The kidney primarily functions to preserve internal homeostasis by regulating effective arterial blood volume (EABV), osmolality, and ionic composition, and to concentrate and excrete the daily endogenous and exogenous load of nitrogenous waste. These actions occur by a complex interplay between glomerular filtration, tubular reabsorption, and tubular secretion. The kidney also plays a vital role as an organ of endocrine function, regulating red blood cell mass by the production of erythropoietin, and calcium and phosphorus homeostasis through the synthesis of vitamin D to its most active form, 1,25-dihydroxycholecalciferol.

The glomerular apparatus is a network of capillaries originating from the afferent arteriole and surrounded by an extension of the basement membrane of the proximal tubule called Bowman’s capsule. The urinary space, or Bowman’s space (BS), separates the capsule from the glomerular tuft. The formation of urine commences with a protein-free ultrafiltrate of plasma passing through the glomerulus into the BS. The rate of formation of tubular fluid (glomerular ultrafiltration rate or GFR) is dependent on the hydraulic permeability of the glomerular capillary and the net ultrafiltration pressures across the capillary wall. GFR is a non-energy-requiring process regulated by the Starling forces, as defined by the following equation:

The difference in hydrostatic pressure between the glomerular capillary (P_GC) and Bowman’s space (P_BS) promotes filtration, whereas colloid osmotic pressure in the capillaries (π_GC) opposes it. As filtration of a protein-free fluid transpires, a progressive increase in π_GC ensues such that by the termination of the capillary the ultrafiltration pressure becomes zero (i.e., P_GC = π_GC + P_BS). This physiologic principle stipulates that GFR is highly dependent on renal plasma flow; the greater the flow rate, the slower the rise in π_GC and hence an increase in GFR.

Normal GFR is approximately 180 L/day and relates to the extensive surface area and permeability of the glomerular tuft. Despite wide fluctuations in mean arterial pressure, only minute changes arise in GFR due to renal autoregulation. Renal autoregulation occurs for both renal blood flow and GFR, and is centered on the afferent arteriole’s inherent ability to sense transmural pressure and alter its wall tension to maintain resistance proportional to pressure. Although renal blood flow begins to decline at mean arterial pressures less than 50 mmHg, autoregulation of GFR occurs at higher pressures (70-80 mmHg). This concept becomes clinically relevant during cardiopulmonary bypass (CPB) when perfusion pressure is decreased below the autoregulatory threshold for GFR, resulting in diminished urine output. The relation of perfusion pressure during CPB to renal blood flow and hence urine output has never been established, as will be discussed later; however, increasing perfusion pressure is often considered when managing intraoperative oliguria to theoretically prevent AKI.

Relative to renal blood flow and perfusion, the kidney receives 20% of the normal cardiac output (Q_t), an amount far in excess of the kidney’s total oxygen and energy requirement. However, this percentage of Q_t is essential to power the processes of filtration, reabsorption, and secretion. Compared with the proportion of overall renal blood flow, a striking disparity exists between the renal cortex and medulla. The cortex receives more than 90% of renal blood flow, creating tissue oxygen tensions of approximately 50 versus 8 to 10 mmHg in the medulla. Although necessary to prevent washout of the hypertonic interstitium and to preserve the osmotic gradient necessary for tubular secretion and reabsorption, preferential blood flow to the cortex results in the medullary thick ascending limb (mTAL) of the loop of Henle being extremely vulnerable to hypoperfusion-induced ischemia, and may explain why AKI can be induced by as little as a 40% decrease in renal blood flow. Prevention of renal ischemia and AKI is therefore dependent on increasing oxygen delivery as well as reducing oxygen demand. An intrinsic tubuloglomerular feedback system provides initial protection in the event of medullary hypoperfusion by renin-mediated afferent arteriolar vasoconstriction, which in turn leads to a decrease in plasma ultrafiltration and hence a decrease in energy expenditure of the cells of the mTAL (4). A decline in urine output is to be expected and is termed prerenal success. Prolonged periods of hypoperfusion may however overwhelm this process, leading to erythrocyte sludging in the medulla and eventual tubular obstruction from necrotic cellular debris. The ensuing rise in intratubular pressure (↑P_BS) in relation to the decline in glomerular capillary pressure (↓P_GC) may result in progressive azotemia not amenable to manipulation of EABV, Q_t, or other extrarenal factors, and ultimately AKI.

