Intravenous Albumin for Mitigating Hypotension and Augmenting Ultrafiltration during Kidney Replacement Therapy

Abstract

Among its many functions, owing to its oversized effect on colloid oncotic pressure, intravascular albumin helps preserve the effective circulatory volume. Hypoalbuminemia is common in hospitalized patients and is found especially frequently in patients who require KRT either for AKI or as maintenance hemodialysis. In such patients, hypoalbuminemia is strongly associated with morbidity, intradialytic hypotension, and mortality. Intravenous albumin may be administered in an effort to prevent or treat hypotension or to augment fluid removal, but this practice is controversial. Theoretically, intravenous albumin administration might prevent or treat hypotension by promoting plasma refilling in response to ultrafiltration. However, clinical trials have demonstrated that albumin administration is not nearly as effective a volume expander as might be assumed according to its oncotic properties. Although intravenous albumin is generally considered to be safe, it is also very expensive. In addition, there are potential risks to using it to prevent or treat intradialytic hypotension. Some recent studies have suggested that hyperoncotic albumin solutions may precipitate or worsen AKI in patients with sepsis or shock; however, the overall evidence supporting this effect is weak. In this review, we explore the theoretical benefits and risks of using intravenous albumin to mitigate intradialytic hypotension and/or enhance ultrafiltration and summarize the current evidence relating to this practice. This includes studies relevant to its use in patients on maintenance hemodialysis and critically ill patients with AKI who require KRT in the intensive care unit. Despite evidence of its frequent use and high costs, at present, there are minimal data that support the routine use of intravenous albumin during KRT. As such, adequately powered trials to evaluate the efficacy of intravenous albumin in this setting are clearly needed.

Introduction

Albumin is an important 66-kD protein that is synthesized solely in the liver and serves many roles within the human body (1,2). The metabolism of albumin is summarized in Figure 1. Although accounting for only approximately half of the plasma protein content, albumin is responsible for generating an even greater proportion (approximately 70%–80%) of intravascular colloid oncotic pressure due to its capacity to attract water (1–3). Beyond that, endogenous albumin serves many other important functions in the human body. These include maintenance of the redox state as well as molecular and drug transport (4,5). Given albumin’s many functions and that its synthesis is dependent on adequate dietary protein intake, it is not surprising that poor outcomes are associated with hypoalbuminemia across a range of medical conditions (6,7). The corollary to this is that administering intravenous albumin could theoretically be beneficial to patients who are hypoalbuminemic or hypotensive due to intravascular volume depletion. To that end, exogenous human albumin products have been used as therapeutic agents since their introduction into the medical field in the 1940s (8).

Figure 1.

Metabolism of albumin and causes of hypoalbuminemia. Overview of albumin metabolism and selected mechanisms for the development of hypovolemia. Not shown is the normal breakdown of albumin, which has a t _1/2 of approximately 3 weeks. Hypoalbuminemia results from some combination of decreased production, increased losses, acute dilution (not shown), or increased shift out of the vascular space. *Increased skin and gastrointestinal losses typically occur in the context of concurrent, increased capillary leak.

Management of hypotension specifically during KRT (i.e., intradialytic hypotension [IDH]) can be challenging in both critically ill patients on KRT and in patients on maintenance hemodialysis (HD). Prompt identification and reversal of IDH during treatment are important, not only for symptomatic relief but also to mitigate the potential consequences of hypoperfusion (9). Both hyperoncotic (20%–25%) and iso-oncotic (4%–5%) intravenous albumin have been used in place of saline for volume resuscitation in this setting, but this practice is controversial. In the context of hemodynamic instability due to intravascular hypovolemia, hyperoncotic albumin can theoretically be expected to shift water into the vasculature, which may in turn improve hemodynamic stability and then permit greater overall fluid removal via ultrafiltration. There are several other properties of albumin that could theoretically be helpful in the settings of dialysis-requiring AKI and maintenance HD, such as its anti-inflammatory action, antioxidant properties, and role in promoting microvascular circulation (10–12). Nonetheless, not all IDH episodes are due to hypovolemia (13,14), and few studies have specifically addressed the role of intravenous albumin in preventing or treating IDH. The objective of this narrative review is to highlight knowledge gaps in the use of intravenous albumin to prevent or treat IDH and/or facilitate ultrafiltration. In this article, we will summarize the physiology of albumin, explore the implications of hypoalbuminemia, and discuss the potential risks and benefits of intravenous albumin use for these indications.

