Fractional excretion of sodium, urea, and other molecules in acute kidney injury

INTRODUCTION — The fractional excretion of sodium (FENa) measures the percent of filtered sodium that is excreted in the urine. This calculation is widely used to help differentiate prerenal disease (decreased kidney perfusion) from acute tubular necrosis (ATN) as the cause of acute kidney injury (AKI, formerly called acute renal failure). (See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Evaluation and diagnosis'.)

As described below, there are a number of limitations to the use of FENa in patients with AKI. These limitations include the small sample size of studies supporting the use of FENa, the variability of the threshold FENa that is used to define prerenal disease, the presence of multiple causes of a low FENa other than prerenal disease, and errors associated with the use of measures of sodium excretion in the setting of renal salt wasting (as with diuretic therapy). Thus, the etiology of AKI should be based upon the totality of the presentation (history, clinical examination, urine microscopy, and, when appropriate, response to volume resuscitation), not just on the FENa. FENa could play a supportive role in the diagnosis of hepatorenal syndrome, a disorder that is characterized by oliguria, avid sodium reabsorption, urine Na <10 mEq/L, and FENa <1 percent [1]. When diuretics may still be active and the FENa is not low, the fractional excretion of urea (FEUrea) may be helpful. (See 'Limitations of fractional excretion of sodium' below and 'Urea' below.)

Although not yet in widespread clinical use, novel urinary and serum biomarkers to assess intrinsic tubular injury are being evaluated [2-5]. (See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Evaluation and diagnosis'.)

Formula — The filtered sodium load can be determined from the product of the glomerular filtration rate (GFR; as estimated from the creatinine clearance) and the serum sodium concentration (SNa). On the other hand, urinary sodium excretion is equal to the product of the urine sodium concentration (UNa) and the urine flow rate (V). Thus:

where urine creatinine concentration (UCr) and serum creatinine concentration (SCr) are the urine and serum creatinine concentrations, respectively. This equation can be simplified since the urine flow rate terms (V) cancel out, allowing the FENa to be estimated from simultaneously obtained serum and urine specimens without measuring the urine volume [6]:

Calculators for the FENa are available using either standard units (calculator 1) or SI units (calculator 2).

OVERVIEW — The FENa is commonly used to assist in differentiating prerenal disease (a transient reduction in glomerular filtration rate [GFR] due solely to decreased kidney perfusion) from acute tubular necrosis (ATN), the two most common causes of AKI [6-9]. However, there is some overlap since severe prerenal disease, particularly if accompanied by hypotension, is also a common cause of ATN. In addition to the laboratory tests, a careful history and physical examination are essential [9]. (See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Causes of acute tubular necrosis'.)

The FEUrea and other molecules, such as lithium and uric acid, have also been evaluated for their utility in the distinction between prerenal disease and ATN [9,10]. The FEUrea may be particularly helpful in patients who are being treated with diuretics, which can transiently raise the FENa above the value used to diagnose prerenal disease. (See 'Fractional excretion of other molecules' below.)

FRACTIONAL EXCRETION OF SODIUM IN ACUTE KIDNEY INJURY — The FENa may be useful in patients suspected of having AKI due to either prerenal disease or acute tubular necrosis (ATN). A variety of studies have confirmed that the FENa more clearly differentiates between these two conditions than other laboratory tests [7,8,11,12]. The urine sodium concentration, the urine-to-serum creatinine ratio, and the urine osmolality all have a much lower predictive value. However, FENa does not have enough accuracy to be used as a sole diagnostic test for this purpose [13]. (See 'Fractional excretion of sodium versus urine sodium concentration' below and "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Other tests that may be helpful'.)

The clinical value of calculating the FENa is highest in patients with oliguric AKI who do not have preexisting chronic kidney disease (CKD) and who are not taking diuretics. In patients with CKD and/or those taking diuretics, the FENa may often be high (ie, above 2 percent) even in the setting of hypovolemia. As an example, in a systematic review of 15 studies and 827 patients with AKI, the sensitivity and specificity of a FENa less than 1 percent to correctly identify prerenal disease were 90 percent (95% CI 81-95 percent) and 82 percent (95% CI 70-90 percent), respectively [14]. Among those patients with oliguria and no preexisting CKD or use of diuretics, the sensitivity and specificity were higher (95 and 91 percent, respectively). By contrast, sensitivity and specificity were lower among those with CKD or diuretic use (83 and 66 percent, respectively).

