of the three most important indices in patients with acute renal failure. Again, when examining the combined prerenal group compared with that of ATN, FEUA wasthe best discriminatory index. For instance, a FENa of less than 1% was present in 75% of prerenal patients and in about 15% of patients with ATN, while FEUA <19.83% was found in 85% of prerenal patients and in 10% of ATN. FEUN appears to be an intermediate differentiator. When a ratio <39.44 was used as the cutoff value, 75% of patients with prerenal azotemia were positive, with only a small number of patients with ATN having such a level (20%). Table (5): Diagnostic performance of the 3 major indices for diagnosis of PR azotemia at the best cut-off value. Diagnostic cut-off (%) Specificity (%) Positive predictive value (%) Negative predictive value (%) Area under the curve FEna (<1.04) FEun (<39.44) FEua (<19.83)

5 Adel Taie, et al ROC Curve % FEna FEun FEua Fig. (2): Diagnostic performance of the 3 major indices for diagnosis of PR azotemia at the best cut-off value. Figs. (3,4) show the ROC curves for these data. We evaluated the equations that describe the relationship between the FEUN and FEUA in our subjects. This was based on previous findings regarding such a relationship (see discussion below for details). We found major differences among the different groups. In prerenal azotemia, FEUN= (U/Pcr), r=0.28, p=0.015, and was similar for the two groups of prerenal patients (prerenal alone or prerenal in the presence of diuretic therapy). In patients with ATN a very different relationship was noted: FEUN=70.3 ± 1.26 (U/PCr), r=0.439, p=0.028, consistent with very different mechanisms of the generation of such a relationship. 1.0 Specficity ROC Curve PPV NPV Specificity Source of the curve BUN/Creat ratio day (0) U/Pcreat day (0) U/Pun day (0) Reference line Fig. (4): Show the ROC curve of sensitivity and specificity in Serum creatinine, BUN, BUN/creatinine ratio. Diagnosis of Prerenal AKI at the Combined Cut-off Value of FENa, FEUN and FEUA showed the better specificity 100% and the highest Positive predictive value (100%) when were combined, Table (6). Table (6): Diagnostic performance for diagnosis of prerenal AKI at the combined cut-off value of FENa, FEUA and FEUN. Diagnostic cut-off Cr.1 (at least 1 positive) Cr.2 (at least 2 positive) Cr.3 (at least 3 positive) (%) Specificity (%) Positive predictive value (%) Negative predictive value (%) % Specificity 10 Source of the curve FEna day (0) FEun day (0) FEua day (0) Reference line 0 Cr 1 Cr 2 Cr 3 PPV Specficity NPV Fig. (3): Show the ROC curve of sensitivity and specificity in FEun, FEua and FEna. Fig. (5): Patients with the most sensitive and specific cut-off number in each AFR group.

6 186 Reliability of FEUA in AKI Finally, the predicted FEUN was examined using Dole sequation as described in the Methodssection. In the prerenal patients, the expected FEUN was 25-4% and 22-4% for the prerenal group without and with diuretic use, respectively. These values were not statistically different from the ones actually found (27-2% and 24-3%, respectively), suggesting that indeed the mechanism for the decreased volume (decreased renal perfusion and increased water reabsorption) also was responsible for the decreased excretion of urea. In contrast, in the cases of ATN the calculated FEUN was very different from the one actually determined (18-3% vs. 58-6%, p=0.0001), suggesting that the oliguria found in this condition is not related to reabsorption of water and urea from functional tubules, but rather due to minimal glomerular filtration through injured nephrons. Discussion When urine excretion decreases as a result of water reabsorption, the level of urinary creatinine increases in inverse proportion to the volume of urine output. Thus, the U/P creatinine ratio logically identifies whether oliguria is the result of increased water reabsorption, as in often appreciated, sodium reabsorption also increases in prerenal states both as a result of an increase in aldosterone as well as the increased sodium and water reabsorption in the proximal tubule because of the increased filtration fraction found in such cases. Acomposite index has been developed and used clinically that incorporates both U/PCr and the FENa. Following the initial report of Espinel [12], this showed that FENa discriminates best between the two most commonly encountered forms of ARF (prerenal and ATN), several subsequent reports substantiated this claim [6,7]. Currently it has become the dominant discriminatory index used for that purpose. On the other hand, coexistence of a prerenal state and natriuretic agents (diuretics) negate the effectiveness of this index [6]. Diuretics