Acute Kidney Injury in the Intensive Care Unit



Acute Kidney Injury in the Intensive Care Unit


Jahan Montague

Konstantin Abramov



Overview of Acute Kidney Injury

Sudden disruption of previously normal or stable kidney function, usually occurring over hours or days, is termed acute kidney injury (AKI), formerly referred to as acute renal failure. The new term underscores the diverse clinical context in which patients with many forms and causes of AKI may present, while the term failure implies an end stage of this clinical spectrum. The pathogenesis of AKI differs from that of chronic kidney disease (CKD), in which nephron loss is more gradual. However, AKI can occur in the setting of antecedent CKD.

AKI is often diagnosed when a patient is noted to have azotemia. This elevation in the blood urea nitrogen (BUN) and serum creatinine typically represents a decline in glomerular filtration rate (GFR), but in certain cases may reflect increased production without any reduction in GFR (Table 73.1).
Oliguria, a reduction in urine output to less than 20 mL per hour, may be present, although many forms of AKI are nonoliguric. When tubular reabsorption of glomerular filtrate is reduced as a result of either tubular dysfunction or diuretic administration, patients may be polyuric even though GFR is markedly reduced.








Table 73.1 Causes of Blood Urea Nitrogen or Serum Creatinine Elevation without Reduction of Glomerular Filtration Rate




Increased biosynthesis of urea
   Gastrointestinal bleeding
   Drug administration
      Corticosteroids
      Tetracycline
   Increased protein intake
   Amino acid administration
   Hypercatabolism and febrile illness
Increased biosynthesis of creatinine
   Increased release of creatine from muscle (rhabdomyolysis)
Drug interference with tubular creatinine secretion
   Cimetidine
   Trimethoprim
Spuriously elevated creatinine colorimetric assay
   Ketoacids (diabetic ketoacidosis)
   Cephalosporins

AKI can occur prior to a significant increase in creatinine. Several recent studies reported that even a small rise in serum creatinine correlated with increased mortality [1,2]. Therefore, the Acute Dialysis Quality Initiative Group has proposed a new classification of AKI based not only on serum creatinine but also on the degree of urinary output reduction and the requirement for renal replacement therapy [3]. This classification is reflected in the RIFLE criteria (Table 73.2), which have been shown to predict renal outcome and mortality in a variety of critically ill and hospitalized patients [4,5].

AKI may stem from any of three general conditions: impaired renal perfusion without parenchymal injury, damage to the renal parenchyma, or obstruction of the urinary tract. These etiologies are referred to as prerenal, renal, or postrenal causes of AKI, respectively, and are summarized in Table 73.3. Although it is helpful to consider the complete array of renal diseases when evaluating AKI, in the inpatient setting two thirds of cases will be due to either acute tubular necrosis (ATN) or prerenal azotemia. Hence, an extensive search for other forms of renal disease is indicated in the intensive care unit (ICU) setting only when suggested by clinical signs or laboratory findings such as urinary abnormalities indicative of glomerular disease.








Table 73.2 Rifle Criteria




























  Serum creatinine (Cr)/glomerular filtration rate (GFR) criteria Urinary output criteria
Risk Cr increase × 1.5 above baseline or GFR decline > 25% < 0.5 mL/kg/h × 6 h
Injury Cr increase × 2 above baseline or GFR decline > 50% < 0.3 mL/kg/h × 24 h
Failure Cr increase × 3 above baseline or Cr ≥ 4 mg/dL or GFR decline > 75% < 0.3 mL/kg/h 24 h or anuria × 12 h
Loss Persistent AKI > 4 wk  
End stage End-stage renal disease (AKI > 3 mo)  
Summary of RIFLE criteria of AKI. Sensitivity for AKI increases toward the top of the chart, while specificity increases toward the bottom of the chart. [Bellomo R, Ronco C, Kellum JA, et al; Acute Dialysis Quality Initiative Workgroup: Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 8:R204–R210, 2004.]








Table 73.3 Causes of Acute Kidney Injury




Prerenal azotemia
   Hypovolemia
   Reduced effective circulating volume
   Autoregulatory failure
Intrinsic renal disease
   Glomerular diseases
   Vascular diseases (main renal artery and microcirculation)
   Tubulointerstitial disease
   Acute tubular necrosis
   Acute cortical necrosis
Postrenal failure
   Ureteric obstruction (bilateral or solitary kidney)
   Lower tract obstruction (bladder neck or urethra)


Prerenal Azotemia and Autoregulatory Failure

When renal perfusion pressure decreases to a point at which GFR falls, prerenal azotemia is said to be present. This is a functional condition that does not represent intrinsic renal disease as such, although it may be superimposed on preexisting renal disease. The causes of prerenal azotemia are listed in Table 73.4. Normalization of renal blood flow, if possible, promptly restores renal function.

Hypovolemia serious enough to cause prerenal azotemia may result from gastrointestinal losses, hemorrhage, venous pooling, sequestering of fluid in “third spaces,” or excessive urinary or skin losses of sodium and water. Patients will usually exhibit signs of hypovolemia, including thirst, diminished skin turgor and mucous membrane moistness, and postural hypotension. Patients whose vascular volume is functionally reduced by the hemodynamic alterations of congestive heart failure, cirrhosis, or hypoalbuminemia may develop prerenal azotemia despite having a normal or even expanded extracellular fluid (ECF) volume. Because the effective circulatory volume is reduced, renal perfusion is impaired just as in true hypovolemia.

When glomerular perfusion is threatened, autoregulatory mechanisms help maintain glomerular capillary pressure. If autoregulatory mechanisms are inoperative, a given reduction
in renal blood flow provokes a sharper decline in GFR. The mechanisms of these processes are shown in Figure 73.1. Use of nonsteroidal anti-inflammatory drugs (NSAIDs) in patients with renal hypoperfusion, for example, can lead to severe AKI [6,7,8]. Likewise, administration of angiotensin-converting enzyme (ACE) inhibitors in patients whose renal blood flow is obstructed by bilateral renovascular renal artery stenoses can cause severe azotemia [9].