In general, a FENa below 1 percent suggests prerenal disease, where the reabsorption of almost all of the filtered sodium represents an appropriate response to decreased renal perfusion. In comparison, a value between 1 and 2 percent may be seen with either disorder, while a value above 2 percent usually indicates ATN. However, a FENa below 1 percent is not diagnostic of prerenal disease, since it can be seen in a variety of other causes of AKI. (See 'Multiple causes of acute kidney injury with FENa less than 1 percent' below and "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults", section on 'Fractional excretion of sodium and urine sodium concentration'.)

The above cut-off values apply only in patients with advanced AKI. The FENa values that define prerenal disease are lower in patients with less severe disease since the filtered sodium load is much higher, being as low as 0.1 percent in patients with normal kidney function. Another potential problem is that the glomerular filtration rate (GFR) cannot be estimated from a single serum creatinine measurement, since the serum creatinine is not stable. (See 'Fractional excretion of sodium varies with glomerular filtration rate' below and 'Single measurements of serum creatinine may be misleading' below.)

The relatively high FENa in ATN can be due to one or both of the following factors: inappropriate sodium wasting due to tubular damage and/or an appropriate response of the remaining, well-preserved nephrons to volume expansion. Inappropriate sodium wasting is most likely to be important early in the disease, when nephrons that are still filtering have impaired tubular function. In comparison, volume expansion may be more important in patients with established ATN and normal tissue perfusion. At this time, most of the damaged nephrons may be nonfunctioning, with the urine output primarily coming from a relatively small number of well-preserved nephrons. In this setting, the maintenance of sodium balance requires a high FENa.

As an example, suppose a patient with established ATN has a GFR that has fallen from the normal level of 180 L/day (125 mL/min) down to 20 L/day. If the plasma water sodium concentration is 150 mEq/L, then the filtered sodium load will be 3000 mEq/day (20 x 150). If sodium intake is 75 mEq/day, then excretion of the ingested sodium will require a FENa of 2.5 percent (75 ÷ 3000).

Fractional excretion of sodium versus urine sodium concentration — The FENa is generally more accurate than the urine sodium concentration in differentiating prerenal disease from ATN because it directly measures sodium handling. In contrast, the urine sodium concentration is also affected by the rate of water reabsorption. As an example, a patient who is both water and sodium avid due to prerenal disease may, because of the low urine volume, have a relatively high urine sodium concentration of 20 to 40 mEq/L or more, despite having little sodium in the urine.

Limitations of fractional excretion of sodium — There are important limitations to the use of FENa in establishing the cause of AKI (table 1) [9]:

●The FENa criterion of less than 1 percent to diagnose prerenal disease applies only to patients with a marked reduction in GFR. (See 'Fractional excretion of sodium varies with glomerular filtration rate' below.)

●Single measurements of serum creatinine may not provide an accurate estimate of the GFR. (See 'Single measurements of serum creatinine may be misleading' below.)

●There are a number of causes of AKI other than prerenal disease in which the FENa can be less than 1 percent. (See 'Multiple causes of acute kidney injury with FENa less than 1 percent' below.)

●The FENa may be above 1 percent when prerenal disease occurs in patients with chronic kidney disease or any cause of sodium wasting, such as diuretic therapy while the diuretic is still acting. (See 'Prerenal disease in patients with chronic kidney disease' below and 'Diuretic therapy' below.)

Fractional excretion of sodium varies with glomerular filtration rate — As mentioned above, the FENa is more reliable in patients with AKI than the urine sodium concentration since it does not vary with the urine volume. However, the FENa value of 1 percent that distinguishes prerenal disease from ATN applies only in patients with a marked reduction in GFR. In patients with less severe disease, the GFR and therefore the filtered sodium load are higher, and, at the same level of urinary sodium excretion, the FENa will be lower. Thus, there is no absolute FENa value that is always diagnostic of prerenal disease, which emphasizes the importance of using FENa in relation to the clinical context.

If, for example, the filtered sodium load is very low at 3000 mEq/day (GFR 20 L/day, which is approximately 14 mL/min [normal 180 L/day or 125 mL/min], and the plasma water sodium concentration is 150 mEq/L), then a FENa below 1 percent is required to appropriately reduce sodium excretion below 20 mEq/day, if prerenal disease were the only problem.