represent the mainstay of treatment for several prerenal states and may themselves produce volume depletion due to their natriuretic effect (even if given to initially euvolemic patients). Furthermore, these agents are used commonly in the Hospital and even more in the intensive care unit, thus significantly decreasing the utility of FE N a in the hospital setting. Hence, FEN a does not give a clearcut distinction in such cases. In addition to the use of diuretics, other situations are known to lead to both prerenal azotemia and increased FE Na. Better understood among them is the case of volume depletion due to vomiting or naso-gastric suction associated with increased FE N a due to bicarbonaturia [9]. On the other hand, a number of clear-cut forms of ATN are associated with decreased FEN a, for example: Myoglobinuria [6,7], sepsis [8], contrast nephropathy [11,13], obstructive uropathy and non-oliguric ATN [6,7], Low FE Na also is found in other forms of acute intrinsic renal failure including acute glomerulonephritis [7,8], urinary tract obstruction, and renal allograft rejection. To emphasize the frequency with which patients with intrinsic acute renal failure have FE Na <1%, suffice it to say that in a five year-period (1980 to 1985) more than 100 cases of documented ATN with FEN a <1% were found in the published literature alone [14]. The relative clearance of urea denoted by the FEUN is affected by decreased renal perfusion as seen in prerenal states. As early as 1921, Austin, Stillman and Van Slyke showed that the clearance of urea in man decreases substantially in dehydration somewhat proportionally to urine output [15]. In 1938, two seminal studies established that FE UN relates inversely to U/P creatinine [16] and U/P inulin [17,18], and therefore proportionally to urine volume. This was the classical teaching in renal physiology during that era [19,20]. More recently, Goldstein, Lenz and Levitt evaluated the relationship of urine volume to urea excretion in humans [20]. Using proximal and distal diuretic agents, they found that the percent of filtered urea excreted increased with proximal (mannitol and acetazolamide) but not with distal agents (mercaptomerin and ethacrynic acid). This suggests that urea is not reabsorbed significantly in the distal nephron. Given that most of the diuretics used clinically work at distal sites, FE UN should not be affected by their use. This is in contrast to FEN a, which is increased by all forms of diuretics. Only in patients with established chronic renal failure does the use of diuretics and the ensuing volume depletion also operate in distal sites [22], a finding with limited significance in the differential diagnosis of ARF. In 1992, Kaplan and Kohn reported six prerenal patients with a discrepancy of FE UN and FEN a in the presence of diuretics [23]. In addition to the small number, most of those patients already had established renal failure, congestive heart failure and various treatments with angiotensin-converting enzyme inhibitors. The authors, in order to better understand the situation, retrospectively evaluated the charts of 87 patients with prerenal azotemia. They noted that the patients that had both low FEN a and FEUN had the usual type of prerenal azotemia. The group with low FE UN but high FE Na

7 Adel Taie, et al. 187 appeared to have prerenal azotemia but had been treated with a diuretic (39 out of 40). The authors concluded that FEUN may be a better index to evaluate possible prerenal status in patients who receive diuretic agents [23]. We prospectively studied two groups of patients: Patients with prerenal azotemia, and patients with acute tubular necrosis (ATN). Patients with prerenal conditions further subdivided into subgroups of patients receiving diuretics and others not receiving diuretics, we found that FEN a failed to discriminate between prerenal patients given diuretics and those with ATN. FEUN and FEUA on the other hand, were excellent in discriminating among all cases of prerenal azotemia and ATN irrespective of the use of diuretics. This has a potentially major practical application, given the frequency of diuretic use in the Hospital setting. For instance, among our 20 unselected prerenal cases, 9 patients (45%) had been given diuretics [23]. Our study found that U/PCr is higher in prerenal cases, as has been shown before and as expected from the underlying physiology. Indeed, it appears that U/PC r is at least as good as FE N a in separating prerenal azotemia from ATN. As was previously reported [16-20], there was a close inverse relationship between U/PCr and FEUN in the prerenal group. This relationship was very different in the group with ATN. Furthermore, the