A rapid decline in glomerular filtration evidenced by the accumulation of nitrogenous waste products (blood urea nitrogen and creatinine) reflects acute kidney injury (AKI), a common major medical problem that occurs in 7%–18% of hospitalized patients including 50%–60% of those admitted to intensive care units. AKI manifesting in the early postoperative phase is particularly common after some surgeries, complicating up to 20%–70% of cardiac, vascular, major trauma, and hepatobiliary procedures. The importance of postoperative AKI lies in its consistent association with worsened outcome, including higher in-hospital mortality, delayed recovery in survivors, and overall increased cost. Even evidence of minor renal impairment carries with it a related risk of major postoperative complications and mortality.
Clinical practice modifications can eliminate or limit renal insult in cohorts and reduce the burden of postoperative AKI. In contrast, once AKI has occurred, beyond dialysis the portfolio of evidence-based interventions that improve outcome is sparse to nil. This unfortunate observation has gained attention from research agencies such as the United States National Institutes of Health and the Cochrane Database of Systematic Reviews. In this context, the aim of this chapter is to review evidence (or lack thereof) for putative renal preservation strategies, including those directed at the postoperative critically-ill patient.
The kidneys constitute less than 0.5% of body weight but receive 25% of cardiac output, making them the most highly perfused major organs. Among their many active roles are regulation of body fluid composition and volume; the kidneys filter equivalent to a 12-oz (355 mL) can of soda every 3 minutes, returning all but 1% (approximately 4 mL of urine) to the circulation. The ability of the kidneys to control the contents of urine precisely is the basis of whole body homeostasis. As highlighted by the renal physiologist, Dr. Homer Smith: “It is no exaggeration to say that the composition of the blood is determined not by what the mouth ingests but by what the kidneys keep; they are the master chemists of our internal environment, which, so to speak, they synthesize in reverse.” Hence, acute derangements that have systemic effects are thought to underpin some of the risk for adverse outcome related to AKI, a notably different state from compensated chronic kidney disease (CKD) where homeostasis is maintained in the context of limited renal reserve.
Although some elements of perioperative AKI pathophysiology are mechanistically well understood and available to guide the clinician, much is still not known. Nonetheless, identified factors can be divided generally into culprits suspected of adding to a cumulative burden of renal insult and those identifying kidney vulnerability. Pathophysiologic contributors to renal insult are discussed in elsewhere in this textbook, and mechanistic understanding of their role in AKI (and potential avoidance) remains the most important overarching renal preservation principle. Sometimes a sole AKI insult is evident (e.g., renal artery atheroembolism), but AKI is most commonly a multifactorial problem. For specific procedures and even individual patients, renal insults may be obvious, and sometimes their effects are avoidable or can be attenuated through thoughtful clinical care. Vulnerabilities such as comorbid disease (e.g., diabetes) are also extensively discussed in elsewhere in this textbook. However, beyond CKD these premorbid conditions only modestly predict variability in postoperative creatinine rise.
To understand the “renal paradox” (that such a highly perfused organ is so susceptible to injury) requires an appreciation of normal kidney vasculature. The most highly studied pathophysiologic model of AKI is cardiac surgery, which has provided many insights about the kidney response to stress that are generalizable.
Normal blood flow is unevenly distributed within the kidney, with the renal medulla disproportionately under-perfused (5% total). Notably, such heterogeneous blood flow facilitates “mission critical” urine concentrating ability of the kidney, but also places the medulla at particular risk of ischemia. To add to this “low-flow” threat, medullary perfusion is notably inefficient; with entering and exiting vessels arranged in parallel, allowing “escape” of oxygen prior to its arrival at tissues. Known as oxygen countercurrent exchange, this oddity is a byproduct of sluggish perfusion that also allows creation of the medullary urea gradient, an essential ingredient for the urine concentrating process. Finally, adequacy of perfusion is further challenged by high oxygen demands of the outer medulla as a result of active solute transport (thick ascending limb [mTAL] of the loop of Henle). These challenges result in a very low medullary pO 2 even in healthy individuals (10–20 mmHg; Fig. 17.1 ), with medullary oxygen needs met through extraction at the highest levels in the body (79%). Originally reported in the 1960s, this important phenomenon is referred to as medullary hypoxia . Local paracrine systems (e.g., nitric oxide and the renin-angiotensin system) mediate the precarious balance between medullary oxygen demand and delivery to maintain local homeostasis. For clinicians, an additional important consideration is that increased renal blood flow does not provide a luxurious supply of oxygen to the medulla, since equivalent increases in glomerular filtration and tubular active transport reabsorption occur, negating any potential gains.
Direct measurement demonstrates that medullary hypoxia is exacerbated by common perioperative occurrences such as dehydration, hemorrhage, low cardiac output, and cardiopulmonary bypass. In addition to the direct effects of such conditions, related homeostatic fluid retention responses may also add to renal risk. Interestingly, urine pO 2 measurement may be useful as a sensitive monitor of medullary perfusion ; one clinical study found that a drop in urine oxygen levels after cardiac surgery predicted postoperative creatinine rise.
Unique Physiology/Vulnerable Populations
The Immature Kidney
Neonates are born with their adult number of nephrons, which mature and grow throughout infancy to reach full function at around the age of 2 years. Glomerular filtration is reduced because of relatively low systemic blood pressure and high renal vascular resistance and in early infancy tubular transport mechanisms are incompletely developed, reducing the reserve for toxin elimination and increasing the risk of (sodium and water) fluid retention. Finally, as a result of the immaturity of the tubule in premature infants and newborns, measures normally used to differentiate pre- from intrarenal injury may be misleading (i.e., urine osmolality, urine sodium, and fractional excretion of sodium).
The Kidney During Pregnancy
During pregnancy, hormonal effects contribute to a systemically vasodilated state. Renal changes also include dilation of the urinary collecting system and increases in renal plasma flow and glomerular filtration (hyperfiltration), which can be complicated by urine protein wasting. In the setting of CKD, the demands of pregnancy can overwhelm renal reserve and increase the risk of accelerated deterioration of renal function and associated adverse maternal and fetal outcomes. Treatment of hypertension, proteinuria, and nephrotic syndrome is particularly important in the management of parturients.
Geriatrics and Renal Function
Beyond the age of 50 years, kidney parenchyma becomes smaller mainly because of accelerated cortical loss and the crowding effects of benign cysts. On a microscopic level, there are fewer functional nephrons and nephrosclerosis is more prevalent. Some functional compensation is achieved through hypertrophy of remaining nephrons, but averaged over cohorts, glomerular filtration falls accordingly, declining by roughly 0.75 mL/min per year after the age of 30 years. Therefore, even in the absence of renal disease geriatric patients commonly meet CKD criteria, and because of limited reserve are at a higher risk of sustaining AKI to levels requiring dialysis.
Cardiac Surgery-Associated AKI as a Generalizable Model for Perioperative AKI
Cardiac surgery is the most common source of postoperative AKI in the United States , occurring following up to 30% of such procedures. Observations in postcardiac surgery AKI cohorts commonly generalize to other surgical settings. Therefore, postcardiac surgery AKI (often referred to as CS-AKI) is a prevalent perioperative AKI research model and often the source of evidence to support or to refute putative renal preservation strategies.
Notably, although the physiologic renal trespass of major surgery is well represented by cardiac surgery, elements such as aortic cross clamping and cardiopulmonary bypass differentiate cardiovascular from most other major surgeries. Although the immediate ischemic consequences of cross clamping on the kidney require little explanation, the complex effects on renal perfusion of cardiopulmonary bypass (and other circulatory assist devices) are reviewed elsewhere.
Diagnosis of Acute Kidney Stress/Injury
Lack of consensus on a “gold standard” definition for AKI previously hampered progress in its research. One review from the mid-1990s identified 28 perioperative studies, with no two using the same criteria. More recently, consensus AKI criteria have allowed for more uniform reporting and comparison of outcomes. Furthermore, the search for reliable early biomarkers that enable prompt AKI detection earlier than the measurement of serum creatinine accumulation alone is a continuously evolving field.
Current Consensus Diagnostic Criteria for AKI and Related Adverse Outcomes
Consensus criteria have been developed to facilitate AKI diagnosis, four of which are worthy of mention ( Table 17.1 ). Limited to surgical populations, the Society of Thoracic Surgeons postoperative AKI criteria include either new need for dialysis or a serum creatinine rise of at least a 50% from baseline to exceed 2 mg/dL (177 μmol/L). However, three more recently developed (related) definitions are the most commonly used to contrast nonsurgical and surgical studies. These include: (1) RIFLE (Risk, Injury, Failure, Loss, End-stage kidney disease), (2) AKIN (Acute Kidney Injury Network), and (3) KDIGO (Kidney Disease: Improving Global Outcomes) criteria. The most current of these (KDIGO) defines AKI as either (1) a serum creatinine rise ≥ 0.3 mg/dL (≥ 26.5 μmol/L) within a 48-hour period; or (2) a ≥ 50% serum creatinine rise within the prior 7 days; or (3) urine volume < 0.5 mL/kg/h for 6 hours. Each definition uses serum creatinine and urine output criteria to define AKI. Although these consensus criteria have brought stability to the definition of AKI, they are not without their problems, as discussed later.