Therefore, GFR, and consequently urine formation, may be diminished for the following reasons: (1) a decrease in Kf from exposure to nephrotoxins; (2) a decrease in P_GC from hypoperfusion; (3) an increase in P_BS from intratubular obstruction due to cellular debris; and (4) an increase in π_GC from concentration of proteins due to dehydration. Although the decline in GFR will produce a decline in urine output, the contrary is not necessarily true; that is, a decrease in urine volume does not always mean a decline in GFR, nor does it imply the diagnosis of AKI.

The additional processes of tubular reabsorption and secretion further refine urine formation. Homeostasis is preserved by transforming the plasma ultrafiltrate into urine of variable volume, osmolarity, and composition through a complex interaction between the renin-angiotensin-aldosterone system, the sympathetic nervous system, and other hormonal and physical factors. The proximal, distal, and collecting tubules each modulate and control various functions. Of primary significance to those caring for patients during cardiac surgery is an understanding of sodium and water homeostasis in relation to safeguarding EABV. Despite the vast amount of sodium that is filtered daily (140 mEq × 180 L = 25,200 mEq), less than 1% is typically excreted in the urine. The bulk of sodium reabsorption, and therefore volume and osmotic control, occurs in the proximal convoluted tubule, the loop of Henle, and the distal tubule. During periods of hypovolemia, these segments fractionally reabsorb more than 99% of the filtered load of sodium and thereby fractionally excrete less than 1%. (This concept of the fractional excretion of sodium will be addressed in more detail later.) Urine is concurrently refined by the effect of antidiuretic hormone (ADH) on the collecting duct to reabsorb water, thereby serving to retain plasma tonicity (normal range, 280-295 mOsm/kg H₂O). Osmoreceptor-mediated and baroreceptor-mediated releases of ADH from the hypothalamus are the result of hypertonicity and low EABV, respectively. Finally, postoperative pain, anxiety, and/or nausea may also stimulate the release of ADH independent of osmolarity or EABV. As would be predicted, these patients characteristically develop oliguria despite having normal renal function. This is a common scenario following cardiac surgery in which efforts to increase urine production by inappropriate volume challenges may result in hyponatremia and/or pulmonary edema.

Multiple models for the prediction of AKI after cardiac surgery have been developed (see Table 14.1). The models in these studies have demonstrated statistical significance in the ability to predict AKI. Although definitions of AKI slightly differ in each of these studies, the similarities of variables were consistent.