prerenal patients appear to follow the prediction of Dole [27]. In reference to expected FEUN based on urine output. In short, these patients behaved the same way as a normal individual with decreased fluid intake and thus decreased urine output. In contrast, urine volume could not account for the FE UN found in the ATN group [27]. We found that FE UA has a high sensitivity (85%) a high specificity (90%) and more importantly a high positive predictive value (89.5%). The latter suggests that if a FE UA of <19.83% is found, 89.5% of such cases will have prerenal azotemia. The relative importance of FE UA and FEUN to FEN a would have been more impressive if our ATN patients had had some of the characteristics described in the literature. For example, low FEN a is often found in patients with myoglobinuric ATN; however, in our study none of the six patients with well-documented myoglobinuria (history and urine analysis positive for myoglobin) had an FE Na value less than 1. If the value had been greater, then the false-positive cases of low FE N a in identifying pre-renal azotemia would have been even higher [6,7]. Fractional excretion of urea nitrogen to a great extent relates inversely to the proximal reabsorption of water. Proximal reabsorption of water (accompanied by urea) increases when renal perfusion decreases and filtration fraction increases. Thus, urea reabsorption leads to a decrease in FEUN and an increase in the BUN/creatinine ratio. The use of indices that reflect proximal tubular function in acute renal failure have been suggested before. A recent textbook of nephrology suggests two such indices, the clearance of lithium and the clearance of uric acid. It is apparent that the use of lithium clearance is quite impractical, given the limited availability of this determination in many laboratories and the need to administer lithium to such patients. Uric acid clearance appears to have more advantage over the use of urea [24]. It is certainly interesting, that if we modify the formula for FE UN=(U/P)UN/(U/P)Cr, it becomes FEUN=U (UN/Cr)/(BUN/Cr), or even, BUN/Cr= U (UN/Cr)/FE UN. This suggests that as proximal urea reabsorption increases and FE UN decreases in prerenal states, the BUN/Cr ratio increases, as is the common teaching and is our own experience in the present study (Table 2). Another cause of increased BUN/Cr is increased urea production (and thus excretion), as in cases of hypercatabolism or consumption of high protein diets. This could be easily differentiated from prerenal azotemia, as in high urea production we have a high urine UN/Cr, much higher than the usual ratio of 5 to 10. The determination of urine urea nitrogen and FEUN may be helpful in a few other conditions. For instance, the presence of a BUN/Cr ratio of 10 in the presence of a low U UN/UCr (< 5 ) suggests the existence of prerenal azotemia in a starving individual [24]. There are situations, however, where FE UN may fail to indicate the ongoing prerenal status. Therefore, such possibilities should be kept in mind when FEUN is used. For instance, in osmotic diuresis the proximal tubular absorption of salt and water is impaired and thus increased FE UN is expected despite renal hypoperfusion. One of us (C.P.C.) has looked into this while working on another study (abstract; Vafadouste R et al., XV International Congress of Nephrology, Buenos Aires, Argentina, p 1999,2001). In ten patients with osmotic diuresis there was a mild increase of BUN (23±3.3mg/dL) with normal serum creatinine (0.9±0.1 mg/dl), increased BUN/Cr to 27 ±3, increased FEUN 49±8%, high normal FEN a 0.7±0.3% and high U/P C r of 41 ± 10. The same phenomenon is expected to happen when osmotic diuretics such as acetazolamide and mannitol are used. Indeed,

8 188 Reliability of FEUA in AKI this was our finding during that study. Similar findings also were reported before. A similar picture emerges in patients given a high protein diet or having excessive catabolism. There, the enormous production of urea acts as the osmotic diuretic agent. On the other hand, FE UN could be helpful in cases with prerenal azotemia in combination with a distal defect in sodium reabsorption [21]. Conclusion: This study suggests that FEUA is a useful index to discriminate PR from ATN. Its diagnostic value seems particularly high in patients with PR who are receiving diuretics, i.e, in a considerable percentage of oliguric renal failure when initially seen. Large studies are, however, needed to confirm its validity and cost/benefit ratio in diagnosis of ARF and to delineate its usefulness in other causes of ARF. Although FEUN and FEUA proved to be the best indices under the clinical conditions encountered in our study, one