Hospital-acquired acute kidney injury (AKI) presents a broad spectrum of clinical manifestations from minimal reductions in glomerular filtration rate to oligo-anuric AKI. It arises from multiple causes in a variety of clinical settings, and it is potentially reversible.
Hospital-acquired AKI is a common condition, and its incidence is rising in critically ill patients. There is solid evidence, that AKI is an independent poor prognostic marker, even under the condition of RRT, and not merely a reflection of the severity of the underlying disease. Even small increases in serum creatinine have important implications on both short- and long-term morbidity and mortality. AKI, severe enough to require RRT, occurs in 4–5 % of patients. This subgroup of AKI patients generally has a dismal prognosis with mortality rates of more than 40 %. Moreover, survivors develop chronic kidney disease (CKD), worsening pre-existing renal function impairment or need of chronic dialysis and have a 10-year survival rate of approximately 20 % [1–3].
Currently, there are no effective pharmacologic drugs that reverse the course of established AKI. The management of established AKI is supportive in nature, with RRT as the mainstay of treatment for renal failure. New approaches to current RRT failed to show an additional mortality benefit . If the rising prevalence of AKI in critically ill patients is most likely due to more aggressive diagnostic and therapeutic interventions in an aged patient population with a high burden of comorbid disorders, then prevention of AKI is of paramount importance to improve patient outcome.
Prevention of AKI by sodium bicarbonate: proposed mechanisms
The renoprotective mechanisms, by which sodium bicarbonate might prevent AKI beyond its volume expanding effects, are poorly defined. Reactive oxygen species-mediated tubular injury appears to play an important role in the pathogenesis of human AKI. A reduction in oxidative stress on renal tubular cells may be a key mechanism of sodium bicarbonate. Bicarbonate is able to slow the Haber–Weiss reaction that generates free radicals. Bicarbonate may also directly scavenge peroxynitrite as well as other reactive species generated from nitric oxide. Urinary alkalinization, as a direct effect of intravenous sodium bicarbonate administration, may reduce the pH-dependent generation of met-hemoglobin containing tubular casts, ferrous-ion catalyzed production of free radicals as wells as proteinuria oxidative damage. Thus, attenuation of oxidative stress through urine alkalinization with sodium bicarbonate may attenuate AKI [5, 6].
Experimental data have shown that higher tubular pH could be protective in the presence of hemoglobinuria and myoglobinuria, especially through inhibition of hydroxyl radical generation . Sodium bicarbonate administration helped prevent AKI in models of ischemic  or doxorubicin-induced AKI  or contrast-medium-induced nephropathy .
Sodium bicarbonate and human AKI: scientific evidence
Today, prophylactic sodium bicarbonate infusion has been used in four forms of human AKI: contrast-medium-induced nephropathy, cardiac surgery-associated AKI, pigment nephropathy and septic AKI.
Contrast-induced nephropathy (CIN)
Pathophysiology of CIN
Contrast-induced nephropathy is defined as an acute impairment of renal function and is measured as a 25 % increase of serum creatinine from baseline or a 0.5 mg/dl increase in absolute values, usually after 48–72 h of intravenous contrast administration. CIN is one of the leading causes of hospital-acquired AKI; its occurrence is associated with a significantly higher risk of in-hospital and 1-year mortality.
Contrast media (CM) act on distinct anatomic sites within the kidneys and exert adverse effects via multiple mechanisms. First, they cause a direct cytotoxic effect on the renal proximal tubular cells, enhance cellular damage by reactive oxygen species and increase resistance to renal blood flow. Second, they exacerbate renal vasoconstriction, especially important in patients with CKD, particularly in the deeper proportion of the outer medulla. CM-mediated vasoconstriction is the result of a direct action of CM on vascular smooth muscle and from vasoconstrictors such as adenosine and endothelin. Third, the osmotic property of CM in the tubular lumen decreases water reabsorption leading to buildup of interstitial pressure. This, along with the increased salt and water load to the distal tubules, reduces GFR and causes local compression of the vasa recta. All mechanisms contribute to worsening medullary hypoxemia and renal vasoconstriction in patients who are already volume depleted .
Prevention of contrast-induced AKI by sodium bicarbonate: clinical evidence
Despite a multitude of prospective controlled clinical trials (PCTs) and meta-analyses, the benefit of sodium bicarbonate over sodium chloride with or without N-acetyl-cysteine (NAC) remains unclear. The first trial of sodium bicarbonate versus saline in the prevention of CIN by Merten et al.  demonstrated a significantly lower rate of CIN in the sodium bicarbonate group and advocated its widespread use. Subsequent trials comparing sodium bicarbonate and sodium chloride found inconclusive results or indicated no additional effect, be it superiority or inferiority of sodium bicarbonate compared to hydration with isotonic saline [13–34]. However, the majority of these investigations had design limitations, such as enrollment of small number of patients, the use of surrogate primary endpoints defined by small increments in serum creatinine concentration, which are associated with, but not necessarily causally related to patient-centered outcomes. The inclusion of low-risk patients with intact baseline kidney function yielded low event rates (need for RRT, in-hospital mortality) and reduced generalizability. Moreover, hydration protocols were variable (dose and duration of sodium bicarbonate administration), and the use of NAC varied substantially between studies. Most, but not all studies used low osmolar or iso-osmolar CM. Clinical settings included predominantly elective or emergency coronary angiography in patients at low risk or high risk (diabetes, CKD).