Notwithstanding inconsistencies in the medical literature in defining the criteria for establishing preoperative renal risk factors (9), virtually all prior studies have established preoperative renal dysfunction, as defined by an elevated serum creatinine or a decrease in creatinine clearance, as the single greatest risk for developing postoperative AKI in patients undergoing cardiac surgery. Creatinine, a breakdown product of muscle creatine, is principally filtered with only nominal reabsorption and secretion and is therefore reflective of eGFR. Prior studies in patients who had undergone cardiac surgery tended to define perioperative renal dysfunction as a serum creatinine level that was greater than 1.35 to 1.5 mg% (120-130 mmol/L). The incidence of postoperative AKI in these studies was less than 1% in patients with normal renal function undergoing routine, elective noncardiac surgery. However, depending on the degree of renal insufficiency (e.g., whether the patient has chronic kidney disease (CKD) requiring preoperative dialysis), the nature of surgery, and ensuing perioperative complications, this incidence exceeded 20% and was associated with a 30% to 80% postoperative mortality (10,11,12,13,14,15,16,17). Recently, a better preoperative estimation of eGFR (and hence better predictive parameter of perioperative renal dysfunction) has been described and employs either serum creatinine and/or cystatin
C as part of a calculated or measured creatinine clearance or eGFR (18). Wang et al. (19) confirmed that the preoperative use of the Cockroft-Gault equation (Table 14.2) was a more accurate predictor of renal outcome in patients undergoing cardiac surgery. In this observational study, it was noted that the odds of AKI requiring dialysis (AKI-D), death, or having major morbidity increased by 52%, 27%, and 18%, respectively, for each 10 mL/min/1.73 m2 decrement in estimated creatinine clearance. This was in comparison to the risk-adjusted odds of having the same adverse outcomes increase by 20%, 8%, and 13% for 0.2 mg% increment increases in serum creatinine levels. Wijeysundera et al. (20) prospectively analyzed 10,751 patients undergoing cardiac surgery, also assessing the value of preoperative screening of creatinine clearance based on the Cockroft-Gault equation. The authors noted a 13% incidence of what they termed occult renal insufficiency (i.e., patients with normal serum creatinine levels but calculated creatinine clearances of ≤60 mL/min), and found them to be at risk for developing AKI-D. Although this approach to eGFR is not novel, it is clearly underutilized by most clinicians in evaluating patients’ preoperative renal risk before cardiac surgery. This formula and more recent versions, the modification of diet in renal disease (MDRD) equation (21), and CKD-EPI (22), offer more precise methods of assessing preoperative renal risk for patients requiring cardiac surgery. In this regard, Inker et al. (23) have further refined the eGFR by employing both serum levels of creatine and cystatin C, a protein less influenced by muscle mass and diet, and solely eliminated by glomerular filtration.

The RIFLE criteria were developed by the Acute Dialysis Quality Initiative group (ADQI) in order to unify the diagnosis and classification of AKI for both research and clinical practice. These criteria utilize biochemical markers (e.g., serum creatinine and eGFR), and urine output and their respective changes over time (Table 14.3). RIFLE criteria classify AKI patients into three severity categories (risk, injury and failure) as well as two outcome categories (loss of function and endstage renal disease).

In patients with chronic, non-dialysis-dependent renal dysfunction, the etiology of the underlying renal disease is far less critical than their level of overall dysfunction. The preoperative renal risk of a patient with a history of lupus nephritis and a creatinine clearance of 25 mL/min undergoing CPB is therefore no different from that of a patient with hypertensive nephrosclerosis with a similar creatinine clearance and undergoing the same procedure.

Advanced age also constitutes a risk factor for developing postoperative renal dysfunction and AKI. Multiple studies have concluded that age more than 63 years is an independent variable for developing postoperative renal failure and may be related to diminished nephron mass as a result of reduced expression of vascular endothelial growth factor (24), as well as loss of autoregulatory ability. The odds ratio of 1.6 for the development of AKI increases with every 10 years of age greater than 60 (8). Exposure to nephrotoxic agents may also contribute to perioperative renal insufficiency. Radiocontrast agents presumably induce calcium-mediated vasoconstriction leading to medullary ischemia, which is accentuated in highrisk individuals (25). Azotemic patients who undergo cardiac or major vascular surgery and who are administered radiocontrast as part of their preoperative evaluation are at particularly high-risk for developing AKI. Higher doses of iodine contrast media are associated with an increased incidence of AKI. Doses of iodine contrast media approaching 5 g/kg are associated with the AKI when cardiac operation is performed after angiography (26). Delaying elective surgery is advisable if the serum creatinine increases by more than 0.5 mg% (25% decrease in creatinine clearance) within 48 hours of exposure, given the additional inherent surgical risk of postoperative AKI-D (27). Perhaps the type of surgery should influence the time interval between angiography and surgical intervention. Hennessey et al. (28) found that patients proceeding to the operative room within 24 hours after catheterization for valve surgery were at increased risk of AKI versus patients with an interval of 48 or 72 hours. However, Ozkaynak et al. (29) found that time to surgery after angiography was not a risk factor for AKI, in a retrospective study that examined mostly coronary artery bypass graft (CABG) patients. Drugs with nephrotoxic side effects, such as aminoglycosides and cyclosporine, may also cause AKI when serum levels are not properly monitored in the perioperative period. The use of nonsteroidal anti-inflammatory agents per se does not seem to influence the incidence of postoperative AKI, provided that preoperative renal function is normal (30).