Incidence and Implications of Hypoalbuminemia

The etiology of hypoalbuminemia in most patients is likely multifactorial. As detailed in Figure 1, several mechanisms are implicated, including increased capillary permeability, exudative losses, breakdown, occult protein wasting enteropathy in critical illness and decreased liver production (4,15). In patients with AKI, the extent to which proteinuria due to tubular damage contributes to hypoalbuminemia is unclear.

Marked hypoalbuminemia (when defined as serum albumin <25 g/L) correlates with a 34% in-hospital mortality rate compared with a 2% mortality rate for hospitalized patients with normal albumin levels of 35–45 g/L (6).The prevalence of hypoalbuminemia in patients on maintenance HD is even greater than in hospitalized patients, reported to be as high as 33% in a study that defined hypoalbuminemia as serum albumin <35 g/L (16). In both patients on incident HD and patients on maintenance HD, hypoalbuminemia is strongly associated with morbidity and mortality (16–18). To a large extent, the correlation between hypoalbuminemia and worse outcomes may simply be due to hypoalbuminemia serving as a marker for poor nutrition and/or inflammation in chronic or severe illness. Notably, the correlation between hypoalbuminemia and increased mortality in patients on maintenance HD is markedly less in those with normal C-reactive protein levels (19). Similarly, hypoalbuminemia is independently associated with IDH (20,21), theoretically due to its effect on oncotic pressure in the circulation; however, residual confounding on the basis of concurrent chronic or severe illness is very likely.

Physiology Relevant to Intravenous Albumin Administration for Intradialytic Hypotension

Endogenous albumin is distributed between the intravascular (40%) and extravascular (60%) space, where it is the most abundant protein in both compartments (3). Serum albumin is responsible for approximately 80% of the colloid oncotic pressure within the vasculature (2). Regarding intravenous albumin, 5% albumin is iso-oncotic with human plasma, whereas 25% albumin is approximately five times more osmotically active.

From a physiologic standpoint, the oncotic action of albumin infusion, which is calculated to attract a total of 18 ml of water per 1 g of protein, would be expected to expand plasma volume to a greater degree than the same volume of normal saline (2). Theoretically, this would be helpful in a setting where a physician desires to minimize volume overload, retaining fluid intravascularly to reduce third spacing. On the basis of its oncotic properties, one would expect that intravenous albumin administration would expand intravascular plasma by a volume that is several times greater than crystalloid alone (3). In intensive care unit (ICU) studies involving critically ill patients, this ratio was found to be much lower than expected, with only approximately 1.4 L saline necessary to result in an equivalent volume of intravascular expansion as 1 L of human albumin product (22,23). Notably, in a trial involving 321 critically ill patients requiring fluid resuscitation, the cumulative volume of resuscitation fluid was reduced by 600 ml over 48 hours through the use of 20% versus 4%–5% albumin fluids, indicating that volume effects depend on the albumin concentration applied (24).

Despite classic teaching, recent studies indicate that fluid movement across the capillary bed into the interstitial space and subsequent reabsorption are not influenced by Starling forces, such as oncotic pressure, in the same manner as previously thought (25). This issue and the role of the endothelial glycocalyx in regulating albumin distribution are detailed in Figure 2 and in Supplemental Material. Supplemental Material also includes additional information regarding the antioxidant and anti-inflammatory roles of albumin in patients with dialysis-requiring AKI or those on maintenance HD.

Figure 2.