Table 73.4 Causes of Prerenal Azotemia




Hypovolemia
   Gastrointestinal losses
      Vomiting
      Diarrhea
      Surgical drainage
   Renal losses
      Osmotic agents
      Diuretics
      Renal salt-wasting disease
      Adrenal insufficiency
   Skin losses
      Burns
      Excessive diaphoresis
   Hemorrhage
   Translocation of fluid (“third spacing”)
      Postoperative
      Pancreatitis
Reduced effective circulating volume
   Hypoalbuminemia
   Hepatic cirrhosis
   Left ventricular cardiac failure
   Peripheral blood pooling (vasodilator therapy, anesthetics, anaphylaxis, sepsis, toxic shock syndrome)
   Renal artery occlusion
   Small vessel disease (malignant hypertension, toxemia, scleroderma)
   Renal vasoconstriction (hypercalcemia, hepatorenal syndrome, cyclosporine, pressor agents)
Autoregulatory failure
   Nonsteroidal anti-inflammatory drugs (preglomerular vasoconstriction)
   Angiotensin-converting enzyme inhibitors (postglomerular vasodilation)






Figure 73.1. Diagrammatic representation of autoregulation and deregulation caused by use of either nonsteroidal anti-inflammatory drugs (NSAIDs), which lead to afferent (Aff.) vasoconstriction, or angiotensin-converting enzyme (ACE) inhibitors, which produce efferent (Eff.) vasodilation. Ang II, angiotensin II; GFR, glomerular filtration rate; PGs, prostaglandins.

Reduced renal perfusion slows down the flow of filtrate through the renal tubules, enhancing the reabsorption of urea. Because creatinine is not reabsorbed in the renal tubules, its clearance is unaffected by these nephronal factors. Thus, the clearance of urea is reduced disproportionately to that of creatinine, explaining the unusually high BUN–creatinine ratio that is often seen in prerenal states. In such situations, the BUN–creatinine ratio typically exceeds 20 to 1.

A high urea–creatinine ratio, however, is not pathognomonic of prerenal azotemia. When urea production is accelerated in catabolic states (as is seen in tetracycline and corticosteroid therapy) or by resorption of a large hematoma or gastrointestinal bleeding, BUN levels rise unless renal urea clearance can increase to meet the augmented urea burden. To establish whether a high BUN–creatinine ratio is due to increased urea production or reduced excretion, calculation of the fractional urea clearance may be useful.

The hallmark of prerenal conditions is the intense renal conservation of salt and water as reflected in the urine composition, which generally shows a low sodium concentration (UNa < 10 mEq per L; fractional excretion of sodium [FENa] < 1%) and a high osmolality (UOsm > 500 mOsm per kg). Renal conservation of sodium involves both proximal and distal tubular mechanisms. A low urinary sodium concentration is expected in these states; its absence signifies a coexisting abnormality of tubular function, the effect of diuretics, or the presence of nonreabsorbable anionic substances in the urine, such as bicarbonate in patients with metabolic alkalosis, or certain penicillins, that obligate the excretion of cations like sodium. Impaired sodium reabsorption is also seen during osmotic diuresis and in certain forms of chronic renal disease.


Intrinsic Renal Disease

Reduced renal function may also result from renal parenchymal injury. Such injury may arise from glomerular, vascular, and tubulointerstitial disorders (Table 73.3) and may represent either primary kidney disease or the renal effects of an underlying systemic illness (e.g., systemic lupus erythematosus).


Glomerular and Vascular Diseases

The GFR may be abruptly reduced in acute glomerulonephritis. In poststreptococcal glomerulonephritis, the prototypic nephritic disorder, patients often present with AKI and oliguria.

Hypertension and edema result from their inability to excrete salt and water normally. The constellation of hypertension, edema, azotemia, and hematuria is known as the acute nephritic syndrome. Although the history of a previous sore throat or streptococcal infection may provide diagnostic clues, the urinalysis is particularly valuable. The urine may be grossly bloody or tea colored. The urinary sediment contains red blood cells (RBCs) and often RBC casts (Fig. 73.2). Similar findings are frequent in patients with other primary nephritic disorders as well as in secondary nephritides, such as those seen in systemic lupus, and bacterial endocarditis. The crescentic glomerulonephritides (rapidly progressive glomerulonephritis) can evoke the acute nephritic syndrome.

Diseases affecting either the main renal arteries or their branches may precipitate AKI. Renal artery occlusion by acute thrombosis or thromboembolism typically only causes AKI if it is bilateral or involves a solitary functioning kidney. These processes may be silent or may produce flank pain and hematuria, particularly if abrupt enough to cause renal infarction. Fever, moderate leukocytosis, and an elevated serum level of
lactate dehydrogenase should raise the suspicion of infarction. With rare exceptions, renal arterial thromboembolism occurs only in the settings of acute myocardial infarction, atrial fibrillation, bacterial endocarditis, hypercoagulable disorders, or other cardiac valvular disease.






Figure 73.2. Typical urinary sediments from patients with parenchymal renal diseases. A: Sediment from patient with acute glomerulonephritis showing free red blood cells and red blood cell casts. B: Sediment from patient with acute interstitial nephritis demonstrating pyuria and white blood cell cast. C: Typical muddy brown, coarse, granular casts in a patient with acute tubular necrosis.

Acute renal vein thrombosis seldom causes renal failure unless both kidneys are simultaneously occluded. Acute flank pain and hematuria are the clinical hallmarks. Renal venous obstruction may occur as a complication of nephrotic syndrome and renal cell carcinoma. Microscopic occlusion of smaller vessels occurs in a variety of disorders, including atheroembolic renal disease, thrombotic thrombocytopenic purpura (TTP) and hemolytic-uremic syndrome, scleroderma, postpartum kidney injury, and malignant hypertension. Scleroderma or malignant hypertension may appear as AKI, with severe blood pressure elevation due to activation of the renin–angiotensin system. These vascular disorders produce renal injury by reducing glomerular blood flow. Because the lesion is proximal to the glomerulus, the urine sediment is usually acellular and bland. Vasculitis produces AKI either through direct involvement of the renal arterial system or by inducing glomerulonephritis. Often, microscopic polyarteritis or Wegener’s granulomatosis may present with evident renal parenchymal disease, as suggested by urinary abnormalities such as microscopic hematuria, RBC casts, and proteinuria. These patients may present to the ICU when there is multiorgan involvement, such as the pulmonary disease that occurs in Wegener’s granulomatosis. Fulminant presentations with severe hypoxemia and pulmonary hemorrhage may be accompanied by rapidly progressive renal dysfunction. In these cases, glomerular involvement may range from focal and segmental necrotizing glomerulitis to severe crescentic glomerulonephritis.