Finally, it has been proposed that there may be a genetic predisposition for developing AKI following cardiac surgery, particularly as it relates to the inflammatory response to CPB. Gaudino et al. (31) studied 111 consecutive patients undergoing single-vessel CABG, and noted a correlation between postoperative interleukin 6 (IL-6) levels and renal dysfunction in those patients with IL-6-174 G/C polymorphism.

Stafford-Smith et al. (32) isolated deoxyribonucleic acid (DNA) from 1,671 patients undergoing CABG and depending upon the patients’ race, noted polymorphism of multiple alleles associated with the development postoperative renal dysfunction. In the future, the ability to determine patients’ preoperative genetic profiles, in order to gauge their degree of renal risk before cardiac surgery, may become commonplace.

Renal dysfunction following CPB continues to be a relatively common occurrence, with AKI-D developing in 1.2% to 13% of patients, depending on their preoperative eGFR (10,11,14,15,16,17). Perhaps more importantly, the development of AKI and AKI-D dramatically increases the risk of mortality. Table 14.4 lists several relevant studies demonstrating the increased mortality risk associated with the development of AKI after cardiac surgery.

Intraoperative renal risk factors include low Q_t, decreased EABV, need for intra-aortic balloon counterpulsation, prolonged CPB times of more than 130 to 180 minutes and the development of a systemic inflammatory response syndrome (SIRS), inappropriate hemodilution, and embolic phenomenon. Cardiac surgery with CPB is associated with the development of a SIRS and associated renal dysfunction. Leukocyte activation, as a component of SIRS, has been suggested as one of the many causes of perioperative renal injury (36). In this regard, leukodepletion instituted during CPB has been shown to reduce markers of renal injury (37) and, in one small prospective, randomized clinical trial, led to a decrease in the incidence of AKI-D (38). Aprotinin, a serine protease inhibitor and antifibrinolytic, has also been postulated to mitigate against the inflammatory response and have a renoprotective
effect (39). Although early reports of the use of aprotinin during CPB seemed to support this contention (40), a recent large prospective, multicenter study noted a profound risk of aprotinin-associated AKI-D (41). In this study of more than 4,300 patients, the intraoperative use of aprotinin during CABG with CPB resulted in a dose-dependent doubling or tripling of AKI-D when compared with the use of aminocaproic acid and tranexamic acid. The use of this drug to limit blood loss during complex or revision cardiac surgeries must be tempered by this reported higher incidence of AKI and other severe end-organ damage, and the use of alternative antifibrinolytic therapy to minimize perioperative bleeding while preserving renal function may be prudent (42).

Relative to the effects of hemodilution on renal function, two recent studies question the role of low hematocrit as a risk factor for postoperative AKI following cardiac surgery with CPB (43,44). Karkouti et al. (43), in a prospective, observational study of more than 9,000 patients, raise the question of optimal hematocrit during hemodilution. The authors found an increase in AKI-D when the hematocrit was less than 21% or more than 25%, suggesting a decrease in renal oxygen-carrying capacity or a decrease in microcirculatory blood flow, respectively. Habib et al. (44) also suggested that a nadir hematocrit of less than 24% during hemodilution for CPB was a risk factor for developing AKI-D.