|Serum Creatinine||Urine Output|
|Risk||Increase in SCr to ≥ 1.5 times baseline or decrease in GFR by > 25% within 7 days||< 0.5 mL/kg/h for > 6 h|
|Injury||Increase in SCr to > 2 times baseline or decrease in GFR by > 50% within 7 days||< 0.5 mL/kg/h for > 12 h|
|Failure||Increase in SCr to > 3 times baseline or decrease in GFR by > 75% within 7 days or increase in SCr to ≥ 4 mg/dL with an acute rise of 0.5 mg/dL||< 0.3 mL/kg/h for > 24 h or anuria for 12 h|
|Loss||Complete loss of kidney function requiring dialysis for > 4 weeks|
|ESRD||Complete loss of kidney function requiring dialysis for > 3 months|
|Stage 1||Increase in SCr by ≥ 0.3 mg/dL or increase in SCr to ≥ 1.5 times baseline within 48 h||< 0.5 mL/kg/h for ≥ 6 h|
|Stage 2||Increase in SCr to > 2 times baseline within 48 h||< 0.5 mL/kg/h for ≥ 12 h|
|Stage 3||Increase in SCr to > 3 times baseline within 48 h or increase in SCr to ≥ 4 mg/dL with a rise of 0.5 mg/dL within 24 h or initiation of dialysis||< 0.3 mL/kg/h for ≥ 24 h or anuria for ≥ 12 h|
|Stage 1||Increase in SCr by ≥ 0.3 mg/dL within 48 h or increase in SCr to ≥ 1.5 times baseline within 7 days||< 0.5 mL/kg/h for ≥ 6 h|
|Stage 2||Increase in SCr to > 2 times baseline within 7 days||< 0.5 mL/kg/h for ≥ 12 h|
|Stage 3||Increase in SCr to > 3 times baseline within 7 days or increase in SCr to ≥ 4 mg/dL or initiation of dialysis||< 0.3 mL/kg/h for ≥ 24 h or anuria for ≥ 12 h|
Notably, all of the above-mentioned consensus definitions use changes in serum creatinine to reflect changes in glomerular filtration rate (GFR) and overall kidney function. One inconvenience of this approach is the nonlinear relationship between changes in GFR and serum creatinine. Additionally, serum creatinine does not generally reliably reflect CKD until GFR rates fall below 50 mL/min, hence serum creatinine may be “normal” in patients even with important underlying chronic renal impairment. Following surgery, creatinine accumulation requires up to 48 hours to confirm AKI and a rise that reflects significant kidney injury may occur while levels remain within the “normal” range (e.g., 50% increase). These factors represent sources of significant delay in the implementation of kidney protective strategies that confound study of their effectiveness. Furthermore, small changes in serum creatinine that never meet the threshold for AKI criteria are still associated with increased mortality and despite their subthreshold degree probably also reflect important AKI episodes.
Controversial (and often also difficult to measure perioperatively) is the significance of urine output as a diagnostic biomarker of postoperative AKI. This contrasts with ICU patients where urine output measurements add explanatory power to AKI predictive outcome models: inclusion of oliguria increases AKI incidence and isolated oliguria (i.e., exclusive of serum creatinine rise criteria) is associated with higher ICU mortality and 1-year mortality or need for renal replacement therapy (RRT). Furthermore, critically ill patients meeting both oliguria and creatinine criteria are at greatest risk. However, the value of immediate perioperative oliguria to define AKI in surgical populations is less clear. Alpert and colleagues documented oliguria in 137 aortic surgery patients given either crystalloid solution, mannitol, furosemide, or no intervention when urinary flow dropped below 0.125mL/kg/h. These authors from the 1980s found no correlation between intraoperative mean urinary output or lowest hourly urinary output and subsequent AKI development by change from pre to peak postoperative levels of serum creatinine. The findings of this study are consistent with data from several other more recent studies that suggest that oliguria in the immediate perioperative period is not as useful as a marker of AKI. For example, in a retrospective study of cardiac surgery patients, the addition of urine output to serum creatinine criteria (AKIN) considerably increased AKI incidence but added no prognostic value for either in-hospital mortality or new requirement for RRT. In contrast, another prospective study of cardiac surgical patients found meeting either creatinine-only or oliguria-only (KDIGO) criteria was related to equally elevated long-term mortality risk when compared with patients without AKI. Furthermore, patients meeting both creatinine and oliguria criteria in this study had additional long-term mortality risk above that of the patients meeting only one such criterion.
Notably, perioperative oliguria is typical during uneventful surgical procedures and simply represents an appropriate homeostatic response to perioperative fasting and the normal response to anesthesia (i.e., acute renal “success”) as opposed to a pathologic response reflecting renal injury. In the absence of consensus, it is recommended that urine output criteria to diagnose AKI should be used cautiously in the immediate postsurgical setting, in contrast to nonsurgical and/or chronic critically ill patient populations. In addition, irrespective of surgical status, it appears that patients meeting both urine output and serum creatinine criteria are at highest risk of poor outcomes.
Long-term kidney outcomes are important and include persistent CKD, temporary or chronic need for dialysis, and death. Although consensus AKI criteria have improved the ability of clinicians to report and to compare acute renal events and their short-term consequences across populations, these do not reflect important long-term kidney outcomes. As a long-term measure of renal outcome, Major Adverse Kidney Events (MAKE) reflect chronic outcomes that may demonstrate improvements with effective renoprotective interventions (e.g., benefits seen in successful phase III renoprotection clinical trials). The MAKE metric is a composite of persistent renal dysfunction (25% or greater decline in eGFR), new hemodialysis, and death to identify poor outcomes following AKI. It is typically reported at 30 (MAKE30), 60 (MAKE60), or 90 (MAKE90) days after AKI diagnosis. . MAKE90 is particularly relevant because this MAKE version aligns with the 90-day criteria typically used in other CKD diagnostic criteria. The MAKE composite is important for clinicians to link AKI with chronic morbidities and for the assessment of meaningful outcomes in AKI intervention trials.
Future (Early) AKI Biomarkers
Several putative biomarkers are postulated to have better sensitivity and specificity than serum creatinine for detecting early kidney stress and AKI; however, these have yet to be widely adopted in the clinical setting. Although still requiring validation, these biomarkers have the advantage of both detecting AKI at earlier time points and discriminating types and degrees of injury as outlined later. AKI biomarkers may be useful to reflect progress along a continuum towards renal injury. In this continuum, an acute preinjury kidney stress phase typically precedes renal structural damage which proceeds overt functional loss. Thus, AKI biomarkers can be divided into those that detect: (1) kidney stress, (2) impaired function without structural damage, (3) structural damage with intact function, and (4) structural damage with loss of function. The framework for how these biomarkers could be useful to detect the phase of renal injury in hopes of guiding more specific interventions (e.g., preventing overt AKI, limiting kidney damage, etc.) is shown in Fig. 17.2 . Furthermore, studies have indicated that such early AKI biomarkers could be useful for long-term risk stratification following AKI. Nonetheless, to date serum creatinine remains a reliable and readily available gold standard test to diagnose AKI. In this context, some newer biomarkers (e.g., TIMP-2 * IGFBP7) may have potential to predict AKI and allow clinicians to intervene earlier. Further validation is needed for most of these potential clinical tools, with the hope that in the future potential biomarker profiles may enable more precise risk stratification and earlier intervention of patients with AKI.
Measures of Kidney Stress
Insulin-like growth factor-binding protein 7 (IGFBP7) and tissue inhibitor of metalloproteinases-2 (TIMP-2) are kidney stress biomarkers. Both substances arrest the G 1 phase of the cell cycle and levels of each can be measured in the urine. These kidney stress biomarkers were identified as the top performing among a collection of related biomarkers for detecting AKI in a variety of high-risk critically ill patients including those with sepsis, shock, major surgery, and trauma. Furthermore, a urinary TIMP-2 * IGFBP7 measurement greater than 0.3 is a strong predictor of AKI in critically ill patients. Several prospective analyses and a recent meta-analysis of cardiac surgery studies also support TIMP2 * IGFBP7 levels measured several hours after cardiac surgery as predictive of subsequent postoperative AKI. Finally, relative to long-term outcomes, TIMP-2 * IGFBP7 levels > 2.0 at ICU admission are associated with death or dialysis at 9 months (HR, 2.16; CI 1.32–3.53).
Measures of Kidney Function
Beyond serum creatinine, steady-state cystatin C level is an alternative biomarker that reflects changes in kidney filtration function. Notably, the use of cystatin C does not appear to add significantly more information regarding AKI compared with serum creatinine. Cystatin C is a cysteine protease inhibitor protein produced by all nucleated cells, and is freely filtered by the glomerulus and completely reabsorbed in renal tubules. In a prospective observational study of septic ICU patients, urinary cystatin C levels were independently associated with sepsis, AKI, and mortality within 30 days. However, in a large meta-analysis, postoperative urinary Cystatin C poorly discriminated for cardiac surgery (CS) of AKI of any stage (composite AUC 0.63 [0.37–0.89]; diagnosis by AKIN), but plasma cystatin C was modestly more useful (composite AUC 0.69 [0.63–0.74]; diagnosis by RIFLE, AKIN, KDIGO).
Measures of Kidney Tubular Damage
The three most studied major biomarkers that indicate renal damage are: (1) neutrophil gelatinase-associated lipocalin (NGAL), (2) kidney injury molecule-1 (KIM-1), and (3) interleukin-18 (IL-18).
NGAL is a small protein with notable upregulation following acute kidney tubular injury, that can be measured either in the urine or plasma. However, in numerous studies both urine and plasma NGAL have modest predictive value for subsequent traditionally diagnosed AKI in critically ill patients. A meta-analysis by Ho et al. also demonstrated modest discrimination for AKI after cardiac surgery for urine NGAL (composite AUC 0.72 [0.66–0.79]) and plasma NGAL (composite AUC 0.71 [0.64–0.77]). Both urine and plasma NGAL have shown only a modest ability to predict AKI severity, renal recovery, and long-term outcomes.
Expression and release of KIM-1 is induced in the proximal tubule after renal injury. A meta-analysis revealed a sensitivity of 74% and specificity of 86% for urinary KIM-1 to diagnose AKI. Blood KIM-1 levels are elevated in patients with AKI and CKD of various etiologies. Urine levels of KIM-1 perform similarly to NGAL in discrimination for CS-AKI (composite AUC 0.72 [0.59–0.84]). Measurement of KIM-1 has not been widely adopted in practice.
IL-18 is a proinflammatory cytokine involved in the evolution of tubular ischemia . As a biomarker for detecting CS-AKI, urine levels of IL-18 perform similarly to both KIM-1 and NGAL. In ICU patients, urinary IL-18 levels may be useful for the early diagnosis of AKI and may predict mortality risk in patients with acute respiratory distress syndrome (ARDS). In a broader ICU population, urinary IL-18 was not reliable in predicting subsequent AKI but was associated with poor clinical outcomes such as death and dialysis. Urinary IL-18 has not been widely adopted in clinical practice.
Mechanisms of Surgery-Related Acute Renal Injury
The sources of postoperative renal insult are extensively reviewed elsewhere. Two primary mechanisms that contribute to perioperative AKI are ischemia-reperfusion and nephrotoxic effects, with three notable sources of insult common to many surgical procedures where postoperative renal dysfunction is prevalent: hypoperfusion, inflammation, and atheroembolism. Several less common sources of renal insult may contribute in selected patients (e.g., rhabdomyolysis, specific drug-related, etc.).