Tubulointerstitial Diseases

Two syndromes are responsible for most cases of parenchymal AKI in hospitalized populations: ATN and acute interstitial nephritis (AIN).


Acute Tubular Necrosis

ATN is a syndrome that may result from renal ischemia or exposure to nephrotoxins such as aminoglycoside antibiotics, radiocontrast agents, heavy metals, and myoglobin. Although historically many other names have been applied to this syndrome, the term acute tubular necrosis prevails even though frank tubular cell necrosis does not appear in all cases. Historically, the pathophysiology of AKI in ATN has been attributed to three processes: (a) obstruction of tubular lumens by sloughed epithelial cells and cellular debris, (b) back-leak of filtered wastes into the circulation through the disrupted tubular epithelium, and (c) sustained reduction in glomerular blood flow following the inciting stimulus. Along these lines, severe cortical vasoconstriction has been noted early in the course of ATN [10], which is likely mediated by endothelial cell injury and locally acting vasoconstrictors, such as endothelin [11]. Afferent arteriolar vasoconstriction has also been described. In ATN, impaired proximal solute reabsorption increases distal chloride delivery to macular densa, which, in turn, mediates afferent constriction via the secretion of adenosine. This process is termed tubuloglomerular feedback [12]. However, it remains unclear whether renal vasoconstriction has a central role in the pathogenesis of ATN, since restoring renal blood flow with vasodilators in a variety of animal models of AKI does not always preserve the GFR [13]. Nonetheless, this process may be important in the initiation of certain forms of ATN such as radiocontrast toxicity (see following discussion).

As already noted, relatively modest hypoperfusion leads to prerenal azotemia, characterized by a modest reduction in urine output and GFR and preservation of tubular function, which is rapidly reversible. However, a more critical decrease in renal perfusion leads to medullary hypoperfusion and ischemic ATN, with greater reductions in GFR, abnormalities of tubular function, and often histologic evidence of tissue injury. Because recovery depends on cellular regeneration, reversal is much slower than in prerenal azotemia [14]. The most extreme form of hypoperfusion injury is cortical ischemia associated with either patchy or diffuse cortical necrosis, typically manifesting the most severe reduction in GFR and a much less certain prognosis for recovery of renal function. The medullary thick ascending limb segment of the loop of Henle is particularly vulnerable to ischemic and nephrotoxic insults because of a combination of low ambient partial pressure of oxygen and intense,
transport-driven oxygen consumption. Other factors, such as adenosine triphosphate depletion activation of phospholipases, cytosolic and mitochondrial calcium overload, and release of free radicals [15], may contribute to cellular damage. In experimental ATN, cytoskeletal reorganization can be demonstrated in proximal tubular cells, leading to loss of normal cell polarity. The sodium–potassium adenosine triphosphatase transport system may thus be translocated from its normal basolateral position to the apical surface of the cell, impairing reabsorptive function.








Table 73.5 Protein Biomarkers for the Early Detection of Acute Kidney Injury




























Biomarker Associated injury
Cystatin C Proximal tubule injury
KIM-1
NGAL
L-FABP
Ischemia and nephrotoxins
Netrin-1 Sepsis, ischemia, nephrotoxins I
NHE3 Prerenal, ischemia, postrenal
α-GST Acute rejection, proximal tubule injury
π-GST Acute rejection, distal tubule injury
Cytokines (IL-6, IL-8, IL-18)
Actin–actin depolymerizing F
Keratin-derived chemokine
Delayed graft function
GST, glutathione S-transferase; IL, interleukin; KIM, kidney injury molecule; L-FABP, L-type fatty acid binding protein; NGAL, neutrophil gelatinase-associated lipocalin; NHE, sodium–hydrogen exchanger.
From Ronco C, Haapio M, House AA, et al: Cardiorenal syndrome. J Am Coll Cardiol. 52:1527, 2008.

There is growing evidence that immune system plays a critical role in pathogenesis of ATN through recruitment of various inflammatory cells, cytokine release, complement activation, and induction of tubular cell apoptosis [16,17]. Adhesion molecules, such as intracellular adhesion molecule 1 (ICAM-1), appear to play a role in the development of postischemic ATN in experimental animal models. However, anti-ICAM antibody failed to protect against ischemic AKI in a clinical trial of kidney transplant patients [18]. Another regulatory molecule expressed in the kidney, the protein neutrophil gelatinase-associated lipocalin (NGAL), is released early in the course of ischemic ATN and appears to attenuate tubular cell injury and apoptosis [19]. There is considerable interest in using NGAL and other molecules as biomarkers of early kidney injury (Table 73.5) with the hope that timely diagnosis of AKI will allow clinicians to make therapeutic interventions that will ultimately improve outcomes (discussed further in “Diagnosis” section).

History of exposure to a predisposing factor, such as prolonged ischemia or toxin, can be elicited in approximately 80% of patients with ATN. Most individuals with this syndrome have the classic findings of sloughed renal tubular epithelial cells, epithelial cell casts, or muddy brown granular casts in the urinary sediment (Fig. 73.2). These findings are not seen in prerenal azotemia. In addition, calculating the fractional excretion of filtered sodium (FENa) may enable the clinician to differentiate between ATN and prerenal azotemia. The FENa expresses urinary sodium excretion as a percentage of filtered load. It provides a more precise representation of tubular sodium avidity than the urinary sodium concentration because it is not influenced by changes in urine concentration or flow rate. The FENa is very low in prerenal azotemia as a result of active sodium reabsorption by the renal tubules. When frank tubular damage has occurred, as in ATN, the tubules can no longer reclaim sodium efficiently, and the FENa is generally high (Fig. 73.3). See “Diagnosis” section for more detail on the FENa. Urinary concentration is also impaired in tubular necrosis; as a result, urinary osmolality approximates that of plasma, and the BUN–creatinine ratio is less than 20.