Numerous early single-center clinical studies and meta-analyses showed a significant risk reduction of CIN by intravenous sodium bicarbonate compared to intravenous saline with or without NAC, even though there was no difference in the need for dialysis or in-hospital mortality [13, 14, 16–19]. However, the majority of the meta-analyses recognized significant heterogeneity and publication bias and clearly demonstrated that the benefit of sodium bicarbonate was limited to small-sized trials of low methodological quality. This subgroup of trials may have been underpowered to detect differences between groups . Generally spoken, early reports probably overestimated the magnitude of any benefit, whereas larger, later reported trials had neutral results . The meta-analysis conducted by Hoste et al.  found a protective effect of sodium bicarbonate with borderline significance on the risk of CIN, especially in patients who underwent emergency coronary procedures and in patients with CKD. However, sodium bicarbonate was not effective in a mix of patients who underwent coronary and non-coronary procedures.
The beneficial effects of sodium bicarbonate for the prevention of CIN have not been sustained in recent trials in patients with CKD undergoing percutaneous coronary or peripheral angiographic interventions. Most of the recent published studies [23–32] found no superiority of intravenous sodium bicarbonate. Amazingly, a multicenter prospective randomized study comparing the efficacy of prophylactic use of saline infusion versus sodium bicarbonate infusion for the prevention of CIN in patients with diabetes mellitus undergoing coronary angiography and/or intervention found that prophylactic hydration with isotonic saline before coronary procedures may reduce the incidence of CIN in diabetic patients not selected according to renal function to a greater extent than sodium bicarbonate . The latest published meta-analyses including 19 clinical trials (10 PCTs showing better CIN prevention with sodium bicarbonate, 9 PCTS with no difference in renal outcome) demonstrated that sodium bicarbonate-based hydration is superior to sodium chloride in patients undergoing exposure to low osmolar CM, but not in those receiving iso-osmolar agents .
The conclusions of recent meta-analyses were that the potential benefit, if any, of bicarbonate-based intravenous fluid administration over normal saline was likely to be very small in clinical practice.
Cardiac surgery-associated acute kidney injury (CSA-AKI)
Pathophysiology of CSA-AKI
Acute kidney injury develops in up to 30 % of patients undergoing cardiac surgery, with up to 3 % of patients requiring dialysis. CSA-AKI is independently associated with substantial short- and long-term morbidity and mortality.
The pathogenetic features of CSA-AKI are complex and multifactorial. Cardio-pulmonary bypass (CPB) contributes to CSA-AKI by activating a systemic inflammatory response altering regional blood flow and vasomotor tone in the kidneys. This CPB-induced systemic inflammatory response syndrome is the result of direct contact of blood cells with the artificial surfaces of the bypass circuit. CPB decreases the effective renal perfusion, thereby contributing to ischemia-reperfusion injury. Furthermore, small microemboli formed during CPB can damage renal capillaries directly. Hemolysis and release of free hemoglobin during CPB is a well-recognized nephrotoxic mechanism. Increased levels of free red blood cell constituents together with exhaustion of their scavengers result in a variety of serious sequelae such as increased vascular resistance, altered coagulation activity, platelet dysfunction and renal tubular damage .
CSA-AKI and sodium bicarbonate infusion
A double-blind randomized controlled pilot trial with 100 patients undergoing cardiac surgery found that intravenous sodium bicarbonate administration led to a significant reduction in postoperative AKI .
Contrary to the positive findings of this pilot study, a recent international multicenter double-blind PCT by the same group of investigators including 350 high-risk patients showed that more patients receiving sodium bicarbonate developed CSA-AKI (unadjusted p = 0.032). Based on these findings, the trial was terminated before planned recruitment was complete. However, a key limitation of this study was that a greater proportion of patients receiving sodium bicarbonate had CKD prior to surgery compared to those receiving saline. After multivariable adjustment, a non-significant unfavorable group difference affecting patients receiving intravenous bicarbonate was found for the primary end-point. A greater postoperative increase in urinary NGAL was observed compared with control patients. The incidence of postoperative AKI requiring RRT was similar, but in-hospital mortality was increased in patients receiving sodium bicarbonate compared with controls (6.3 vs. 1.7 %, p = 0.031). In the multicenter PCT, the authors used slightly larger doses of bicarbonate (5.1 vs. 4 mmol/kg during 24 h). Whether or not the difference in intravenous sodium bicarbonate dose used in these studies might have a negative impact on renal outcomes remains unknown .
Moreover, a phase II b multicenter double-blind PCT conducted in New Zealand and Australia enrolled 427 patients at risk for CAS-AKI scheduled to undergo elective cardiac surgery. Preoperative alkalinization of blood and urine using sodium bicarbonate infusions did not result in significant different in CAS-AKI (within 5 days), in the need for RRT, in the hours on ventilation, ICU or hospital length of stay, or mortality . The study was stopped prematurely before full enrollment (based on preliminary results communicated to the investigators from another concomitant similar trial). Moreover, a high proportion of the study patients underwent valvular surgery either alone or combined with coronary artery bypass grafting. These patient groups could be expected to have more microembolic events and greater hemolysis.