Abnormalities of renal perfusion that exceed the autoregulatory reserve of the renal circulation, such as cardiopulmonary bypass, are believed to be a major source of perioperative ischemia-reperfusion injury. Other conditions, including low output states, hypovolemic shock ( Fig. 17.1 ), vasoconstrictor use, and circulatory arrest, may all contribute to the renal ischemic burden of specific surgical procedures. As described earlier, the normally low medullary pO 2 (10–20 mmHg) makes this tissue particularly vulnerable to hypoperfusion. Paracrine systems, such as the renin-angiotensin system and nitric oxide synthase, are key to regulation of renal blood flow and modulation of microvascular function and oxygen delivery in the renal medulla. Autonomic influences are also important, with the alpha-1 adrenergic receptor-mediated vasoconstriction and alpha-2 adrenergic receptor-mediated vasodilation contributing to modulation of renal perfusion. Lastly, it is increasingly recognized that venous congestion caused by elevated central venous pressure may contribute to postsurgical AKI pathophysiology.
Both systemically and locally in the kidney, surgery provides a consistent trigger for the generation of proinflammatory cytokines (e.g., tumor necrosis factor alpha [TNFα] and interleukin 6 [IL-6]) and an inflammatory response caused by the disruption of tissues and the potential for perioperative endotoxemia. Preexisting infection may further predispose surgical patients to postoperative renal dysfunction. Other postoperative factors, such as infectious and septic complications, transfusion, and interventions such as cardiopulmonary bypass, contribute to the generation of cytokines that have major effects on the renal microcirculation and may lead to tubular dysfunction through a common path of apoptosis.
Showers of emboli are common during certain surgical procedures. Although some types of emboli do not appear to have an important role in postoperative AKI (e.g., air) others are significant contributors to renal injury. Atheroembolism is common during some surgical procedures, particularly those involving operative aortic manipulation, and is highly associated with AKI. The strong association of intra-aortic balloon counterpulsation with AKI may also be related to the dislodgement of aortic plaque. Thromboembolism and infectious emboli from endocarditis may also contribute to renal insult ( Figs. 17.3 and 17.4 ). Atheroembolism can produce all degrees of renal injury, ranging from the obstruction of major renal vessels from large fragments of plaque to the occlusion of multiple small renal vessels by cholesterol microcrystals.
Pigment (Hemoglobin and Myoglobin)
Hemoglobinuria occurs when intravascular hemolysis releases sufficient hemoglobin to exceed the adsorptive capacity of circulating haptoglobin and renal tubular reuptake mechanisms. Only approximately 25% of circulating free hemoglobin is in a form readily filtered by the kidney. Myoglobinuria occurs with muscle cell necrosis in conditions such as ischemic limb reperfusion and major trauma (crush syndrome). The mechanisms underlying myoglobin- and hemoglobin-related AKI are similar. Deterioration in kidney function comes from the combined effects of renal vasoconstriction, tubular obstruction by casts, and direct cytotoxicity. The nitric oxide scavenging potential of these pigments contributes to renal vasoconstriction.
Several well-known nephrotoxins are relevant to the perioperative period. Aminoglycoside antibiotics have potent nephrotoxic effects. Non-steroidal anti-inflammatory drugs (NSAIDs), such as aspirin and indomethacin, are nephrotoxic, and cause renal vasoconstriction through inhibition of endogenous locally synthesized vasodilator prostaglandins. Perioperative treatment with cyclosporin as part of transplant surgery can have acute effects on renal function in addition to the well-known long-term adverse effects.
Although more controversial, intravenous contrast has been associated with AKI. With the use of less toxic contrast agents and lower dosing strategies, evidence suggests that use of intravenous contrast for diagnostic imaging studies (i.e., computed tomography [CT] scans) is not associated with AKI. However, it is still unclear whether higher doses of intravenous contrast agents (i.e., cardiac catheterization, interventional radiology procedures) cause contrast-induced nephropathy. The judicious use of intravenous contrast agents in these settings, particularly in those with preexisting renal dysfunction, is still warranted.
Renal Preservation: Best Practices
There is good evidence to support the role of perioperative clinical management decisions in affecting important renal outcomes. A vast majority of these best practices involve kidney insult reduction or avoidance at all stages throughout the perioperative period. Notably, there is little evidence to support significant value in therapeutic interventions to accelerate recovery once AKI has occurred (short of temporary RRT); rather management of patients with established kidney injury involves mostly minimizing further AKI to facilitate normal recovery. Potential interventions and best practices are outlined later and discussed by time period (preoperative, intraoperative and postoperative), and topics are also organized for practical purposes into three tables by their value ( Table 17.2 : recommended practices to avoid; Table 17.3 : limited value probably no harm; and Table 17.4 : recommended practices).
|Class III: Risk > Benefit|
|Level A (consistent evidence)|
|Level B (some evidence)|
|Level C (limited evidence)|
|Level A (consistent evidence)|
|Level B (some evidence)|
|Level C (limited evidence)|
|Class I: Benefit >> Risk|
|Level A (consistent evidence)|
|Level B (some evidence)|
|Level C (limited evidence)|
Individualized for specific patient level concerns, the context to apply this outlined “optimal renal practices” approach to clinical management can be gained from consideration of renal risk stratification, as outlined elsewhere in this textbook , , , However, AKI prediction beyond knowledge of the risk associated with specific procedures, and the greater likelihood of need for dialysis with chronic kidney disease is notoriously poor, particularly at pre-identifying the victim of severe AKI. This inability to pre-identify at-risk patients with confidence further highlights the importance of developing a safe “renal best practices” approach to reduce the overall burden of AKI as part of routine patient management.
Preoperative Management and Planning
Renal preservation strategies with the goal of avoiding, preventing, or attenuating renal insult before it occurs include the use of preoperative risk stratification to influence procedure selection, procedure modification, and management of chronic medication administration. Procedure modifications with lowered AKI risk have already been part of the justification for transitions in routine care such as stent grafting (versus open surgical procedures) to treat aortic aneurysmal disease. Potential future tools also include preoperative genetic screening to identify subgroups at high renal risk. The following section focuses on procedural choice, nonpharmacologic interventions, and pharmacologic agents.
Recent surgical innovations have embraced the philosophy that fewer invasive procedures requiring smaller incisions and less physiologic disturbance can safely achieve surgical goals and may reduce postoperative organ dysfunction and improve overall outcome. Examples include the use of stent-graft technology, videoscopic-assisted procedures, avoidance or modification of cardiopulmonary bypass, and reduced aortic manipulation. Although some innovations appear to confer renal benefit, this is not the case for all of these minimally invasive procedure modifications.
Cardiac Surgery Modifications
Modifications of coronary artery bypass graft (CABG) surgery have been studied extensively. Although the use of extracorporeal circulation during cardiac surgery has been speculated as a major pathophysiologic driver of postoperative AKI, the value of off-pump procedures in studies has been inconsistent in mitigating renal risk. One reason for this could be that comparisons of off-pump and on-pump approaches may be limited by baseline imbalance between groups. In particular, patients receiving off-pump procedures are often healthier at baseline, with reduced preoperative renal risk. Thus, stratification of patients by preoperative renal risk in such studies may improve interpretability. Nonetheless, a meta-analysis of several randomized trials comparing on-pump versus off-pump CABG procedures did not demonstrate a protective effect with regard to risk of postoperative RRT. However, some more recent studies report a reduction in AKI rate with off-pump procedures. In particular, studies involving the subgroup of patients with preoperative CKD stage 4, but not those with CKD stage 5 (i.e., end-stage renal disease), may benefit from off-pump procedures with respect to both mortality and need for postoperative RRT.
Reduction of Tissue Trauma
Similar to the potential value of avoiding extracorporeal circulation, avoidance of tissue trauma from large incisions has been proposed as a strategy to reduce postoperative renal dysfunction. Hence, minimally invasive surgical approaches that use smaller incisions and alternative catheter-based circulatory support strategies that avoid major tissue disturbance (such as median sternotomy) have been studied as methods to reduce inflammatory renal insult and AKI.
Smaller incisions: mini-thoracotomy/mini-sternotomy
Both mitral and aortic valve procedures (i.e., valve repair or replacement) can be performed using peripheral catheter-based cardiopulmonary bypass (CPB) and a mini-thoracotomy or mini-sternotomy approach in place of the traditional median sternotomy. With respect to AKI, although initial data suggested a benefit with mini-sternotomy in these procedures, some subsequent retrospective studies have not confirmed these findings. Similar to off-pump CABG procedures, the use of minimally invasive valve procedures may preferentially provide benefit to patients with CKD prior to surgery.
For major aortic procedures, endovascular stent graft rather than surgical approaches to repair represent another minimally-invasive strategy to reduce tissue trauma. Potential advantages for the kidney of endovascular stenting of abdominal aortic aneurysm (EVAR) have been well studied and yield conflicting results. A number of well-designed retrospective studies support a reduced incidence of postoperative AKI with EVAR compared with an open surgical approach. However, this does not appear to translate into any benefit with respect to requirement for postoperative RRT. As with retrospective studies of off-pump cardiac surgical procedures, identifying renal benefit with EVAR is complicated by baseline patient characteristics. Thus, while the use of endovascular approaches for aortic aneurysm repair may provide a benefit with respect to AKI, selection of this approach in patients with high-risk aortic lesions should not be based solely on reducing renal risk.
Fewer studies have compared open surgical repair and thoracic endovascular aortic aneurysm repair (TEVAR) relative to postoperative AKI. In the largest available retrospective study to date, there was no difference in the renal complication rate between TEVAR and an open surgical approach. However, several smaller studies have reported reduced rates of postoperative renal injury with either endovascular repair or a hybrid approach (aortic debranching and endovascular repair). In patients with aortic arch aneurysmal disease, several studies have failed to show a difference in AKI incidence or requirement for new RRT among the various options.