Acute Interstitial Nephritis

The term acute interstitial (or tubulointerstitial) nephritis encompasses a collection of disorders characterized by acute inflammation of the renal interstitium and tubules. Depending on the specific nature of the condition, the inflammatory infiltrate may consist of a combination of neutrophils, eosinophils, and lymphocytes or plasma cells. Most cases of AIN represent an allergic reaction with eosinophilia and skin eruptions, usually induced by medication. Interstitial disease can also occur as a result of infectious agents, including brucellosis, leptospirosis, legionella, toxoplasmosis, and Epstein–Barr virus.

These disorders are to be distinguished from the familiar entity acute pyelonephritis. Acute pyelonephritis is a suppurative
disease of the tubulointerstitium, usually caused by bacterial infection ascending from the urinary bladder. Acute pyelonephritis rarely causes renal dysfunction. AIN, however, is an important cause of AKI and is discussed in detail later.






Figure 73.3. Diagnostic parameters in acute renal failure. Two laboratory tests used to distinguish prerenal (PR) azotemia from acute tubular necrosis (ATN) are shown. Left: Urinary sodium concentration (UNa, mEq per L). Right: Fractional excretion of sodium (FENa, %). Area within each symbol denotes the proportion of patients with each condition correlated with the laboratory parameter. Note that although considerable numbers of patients with PR and ATN fall in an intermediate zone of UNa (20 to 40 mEq per L), the FENa almost completely differentiates the two groups. [Adapted from Rudnick MR, Bastl CP, Elfinbein IB, et al: The differential diagnosis of acute renal failure, in Brenner BM, Lazarus JM (eds): Acute Renal Failure. New York, Churchill Livingstone, 1988, p 177.]


Postrenal Azotemia

The term postrenal azotemia, or obstructive uropathy, refers to azotemia caused by obstruction of urine flow from the kidneys. Renal outflow obstruction has many causes, but the most common causes are prostatic enlargement, nephrolithiasis, and genitourinary tumors. For obstruction to produce azotemia, both kidneys must be involved because one normally functioning kidney is sufficient to maintain a near normal GFR. AKI may occur with unilateral obstruction in a patient who has one functioning kidney or in whom unilateral obstruction is superimposed on underlying CKD.

The clinical history often helps in the diagnosis of obstructive uropathy. Prior kidney stones should raise the index of suspicion for obstruction, particularly in the setting of symptoms of renal colic. AKI in an elderly man who has been experiencing urinary hesitancy most likely represents obstruction of the bladder outlet by an enlarged prostate. A history of genitourinary malignancy in an azotemic patient also makes obstruction the most likely diagnosis. AKI in the setting of painless gross hematuria and a history of NSAID use should prompt a suspicion of papillary necrosis, a condition in which sloughed-off renal papilla can cause bilateral ureteral obstruction. Finally, renal failure in a newborn infant is likely to be due to congenital anatomic ureteral obstruction.

When urine output declines precipitously or ceases entirely (anuria), complete obstruction of the urinary tract must be ruled out. Such an obstruction is likely to be located at the bladder outlet because the probability of simultaneous obstruction in both ureters from any cause is remote. If the patient has only one kidney (e.g., due to previous nephrectomy, unilateral renal disease, or congenital solitary kidney), however, anuria may occur with unilateral ureteral obstruction. Even though complete obstruction is a common cause of anuria, partial obstruction is not always associated with a decline in urine output. With partial obstruction, damage to the kidney may impair the ability to concentrate urine, resulting in a polyuric state (acquired nephrogenic diabetes insipidus) [20,21]. In patients with complete unilateral obstruction of a ureter, the contralateral kidney often sustains a normal urine output.

As discussed later, urologic causes of AKI are best diagnosed by renal imaging techniques. The urine chemistry is generally of little help in diagnosing obstructive uropathy. Likewise, the urinalysis provides only indirect evidence of a possible cause of AKI. Hematuria reflects trauma to the urinary epithelium caused by the obstructing lesion. Crystals (calcium oxalate or uric acid) in the urine sediment may suggest a kidney stone.


Clinical Syndromes Associated with AKI in the Intensive Care Setting

With the higher level of acuity of illness, and more radical approaches to surgical and pharmacologic therapeutics, the incidence of AKI is increasing. As with other areas of clinical medicine, patterns of presentation often can be recognized and can lead the physician to the most likely diagnoses. The following section explores in greater detail the specific AKI syndromes most commonly encountered in the ICU (listed in Table 73.6).








Table 73.6 Intensive Care Syndromes Associated with Acute Kidney injury (AKI)






Ischemic AKI
   Extracellular volume depletion
   Postoperative (particularly cardiac surgery)
   Severe ventricular dysfunction or cardiogenic shock
   Sepsis
   Pancreatitis
   Trauma
   Burns
Acute bilateral cortical necrosis
Nephrotoxicity and drug-induced AKI
   Myoglobinuric AKI
   Radiocontrast nephropathy
   Drugs (see Table 73.13)
Renal vascular disease
   Major vessel disease
      Renal artery embolism or thrombosis
      Renal vein thrombosis
   Microvascular disease
      Atheroembolism
      Vasculitis
      Scleroderma
Cancer related
   Obstructive uropathy
   Hypercalcemia
   Tumor-lysis syndrome
   ATN secondary to chemotherapy
Renal dysfunction with liver disease
   Prerenal azotemia
   ATN
   Hepatorenal syndrome
ATN, acute tubular necrosis.


Ischemic Acute Kidney Injury

The most common forms of AKI in the ICU result from renal hypoperfusion. Because frank hypotension is documented in fewer than half of these cases, the causal events may often be overlooked or obscured by multiple factors. Frequently, more than one causal factor is necessary to provoke AKI. For example, the presence of hypovolemia enhances the risk for AKI due to nephrotoxic insults.