Finally, as it relates to thromboembolic phenomenon and the incidence of AKI after CPB, Davilla-Roman et al. (45), using intraoperative epiaortic ultrasound, identified ascending aortic atherosclerosis as an independent risk factor for developing postoperative renal dysfunction. The authors studied 978 patients with normal preoperative renal function undergoing open-heart surgery and noted that the incidence of postoperative renal dysfunction increased with the degree of ascending aorta atherosclerosis from 4.1% in normal-mild disease to 9.0% in moderate disease, and to 17.1% in those with severe disease. Although not assessed in this study, it is postulated that this increased risk may relate to renal thromboembolism. This atheroma burden may be associated with a genetic predisposition as noted by MacKensen et al. (46), who found a greater susceptibility of renal dysfunction in patients with aortic atheromatous disease who lacked the apolipoprotein E ε4 allele. Other factors such as pulsatile versus nonpulsatile flow and hypothermic versus normothermic CPB are discussed in later chapters, but in general they have minimal, if any, effect on perioperative renal function (47,48).

Despite the observations that CPB duration is independently associated with AKI-D, the benefits of off-pump coronary bypass (OPCAB) surgery on renal function remain controversial. An early retrospective analysis by Gamaso et al. (49) failed to show a difference in renal outcome between OPCAB and CPB surgery, whereas a similar retrospective study by Hayashida et al. (50) concluded that OPCAB surgery improved renal function as measured by creatinine clearance, but without a difference in the incidence of AKI-D when compared with the CPB group. Subsequently, Ascione et al. (51), in a small prospective, randomized controlled trial of 50 patients undergoing CABG, analyzed creatinine clearance, urinary microalbumin/creatinine ratio, and urinary N-acetyl-glucosaminidase activity, and concluded that OPCAB offered “superior” renal protection versus conventional CABG employing CPB. (It should be pointed out that patients with preoperative renal risk factors were excluded from the study and that no patient in either group developed AKI.) Several other small, prospective studies assessing levels of cystatin C, urinary by-products of oxidation, and other markers of inflammation have concluded that OPCAB is less likely to promote renal ischemia than on-pump procedures (52,53,54,55). However, the clinical relevance of these findings is lacking. Given that urinary microenzymes and other urinary or serum inflammatory mediators do not correlate with the development of AKI-D, larger prospective trials appear to be necessary to determine whether OPCAB surgery truly offers renal protection for patients needing CABG.

In addressing this issue, two recent prospective observational studies failed to show any difference in the incidence of AKI-D in patients undergoing OPCAB versus CPB, as did a recent meta-analysis (56,57,58). However, Bucerius et al. (59), in a very large prospective study of 9,631 patients (8,870 CPB and 761 OPCAB), found a 4.1% incidence of postoperative AKI-D in patients undergoing CPB versus 1.8% in patients undergoing OPCAB procedures. Patients in the OPCAB group who sustained AKI-D were likely to have longer perioperative length of stay (LOS), high transfusion requirement, and the need for urgent surgery. Although the benefits of OPCAB on renal function remain under continued study, implications of CPB-associated AKI continue to raise concern in regards to outcomes. In a study comparing propensity-matched patients undergoing OPCAB versus CPB, relative risk of death was greater in patients undergoing CPB. Hix et al. (34) discussed survival benefit of OPCAB having been a reflection of AKI risk reduction.

Thoraco-abdominal aortic surgery yields the greatest risk of developing AKI among the types of cardiac surgeries. Over 50% of patients undergoing thoracic aortic surgery exhibit AKI (60). Prospective studies describing the incidence of
AKI-D following thoracoabdominal aortic surgery ranges from 2.7% to 15% (61,62,63,64). In addition to direct renal hypoperfusion from suprarenal aortic cross-clamping, the pathophysiology of renal dysfunction from aortic cross-clamping and unclamping at any level appears to be related to an intricate relation between the renin-angiotensin system, the sympathetic nervous system, prostaglandin pathways, oxygen-free radicals, the complement cascade, and the release of cytokines and other inflammatory mediators (65,66). As with patients undergoing cardiac surgery, preoperative renal function and age contribute considerably to renal outcome, but a duration of more than 30 minutes of left renal ischemia and an intraoperative blood loss requiring more than 5 units of packed red blood cells are also predictive of developing AKI-D following aortic surgery (61,62,63

Stay updated, free articles. Join our Telegram channel

Join