For patients with aortic valve stenosis, transcatheter aortic valve replacement (TAVR) is increasingly used. While early studies were limited to high surgical-risk patients, TAVR is now an alternative to surgical aortic valve replacement (SAVR) for intermediate surgical-risk patients. In a large trial of randomized intermediate-risk patients, investigators noted significantly lower AKI rates with TAVR compared with SAVR procedures. Other reports suggest that differential renal risk among patients undergoing TAVR are related to both the TAVR procedural characteristics and preoperative patient characteristics. In a meta-analysis among TAVR approaches, transapical TAVR has been associated with increased risk for postoperative AKI compared with the more common transfemoral approach. This result may be, at least partially, an artifact of selection bias: transapical TAVR is often used in patients with worse peripheral vascular disease and unfavorable femoral artery anatomy, a factor known to be associated with elevated renal risk. Notably, postoperative mortality following SAVR, but not TAVR, are directly related to baseline renal function.
In summary, when considering renal risk, the current literature supports TAVR as the favorable approach for aortic valve replacement for aortic stenosis, particularly when a transfemoral approach is used. With regard to aortocoronary bypass procedures, off-pump procedures may provide benefit for patients with CKD.
Relative to procedure-planning, combined neuraxial and general anesthetics have been postulated to reduce the risk of postoperative AKI. Beyond improved analgesia, a supporting rationale comes from the potential for epidural analgesia to block renal autonomic alpha adrenergic innervation and the systemic adrenergic stress-response to surgery. In a randomized study of cardiac surgery patients, thoracic epidural analgesia (TEA), in combination with general anesthesia (GA), reduced AKI risk (unadjusted odds ratio [OR] for GA versus GA + TEA 3.69, 95% confidence interval [CI] 1.34–10.2, P = 0.016). Other retrospective studies in cohorts of thoracotomy patients report similar kidney protective benefits with epidural analgesia.
When considering the potential value of epidural anesthesia, renal benefits must be evaluated relative to other concerns (e.g., risk of spinal hematoma, intraoperative hypotension). Available studies support the potential renal benefit of epidural analgesia in cardiac and thoracic procedures, but further investigations are needed to determine whether epidural analgesia is helpful in other settings.
Procedure Delay Versus Avoidance of Contrast Agents
Historically, iodinated contrast exposure has been a major source of in-hospital kidney injury; however, with the advent of less toxic contrast agents and lower dosing strategies, this concept has more recently come under scrutiny and is currently debated for its significance as an AKI risk factor. In emergency room patients undergoing diagnostic computed tomography (CT) scans, a meta-analysis and a large retrospective propensity-matched cohort analysis found no association with contrast administration and AKI incidence ( Fig. 17.5 ). However, these findings may not be transferable to patients receiving high contrast loads for angiography procedures such as coronary artery catheterization or endovascular procedures.
A meta-analysis of retrospective cardiac surgery studies relating renal risk to pre contrast dye exposure including more than 11,500 procedures with cardiopulmonary bypass found lower AKI rates when the time between surgery and coronary angiography was more than 1 day, but little added “benefit” when surgery was more than 3 days after the contrast exposure. A more recent retrospective study suggests there is increased risk when surgery is separated from coronary angiography by less than 7 days.
In summary, administration of limited doses of newer contrast agents for diagnostic CT scans appears to be relatively AKI risk-free; however, evidence coming mostly from retrospective study of cohorts of cardiac surgery patients suggests caution should be exercised in giving large contrast loads during angiography and ideally such exposures should be separated from surgery by at least 24 hours. Of note, if contrast exposure does occur, the literature suggests that sodium bicarbonate and oral acetylcysteine do not appear effective at preventing contrast nephropathy.
Ischemic Preconditioning Reflex
A brief exposure to ischemia attenuates injury resulting from subsequent ischemic episodes in many organs; this reflex phenomenon, known as ischemic preconditioning, is present in the kidney. In preclinical studies in rats, Cochrane and colleagues observed that 2 minutes of renal ischemia, or three such episodes separated by 3 minutes of reperfusion, protects the kidney from subsequent prolonged (45 minutes) renal ischemia. An example of remote ischemic preconditioning (RIPC) involves occluding the blood supply to the arm for short periods of time in an effort to protect the kidneys. The mechanisms of RIPC for kidney protection are well understood and have been extensively discussed elsewhere. In humans, RIPC for the kidney has been most studied in the setting of cardiac surgery, as summarized in a meta-analysis of 12 cardiac surgery trials, which reported no difference in AKI risk between patients randomized to preincision RIPC or sham. Notably, in a subgroup analysis of three trials in which the anesthetic regimen did not include propofol, RIPC was associated with significantly reduced AKI risk, suggesting that propofol may attenuate the potential renal benefit of RIPC. Furthermore, in high-risk patients, RIPC significantly reduced 3-month risk for major adverse kidney events (MAKE: a composite outcome consisting of persistent renal dysfunction, new RRT, and mortality). Thus, although RIPC does not appear to improve early (less than 90 days) mortality and renal outcomes, this practice appears safe and is worthy of more study for high-risk patients.
Preoperative management decisions regarding chronic therapies may contribute to perioperative renal risk. However, the retrospective nature of most available studies means that there is limited high-quality evidence to guide decisions regarding which drugs to continue or to withhold prior to surgery. Evidence comes mostly from retrospective assessments of cardiac surgery populations and the majority of these reports focus on the management of a limited number of agents, including renin-angiotensin system (RAS) antagonists, diuretics, statins, and beta blockers.
Evidence to guide perioperative management of chronic RAS blocker therapy is inconclusive and even confusing regarding whether to continue or to withhold. Angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin I receptor blockers are prescribed for various reasons that relate to kidney disease, including to delay progression of chronic kidney disease; however, these agents notoriously precipitate acute renal deterioration in situations where angiotensin is critical to the regulation of renal filtration (e.g., renal artery stenosis, volume depletion). In a large meta-analysis of retrospective cardiac surgery studies; the authors found an increased risk for both CS-AKI (OR 1.17, 95% CI 1.01–1.36, P = 0.04) and mortality (OR 1.20, 95% CI 1.06–1.35, P = 0.005) for patients who continued their chronic RAS-blocker therapy up to the day of surgery. In contrast, a second meta-analysis involving both cardiac and general surgical patients found no association. However, in the second study, an analysis limited to one randomized control trial (RCT) and several propensity-matched retrospective studies found RAS-blockers had a net-protective effect (relative risk [RR] of 0.92, 95% CI 0.85–0.99). Although no firm conclusions are warranted, the authors from the first meta-analysis suggest it may be wise to withhold pre RAS blockers, at least in cardiac surgical patients.
Regarding preoperative diuretics, retrospective studies in cardiac surgery patients have shown an increased risk of kidney insult for patients taking preoperative diuretics. In noncardiac surgery patients, preoperative diuretic use was independently associated with postoperative AKI; loop diuretics had the strongest association with AKI. Thus, although the evidence is very limited, it would be prudent to withhold these diuretics prior to surgery, especially loop diuretics.
The preoperative use of statins and beta-blockers has been extensively studied, again mainly in the cardiac surgery population. A meta-analysis of nine RCTs examining the use of preoperative statins found no net benefit with respect to either CS-AKI or RRT. A large retrospective analysis also found no association between preoperative beta-blocker therapy and CS-AKI. Thus, these agents probably have limited effects on perioperative kidney function or injury.
Intraoperative Fluid Choice
Fluid choice has been much better studied in the context of critical illness and will be discussed at length in the postoperative section. However, there have been a number of studies to guide fluid choice specific to the intraoperative environment. A meta-analysis of randomized trials in surgical patients showed that compared with normal saline, balanced crystalloids were associated with fewer minor metabolic derangements but were not associated with a lower need for RRT. Colloid solutions include hydroxyethyl starch (HES) solutions and albumin preparations. A trial using a closed-loop goal directed fluid therapy system randomized patients to receive boluses of either intraoperative colloid (HES) or balanced crystalloid in addition to a background crystalloid infusion for maintenance fluid therapy. Although these investigators reported fewer major complications in patients receiving intraoperative HES boluses, there was no difference in kidney injury. In clinical trials, colloids have never shown any benefit over crystalloids for kidney protection. Thus, when considering kidney protection, for the majority of patients a balanced crystalloid should be used as the primary intraoperative fluid.
The use of sodium bicarbonate has been studied for kidney protection in the setting of cardiac surgery since a small randomized trial reported promising findings. However, a larger trial by the same study group, a second randomized trial, and a retrospective study all failed to demonstrate sodium bicarbonate-mediated kidney protection. In the context of cardiac surgery, sodium bicarbonate should not be used for the purpose of kidney protection.
Intraoperative Volume Management
Goal-directed fluid therapy (GDFT) as a part of an enhanced recovery after surgery (ERAS) protocol has become the standard of care at many institutions. Typically, GDFT involves an algorithm where fluid administration is targeted to achieve a continuously-measured hemodynamic variable such as cardiac output, stroke volume, pulse pressure variation, or stroke volume variation. GDFT is meant to optimize oxygen delivery to organs such as the kidney, limiting injury. Unfortunately, although it has been demonstrated to improve overall recovery and to reduce complications after major surgery, a beneficial renoprotective role has not been validated. In a prospective, randomized control trial of 180 patients undergoing major abdominal surgery, GDFT was not superior to standard care for reducing AKI. The large, randomized OPTIMISE trial also did not demonstrate a difference in postoperative AKI, although it was not intended for this purpose. It can be debated whether the widespread adoption of ERAS protocols has raised the standard of care to where it will be difficult for GDFT to show benefit. Nevertheless, one proposed beneficial component of GDFT is to restrict fluid administration and avoid fluid overload, which is also associated with kidney injury. However, overly restrictive fluid regimens should also be avoided, as the RELIEF trial showed higher rates of RRT in restrictive fluid regimens compared with traditional liberal fluid regimens (median fluid balance at 24 hours for restrictive versus liberal: + 1.38L versus + 3.09L; RR 3.27, 95% CI 1.01–13.8, P = 0.048). Further studies are needed to determine whether this is because of restrictive fluid management, as would be supported by the RELIEF trial, or caused by other factors, such as increased NSAID use as a component of multimodal analgesia. Nevertheless, it seems prudent to avoid gross fluid overload, but also not to be overly restrictive in fluid administration, particularly in the context of renal protection and in patients at high risk of renal injury. The old adage “everything in moderation” appears to hold true for perioperative fluid administration.