Extracellular Volume Depletion

Extracellular volume depletion accounted for approximately 17% of cases of AKI in a prospective study of AKI in a major hospital [14]. In most instances, urinary losses are the cause of hypovolemia. Injudicious use of diuretics and the osmotic diuresis that accompanies diabetic hyperglycemia are the most common etiologies. Cessation of diuretic therapy and volume repletion lead to rapid recovery; consequently, the mortality is quite low [14].

In rare instances, gastrointestinal losses of substantial magnitude may lead to AKI. (In the developing world, however, this is one of the most common causes of AKI and the major cause of morbidity and mortality in epidemic cholera.) In such cases, the source of the gastrointestinal losses, either gastric or intestinal, may lead to distinctive electrolyte abnormalities. In the former, metabolic alkalosis mandates repletion with
chloride-rich replacement solutions (normal saline, usually with potassium chloride, as most patients are also hypokalemic). With intestinal losses of fluid, metabolic acidosis often ensues, and appropriate replacement may consist of a buffer solution of either isotonic bicarbonate or lactate-containing (Ringer’s) solution in combination with saline. (This is more fully discussed in Chapter 71.)

Transdermal fluid losses usually occur in the setting of major burns, with the degree of hypovolemia and the severity of AKI corresponding to the extent of thermal injury (body surface-area involvement). Significant burns can lead to severe hypovolemia as a result of massive evaporative and exudative fluid loss across the damaged epidermis as well from redistribution of fluid due to edema in the injured tissues. This hypovolemia can stimulate a sympathetic nerve-mediated response with resultant renal vasoconstriction. In some instances, the severity of renal vasoconstriction, superimposed on hypovolemia, culminates in ATN. In addition, deeper thermal injury with skeletal muscle involvement may induce myoglobinuric AKI (see following discussion). Dermal losses of fluid are also seen in the setting of hyperthermia and heat stroke. The evaporative loss of sweat, which is hypotonic, leads to a hypertonic dehydration in these cases. Replacement with half-normal saline corrects free water and sodium deficits.


Postoperative

Postoperative AKI has long been recognized as a common complication of major vascular, abdominal, and open-heart surgery. The pathogenesis of postoperative renal dysfunction varies with the type of the surgery and the preoperative condition of the patient. AKI following abdominal surgery is often the result of translocation of fluid into the peritoneal cavity. In this phenomenon, third spacing causes intravascular hypovolemia and subsequent renal hypoperfusion. AKI is uncommon in patients undergoing routine abdominal surgery, but the risk is substantial in surgery for obstructive jaundice; this complication may develop in approximately 10% of patients [22]. Major vascular surgery, particularly aortic repairs, is also frequently complicated by AKI, especially in the setting of a ruptured aortic aneurysm [23]. Elective repair of an aortic aneurysm is seldom associated with AKI unless cross clamping is placed above the renal arteries. Although the definition of AKI and its incidence varies among studies, cardiac surgery appears to generate most of the cases in the acute care hospital. In a prospective study, cardiac surgery accounted for nearly two thirds of the postoperative AKI with an overall incidence of 15% [14]. Repeated episodes of AKI and sepsis complicating AKI are associated with substantially higher mortality (85%) after surgery.

Myers and Moran [10] have described three distinct clinical patterns of AKI following open-heart surgery. The abbreviated pattern, observed in 80% to 90% of patients, usually has an abrupt onset after surgery, followed by a brief and mild rise in the serum creatinine, peaking by 3 to 4 days, after which recovery is rapid. This form of AKI is frequently associated with the use of vasoconstrictors, such as norepinephrine and epinephrine, in the immediate postoperative period. The overt form is associated with a more severe reduction in GFR and rise in serum creatinine, which peaks 1 to 2 weeks after surgery. This is generally associated with poor cardiac performance following surgery, whereas recovery is associated with improved ventricular function. The protracted form of AKI generally follows a second insult after surgery, such as sepsis or pericardial tamponade, and is associated with prolonged AKI and a poor prognosis.

Several predisposing factors for the development of postoperative AKI have been identified, which can be used to stratify risk (Table 73.7). These include emergent surgery, an elevated preoperative serum creatinine, the use of an intra-aortic balloon pump, and combined coronary artery bypass graft and valvular surgery [24]. Still, it is often difficult to prospectively identify those patients at heightened risk for perioperative AKI, especially since 40% of patients who develop AKI do not have frank perioperative hypotension or evidence of shock. Other factors appear to be important in the development of AKI. For example, aprotinin, an antifibrinolytic agent used until recently to decrease perioperative blood loss in cardiac surgery patients, tends to increase AKI and postoperative mortality [25]. Prolonged cardiopulmonary bypass appears to induce oxidative stress, embolism, and systemic inflammation, thus contributing to AKI [26]. An improved mortality and reduced incidence of AKI is observed with the use of “off-pump” technology in one large observational study [27]. Prospective, randomized trials are underway to study the effects of the off pump cardiac surgery.








Table 73.7 A Clinical Score to Predict Aki Requiring Dialysis after Cardiac Surgery














































Risk factor Points
Female gender 1
Congestive heart failure 1
Left ventricular ejection fraction < 35% 1
Preoperative use of intra-aortic balloon pump 2
Chronic obstructive pulmonary disease 1
Diabetes requiring insulin 1
Previous cardiac surgery 1
Emergency surgery 2
Valvular surgery only 1
Coronary artery bypass graft surgery plus valvular surgery 2
Other cardiac surgeries 2
Preoperative serum creatinine 1.2 to < 2.1 mg/dL 2
Preoperative serum creatinine ≥ 2.1 mg/dL 5
Minimum score, 0; maximum score, 17. Risk of development of AKI requiring dialysis increases with higher score. Frequency of AKI requiring dialysis for score of 0–2 point is 0.5%, 3–5 points is 2%, 6–8 points is 8%, 9–13 points is 22%.
From Thakar CV, Arrigain S, Worley S, et al: A clinical score to predict acute renal failure after cardiac surgery. J Am Soc Nephrol 6:162, 2005.