must remember that none of the indices offers a 100% discriminatory ability. The three major indices allow evaluation of different sites of action, and differ in the technology needed for their determination, thus making it unlikely to have simultaneous laboratory errors in all of them. FEUN allows insight to proximal peritubular forces, FEN a mostly reflects distal forces (mostly aldosterone-dependent sites), and FE UA indicates water reabsorption through the length of the nephron. Indices are tools and do not obviate the importance of history, physical examination, and direct urinalysis in making a diagnosis of acute renal failure. However in our study weconfirmed the highest sensitivity and specificity of (FEua) in diagnosis of PR Azotemia, and we elevated its sensitivity and specificity when we combined FENA with one or more of the other 2 indices (FEUN and FEUA) together and it is proved that the highest sensitivity, specificity and positive predictive value (60%, 100% and 100% respectively) were achieved when combination of the 3 indices FENA, FEUN and the FEUA together [21]. References 1- CONSTANTIN J.M., FUTIER E., et al.: Plasma neutrophil gelatinase-associated lipocalin is an early marker of acute kidney injury in adult critically ill patients. Journal of Critical Care, 25: 176.e1-176.e6, UCHINO S., KELLUM J.A., et al.: Acute renal failure in critically ill patients, multinational, multicenter study. JAMA, 294 (7): , KÜMPERS P., HAFER C., et al.: Serum neutrophil gelatinase-associated lipocalin at inception of renal replacement therapy predicts survival in critically ill patients with AKI. Critical Care, 14: 1-9, GRANDE A D.I., GIUFFRIDA C., et al.: Neutrophil gelatinase-associated lipocalin: A novel biomarker for the early diagnosis of acute kidney injury in the emergency department. European Review for Medical and Pharmacological Sciences, 13: , PARIKH C.R., JANI A., et al.: Urine NGAL and IL-18 are predictive biomarkers for delayed graft function following kidney transplantation. American journal of Transplantation, 6: , BRADY H.R., BRENNER B.M. and LIEBERTHAL W.: Acute renal failure, in Brenner and Rector s The Kidney (5th ed), edited by Brenner BM, Philadelphia, W.B. Saunders, pp , DANOVITCH G., CARVOUNIS C., WEINSTEIN E. and LEVENSON S.: Non-oliguric acute renal failure. Isr. J. Med. Sci., 15: 5-8, PRU C. and KJELLSTRAND C.M.: The FEN a test is of no prognostic value in acute renal failure. Nephron, 36: 20-23, NANJI A.J.: Increased fractional excretion of sodium in prerenal azotemia: Need for careful interpretation. Clin. Chem., 27: , CHARLES J. DISKIN, THOMAS J. STOKES, LINDA M. DANSBY, LAUTREC RADCLIFF and THOMAS B. CARTER: The Comparative Benefits of the Fractional Excretion of Urea and Sodium in Various Azotemic Oliguric States. 11- JANG J.S., KIM H.S., LEE S.Y., LEE S.H., KIM S.J., KWON S.K. and KIM H.Y.: The Clinical Significance of the Fractional Excretion of Uric Acid and Urea in the Differentiation of Acute Kidney Injury. 12- ESPINEL C.H.: The FEN a test: Use in the differential diagnosis of acute renal failure. JAMA, 236: , VANZEE B., HOY W., TALLEY T.E., et al: Renal injury associated with intravenous pyelography in non-diabetic and diabetic patients. Ann. Intern. Med., 89: 51-54, PRU C. and KJELLSTRAND C.M.: Urinary indices and chemistries in the differential diagnosis of prerenal failure and acute tubular necrosis. Semin Nephrol., 5: , AUSTIN J.H., STILLMAN E. and VAN SLYKE D.D.: Factors governing the excretion of urea. J. Biol. Chem., 46: , CHASIS H. and SMITH H.W.: The excretion of urea in normal man and in subjects with glomerulonephritis. J. Clin. Invest., : BULL G.M., JOEKES A.M. and LOWE K.G.: Renal function studies in acute tubular necrosis. Clin. Sci., 9: , SHANNON J.A.: The renal reabsorption and excretion of urea under conditions of extreme diuresis. Am. J. Physiol., 123: , 1938.

9 Adel Taie, et al SMITH H.W.: The Kidney. Structure and function in health and disease. New York, Oxford University Press, pp 63-80, PITTS R.F.: Kidney and Body Fluids (2 nd ed). Chicago, Year Book Medical Publishers Inc., pp 88-93, GOLDSTEIN M.H., LENZ P.R. and LEVITT M.F.: Effect of urine flow rate on urea reabsorption in man. Urea as a 'tubular marker'. J. Appl. Physiol., 26: , DAL CANTON A., FUIANO G., CONTE G., et al.: Mechanism of increased plasma urea after diuretic therapy in uremic patients. Clin. Sci., 68: , KAPLAN A.A. and KOHN O.F.: Fractional excretion of urea as a guide to renal dysfunction. Am. J. Nephrol., 12: 49-54, DWINNELL B.G. and ANDERSON R.J: Diagnostic evaluation of the patient with acute renal failure, in Atlas of Diseases of Kidney, edited by Schrier R.W., Philadelphia, Current Medicine Inc., pp , 1999.