Blood Pressure Management
Intraoperative blood pressure management has received considerable attention relative to perioperative outcomes and AKI in both cardiac and noncardiac surgical patients. A number of large retrospective analyses have shown no association between hypotension (< 50–55 mmHg) during CPB and development of CS-AKI. This is additionally supported by results from a randomized trial comparing two CPB blood pressure strategies (target of 75–85 versus 50–60 mmHg), which found no difference in rate of CS-AKI. Notably, CPB blood flow is typically maintained regardless of blood pressure, unlike during noncardiac surgery where blood pressure fluctuations often reflect parallel changes in flow. Thus, in retrospective cohort studies of noncardiac surgery procedures (where blood pressure may be a useful surrogate for perfusion), both degree and duration of intraoperative hypotension are independent predictors of AKI. In summary, maintenance of intraoperative blood pressure within 10%–20% of baseline may be prudent during procedures when CPB cannot be relied upon to maintain perfusion during episodes of low blood pressure.
Cardiopulmonary Bypass Management
Although CPB represents only a small portion of the total perioperative period, research suggests that this period is associated with significant renal risk. Animal studies indicate that oxygen supply demand inequalities are exaggerated and medullary hypoxia is extreme during CPB; effects that last well beyond separation from circulatory support. In humans, changes known to occur at the initiation of CPB include: greater reduction in renal than systemic perfusion, loss of renal blood flow autoregulation, and stress hormone and inflammatory responses known to be harmful to the kidney. These effects may explain why the duration of CPB independently predicts CS-AKI.
Modifiers of Renal Oxygen Delivery
Organ perfusion and oxygen delivery are primary goals of CPB; however, the relationship between standard CPB management guidelines and renal complications is only partly understood. Renal blood flow during CPB is not autoregulated and varies with pump flow rates and blood pressure. Fischer and colleagues retrospectively compared CPB flow rates in a case control analysis of three groups of patients with normal baseline renal function that postoperatively either required dialysis ( n = 44), sustained a renal injury without requiring dialysis ( n = 51), or had no renal impairment ( n = 48). These authors noted that, on average, greater renal injury was associated with longer bypass durations and lower flow rates. A serious limitation of this study is the potential for confounding surgical or CPB variables and unknown or undocumented renal risk factors.
Pulsatile blood flow has been postulated to mediate kidney protection by improving renal microcirculation and lowering vascular resistance. Unfortunately, studies examining pulsatile versus nonpulsatile blood flow on CBP have been inconsistent on their effect on postoperative kidney function: a propensity-matched study did not show benefit from pulsatile flow, but a small, randomized trial noted improvements in perioperative GFR. Thus, pulsatile flow during CPB is not considered an effective renoprotective strategy.
Renal perfusion is affected by renal artery stenosis. In a retrospective study of 798 aortocoronary bypass patients whose cardiac catheterization procedures routinely included renal angiogram, Conlon and colleagues found that 18.7% of patients had at least 50% stenosis of one renal artery (nine patients had > 95% renal artery stenosis bilaterally). However, the presence or severity of renal artery stenosis was not associated with postoperative AKI. However, another propensity-score adjusted retrospective study found that the occurrence of postoperative AKI was higher in patients with renal artery stenosis compared to those without (OR 2.858, 95% CI 1.26-6.48, P = 0.011). Further evidence is needed to determine whether renal angiograms should be a universal preoperative study.
The relationship of anemia and hemodilution with CS-AKI is an area of high clinical interest and highly-investigated, although mostly through retrospective studies. Hemodilution occurs during initiation of CPB because crystalloid or colloid solutions are typically the mainstay of prime for the extracorporeal circuit. Thus, CPB initiation is associated with an acute drop in oxygen-carrying capacity. Animal studies indicate that moderate hemodilution reduces the risk of kidney injury during CPB through improved regional blood flow and reduction of blood viscosity. However, in the clinical setting, where extreme levels of CPB anemia are sometimes tolerated (e.g., hematocrit < 20%), hemodilution has been linked to adverse outcomes. Several large retrospective studies have linked low hematocrit levels during CPB with CS-AKI. However, to complicate the matter, a first line alternative to tolerating extreme CPB anemia, transfusion, also confers AKI risk.
A large retrospective study by Kertai et al. found anemia at times other than CPB to be independently relevant to renal risk, with preoperative anemia (hazard ratio of 1.23) and a combination of preoperative and postoperative anemia (hazard ratio 1.24) associated with CS-AKI. As previously mentioned, red cell transfusion but also the need for surgical reexploration for bleeding are also independently associated with risk of CS-AKI. To date no trials have been examined for any benefit from addressing preoperative anemia (e.g., through iron supplementation).
Several studies have examined the relationship between degrees of extreme anemia and likelihood of CS-AKI. An “inflection point” where worsening AKI risk escalates is present at a threshold lowest CPB hematocrit below 21% ( Fig. 17.6 ). Notably, although extreme anemia is linked to CS-AKI, the occurrence of anemia with low blood pressure during CPB is not associated with elevated risk beyond that of anemia alone. Although there is consensus that avoiding anemia, transfusion and the need for reexploration is ideal (i.e., meticulous hemostasis, limiting hemodilution), evidence regarding transfusion thresholds or “trigger” is just emerging. The TRICS III cardiac surgery trial compared transfusion thresholds, demonstrating that a threshold of 7.5 g/dL was noninferior to a 9.5 g/dL transfusion threshold in all endpoints studied, including renal failure. Thus, while extreme anemia should be avoided, it appears that low transfusion thresholds are an acceptable way to avoid transfusion risk. Similar findings have been reported in ICU patients.
The storage age of transfused red blood cells (RBCs) has been interrogated as an AKI risk factor. With current preservatives, the accepted storage life of RBCs is 42 days. Early retrospective studies associated older RBCs transfused in cardiac surgery patients with increased mortality and postoperative AKI. However, multiple subsequent high-profile cardiac surgery and randomized trials of critically ill patients have not confirmed that RBC age affects mortality or other secondary outcomes such as kidney injury and renal failure. Thus, the age of transfused RBC is not thought to play a large role in postoperative AKI risk.
Utilization of cell saver technology may reduce red cell transfusion, particularly intraoperatively. However, in a large meta-analysis of randomized cardiac surgery trials, despite reducing transfusion, cell saver use did not reduce AKI or mortality risk. In contrast, a small meta-analysis of randomized cardiac surgery trials, comparing transfusion with leukocyte filtered and unfiltered blood there was a large (five-fold) reduction in AKI risk. Because there is no suggestion of harm with cell saver use, it should be routinely used to minimize allogeneic transfusion, and the use of leukocyte filtered red blood cells is also recommended.
In summary, optimal hematocrit management during CPB is still being defined, but both extreme anemia and transfusion may contribute to the development of CS-AKI and transfusion should only be considered after all sources of hemodilution have been minimized.
Modifiers of Renal Oxygen Demand
Because of its well-known association with organ transplant, hypothermia during CPB has been proposed as a strategy to reduce CS-AKI. However, in contrast to the protection conferred by extreme cold (e.g., 0–4°C), several randomized trials involving mild hypothermia (e.g., 32–34°C) have failed to show a protective effect. Regarding temperature, however, multiple retrospective analyses suggest that cumulative duration of hyperthermia during CPB (arterial outlet temperature > 37°C) and at the time of ICU arrival are independent CS-AKI risk factors. Overall, these data discourage clinicians both from the use of aggressive CPB rewarming strategies and from targeting hypothermic temperatures to reduce renal risk meaningfully in cardiac surgical populations.
Hyperglycemia increases renal O 2 demand through energy consumption to achieve glucose reuptake in the proximal tubule. Several retrospective studies linking elevated pre- and/or intraoperative serum glucose levels with increased postcardiac surgery mortality and AKI made intensive intraoperative glucose management protocols seem a logical approach to reducing renal risk. However, it is likely this link was mainly the result of the corelationship of a patient’s glucose intolerance with AKI risk, because in a well conducted randomized trial, intensive intraoperative and postoperative glucose control (i.e., target 80–100 mg/dL) was ineffective at reducing rates of CS-AKI. Furthermore, this regimen was associated with episodes of significant hypoglycemia, an increased risk of stroke, and mortality. Thus, current recommendations advise the administration of insulin therapy in the perioperative period targeting more modest glucose levels of 150–200 mg/dL.
Avoidance of Ascending Aortic Atheroembolism
Embolic arterial obstruction is poorly compensated for by collateral flow in the kidney, partly because of the organization of renal vasculature. Embolic vessel occlusions typically cause ischemic wedge-shaped infarcts in the cortex and medulla. Cardiovascular surgical procedures involving atherosclerotic vessels are known to have high particulate emboli rates (57%–77%).
Clamp occlusion and cannulation of the ascending aorta are actions that can disrupt atheromatous plaque and cause embolization. Measures of atherosclerosis of the ascending aorta and intraoperative arterial emboli counts are strong independent predictors of CS-AKI. Technologies have been introduced to reduce embolization by limiting aortic manipulation or trapping emboli. The Symmetry aortic connector system (ACD) is a technology developed to implant saphenous vein coronary bypass grafts onto the aortic wall with less manipulation of the aorta, with the hope that this might reduce embolization and organ dysfunction. In a large retrospective analysis comparing AKI after off-pump coronary artery bypass (OPCAB) procedures using Symmetry devices with standard OPCAB and standard CABG procedures, there was no difference in AKI rates. Another emboli-reducing device is the Embol-X intra-aortic filtration system, which catches emboli during aortic cross-clamp release. In a randomized trial of the Embol-X, there was no difference in a composite outcome of ischemic events; however, in a post hoc assessment of high-risk patients, Embol-X use was marginally associated with reduced renal complications (17/124 = 14% versus 28/117 = 24%, P = 0.04). In summary, there is no strong evidence that atheroembolism-reducing devices reduce AKI.