A number of methods have been used to try to protect kidney function in patients undergoing surgery. Administration of “low-dose” dopamine has long been advocated for the prevention of AKI but has fallen out of favor due to the lack of efficacy [28]. A meta-analysis of fenoldopam, a selective dopamine receptor agonist, demonstrated a reduction in mortality and the need for dialysis in cardiac surgery patients [29]. A large, randomized, prospective trial of preventive role of fenoldopam is underway. Nesiritide, a recombinant human B-type natriuretic peptide, which increases diuresis, natriuresis, and afterload reduction, was studied in a double-blind, randomized trial of 300 patients with mostly preserved renal function undergoing coronary artery bypass graft surgery [30]. The use of nesiritide was associated with improved postoperative serum creatinine and reduced length of hospital stay as compared with the placebo. Nevertheless, significant concerns remain regarding the safety of nesiritide, especially in patients with acute decompensated heart failure (ADHF) and reduced renal function (discussed later in the chapter). Several other strategies, including perioperative N-acetylcysteine and mannitol administration, have not been shown to be effective for the prevention of postoperative AKI [31,32]. The use of calcium channel blockers, ACE
inhibitors, or diuretics has been disappointing in this context as well [33]. In fact, perioperative furosemide use has been associated with detrimental effect on renal function after cardiac surgery [34]. Furosemide should be used only in patients with definite volume overload.

Prevention of postoperative renal failure still hinges on withdrawal of vasopressors as early as safely possible and maintenance of adequate perioperative intravascular volume. Identification of modifiable risk factors for the prevention of postoperative AKI is paramount. A retrospective study of 3,500 patients identified three potentially modifiable risk factors associated with AKI after cardiac surgery, such as preoperative anemia, perioperative RBC transfusions, and the need for surgical reexploration [35].


Cardiogenic Shock and Acute Decompensated Heart Failure

Our understanding of “cardiorenal syndrome” is evolving beyond the concept of low cardiac output causing renal dysfunction. We know that approximately half of patients with ADHF have preserved left ventricular function [36]. In the Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheterization Effectiveness (ESCAPE), optimization of hemodynamics did not prevent AKI, further suggesting that reduction in cardiac output does not fully explain the development of impaired renal function [36].

A complex bidirectional relationship emerges, whereby heart failure and associated renal dysfunction affect each other. In addition to the traditional hemodynamically mediated AKI due to low cardiac output, the heart and the kidney are simultaneously affected by activation of the sympathetic nervous and renin–angiotensin–aldosterone systems. Such activation results in systemic vasoconstriction, salt and water retention, and volume overload, further exacerbating kidney dysfunction and heart failure. In addition, the immune system affects renal and cardiac function through monocyte-mediated endothelial activation, cytokine release, and apoptosis induction [16,17,37]. Furthermore, CKD, a risk factor for coronary artery disease [38], contributes to volume overload, diuretic resistance, and poor prognosis in CHF [39] and ADHF [36]. Here, we will focus on acute ADHF with AKI as a common clinical problem in the ICU setting.

ADHF is frequently complicated by AKI. The Acute Decompensated Heart Failure Registry (ADHERE) of more than 30,000 patients with ADHF suggests that AKI has poor prognostic implications and predicts mortality in this patient group [40]. In one study, more than a quarter of patients with ADHF developed AKI as defined by a rise in serum creatinine of 0.3 mg per dL. However, even this relatively small rise was associated with 7.5-fold increase in hospital mortality [41].

Traditional treatment strategies of ADHF include diuresis, afterload reduction, and administration of inotropic agents. However, these patients are frequently diuretic resistant and hypotensive. The use of ACE inhibitors is often limited by AKI and hyperkalemia. Recently, nesiritide, a recombinant B-type natriuretic peptide, was approved for management of symptomatic ADHF [42]. However, a meta-analysis of five randomized clinical trials in more than 1,200 patients with ADHF suggested nesiritide was associated with worsening of renal function [43]. New trials are ongoing to further investigate the risk and benefit of nesiritide. Tolvaptan, a vasopressin antagonist, has been studied in a large international randomized trial of more than 4,000 heart failure patients [44]. Tolvaptan was statistically better then placebo at improving dyspnea, edema, and weight loss but did not significantly improve the rate of death and rehospitalization for heart failure.

Inotropic agents such as dobutamine and milrinone are used in the treatment of ADHF. Dobutamine acts primarily on β1-adrenergic receptors, with minimal effects on β2 and α1 receptors. Dobutamine increases cardiac output and stroke volume and decreases systemic vascular resistance and pulmonary capillary wedge pressure. The 2004 ACC/AHA STEMI guidelines suggest using dobutamine in patients with hypotension who do not have clinical evidence of shock [45]. Milrinone is a phosphodiesterase inhibitor that increases myocardial contractility, reduces systemic vascular resistance, and improves left ventricular diastolic relaxation. However, inotropes increase myocardial oxygen consumption and can worsen myocardial ischemia, and their use has been limited by arrhythmia development and adverse outcomes. In a large prospective, randomized trial, milrinone infusion was associated with increased hypotension and atrial arrhythmias as well as a trend toward increased mortality [46]. The use of inotropes is limited to patients with ADHF and low cardiac output who fail or cannot tolerate diuretic and vasodilator therapy. Additional information on the use of inotropes can be found in Chapter 33 of the “Cardiovascular Problems and Coronary Care” section.

Other strategies for diuretic-resistant patients with ADHF include mechanical fluid removal with ultrafiltration or paracentesis. The use of ultrafiltration in the ICU setting is discussed in “Renal Replacement Therapies” section of the text. Mullens et al. suggested that elevated intra-abdominal pressures in ADHF may play a role in the pathogenesis of renal dysfunction [47]. The reduction of intra-abdominal pressure from approximately 13 to 7 mm Hg by paracentesis was associated with a reduction in serum creatinine from 3.4 to 2.4 mg per dL in diuretic-resistant patients.


Sepsis

Sepsis is among the most common causes of AKI. In one large series of patients with ATN, sepsis was believed to be the cause in 15%, with a mortality rate of 40% [48]. The association between septicemia and AKI is confounded by the experience that renal dysfunction due to other causes is often complicated by infection. Although the incidence of sepsis in patients with AKI has been reported as high as 75% [49], only one third of patients have clinically apparent septicemia at the outset of renal dysfunction [50].