Albumin movement across the capillary endothelium in health and disease. (A) In the historical model of capillary fluid exchange on the basis of Starlings forces, net plasma filtration was believed to occur at the arteriolar end of the capillary due to predominant hydrostatic force (blue arrows), which then transitioned to net reabsorption as plasma reached the venular end due to increasing oncotic drive from the interstitium, back into the capillary lumen (green arrows) (additional details are in Supplemental Material). Some research now suggests that there is, in fact, little to no reabsorption at the level of the capillary in the steady state (but it still may occur during acute hypovolemia). (B and C) A simplified version of the endothelial glycocalyx in its role as a selectively permeable barrier for fluid resorption. Hydrostatic pressure is the predominant force in this new model regulated by the glycocalyx, with little to no effect from osmotic forces between the vascular lumen and the interstitium. Instead, the endothelial glycocalyx functions to create an osmotic gradient (orange arrow) between the capillary lumen and the subglycocalyx space that opposes outward shift of fluid but does not reverse the direction of fluid flow. Both the interstitium and the subglycocalyx space have low oncotic content, resulting in negligible oncotic forces between the two spaces and overall net ultrafiltration throughout plasma transit in capillary microcirculation. Fluid that has shifted into the interstitium is resorbed solely through the lymphatic system, with exceptions (including the kidneys). (D) Fluid shift in the setting of a damaged endothelial glycocalyx from stressors, including inflammation, uremia, arteriosclerosis, hyperglycemia, and hypovolemia, among many others. With a damaged endothelial glycocalyx, capillaries have increased permeability to both serum proteins and fluid, leading to increased net fluid movement into the interstitium and interstitial edema when the lymphatic drainage capacity is exceeded.

Albumin Usage for Intradialytic Hypotension

Currently, there are few generally agreed upon indications for intravenous albumin (3). These indications include spontaneous bacterial peritonitis, large volume paracentesis, and hepatorenal syndrome (3). Despite the apparent effectiveness of albumin in the setting of end stage liver disease, it is unclear the extent to which that hinges on pharmacologic cointerventions or other disease-specific factors. Newly reported but not yet published data from a randomized controlled trial (n=828) that assessed albumin normalization using 20% intravenous albumin in hospitalized patients with cirrhosis and serum albumin <30 g/L indicates that this strategy did not significantly improve 28-day mortality. Furthermore, the utility of intravenous albumin products is widely debated for primary volume expansion and specific to nephrology, nephrotic syndrome, enhancement of ultrafiltration/prevention of IDH, or treatment of refractory IDH.

IDH is multifactorial (i.e., not always due to volume depletion) (14), and thus, fluid administration often may not be the most appropriate response. The initial clinical response to IDH is usually to place the patient in Trendelenburg position, pause ultrafiltration, or stop treatment altogether, prior to administering any type of fluid (9,15,26). In a study that included 2559 inpatient HD sessions, of the 433 sessions complicated by IDH, a protocolized approach that used albumin administration as the last line of treatment resulted in reversal of hypotension for 91% of patients who did not receive albumin and in another 2% overall who did get it (27). Currently, normal saline is generally advised as first line if fluid resuscitation is required to restore BP in this setting (28). The potential benefits and risks of using intravenous albumin to enhance ultrafiltration and/or prevent or treat IDH are detailed in Table 1. One unique consideration of the use of albumin in this setting is that its use runs counter to the usual overall goal of net fluid removal with KRT. This is especially the case when considering the use of albumin for prevention of IDH (or to facilitate overall fluid removal) where the alternative would be to not give any fluid. One must also consider that there is an accompanying sodium load as albumin formulations typically have a sodium concentration similar to that of normal saline (as detailed in Table 2). This is a particularly relevant concern given that worse outcomes are found in both critically ill patients and patients on maintenance HD with more fluid overload (29–32). The basis for intravenous albumin use in the setting of KRT (including maintenance HD) is generally extrapolated from the results of studies in critical care but, as further discussed below, is not backed by high-quality trials designed specifically to assess its effect on KRT-related outcomes. Nonetheless, recent data from three large academic medical centers in Ontario, Canada provide ancillary evidence that albumin is frequently used for patients for the prevention or treatment of IDH: approximately 30% of all inpatients who received any form of KRT while hospitalized were also administered albumin (25%) on at least 1 day that they underwent KRT treatment (33).