Selection of Antifibrinolytic Agent
The use of serine protease inhibitors (e.g., aprotinin) or lysine analogue antifibrinolytic agents (e.g., tranexamic acid, ɛ-aminocaproic acid) is associated with reduced bleeding and transfusion in cardiac, orthopedic, and trauma surgery patients, but questions have been raised regarding their effects on kidney function. Both groups of drugs are filtered by the kidney and saturate brush border binding sites of the low molecular weight protein transport system in the proximal renal tubule. Saturation by these agents prevents the transport system from processing other small proteins, which pass into the urine, causing “tubular proteinuria” that is presumed benign. This phenomenon has confused study of the safety of antifibrinolytic agents, because in other settings impaired protein reuptake is used as evidence of subtle renal injury.
Several studies have looked for clinical effects of antifibrinolytic agents on postoperative kidney outcomes. The best studied is aprotinin. As a result of earlier clinical trials that associated aprotinin use with increased mortality, aprotinin is not available in the United States. It has been approved for use in Canada and Europe but only for CABG surgery. In these countries aprotinin carries a warning to avoid use in patients with preoperative kidney dysfunction because of its known strong association with AKI. One retrospective study of ɛ-aminocaproic acid use reported no change in the incidence of CS-AKI with its introduction at a single institution. However, its use has been associated with kidney failure both in a prospective randomized study and another retrospective study. In contrast, tranexamic acid has not been associated with risk of kidney injury and may be the antifibrinolytic agent of choice in high-risk patients.
Selection of Vasoactive Agents
In addition to the indirect hemodynamic, humoral, and autonomic effects that typically influence intraoperative renal blood flow, use of vasoactive agents may have important additional direct effects on kidney perfusion. This section will cover vasopressor agents in both the intraoperative and postoperative settings.
Adrenergic receptor-mediated drugs are a mainstay among the agents available to the anesthesiologist and intensivist. Catecholamine-mediated renal effects include vasoconstriction through alpha-1 adrenergic receptors and vasodilation through dopaminergic, beta-2, and alpha-2 adrenergic receptors. In early animal models, long-lasting severe renal vasoconstriction and reductions in glomerular filtration resulted from brief high-dose infusions of norepinephrine. However, norepinephrine-induced kidney injury was queried because of subsequent animal models of sepsis, in which norepinephrine increased both global and medullary renal blood flow. Some authors have suggested that the idea that norepinephrine induces kidney injury in shock is “flawed, not evidence-based, contradicted by the available observations, and misleading.” This was supported by a high-profile randomized control trial demonstrating fewer adverse events overall and a trend toward fewer days of renal support ( P = 0.07) in patients receiving norepinephrine versus dopamine for circulatory shock. Overall evidence supports norepinephrine being a safe vasopressor for patients at high risk of AKI.
Arginine vasopressin (also called antidiuretic hormone) is a peptide secreted by the posterior pituitary that is used to treat vasodilatory hypotension and has widespread effects mediated by V 1 and V 2 receptors. Low-dose vasopressin activates baroreceptor reflexes, which contributes to this agent’s clinical usefulness when baroreceptor reflexes are impaired, such as during septic shock. Higher doses activate vascular smooth muscle V 1a receptors that mediate direct vasoconstrictor effects and increase systemic vascular resistance. In animal models of septic shock, vasopressin increases perfusion pressure while preserving renal blood flow. In a small double-blind trial, 24 septic shock patients were randomized to a 4-hour infusion of either norepinephrine or low-dose vasopressin, and open-label norepinephrine was available to both groups to maintain blood pressure. Urine output increased substantially in the vasopressin group but was unchanged in the norepinephrine group. Similarly, creatinine clearance was increased by 75% in the vasopressin group but did not change in the norepinephrine group ( P < 0.05). As such, some have speculated that vasopressin is a useful renoprotective agent rather than conventional catecholamines for the treatment of vasodilatory shock. However, a high-profile, large, randomized control trial did not confirm this, although this study did not aim to detect AKI rate changes. Regarding intraoperative vasopressin use, a retrospective review found that its administration was independently associated with CS-AKI (OR 3.60, 95% CI 1.22–10.62, P = 0.02). Thus, although vasopressin may be the preferable vasopressor in postoperative and critically ill patients, it remains unclear whether intraoperative vasopressin use should be avoided in patients at high renal risk.
The longstanding controversy surrounding dopamine infusion and kidney preservation highlights the importance of careful evaluation of a therapy before widespread adoption occurs. Dopamine infusion at rates less than 5 μg/kg/min selectively stimulates mesenteric dopamine-1 (DA1) receptors, causing increased renal blood flow, reduced renal vascular resistance, natriuresis, and diuresis. Although animal studies promoted the kidney protective potential of dopamine 40 years ago, numerous subsequent randomized studies in surgical and nonsurgical settings have not substantiated this claim. Meta-analyses of randomized trials have also failed to support dopamine as a kidney protective agent, without benefit regarding mortality, dialysis, or adverse events. A number of strongly worded editorials and reviews discourage the use of dopamine for kidney protection; with articles entitled “Bad Medicine: low-dose dopamine in the ICU” and “Renal-dose dopamine: from hypothesis to paradigm to dogma to myth and, finally, superstition?” These articles highlight the numerous negative studies and catalogue undesirable consequences of low-dose dopamine, including worsened splanchnic oxygenation, impaired gastrointestinal function, impaired endocrine and immunologic system function, blunting of ventilatory drive, and increased risk of postcardiac surgery atrial fibrillation. A body of literature cumulatively strongly discourages the routine use of dopamine infusion for kidney protection, but this therapy remains common.
Other vasoconstrictor agents have been less well-studied for their relationship to postoperative kidney impairment. A small study of 20 patients taking preoperative ACE inhibitor agents randomized to receive either 24-hour perioperative phenylephrine or angiotensin II for the control of systemic vascular resistance found no CS-AKI and concluded that angiotensin II is a safe alternative to phenylephrine. A high-profile randomized trial reported that angiotensin II infusion effectively increased blood pressure in critically ill patients with refractory vasodilatory shock. Post hoc analysis of this study also found that 28-day survival and RRT discontinuation was improved in patients receiving angiotensin II compared with placebo, among those with AKI requiring RRT at time of randomization. Although further studies are needed to support the routine use of angiotensin II, these findings indicate that patients with both severe vasodilatory shock and kidney injury might benefit from this agent.
Phosphodiesterase III inhibitors (milrinone, amrinone, enoximone) have positive inotropic and vasodilatory effects. The small number of studies addressing their effects on kidney function draw conflicting conclusions. In an animal model, milrinone exacerbates endotoxin-induced renal failure and anecdotal cases of renal dysfunction and failure in humans have been associated with the use of milrinone. In a randomized study of 40 CABG patients receiving enoximone or placebo, there was a significant rise in urinary α 1 -microglobulin in controls, compared with the enoximone group. Currently, there is insufficient data to characterize the kidney effects of phosphodiesterase III inhibitors; however, as they are excreted mainly through the kidney, their use is cautioned in patients with kidney failure.
Fenoldopam mesylate, the first clinically available selective agonist of DA1 receptors, is an approved therapy for the treatment of hypertension. Much like dopamine, its less DA1-selective predecessor, fenoldopam showed preclinical promise as a kidney protective agent. A randomized prospective study involving 160 cardiac surgery patients with baseline renal dysfunction noted lower postoperative serum creatinine values and higher creatinine clearance values compared with baseline with fenoldopam, but not placebo. However, a study involving 80 high-risk cardiac surgery patients found no benefit. Similarly, a prospective randomized double-blind study of 155 critically ill patients with established renal injury (including postcardiac surgery patients) found no benefit and even discussed possible increased adverse outcomes in diabetic patients. In liver transplant surgery, data from two studies identified improvements in kidney function in the first 3–5 postoperative days with fenoldopam (versus dopamine and placebo). Similarly, in aortic surgery patients, Halpenny and colleagues found a decline in creatinine clearance with aortic cross-clamp application and higher postoperative day 1 serum creatinine values in patients receiving placebo, but not in the fenoldopam group.
A subsequent meta-analysis of randomized control trials evaluating fenoldopam for kidney protection after major surgery noted that larger trials are needed because there was heterogeneity in surgical type, AKI definitions, and a high risk of bias in the majority of available studies, although they did report a reduction in postoperative AKI without changes in RRT or hospital mortality. Overall, current studies provide insufficient data to draw definitive conclusions on fenoldopam for kidney preservation.
Dopexamine hydrochloride is a DA-1 receptor and beta 2-adrenoceptor agonist with renal vasodilatory, natriuretic, and diuretic effects that has shown promise in animal studies as a kidney protective agent. In a study of 44 CABG surgery patients randomized to three different infusion doses of dopexamine or placebo, Berendes and colleagues noted improved systemic oxygen delivery, increased postoperative creatinine clearance, and reductions in markers of perioperative inflammatory response with dopexamine. In contrast, in a study of CABG surgery patients with normal ( n = 24) or impaired ( n = 24) baseline renal function randomized to dopexamine or placebo, Dehne and colleagues found no evidence of kidney protection. In a randomized comparison of dopexamine with dopamine in 24 liver transplant patients, Gray and colleagues found a trend toward better kidney function with dopexamine, but no overall outcome difference between the two therapies. A notable property of dopexamine is that its metabolism is significantly reduced in the presence of impaired liver function. A systematic review of 21 randomized-controlled trials involving dopexamine by Renton and colleagues concluded that existing evidence is inconsistent and insufficient to recommend dopexamine for kidney protection for either high-risk surgical or critically ill patients.
Mannitol is an osmotic diuretic with kidney protective potential in animal models and effects that include augmentation of renal blood flow and increased glomerular filtration rate. Because of these properties and augmented urine output, mannitol has commonly been used as a part of CPB priming solutions. However, mannitol is no longer commonly used for this purpose because randomized studies have failed to show a renoprotective benefit. A meta-analysis of nine trials relating mannitol use with AKI prevention found no benefit following cardiac or major noncardiac surgery, and even potentially added risk with radiocontrast agents. As an exception, kidney transplant patients may have a reduction in rates of kidney failure when mannitol is administered prior to cross-clamp removal. More research is needed in this area. Thus, although mannitol is useful for renoprotection during kidney transplant, it should not be used routinely for AKI prevention in other surgeries.
N-acetylcysteine has antioxidant properties, is a vasodilator, enhances the endogenous glutathione scavenging system, and in animals counteracts renal ischemia and hypoxia. Although early clinical studies comparing N-acetylcysteine with mannitol or placebo suggested benefit, this was not supported in subsequent investigations. A meta-analysis of cardiac surgery studies reported that N-acetylcysteine does not reduce the risk of CS-AKI. There is no evidence to support its use to reduce postoperative renal dysfunction.