The results are quite different when the GFR is normal or only modestly reduced. At a normal GFR of 180 L/day and plasma water sodium concentration of 150 mEq/L, the filtered sodium load is 27,000 mEq/day. A FENa of 1 percent in this setting represents the excretion of 270 mEq/day, which is greater than the average sodium intake of 80 to 250 mEq/day. Thus, almost all individuals with a normal GFR have a FENa below 1 percent. To reduce sodium excretion to less than 20 mEq/day in the presence of volume depletion requires a FENa below 0.1 percent. This value increases to 0.2 percent if the GFR is 50 percent of normal, 0.4 percent if the GFR is 25 percent of normal, and 1 percent if the GFR is 10 percent of normal.

These examples illustrate the difficulty of using the FENa to diagnose prerenal disease in settings other than severe AKI.

Single measurements of serum creatinine may be misleading — The GFR cannot be estimated from a single serum creatinine measurement, since the serum creatinine is not stable.

Suppose, for example, that a patient with septic shock is admitted to the hospital. The patient may be oliguric and have little or no kidney function. However, if the disease duration is only a few hours, the serum creatinine will still be close to the patient's baseline value since there has not been time for creatinine to be retained. Thus, a serum creatinine of 1 mg/dL (88 micromol/L) could be present in a patient with a normal baseline GFR who now has little or no glomerular filtration. The presence of AKI will be apparent from serial measurements of the serum creatinine, which can initially rise by as much as 2 mg/dL (177 micromol/L) or more per day.

The same principle applies to patients with underlying chronic kidney disease. A patient with a baseline serum creatinine of 2.5 mg/dL (221 micromol/L) may initially have the same value shortly after the onset of severe AKI even though the GFR is markedly reduced. (See 'Prerenal disease in patients with chronic kidney disease' below.)

Multiple causes of acute kidney injury with FENa less than 1 percent — The history, physical examination, and urinary findings are important in evaluating patients with AKI because the FENa can be less than 1 percent in patients with a variety of conditions other than prerenal disease [6-8,15,16].

●Some patients with postischemic ATN who have less severe disease and are evaluated early in the course, as they are making the transition from prerenal disease to postischemic ATN, or remain ischemic with established ATN [15]. In the latter setting, most of the urine output may be coming from a small number of more normal nephrons. A low FENa is appropriate in these nephrons due to the continued kidney ischemia and can be achieved since tubular function is relatively intact. The injured nephrons do not prevent sodium conservation, since they are largely nonfunctioning.

●ATN superimposed upon a chronic prerenal disease, such as cirrhosis or heart failure, due to the mechanism described in the preceding bullet [17].

●Approximately 10 percent of patients with nonoliguric ATN, presumably due to persistent kidney ischemia and less severe disease [15].

●Some patients with ATN associated with (early) sepsis [12,18] and in animal models of septic AKI [19,20].

●AKI due to radiocontrast media or heme pigments (myoglobinuria or hemoglobinuria) [21,22]. (See "Clinical features and diagnosis of heme pigment-induced acute kidney injury", section on 'Urinalysis' and "Contrast-associated and contrast-induced acute kidney injury: Clinical features, diagnosis, and management", section on 'Clinical features'.)

Thus, a FENa less than 1 percent may be present in many patients with ATN; conversely, the presence of "muddy brown" granular casts appear to be a stronger predictor of ATN than FENa. As an example, in one cohort study of 270 patients with AKI, the accuracies of the FENa and the muddy brown granular casts in the urine sediment were evaluated for the prediction of a final diagnosis of ATN (which was confirmed by kidney biopsy in a subset) [23]. ATN was the final diagnosis in 186 patients; 115 had abundant muddy brown granular casts in the urine sediment. The presence of abundant muddy brown granular casts in the urine had a positive predictive value of 100 percent for a final diagnosis ATN. However, 33 percent of those with abundant muddy brown granular casts also had a FENa less than 1 percent. The low FENa with muddy brown granular casts could be due to varied responses across nephron segments; some nephron segments are sodium avid while others are necrotic; particularly the distal portion of the proximal tubule, the S3 segment of the outer medulla, and the thick ascending limb of the loop of Henle are susceptible to injury during hypoperfusion. (See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults".)