Table 1.

Potential benefits and risks of intravenous albumin for patients on KRT

Benefits/Risks	Related Evidence	Comments
Potential benefits
Reduced hypotension, enhanced ultrafiltration	No data support use of iso-oncotic albumin. Hyperoncotic albumin trials report hemodynamic benefits (Table 3)	<150 patients included across all trials
Anti-inflammatory effects	Patients on maintenance treated with 20% albumin have reduced markers of inflammation (Supplemental Table 1)	Potential effect on patient outcomes is unclear
Potential risks
Unnecessary fluid accumulation	Observational data: less net fluid removal reported for inpatient HD sessions when 25% albumin was given	This finding likely confounded according to indication for albumin use
AKI due to hyperoncotic albumin administration	Finding of recent observational studies but hyperoncotic albumin never found to cause AKI across multiple trials	Implications for kidney recovery in AKI, preservation of residual function in HD. Trial data are very reassuring
Aluminum toxicity	Historically, trace metals were introduced during manufacturing	Very unlikely given manufacturing improvements
Infection	Theoretical risk of transmission of viruses, prions	No cases have been reported
Other	Anaphylaxis and hypotension have been reported	Extremely rare relative to other blood products

HD, hemodialysis.

Table 2.

Fluids used in the treatment or prevention of intradialytic hypotension

Fluid	Sodium Content, mEq/L	Protein Content, g/L	pH	Cost per 100 ml, US Dollars
Iso-oncotic albumin a	140–145	4–5	7	15.91
Hyperoncotic albumin a	140–145	20–25	7	108.75
Normal saline	154	0	5.5	0.10
Hydroxyethyl starch	Contraindicated in AKI and maintenance HD

HD, hemodialysis.

^aAverage values are reported for albumin products (57). Specific manufacturer-reported details for albumin brands available in the United States and Canada are reported in Supplemental Table 1.

Safety of Intravenous Albumin Administration

The safety and efficacy of intravenous albumin administration have been scrutinized heavily in the past due to its cost in comparison with crystalloid solutions and concerns raised by a flawed 1998 metanalysis (34) that included studies done prior to significant quality improvements in the production of albumin (35). Table 1 includes additional details regarding albumin safety. The issue of intravenous iso-osmolar albumin safety, specifically in the setting of fluid resuscitation, was largely resolved with the publication of the Saline versus Albumin Fluid Evaluation (SAFE) study in 2004 (36). A heterogenous group of 6997 patients in the ICU were randomized to receive either 4% albumin or normal saline for intravascular fluid resuscitation, with the primary outcome being all-cause mortality. There was no significant difference in death at 28 days between the two groups. Furthermore, there was no difference in mechanical ventilation, need for KRT, or length of ICU stay. Subgroup and post hoc analyses of 460 patients suggested possible increased risk of death with albumin administration in patients with traumatic brain injury (Relative Risk, 1.63; 95% confidence interval [95% CI], 1.17 to 2.26; P=0.003), as well as increased intracranial pressure in those with intracranial pressure monitoring (36–38).