AKI may develop in a setting of sepsis through multiple mechanisms. As discussed previously, inflammation appears to play a significant role. In animal models, sepsis can cause renal impairment even in the absence of hypotension [51]. Clinically, it is likely that endotoxin causes a reduction in GFR through hemodynamic mechanisms, including vascular pooling and renal vasoconstriction, which are mediated by local vasoconstrictors such as thromboxane and endothelin. Although cardiac output is often elevated in patients with sepsis, systemic vasodilation coupled with renal vasoconstriction can shunt perfusion away from the kidneys. Vascular pooling and third spacing generally necessitate volume expansion with isotonic saline. Because myocardial suppression, oliguria, and capillary leakage may accompany sepsis, it is essential to monitor the administration of fluids closely.


Pancreatitis

Pancreatitis may occur in association with various causes of AKI but can itself induce ATN. This is a rare phenomenon and is generally seen in patients with severe or hemorrhagic pancreatitis with serum amylase values of more than 1,000 U per L. Mortality may approach 70% to 80% in this setting, especially in those with multiorgan failure.


Trauma

AKI associated with severe trauma generally reflects the combination of acute volume depletion, hemorrhage, and
myoglobinuria (see following discussion). Survival after trauma is markedly reduced when complicated by AKI.


Acute Bilateral Cortical Necrosis

Acute bilateral cortical necrosis is rare. Unlike ATN, in which only tubular elements are involved, in acute cortical necrosis, glomeruli and tubules are destroyed by a process in which cortical vessels may be occluded with fibrin thrombi. Cortical necrosis usually occurs after profound hypotension. Approximately two thirds of cases are related to obstetric complications, including abruptio placentae, preeclampsia and eclampsia, septic abortion, and amniotic fluid embolism [52]. Nonobstetric cortical necrosis is most common in shock, sepsis, and disseminated intravascular coagulopathy, but isolated cases have been reported with snakebites [53], arsenic ingestion [54], and hyperacute renal allograft rejection [55]. The pathogenesis of AKI in these conditions involves the hemodynamic insults of hypoperfusion and renal vasoconstriction and formation of fibrin thrombi in the renal microvasculature.

Typically, patients with bilateral cortical necrosis have anuric AKI. Although the diagnosis may be suspected early in the course of renal injury, ATN remains far more likely. When renal function fails to recover after several weeks, cortical necrosis may be confirmed by a renal biopsy. Other diagnostic tests are less specific. Renal scintigraphy most often demonstrates complete absence of isotope in the region of the kidneys. Computed tomography (CT) with contrast enhancement may demonstrate similar findings, indicating absence of perfusion to the renal cortex. Renal angiography shows patency of the main renal arteries and either a complete absence of cortical filling or a mottled nephrogram. Given the severity of the inciting disorder, mortality is high in acute cortical necrosis, with fewer than 20% of patients surviving. At least 25% of survivors eventually require maintenance dialysis [50].


Nephrotoxicity and Drug-Induced Acute Kidney Injury

Many cases of AKI in the ICU can be linked to the effects of endogenous and exogenous nephrotoxins.


Myoglobinuria and Hemoglobinuria

Rhabdomyolysis is often associated with leakage of myocyte contents, particularly the pigment protein myoglobin, into the plasma. Myoglobin, with a molecular weight of approximately 17,000 daltons, is freely filtered by the glomerulus. In the distal nephron, myoglobin forms proteinaceous casts that obstruct urine flow. Myoglobin may also exert direct cytotoxic effects on tubular epithelium through the generation of reactive oxygen species.

Myoglobinuric AKI is a consequence of massive skeletal muscle injury of diverse causes. Traumatic rhabdomyolysis occurs in the setting of direct mechanical injury (crush syndrome), burns, or prolonged pressure. Myoglobinuric renal failure is an important cause of morbidity in virtually all wide-scale human catastrophes. Indeed, much of what is known about the syndrome derives from experiences with victims of wars and natural disasters. Crush injuries during the Armenian earthquake of 1988 necessitated emergent mobilization of dialysis resources on a massive scale [56].

Nontraumatic rhabdomyolysis can occur with toxic, metabolic, and inflammatory myopathies, vigorous exercise, severe potassium and phosphate depletion, and hyperthermic states such as the neuroleptic malignant syndrome and malignant hyperthermia. Lipid-lowering drugs currently represent one of the most common causes of rhabdomyolysis. The use of heroin and amphetamines [57,58] has been reported in association with rhabdomyolysis.

As with other forms of AKI, the prognosis depends largely on the gravity of the predisposing condition; AKI following massive trauma can be expected to run a longer course than that associated with nontraumatic causes, such as drugs. In particularly severe cases, oliguria and dialysis dependence may persist for weeks.

Clinical signs and symptoms of muscle injury, such as muscle tenderness, are absent in at least half of cases of significant nontraumatic rhabdomyolysis. The diagnosis is suggested by markedly elevated serum levels of muscle enzymes with serum creatine kinase levels usually higher than 5,000. The serum levels of phosphate and potassium are also typically elevated in rhabdomyolysis because lysis of muscle cells causes release of intracellular contents into the blood. A fall in the serum calcium is quite common. Rebound hypercalcemia often occurs during the recovery phase.

The therapy of myoglobinuria is similar to that of other forms of AKI, but there are several particular considerations. The tubular toxicity of myoglobin is enhanced when urine flow rates are low, urine is concentrated, and urinary acidification is maximal. It is therefore important in the early phases of the illness to ensure that the patient is in a volume-replete state and maintaining a rapid diuresis (i.e., urine output of at least 150 mL per hour). To this end, isotonic fluids may be administered. Most experts recommend the administration of bicarbonate-rich fluids to alkalinize the urine above a pH of 6.5 so as to improve the solubility of myoglobin. Diuresis may be enhanced with concurrent administration of loop diuretics. Some have argued that loop diuretics may introduce the potentially adverse effect of increasing urinary acid excretion and have advocated the use of osmotic diuretic agents such as mannitol [59]. Mannitol, however, has the potential drawback of causing intravascular volume overload in patients whose kidneys may already have impaired urine output.