In addition to ATN, a FENa less than 1 percent can be seen in other causes of AKI:

●Acute glomerulonephritis or vasculitis in which the diagnosis is typically suspected from the active urine sediment (red cells with or without red cell casts) and variable degrees of proteinuria [11,16].

●Some cases of acute interstitial nephritis, particularly if nonoliguric and if the kidney failure is less severe [24,25]. The diagnosis may be suspected from the history (eg, exposure to a drug known to cause interstitial nephritis) and from laboratory findings such as eosinophilia and, in the urine sediment, white cells with or without white cell casts and, in many cases, eosinophiluria [16]. (See "Clinical manifestations and diagnosis of acute interstitial nephritis", section on 'Laboratory and radiographic findings'.)

●Acute urinary tract obstruction (rare).

In each of the last four conditions, it is presumed that the low FENa reflects well-preserved tubular function in the face of a reduction in GFR [8].

Prerenal disease in patients with chronic kidney disease — In patients with chronic kidney disease, sodium conservation in response to any acute cause of prerenal disease or an abrupt decrease in dietary sodium intake is often impaired [26-28]. As a result, the FENa may initially be greater than 1 percent.

One or more of the following three factors may contribute to the inability to conserve sodium in patients with chronic kidney disease [26-28]:

●The osmotic diuresis induced by increased urea excretion per nephron due to the decreased number of functioning nephrons

●Tubular injury due to the underlying disease

●Probably most important, an inability to acutely shut off natriuretic forces

Support for the last mechanism was provided by a study of patients with chronic kidney disease and apparent salt wasting (with acute obligatory losses as high as 300 mEq/day) [28]. All of the patients were able to maintain sodium balance at a sodium intake of only 5 mEq/day when intake was gradually reduced over a period of weeks, rather than acutely.

Diuretic therapy — A FENa greater than 1 percent can occur in patients with prerenal disease who are treated with diuretics if the measurement is made while the diuretic is still acting. In such patients, the fractional excretion of urea (FEUrea) may be helpful, and the FENa can be measured again after the diuretic effect has worn off. (See 'Urea' below.)

FRACTIONAL EXCRETION OF OTHER MOLECULES — Values for the FENa and urine sodium concentration that are not low are difficult to interpret when there is concurrent use of diuretics, which are commonly used in patients with AKI. The ensuing natriuresis will raise the FENa, even in patients with prerenal disease. In older patients, the natriuresis may be delayed following diuretic administration. This was illustrated in a study of patients given a single dose of furosemide (1 mg/kg) [29]. The peak FENa was achieved at 30 minutes following injection in younger patients (17 to 40 years old) and at 120 minutes following injection in older patients (75 to 80 years old). Thus, timing of the urine sample can lead to variability in the measurement of FENa, rendering this test less reliable after diuretic use.

Because of this variability, the clinical utility of the FEUrea, lithium, and uric acid have been studied since these substances are primarily reabsorbed and, with uric acid, secreted in the proximal tubule, which is proximal to the sites of action of loop diuretics (loop of Henle) or thiazide diuretics (distal tubule). Of these, the FEUrea may be most useful in patients being treated with diuretics, although inconsistent results have been reported.

The accuracy of all of the proximal markers requires intact proximal function. As an example, the fractional excretion of these markers will increase, independent of volume status, when proximal reabsorption is reduced, as occurs during a glucose osmotic diuresis in uncontrolled diabetes mellitus [30].

Urea — The measurement of the FEUrea may be useful in patients with AKI. FEUrea is generally 50 to 65 percent in acute tubular necrosis (ATN) and below 35 percent in prerenal disease [30-36], and the effect of diuretics on the FEUrea is less than the effect on FENa [36]. Some [30,33], but not all [31] prospective studies have shown that FEUrea is superior to FENa for distinguishing between prerenal and intrinsic disease among patients on diuretics. However, FENa retains superior accuracy for discriminating between prerenal and intrinsic disease in patients not on diuretics [31,33] and in those for whom the diuretic effect has worn off.

Lithium — Endogenous lithium is present in the circulation in trace amounts. In the kidney, filtered lithium is primarily reabsorbed in the proximal tubule. Thus, proximal reabsorption of lithium is increased in prerenal disease, which will lower lithium excretion. This proximal tubular reabsorption of lithium is not influenced by loop diuretics or by thiazide diuretics other than those such as chlorothiazide that also act in the proximal tubule because they inhibit carbonic anhydrase [37].