Some recent studies have suggested a risk of AKI associated with the use of hyperoncotic albumin. A cohort study of 984 patients in the postoperative cardiac surgery period found a dose-dependent increase in AKI with hyperoncotic albumin administration according to a propensity score analysis (39). Similarly, a retrospective cohort study of 11,512 patients in postoperative shock, the majority of whom had undergone cardiac surgery, found a greater risk of AKI (odds ratio [OR], 1.10; 95% CI, 1.04 to 1.17; P=0.002) in those exposed to hyperoncotic albumin (40). Generally, observational studies that have assessed the relationship between hyperoncotic albumin and AKI are significantly limited by the potential for residual confounding according to the indication for hyperoncotic albumin administration. In contrast to the observational data reviewed above, a recent randomized control trial of 220 patients undergoing off-pump coronary artery bypass surgery revealed that preoperative treatment with 20% hyperoncotic albumin for patients with albumin <40 g/L was associated with significantly less postoperative AKI than occurred in patients given the same volume of normal saline (13.7% versus 25.7%, respectively; P=0.05) (41). Furthermore, a 2010 metanalysis evaluating the use of seven trials utilizing hyperoncotic albumin found that albumin was protective against kidney insult (OR, 0.24; 95% CI, 0.12 to 0.48; P<0.001) and mortality (OR, 0.52; 95% CI, 0.28 to 0.95; P=0.05), but the indication for albumin administration in the included studies was quite variable, with many patients with liver cirrhosis included (42). The potential occurrence of renal tubular damage in response to administration of other hyperosmolar fluids is well established (43); however, albumin is not an exogenous colloid and was not found to have accumulated in early kidney autopsy studies of a small number of patients who had received very high volumes of 25% albumin (44). More broadly, a recent network metanalysis including ten studies (6664 patients) with direct comparison could not establish an increased risk for KRT when comparing albumin with crystalloids with moderate certainty (although this included studies of iso-oncotic albumin) (45). In a trial of patients with severe sepsis (n=1818), those randomized to 20% albumin plus crystalloid (targeting albumin >30 g/L) were not at increased risk of severe AKI as a predefined secondary outcome (serum creatinine >300 μmol/L) compared with those randomized to crystalloid alone (21). A post hoc analysis also found no difference between the groups when AKI was defined according to RIFLE criteria.

On balance, we conclude that hyperoncotic intravenous albumin is very unlikely to be directly harmful to the kidneys. This is reassuring with respect to the potential effect on kidney recovery in albumin-treated patients with AKI and preservation of residual kidney function in those on maintenance HD.

Effectiveness of Intravenous Albumin for Patients Receiving Kidney Replacement Therapy

A recent large observational study assessed the use of intravenous 25% albumin for improving ultrafiltration in 238 inpatients (973 HD sessions) at a single large US center (46). Contrary to what might be expected, albumin use (compared with no albumin use) was associated with less ultrafiltration achieved (1242 versus 1899 ml, respectively; P<0.001). Despite adjustment, this result may have been confounded by additional factors related to the indication for which albumin was given, as this study sought to specifically assess instances when albumin had been given to enhance ultrafiltration.

As summarized in Table 3, clinical trials that have specifically assessed the use of intravenous albumin for mitigating hypotension and/or augmenting ultrafiltration (47–51) have generally been small (with all having been conducted at single centers) and are heterogeneous with respect to the intervention (iso- or hyperoncotic albumin, volume, and timing/indications for administration) and outcomes assessed. In the largest trial involving maintenance HD, Knoll et al. (48) sought to assess the use of iso-oncotic (5%) albumin infusion versus normal saline for the acute management of IDH during maintenance HD. Participants were randomized to one of two possible sequences in a crossover study design following three dialysis sessions affected by IDH in a given patient. Individuals in one sequence were treated with 5% albumin infusion for the first dialysis session complicated by IDH, followed by the use of normal saline for the remaining two dialysis sessions where IDH occurred. In the other sequence, treatment was reversed, and patients were treated with normal saline during the first dialysis session complicated by IDH and subsequently, 5% albumin in the remaining sessions. Patients were placed in Trendelenberg position, and ultrafiltration was stopped prior to the administration of study fluid; however, the number of patients whose hypotension resolved prior to any fluid administration was not reported. The primary outcome of the study was percentage of target ultrafiltration achieved on the basis of pre- and postdialysis weights. Approximately 84% and 80% ultrafiltration values were achieved in the albumin and normal saline treatment groups, respectively, with no significant difference detected between the two groups (P=0.14). Secondary outcomes, including postdialysis BP, volume of fluid used in the resuscitation effort, time required to restore BP, or frequency of recurrent hypotensive episodes, were also not significantly different.

Table 3.