Finally, Heringlake et al.  reported the absence of superiority of intravenous bicarbonate for the prevention of CSA-AKI in comparison with saline. These authors retrospectively compared a cohort of 280 patients who received 4 mmol sodium bicarbonate per kilogram to a control cohort. The proportion of patients experiencing an increase in serum creatinine by 25 % was similar in both groups (43.1 % sodium bicarbonate vs. 32.9 % control group, difference not significant). This cohort study had the advantage of including almost all patients admitted to their center, which increases external validity of the results.
The results of these studies provide most current and best available evidence to suggest that preoperative sodium bicarbonate does not reliably reduce AKI following cardiac surgery in patients at high risk.
Heme pigment nephropathy
Pigment nephropathy implies AKI due to the nephrotoxic effect of the heme pigments myoglobin, hemoglobin and bilirubin.
The pathogenesis of heme pigment nephropathy is complex, and due to a combination of multiple factors, and facilitated by hypovolemia and acidosis. The main mechanisms are renal vasoconstriction, renal cytokine production, direct cytotoxic effects of heme proteins and formation of obstructive casts with Tamm-Horsfall protein .
Pigment nephropathy and sodium bicarbonate
Aggressive hydration is effective to prevent or to treat pigment nephropathy. Currently, it is not known whether or not alkalinization of the urine is beneficial once a brisk diuresis with saline or diuretics is established. Theoretically, alkalinization of the urine has been advocated to increase the solubility of heme proteins in tubular fluid, with the goal of achieving a urine pH of more than 6.5. Alkalinization may also reduce the production of reactive oxygen species, thus reducing the oxidant stress. These proposed effects of sodium bicarbonate infusion are exclusively generated from animal studies [7, 41].
While many articles on rhabdomyolysis treatment focus on management of established AKI, only two systematic reviews have been published focusing on the therapies used to prevent AKI in patients with rhabdomyolysis. The European Renal Best Practice and Renal Disaster Relief Task Force addressed the prevention of AKI in crush victims following mass disaster . Sharman and Troutman conducted a systematic literature review to evaluate evidence-based recommendations for the prevention of rhabdomyolysis-associated AKI . Twenty-seven studies met the inclusion criteria. With the exception of seven level 2 b studies (low quality-randomized clinical trials, cohort studies), most studies were level 4 (case series, poor quality cohort or case control studies).
There was no standardized definition of the syndrome, generally accepted diagnosis and grading system. The number of participating patients was mostly small to very small. Ten of the 27 articles did not provide information on the type of fluid used. There were no controlled studies comparing fluid administration plus sodium bicarbonate (without the use of diuretics) to fluid administration. One level 2 b and two level 4 studies concluded that there was no difference in the rates of development of AKI between patients receiving sodium bicarbonate or not receiving sodium bicarbonate. However, documentation of urinary pH was not provided. Moreover, with the exception of one report (number of patients 382), the populations were very small (24 and 10, respectively). Therefore, these studies might not have been adequately powered to detect differences between patient groups. Of clinical importance, in one level 2 b study, a greater percentage of patients receiving bicarbonate developed AKI compared with those not receiving bicarbonate. The authors of the systematic review recommend that intravenous sodium bicarbonate should be administered only to correct systemic acidosis, and the amount of and rate of sodium bicarbonate should be individualized based on the patients` size and medical condition.
Prevention of septic AKI by sodium bicarbonate: clinical evidence
Administration of sodium bicarbonate to patients at risk for septic AKI was associated with frequent major electrolyte abnormalities compared to sodium chloride administration and early protocol cessation . Twenty-five patients with SIRS, oliguria and elevated serum neutrophil gelatinase-associated lipocalin (sNGAL) received either intravenous sodium bicarbonate (13 patients) or sodium chloride (12 patients). Adverse events (serum bicarbonate greater than 30 mmol/L in the sodium bicarbonate group, serum chloride greater than 111 mmol/L) occurred in 7 patients. There was no difference between the two groups in occurrence of AKI, requirement of RRT, hospital length of stay or mortality.
Prevention of AKI by sodium bicarbonate infusion: harms
Retrospective cohort analyses suggested that intravenous sodium bicarbonate was associated with an increased risk of CIN  or pigment nephropathy [44, 45]. The multicenter trial of prophylactic perioperative sodium bicarbonate administration to prevent AKI following open heart surgery found a non-significant increase in AKI and a possible increase in in-hospital mortality . Whether or not some patients as those with clinically latent congestive heart failure and/or left ventricular dysfunction, may have had adverse cardiac events, if they were given large volumes of sodium bicarbonate inducing volume overload, hypokalemia or alkalemia, remains unknown. However, cardiac surgery patients receiving sodium bicarbonate had lower mean arterial blood pressure values after induction of anesthesia, and they required more fluids and vasopressors and spent more time in hospital .