Hemoglobinuria can also result in AKI. The pathophysiologic mechanisms are similar to those involved in myoglobinuric AKI. Hemoglobinuric renal failure is relatively rare. Hemoglobin, with a molecular weight almost four times that of myoglobin, is less readily filtered. Furthermore, when hemoglobin is released into the plasma, it binds to haptoglobin, forming a bulky, nonfilterable molecular complex. Only when the haptoglobin binding capacity is saturated (at plasma hemoglobin concentrations > 100 mg per dL) does hemoglobin appear in the tubular fluid. Thus, only massive intravascular hemolysis, as may occur with fulminant transfusion reactions, autoimmune hemolytic crises, and mechanical hemolysis [60] from a dysfunctional prosthetic heart valve (Waring blender syndrome), can induce AKI.


Radiocontrast-Induced Nephropathy

The administration of intravascular radiocontrast agents leads to a syndrome of rapidly developing AKI. Contrast-induced nephropathy (CIN) is commonly defined as an absolute increase in serum creatinine of 0.5 mg per dL or a relative increase of 25% from the baseline within 48 to 72 hours of contrast exposure. The serum creatinine level begins to rise 12 to 24 hours and peaks approximately 4 days after the procedure [61]. Some patients develop a transient increase in urine output as a result of contrast-induced osmotic diuresis, followed by oliguria. The majority of patients are nonoliguric and do not require dialysis. Some, but not all, studies identified an increased mortality risk in patients with CIN [62,63,64]. In patients who have undergone endovascular procedures, CIN must be differentiated from atheroembolic disease, which has a significantly worse prognosis (discussed below).









Table 73.8 Risk Factors Associated with Radiocontrast Nephropathy




Preexisting renal insufficiency
Diabetic nephropathy, with renal insufficiency
Volume depletion
Diuretic use
Large contrast dose (> 2 mL/kg)
Age > 60 y
CHF
Hepatic failure
Multiple myeloma (with high osmolar contrast agent)
Use of intra-aortic balloon pump

Prospective studies report that the incidence of radiocontrast-induced kidney injury ranges from 1% to more than 50%. Some of this variance in frequency can be attributed to the disparity in the definitions of AKI, the number of associated risk factors, and the type of procedure performed [65]. The incidence appears to be low in patients with normal renal function, even in the presence of diabetes. Preexisting CKD, however, particularly in patients with diabetes, confers a 6- to 10-fold increased likelihood of radiocontrast-induced AKI [66,67]. Contrast-enhanced CT is associated with a lower risk of CIN as compared with coronary angiography. Noncoronary angiography had the highest incidence of CIN in a study of 660 military veterans, reaching 15% in patients with GFR less than 60 mL per minute per 1.73 m2 [68]. Other risk factors are listed in Table 73.8.

There are several mechanisms by which radiocontrast-induced renal injury may develop. Hemodynamic factors are believed important, as contrast exposure causes initial vasodilation followed by prolonged vasoconstriction of the renal circulation. The finding of a low FENa in some patients with contrast-induced AKI and the tendency toward rapid recovery suggest a role for reversible vasoconstriction. The intensity and duration of the vasoconstriction may be influenced by the underlying characteristics of the renal microcirculation. Endothelial factors that promote vasoconstriction of preglomerular vessels, such as endothelin, may participate in the pathogenesis of radiocontrast-induced nephropathy [11,69]. Tubular adenosine receptors appear to be stimulated by radiocontrast agents in animal models [70,71]. Adenosine induces afferent glomerular vasoconstriction and reduction in GFR. However, the protective effect of theophylline, an antagonist of adenosine receptors, was not statistically significant in the meta-analysis of available trials of CIN prevention [65]. Other postulated mechanisms of radiocontrast-induced nephropathy include the generation of reactive oxygen species and the direct cytotoxic effect of the contrast media, especially with highly osmolar agents [72].

Because there is no specific treatment for radiocontrast-induced nephropathy other than supportive measures, attention has focused on methods of prevention. The best preventive measure is avoidance of radiocontrast and use of an alternative noncontrast imaging procedure if at all possible.

A number of prophylactic measures (listed in Table 73.9) have been promoted. Experimental data and retrospective clinical studies suggest that radiocontrast injury is augmented by preexisting hypovolemia, particularly in the presence of prostaglandin inhibitors [73,74,75,76]. Therefore, modest hydration before the procedure and avoidance of diuretics and NSAIDs are justifiable.

Reports suggest that pharmacoprophylaxis of radiocontrast nephropathy may be possible. In a German study, N-acetylcysteine (NAC) was given to patients with CKD once before and once after they underwent angiography. The incidence of radiocontrast nephropathy in these patients was 2%, compared with 9% in an untreated control group. Both groups received concomitant hydration, and low-osmolality contrast medium was used [77]. Some, but not all, subsequent studies have confirmed these results [78,79]. Maranzi et al. [80] compared two different regimen of NAC in patient undergoing emergent angioplasty for STEMI. Patients receiving a high dose of NAC (1,200 mg IV prior to procedure, followed by 1,200 mg orally for four more doses) had a significantly lower incidence of CIN. The benefit of NAC remains controversial as conflicting results continue to emerge from other studies [65,81]. However, NAC appears to be a low risk intervention and is frequently used. We typically give NAC 1,200 mg orally for two doses before and two doses after the contrast exposure.








Table 73.9 Preventive Measures for Radiocontrast Nephropathy




Volume expansion with normal saline or isotonic bicarbonate (3 mL/kg bolus over 1 h prior to the procedure, followed by 1 mL/kg/h for 6 h postexposure
Limit radiocontrast load to ≤ 1 mL/kg in high-risk patients
Avoid high osmolar contrast agents
Discontinue diuretics, ACE inhibitors, and nonsteroidal anti-inflammatory drugs for 24 h postprocedure
N-acetylcysteine (4 doses of 600–1,200 mg PO every 12 h with 2 doses before and 2 doses after procedure)

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Sep 5, 2016 | Posted by in CRITICAL CARE | Comments Off on Acute Kidney Injury in the Intensive Care Unit

Full access? Get Clinical Tree

Get Clinical Tree app for offline access