Albumin level and long-term treatment

A post hoc analysis of the ANSWER study investigated whether serum albumin levels during treatment could predict patient outcomes.¹

What was the rationale?

Decompensated cirrhosis has a median survival time of approximately 2 years and imposes a large economic burden.^2,3

Ascites is the most common cause of decompensated cirrhosis and is associated with reduced quality of life, increased susceptibility to infections and increased mortality^4,5Albumin is used to expand plasma volume and prevent or improve complications of decompensated cirrhosis with ascites.⁴ Albumin also exerts non-oncotic properties that may prevent other complications of decompensated cirrhosis.⁶ However, it has been unclear whether long-term albumin treatment for people with ascites provides a survival benefit, and two randomised controlled trials that sought to address this question – the open-label ANSWER study and the placebo-controlled MACHT study – produced conflicting results.^6,7

In the ANSWER study, albumin treatment significantly improved 18-month overall survival and management of ascites, whereas no difference was seen between treatment and placebo groups in the MACHT study.^6,7 The ANSWER study investigators suggested that a key difference between the two studies that may have influenced the outcomes was the different dose of albumin used and subsequent difference in serum albumin levels.¹ Therefore, a post hoc analysis of the ANSWER trial was undertaken to investigate whether on-treatment serum albumin levels are associated with overall survival and incidence of complications, and whether this can be used to guide long-term treatment.¹

What were the methods?

Study population

In the ANSWER study, 431 people with cirrhosis and uncomplicated ascites were randomly assigned to receive standard medical treatment (SMT) or SMT plus 40 g human albumin (HA) twice weekly for the initial 2 weeks, then 40 g weekly for up to 18 months.⁶ All 431 participants from the intention-to-treat (ITT) population in the ANSWER study were included in this post hoc analysis (SMT group, n=213; SMT plus HA group, n=218).

Analysis

Serum albumin was determined at baseline and 1 month after treatment. This timepoint was chosen because a significant increase in serum albumin occurred at 1 month in the ANSWER study and remained stable thereafter.⁶ In addition, a prognostic marker at 1 month would allow for early intervention to adjust treatment.

Survival and complications were assessed 1–18 months after randomisation. Secondary endpoints of the ANSWER study were also included in the post hoc analysis.

Logistic regressions, Kaplan–Meier estimation and Cox regression were used in the statistical analyses.

What were the results?

Baseline serum albumin

• Baseline serum albumin was 3.10 ± 0.49 g/dL in the SMT group and 3.09 ± 0.55 g/dL in the SMT + HA group (P=0.86)
• In the SMT group, baseline serum albumin was associated with 18-month mortality, with mortality progressively increasing with every decrease of 0.1 g/dL (HR, 4.42; 95% CI, 2.34–8.33; P<0.001)
• The risk of death was more than halved in the SMT + HA group, and no association between baseline serum albumin and 18-month mortality was seen (HR, 1.85; 95% CI, 0.99–3.49; P=0.055)

Albumin administration

• At 1 month of treatment, serum albumin increased significantly to 3.83 ± 0.60 g/dL (n=193; P<0.001) in the SMT + HA group
• Normal serum albumin (≥3.5 g/dL) at 1 month was observed in 79.3% of people (n=153) in the SMT + HA group

Relationship between 1-month on-treatment serum albumin and clinical outcomes

• Cox regression analysis of factors associated with 18-month mortality showed that baseline serum albumin and model for end-stage liver disease (MELD) score incorporating serum sodium concentration (MELD-Na) were the only independent predictors of mortality
• Second-order polynomial regression showed an association between increased serum albumin at 1 month of treatment and improved overall survival
• Kaplan–Meier estimates of 18-month overall survival for different levels of serum albumin at 1 month of treatment identified cut-off values of 2.5–4.5 g/dL associated with improved overall survival
• A value of 4.0 g/dL was associated with an 80% reduction in mortality, which was significantly greater than that seen in people with values <4.0 g/dL (HR, 0.20; 95% CI, 0.08–0.52; P<0.001)
• Compared with those in the SMT group, participants in the SMT + HA group whose serum albumin levels were below the lower normal limit (<3.5 g/dL) had a higher probability of overall survival at 18 months (61.3% SMT + HA versus 36.7% SMT; P=0.032) with a 57% reduction in mortality (HR, 0.43; 95% CI, 0.19–0.95; P=0.036)

What are the implications of the results?

The authors propose that long-term albumin treatment for people with decompensated cirrhosis should aim to fill a gap that exists between baseline serum albumin and an on-treatment target associated with optimal outcomes, which they identified as 4.0 g/dL. This gap varies between individuals depending on their baseline serum albumin level and disease severity. This means that instead of administering a fixed dose of albumin, treatment should be individualised according to the extent of the albumin gap (Figure 1).

Figure 1. Implications of associations between serum albumin levels and clinical outcomes. The graph illustrating the albumin gap hypothesis is reprinted from Caraceni P et al¹ with permission from Elsevier. MELD-Na, model for end-stage liver disease score incorporating serum sodium concentration.

What were the limitations of the study?

The analysis was performed on a single study; therefore, the predictive value of the findings would need external validation and calibration before being utilised in clinical practice.¹

Conclusions

The beneficial effects of long-term albumin treatment seen in the ANSWER study are directly related to serum albumin levels.¹ Notably, 1-month on-treatment serum albumin levels can predict overall survival in people with decompensated cirrhosis and can be used to guide treatment.¹

A target serum albumin level of 4.0 g/dL provides optimal outcomes for people with decompensated cirrhosis with ascites, but people with levels below normal can also benefit from long-term albumin treatment¹

References

1. Caraceni P, Tufoni M, Zaccherini G, Riggio O, Angeli P, Alessandria C, et al. On-treatment serum albumin level can guide long-term treatment in patients with cirrhosis and uncomplicated ascites. J Hepatol. 2021;74(2):340–349.
2. D'Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J Hepatol. 2006;44(1):217–231.
3. Stepanova M, De Avila L, Afendy M, Younossi I, Pham H, Cable R, Younossi ZM. Direct and indirect economic burden of chronic liver disease in the United States. Clin Gastroenterol Hepatol. 2017;15(5):759–766 e755.
4. Tufoni M, Zaccherini G, Caraceni P, Bernardi M. Albumin: Indications in chronic liver disease. United European Gastroenterology Journal. 2020;8(5):528–535.
5. Zaccherini G, Tufoni M, Bernardi M. Albumin administration is efficacious in the management of patients with cirrhosis: A systematic review of the literature. Hepatic Medicine: Evidence and Research. 2020;12(null):153–172.
6. Caraceni P, Riggio O, Angeli P, Alessandria C, Neri S, Foschi FG, et al. Long-term albumin administration in decompensated cirrhosis (ANSWER): an open-label randomised trial. The Lancet. 2018;391(10138):2417–2429.
7. Solà E, Solé C, Simón-Talero M, Martín-Llahí M, Castellote J, Garcia-Martínez R, et al. Midodrine and albumin for prevention of complications in patients with cirrhosis awaiting liver transplantation. A randomized placebo-controlled trial. J Hepatol. 2018;69(6):1250–1259.