Summary of trials assessing intravenous albumin for intradialytic hypotension

Author [Year]	Study Design a	Study Size	Population	Intervention	Main Results	Study Conclusions
Macedo et al. b	Randomized, crossover	n=65 (249 HD sessions)	Inpatients (maintenance HD or for AKI) with serum albumin <30 g/L	25% albumin (100 ml) versus NS (100 ml) prior to start of HD session	Albumin resulted in significantly less IDH (defined according to SBP decrease ≥30 or ≥20 mm Hg or nadir SBP <90 mm Hg	Pretreatment with 25% albumin resulted in less frequent IDH
Rostoker et al. [2011] (49)	Randomized, single-blind, crossover study	n=10 (HD sessions not specified)	Outpatients on maintenance HD “prone to IDH”	20% albumin (200 ml) versus 4% gelatin (200 ml) given throughout HD	Fewer episodes of SBP<100 mm Hg in six patients (60%) with albumin	20% albumin improved hemodynamic parameters
Knoll et al. [2004] (48)	Randomized, blinded, crossover study	n=45 (135 HD sessions)	Outpatients on maintenance HD	5% albumin (250 ml) versus NS (250-ml) boluses (up to 3× per session) for symptomatic IDH	No significant differences in primary outcome (percentage of target UF achieved) or other hemodynamic measures	5% albumin was not superior to NS for treatment of symptomatic IDH
van der Sande et al. [2000] (50)	Randomized, nonblinded, crossover study	n=9 (27 HD sessions)	Outpatients on maintenance HD with recurrent IDH and CHF	20% albumin (100 ml) versus 3% saline (33 ml) versus 10% HES (100 ml) for BP drop	20% albumin (and HES) sessions: lower drop in BV and lower drop in end treatment SBP versus 3% saline	20% albumin was superior to 3% saline for preventing IDH
van der Sande et al. [1999] (51)	Randomized, nonblinded, crossover study	n=10 (30 HD sessions)	Outpatients on maintenance HD; CHF excluded	20% albumin (100 ml) versus NS (100 ml) versus 10% HES (100 ml) for 10% BV decline	Albumin and HES both maintained a significantly smaller change in BV than NS (P=0.05). No significant differences in SBP	20% albumin was superior to NS for preserving BV during HD
Jardin et al. [1982] (47)	Nonblinded, crossover study	n=8 (53 HD sessions)	Inpatients with sepsis and anuric AKI	HD circuit primed with either 300 ml 17.5% albumin or 300 ml NS	Albumin sessions: more UF achieved and higher MAP during treatment versus NS	Priming HD circuit with 17.5% albumin limited IDH, facilitated ultrafiltration

n, number of patients; HD, hemodialysis; NS, normal saline; IDH, intradialytic hypotension; SBP, systolic BP; UF, ultrafiltration; CHF, congestive heart failure; HES, hydroxyethyl starch; BV, blood volume; MAP, mean arterial pressure.

^aAll single-center studies.

^bMacedo E, Karl BE, Jacinto LD, Lee E, Mehta RL: A randomized controlled trial of albumin versus saline for the prevention of intradialytic hypotension in hypoalbuminemic patients [Abstract]. J Am Soc Nephrol, 29: 97, 2018

When the Cochrane Group performed a meta-analysis in 2010 to evaluate the use of albumin for IDH in maintenance HD, the study by Knoll et al. (48) in 2004 was the only study to meet inclusion criteria for the analysis (28). No randomized control trials or other randomized crossover studies of sufficient quality were identified, and they concluded that the evidence to support albumin use over normal saline in the setting of IDH was insufficient (28).

In theory, the use of hyperoncotic albumin has a stronger theoretical foundation for use in KRT than iso-oncotic albumin given its potential to result in greater fluid shifts from the interstitium to the intravascular space. In general, the small studies that assessed the use of hyperoncotic albumin did suggest that it led to significant hemodynamic improvements (47,49–51) (see Table 3). The largest trial that assessed the effect of hyperoncotic albumin for prevention (rather than treatment) of IDH included 65 patients starting HD for either AKI or as maintenance HD who were randomized to 100 ml of normal saline versus 25% albumin prior to HD and then alternated between the two fluids for up to six subsequent HD sessions (249 sessions were included). According to some of the IDH definitions assessed (including drop in systolic BP to minimum BP during treatment of >20 mm Hg or nadir systolic BP during treatment <90 mm Hg), albumin administration resulted in significantly fewer episodes of IDH.