Many issues related to postoperative kidney protection mirror those discussed in the intraoperative period (e.g., vasoactive agents, pharmacologic agents). This section focuses on fluid management, glucose management, and other select pharmacologic agents.
Intravenous Fluid Choice
The best known link between intravenous fluid choice and kidney injury is for HES preparations. In randomized trials of septic critically ill patients, the use of HES was strongly associated with kidney failure and the need for renal replacement therapy. After these trials in 2012, HES preparations were largely pulled from the market and unavailable in the United States.
The administration of normal saline (0.9% sodium chloride) or saline-based colloid solutions results in an acid-base abnormality frequently observed in postoperative and critically ill patients: metabolic hyperchloremic acidosis. Hyperchloremic acidosis promotes AKI because of the adverse effects of elevated chloride levels and acidosis on kidney homeostasis, including reduced blood flow and glomerular filtration rate, increased afferent arteriolar tone, and altered renin release. The propensity of studies in this area seem to support the notion that administration of chloride-liberal fluids (i.e., normal saline) are associated with AKI. In one study of critically ill patients, a chloride-restrictive fluid policy (versus chloride-liberal) was associated with a reduced rate of AKI. Interestingly, the SPLIT randomized trial comparing buffered crystalloid with normal saline did not confirm this finding; however, it was argued that the fluid volume administered in this trial (~ 2 L) was insufficient. Two subsequent large pragmatic, cluster-randomized, cross-over trials have since demonstrated that balanced crystalloids are safer than normal saline with regard to kidney injury. The SMART trial examined the effect of balanced crystalloids versus normal saline in critically ill patients and demonstrated that balanced crystalloids were associated with lower rates of MAKE30 (OR 0.90, 95% CI 0.82–0.99, P = 0.04) and a trend toward lower mortality at 30 days (OR 0.90, 95% CI 0.80–1.01, P = 0.06). Furthermore, in the SALT-ED study, the lower rate of MAKE30 for balanced crystalloids compared with normal saline extended to a heterogenous emergency department population. These findings mirror those of a very large, retrospective, propensity matched cohort study on this topic. Thus, except where saline is specifically clinically indicated (i.e., significant hyponatremia, cerebral edema), balanced crystalloids are preferable over normal saline for kidney protection.
Regarding GDFT, early studies demonstrated outcome benefits related to septic shock, organ dysfunction, and mortality. However, a high-profile, randomized trial subsequently did not confirm these benefits from GDFT compared with usual care, although some argue that the original Rivers trial had already improved baseline standard of care for patients with sepsis. Notably GDFT did not decrease kidney injury in either trial. Nevertheless, the debate continues whether GDFT should be used in patients with septic shock; however, existing data appear to indicate that it should not be used for the purpose of kidney protection.
Fluid overload has been consistently linked to worse outcomes in critically ill patients. Diuretic agents are frequently used to promote fluid clearance after the initial resuscitation period. Diuretic agents increase urine generation by reducing reuptake of tubular contents; this can be achieved by numerous mechanisms, including blocking tubular solute reuptake through active transport mechanisms (e.g., loop diuretics), altering the tubular osmotic gradient to favor solute remaining in the tubule (e.g., mannitol), or by affecting the hormonal signaling to the tubule to increase urine generation (e.g., atrial natriuretic peptide [ANP]). In general, the rationale underlying kidney protection from diuretic agents relates to the decreased likelihood of tubular obstruction by casts with increased solute flow through injured renal tubules, thus retaining tubular patency and avoiding oligo/anuria and the need for dialysis caused by fluid overload. Importantly, despite the clinician’s satisfaction at seeing increased urine output in response to diuretics, this increase does not insure improved renal function. The classes of diuretics are discussed in the following sections.
Loop diuretics, including furosemide, ethacrynic acid, and bumetanide, are also called “loop inhibitors” because they inhibit active solute reabsorption in the mTAL of the loop of Henle, causing more solute to remain in the kidney tubule and increasing urine generation. Furosemide also induces renal cortical vasodilation. In animal models, furosemide raises oxygen levels in the medulla, and in some settings even protects tubules from damage after ischemia-reperfusion or nephrotoxic insult.
However, in surgical and critically ill patients, numerous retrospective studies have shown no benefit and have even suffered harm with loop diuretic use. A randomized study of patients undergoing major thoraco-abdominal or vascular surgery procedures found no kidney protective benefit of an extended postoperative furosemide infusion compared with placebo. Another double-blind, randomized-controlled trial comparing infusions of low-dose dopamine, furosemide, and placebo in cardiac surgery patients found a two-fold greater postoperative rise in serum creatinine in the group receiving furosemide relative to dopamine and placebo groups. In addition, multiple meta-analyses have been performed on this topic. Some meta-analyses have demonstrated improvements in urine output and possible shorter duration of RRT, but no meta-analysis has shown improvements in AKI. The sum of evidence does not support the use of loop diuretics as perioperative kidney protective agents and suggests they may even have nephrotoxic effects, especially in cardiac surgery patients.
The potential harm of loop diuretics to the kidney may be explained by inhibition of loop of Henle solute transport, which causes greater solute delivery and active transport demands downstream in the distal renal tubule, where higher oxygen expenditure is required per unit sodium reabsorbed. Restated, loop diuretics may spare injury to the loop of Henle at the expense of insult to the metabolically less-efficient distal renal tubule. Rats were chronically administered loop diuretics and showed hypertrophied distal tubular cells with increased numbers of mitochondria. Thus, loop diuretics may displace injury from the loop of Henle to more distal tubular locations.
The natriuretic peptides are hormones that interact with a specific signal transmission system involved in the regulation of volume homeostasis. In response to volume expansion, the release of these hormones (atrial natriuretic peptide [ANP], renal natriuretic peptide [urodilatin], and brain natriuretic peptide [BNP]) is associated with receptor-mediated vasodilation and natriuresis.
ANP is normally synthesized by the atria in response to atrial wall tension; anaritide is the human recombinant form of ANP. ANP increases glomerular filtration and urinary output by constricting efferent while dilating afferent arterioles and is associated with attenuation of renal cell injury in animal models of AKI. Unfortunately, human trials of ANP as a kidney protective agent have not been conclusive. In a multi-center randomized double-blind, placebo-controlled clinical trial of ANP as a 24-hour intravenous infusion (0.2 μg/kg/min) in critically ill patients with established acute renal injury, the primary end point of improved dialysis-free survival after 21 days was not achieved. A secondary analysis revealed that dialysis-free survival was higher in the ANP group for oliguric patients. Conversely, in nonoliguric patients, dialysis-free survival was higher in the placebo group. Some have speculated that the disparity of outcomes in this study is a result of the vasodilating properties of ANP; hypotension was more frequent in the nonoliguric patients and may have overwhelmed any kidney protective benefit. However, a second similar study was designed to reproduce the favorable findings from the first study. In critically ill patients with established oliguric renal dysfunction, there was no benefit. Two randomized trials in cardiac surgery patients have been performed examining the effect of ANP on perioperative kidney function. Both studies examined low-dose (0.02 μg/kg/min) starting on the initiation of CPB and continuing until 12 hours after oral intake. The authors demonstrated improvements in postoperative creatinine levels and serum creatinine levels up to 1 year. However, both trials had concerns regarding study methodology (randomization, blinding, etc.), and it is therefore difficult to draw firm conclusions from these studies. Further trials are needed to determine whether ANP can be used for AKI prevention or treatment.
Urodilatin differs from ANP only by the addition of four amino acids to the N terminus end of the peptide but has more potent natriuretic properties. Three small randomized trials indicate kidney protective benefit from this agent in patients with established kidney dysfunction, including reduced duration of hemofiltration and frequency of hemodialysis following heart transplant ( n = 24) and reduced incidence of dialysis after cardiac ( n = 14) and liver transplant ( n = 9) surgery. However, interpretation of these data is complicated by very high (up to 86% in the control group) dialysis rates. A larger randomized, double-blind trial including 176 critically ill patients with oliguric acute renal failure did not demonstrate a benefit from any of four different urodilatin dose regimens compared with placebo.
BNP is normally synthesized by the left and right ventricles in response to ventricular dilatation and BNP assay has become a diagnostic tool for congestive heart failure. The human recombinant form of BNP is nesiritide. Nesiritide has potent vasodilating properties and has been approved by the US Food and Drug Administration (FDA) as a treatment for acutely decompensated heart failure, partly because of its ability to reduce ventricular filling pressures rapidly, to relieve dyspnea, and to induce sustained diuresis. Studies have indicated that BNP treatment may worsen kidney function in heart failure patients. Nesiritide has been tested in cardiac surgery patients with reduced left ventricular function; however, results have been inconsistent, with nesiritide showing some benefit for AKI prevention, but no benefit on death or dialysis. With the current evidence, the routine use of nesiritide for AKI prevention is not encouraged for cardiac surgery patients and should be discouraged for those with nonsurgical heart failure.
Early clinical ICU studies suggested that improved glucose management resulted in improved survival and better kidney outcome (e.g., target in ICU: 80–110; post-ICU: 180–200 mg/dL). Unfortunately this has not been confirmed, with even increased mortality from over-aggressive insulin therapy, in numerous subsequent studies including cohorts of medical and surgical critically ill patients, including multiple intraoperative randomized studies of cardiac surgery patients. Consequently, although intensive glucose control (target < 110 mg/dL) may provide some kidney protection, the risk of mortality outweighs the benefit and a more modest glucose level of < 180 mg/dL should be targeted.
Corticosteroids attenuate the inflammatory response and provide kidney protective effects in animal models. Clinically, in a small, randomized, placebo-controlled, double-blind trial of 20 patients undergoing cardiac surgery with cardiopulmonary bypass, there was no evidence of kidney protection in patients receiving dexamethasone 1 mg/kg before induction of anesthesia and 0.5 mg/kg 8 hours later. The DECS randomized control trial also found no kidney protective effects of high-dose dexamethasone (1 mg/kg) when compared with placebo in cardiac surgery patients. In a post hoc analysis of the DECS trial, Jacob et al. reported that high-dose dexamethasone did decrease the rate of severe AKI as defined by need for RRT, with 0.4% of patients requiring RRT in the dexamethasone group and 1.0% in the placebo group. The greatest benefit was for those with CKD, especially those with eGFR < 15. However, another very large randomized control trial, the SIRS trial, also failed to show a kidney protective effect for methylprednisolone in cardiac surgery patients. Thus, although steroids were safe in these trials apart from an increased rate of hyperglycemia, steroids should not be used for kidney protection as they have shown no convincing evidence of benefit.