The fractional excretion of lithium (FELi) is approximately 20 percent in healthy controls. The potential utility of the FELi was illustrated in a study of 46 patients with AKI (prerenal disease in 26 and ATN in 18) [38]. A FELi below 15 percent (and usually below 10 percent) was highly suggestive of prerenal disease, independent of diuretic therapy. In comparison, the mean FELi was 26 percent in ATN. A FELi below 15 percent or above 25 percent had a sensitivity of 93 to 96 percent but a much lower specificity of 56 to 72 percent. The sensitivity was higher than seen with the FENa (75 percent) in patients treated with diuretic therapy. The clinical utility of the FELi is limited by the lack of widespread availability of trace lithium measurements.

Uric acid — Uric acid is another marker that is almost entirely handled in the proximal tubule. As a result, the fractional excretion of uric acid is not affected by loop or thiazide diuretics.

In the study of 46 patients with AKI cited in the preceding section, a fractional excretion of uric acid below 12 percent was suggestive of prerenal disease (sensitivity 68 percent [much lower than the sensitivity of the FELi in the same study], specificity 78 percent), while values above 20 percent were highly suggestive of ATN, although the specificity was low (sensitivity 96 percent, specificity 33 percent) [38].

INVESTIGATIONAL BIOMARKERS — Although not yet in widespread clinical use, novel urinary and serum biomarkers are being evaluated for the possible early detection of acute tubular necrosis (ATN) that might permit therapy that would minimize the degree of AKI. (See "Investigational biomarkers and the evaluation of acute kidney injury".)

SUMMARY AND RECOMMENDATIONS

●Formula for the fractional excretion of sodium (FENa) – The FENa measures the percent of filtered sodium that is excreted in the urine. In the following formula, UNa and UCr represent concentrations of urine sodium and creatinine, respectively, and SNa and SCr represent concentrations of serum sodium and creatinine, respectively. Calculators for the FENa are available using either standard units (calculator 1) or SI units (calculator 2). (See 'Formula' above.)

●Overview – The FENa is commonly used to assist in differentiating prerenal disease (a reduction in glomerular filtration rate [GFR] due solely to decreased kidney perfusion) from acute tubular necrosis (ATN), the two most common causes of acute kidney injury (AKI). (See 'Overview' above.)

●FENa values and AKI – Among patients with suspected prerenal disease or ATN, we typically measure the FENa. A value of the FENa below 1 percent commonly indicates prerenal disease; in comparison, a value between 1 and 2 percent may be seen with either disorder, while a value above 2 percent usually indicates ATN. (See 'Fractional excretion of sodium in acute kidney injury' above and "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults".)

●Limitations of FENa and use of fractional excretion of urea (FEUrea) – There are a number of limitations to use of the FENa in patients with AKI:

•The FENa defining prerenal disease varies with the GFR, ranging from less than 0.1 percent in patients with normal kidney function to more than 1 percent in AKI. (See 'Fractional excretion of sodium varies with glomerular filtration rate' above.)

•Single measurements of the serum creatinine are not a good measure of GFR, since the serum creatinine is not stable in early AKI. (See 'Single measurements of serum creatinine may be misleading' above.)

•In addition to prerenal disease, the FENa can be less than 1 percent in patients who have postischemic ATN, ATN superimposed upon a chronic prerenal disease, nonoliguric ATN, AKI due to radiocontrast media or heme pigments, acute glomerulonephritis or vasculitis, acute interstitial nephritis, or (rarely) acute urinary tract obstruction. (See 'Multiple causes of acute kidney injury with FENa less than 1 percent' above.)

•The FENa may initially be greater than 1 percent when prerenal disease occurs in patients with underlying chronic kidney disease. (See 'Prerenal disease in patients with chronic kidney disease' above.)

•The FENa and the urine sodium concentration may be difficult to interpret with concurrent diuretic therapy. In this setting, measuring the FEUrea may be helpful. (See 'Diuretic therapy' above and 'Urea' above.)

●Fractional excretion of other molecules – The fractional excretion of lithium (FELi) and uric acid are not commonly measured, since data are more limited than with FENa and FEUrea. Other biomarkers of AKI are being evaluated. (See 'Fractional excretion of other molecules' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Rasheed Balogun, MD, who contributed to earlier versions of this topic review.