There is almost no evidence regarding administration of intravenous albumin for IDH (hemodynamic instability during KRT [14]) in critically ill patients. The only trial in critically ill patients was a study that included eight patients who were septic (47) (Table 3). A 2018 systematic review of interventions to prevent hemodynamic instability during KRT in critically ill patients did not uncover any new studies assessing the use of albumin for IDH (52). The saline versus albumin for extracorporeal fluid removal in slow low-efficiency dialysis pilot study (53) recently completed enrollment of 60 critically ill patients randomized to receive 100 ml of 25% albumin versus normal saline twice during slow low-efficiency dialysis treatments; however, the results have not yet been reported.

Costs of Intravenous Albumin Treatment

Establishing evidence-based indications for intravenous albumin is important given the potential clinical and societal implications, including the high cost of intravenous albumin preparations (Table 2) relative to crystalloid fluids (or relative to not giving anything, if considering its use for prevention of IDH/augmenting ultrafiltration). Analysis over a 3-month period at a single center revealed that the cost associated with albumin use for nonevidence-based indications was US $48,900 (54). In the SAFE trial (described in more detail above), the difference in the cost of treatment was 75-fold higher for patients randomized to 4% albumin versus crystalloid (36), highlighting the need to establish whether the theoretical benefits of albumin use translate into real-world benefits with respect to patient outcomes. One review from 2004 reported that the base cost of a single liter of 5% albumin was $232 compared with <$2.32 for the same volume of normal saline (55). A study published in 2020 estimated that the median cost per patient in the ICU for 25% albumin is $417 (56). Another recently published study evaluated the costs of resuscitation fluids internationally and reported that, after taking into account the greater volume required to achieve an equally effective volume challenge, the cost of normal saline was approximately 27-fold less than for albumin (57). In our opinion, albumin should be thought of as an expensive medication whose routine use should be restricted to evidence-based indications, which do not currently include the prevention or treatment of IDH or facilitation of ultrafiltration during KRT.

Conclusion

More rigorous studies are needed to determine whether there is a role for the routine use of intravenous albumin during KRT. It is possible that hyperoncotic intravenous albumin use during KRT can improve patient outcomes by minimizing fluid overload and limiting cardiac, kidney, and other organ damage potentiated by IDH. Studies to definitively establish if this is true are not only needed, but also, they are readily justifiable on the following basis: there are well-grounded theoretical benefits, and, although generally safe, there are also potential risks. Beyond that, intravenous albumin is very expensive, and there is good evidence that this intervention, tested in fewer than 150 patients in all trials combined, is likely already being given to tens of thousands of patients for these indications on an annual basis. If shown to be ineffective in improving clinically meaningful outcomes, significant resources could be saved, but if shown to be effective, intravenous albumin is widely available, and its use could immediately be more broadly implemented.

Disclosures

J. Callum reports employment at Sunnybrook Health Sciences Centre and has received program support funding from Canadian Blood Services and a clinical trial grant from Octapharma. E. Clark reports receiving research salary support from the University of Ottawa, outside the submitted work. S. Hiremath reports receiving research salary support from the University of Ottawa and serving on the boards for Hypertension Canada and NephJC, outside the submitted work. M. Joannidis reports receiving grants and personal fees from Baxter Healthcare, grants from Fresenius Kabi, and speakers’ fees from CLS Behring, outside the submitted work. All remaining authors have nothing to disclose.

Funding

None.

Supplemental Material

This article contains the following supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.09670620/-/DCSupplemental.

Supplemental Material. Additional details regarding albumin physiology in patients on KRT.

Supplemental Table 1. Albumin formulations available in the United States and Canada.