Non-steroidal Anti-Inflammatory Drugs: Aspirin and Cyclo-Oxygenase-2 Inhibitors
Although there is little evidence or rationale for the use of many nonsteroidal anti-inflammatory agents (NSAIDs) for kidney protection, in a retrospective study of 5065 coronary artery bypass surgery patients, Mangano and colleagues reported that early re-institution (< 48 hours) of aspirin therapy after surgery was associated with a three-fold reduction in mortality and a 74% reduction in the incidence of renal failure. In contrast, the POISE-2 trial examined the effects of perioperative aspirin in patients undergoing major noncardiac surgery who were at high risk of vascular complications. Although it was a tertiary outcome of the study, the authors reported that patients receiving aspirin were at higher risk of AKI with RRT. The kidney injury substudy of the POISE-2 trial demonstrated that one dose of preoperative aspirin and then aspirin for up to 30 days postoperatively did not reduce the risk of kidney injury compared with placebo. Regarding other NSAIDs, a review by Lee et al. reported the combined findings from randomized controlled postoperative NSAID analgesia trials that recorded renal function, including over 1200 patients receiving ketorolac, diclofenac, indomethacin, or placebo. These authors noted a 16 mL/min decline in creatinine clearance and a reduction in potassium clearance in NSAID patients at 24 hours after surgery, but no episodes of acute renal failure requiring dialysis. Lee and colleagues concluded that postoperative NSAID analgesia has only small temporary negative effects on kidney function in adults with normal kidneys at baseline, stressing that these findings might not apply to children or adults with already impaired kidney function. Although similar analyses have not been performed for cyclo-oxygenase (COX)-2 selective NSAIDs, a randomized, double-blind, placebo-controlled postoperative analgesia trial involving 462 CABG patients showed a greater incidence of serious adverse events in patients receiving valdecoxib/paracoxib, including a trend toward more postoperative kidney dysfunction. The interim findings from this trial prompted the investigators to terminate the study for safety reasons. A second similar randomized trial including 1671 CABG patients or placebo also noted increased adverse outcomes, in particular an increase in cardiovascular events in patients receiving valdecoxib/paracoxib. The incidence of kidney failure or dysfunction in this study was insignificantly higher in the valdecoxib/paracoxib groups than in the placebo group. Taken together, NSAIDs at best have negligible effects on perioperative kidney function but might be associated with increased risk of kidney injury, therefore they should not be used for kidney protection. In the case of aspirin, however, the current recommendation is to restart aspirin within 48 hours of cardiac surgery to maintain graft patency.
Alpha 2 Adrenergic Agonist Agents
The normal physiology of the kidney includes a role for adrenergic receptors in modulating vasoconstrictor (alpha 1) and vasodilating (alpha 2) effects, respectively. Vasoconstriction contributes to the pathophysiology of acute renal injury, and several preclinical studies confirm that clonidine (an alpha 2 agonist) attenuates experimental acute renal injury. A double-blind, randomized placebo-controlled trial in 48 CABG surgery patients evaluated preoperative clonidine for kidney preservation and found a significant benefit on kidney injury in the first postoperative day but no difference 3 days after surgery. However, the kidney injury substudy of the POISE-2 trial showed no benefit for perioperative administration of clonidine to prevent AKI. Dexmedetomidine has shown some promise for kidney protection. A meta-analysis of 10 randomized control trials reported improvement in AKI in cardiac surgery patients receiving dexmedetomidine; however, the trials included were small and did not have uniform definitions of AKI. Taken together, insufficient data are available to recommend these agents and currently they are not commonly used for kidney protection.
Calcium Channel Blockers
Three major classes of calcium blockers exist with varying pharmacologic properties: benzothiazepines (e.g., diltiazem), phenylalkylamines (e.g., verapamil), and dihydropyridines (e.g., nifedipine, nimodipine). Calcium channel blockers decrease renal vascular resistance and increase glomerular filtration and have been reported to exert beneficial effects in experimental models of toxic and ischemic acute renal failure. A meta-analysis of randomized studies comparing perioperative use of calcium channel antagonist agents with control or other agents (e.g., nitroglycerin, dopamine) in cardiac surgery demonstrated no overall effect, but post hoc analysis identified some benefit if preoperative creatinine clearance was less than 95 mL/min. The overall inconclusive findings from trials of calcium channel blockers do not support the use of these agents for perioperative kidney preservation.
Renin-angiotensin System (RAS) Blockers
The RAS mediates vasoconstriction and contributes to the paracrine regulation of the renal microcirculation. AT1 blockers and ACE inhibitors act by decreasing activation of the RAS and animal studies suggest that these agents may have protective properties in experimental AKI. While both drug classes are recognized for their ability to slow the progression of chronic renal disease, their perioperative use is controversial. Pertinent to the renal effects of these agents is their potential to precipitate acute kidney deterioration in clinical conditions in which RAS activation is critical to the regulation of renal filtration, such as with renal artery stenosis or volume depletion. These agents can also lead to hemodynamic instability when initiated soon after cardiac surgery. Two retrospective studies have not found perioperative ACE inhibitor therapy to be an independent predictor of kidney outcome after cardiac surgery. However, in a prospective study of aortic surgery patients, Cittanova and colleagues reported an increased risk of postoperative kidney dysfunction in patients receiving chronic ACE inhibitor therapy. In a retrospective review of cardiac surgery patients, Chou et al. reported an independent association of RAS blocker therapy with a lowered risk of ensuing CKD. More studies are needed to determine whether RAS antagonists provide harm or benefit in the perioperative period and, if benefit, the timing of their administration.
Endothelin Receptor Antagonists
Endothelin is a potent vasoconstrictor that severely limits renal blood flow and may contribute to the pathophysiology of acute renal failure. Two types of endothelin receptors have been identified: ET A receptors are located on vascular smooth muscle cells and mediate endothelin-induced vasoconstriction, and ET B receptors are on endothelial cells and mediate release of nitric oxide and prostacyclin. The development of endothelin receptor antagonists has provided the opportunity to assess the kidney protective potential of this class of drug. Most studies to date have been preclinical. However, endothelin antagonists are being studied in the setting of diabetic nephropathy and glomerular disease. It is possible that these agents could be used to prevent the AKI to CKD transition, but no studies would support their clinical use for this at the present time.
Recombinant human erythropoietin is typically prescribed for anemia associated with cancer chemotherapy or end-stage renal failure; however, this agent has also been shown to have kidney protective properties in at least nine studies in animal models of acute ischemic, hypovolemic, endotoxic, and nephrotoxic renal injury. Although it appears safe, a meta-analysis of 10 randomized control trials did not show a benefit for erythropoietin to prevent AKI.
Growth factors (IGF-I, EGF, HGF)
When the kidney is acutely injured, animal models indicate that growth factors are involved in tubular recovery, such as insulin-like growth factor-I (IGF-I), epidermal growth factor (EGF), and hepatocyte growth factor (HGF). Therefore, providing an exogenous source of these substances has been identified as potential kidney protective strategy. However, human studies do not support the effectiveness of this strategy. In a multi-center, randomized-controlled double-blind study of 72 critically-ill dialysis patients receiving recombinant human IGF-I or placebo, there was no difference in the rate of recovery of kidney function or mortality. In a second study involving 44 post-kidney transplant patients with established kidney dysfunction, there was no difference in the return of kidney function or subsequent requirement for dialysis with IGF-1 compared with placebo.
Pulmonary Vasodilators (PGE1, PGI2, NO)
Prostaglandins (PGs) are mediators of microvascular blood flow distribution and there is rationale for the potential kidney benefit of prostaglandins PGE 1 and PGI 2 . Studies examining the kidney protective benefit of PGE 1 (also known as alprostadil) were not conclusive. A small study randomized CABG patients with poor ventricular function randomized to either low-dose PGI 2 (2 ng/kg/min) or placebo and found 24 hour postoperative creatinine clearance to have dropped 32% in controls, but a rise of 13% in patients receiving PGI 2 . PGI 2 therapy was also associated with significantly higher 24 hour postoperative creatinine clearance values ( P < 0.01). These authors concluded that PGI 2 (also known as prostacyclin, epoprostanol, PGX, and Flolan) may have kidney protective properties.
Nitric oxide (NO) has vasodilating properties that could be beneficial in preventing AKI. Furthermore, hemolysis is common during CPB and results in release of oxyhemoglobin. Oxyhemoglobin promotes oxidative stress in the kidney and can deplete endogenous NO through formation of methemoglobin. Thus, it was postulated that supplemental NO during cardiac surgery could prevent postoperative AKI. Lei et al. randomized 244 patients undergoing multiple valve surgery (mostly caused by rheumatic fever) to 80 ppm of inhaled NO or N 2 . They found that NO reduced postoperative AKI, MAKE30, MAKE90, and MAKE at 1 year. These results are very promising and warrant future trials to assess the efficacy of NO to prevent AKI in other types of cardiac surgery, cardiac surgery on patients with multiple comorbidities, and other disease states with elevated hemolysis.
Kidney Protective Bundle
Because there are currently no pharmacologic agents proven to attenuate ongoing AKI, the “KDIGO bundle” was designed. The KDIGO bundle is a group of treatments aimed at precise hemodynamic monitoring/support and avoidance of nephrotoxic agents ( Fig. 17.7 ). Meersch et al. randomized cardiac surgical at high risk of postoperative AKI (TIMP-2 * IGFBP7 > 0.3 at 4 hours postoperatively) to usual care or the KDIGO bundle. They demonstrated improvements in hemodynamics, glycemic control, and the frequency and severity of CS-AKI. It appears that the most effective kidney protective strategy currently available for stressed kidneys is an astute clinician who provides optimal, timely, and goal-directed care with particular attention to minimizing renal related insults.