Hepatorenal syndrome (HRS) is a morbid and often fatal complication of decompensated cirrhosis. HRS has an 8% annual incidence among patients with ascites with a dismal 1-month survival of 34.6% that has not improved over time (Thomson et al, Dig Dis Sci 2020). However, for the lucky patient with reversal of HRS, survival is significantly improved.

How can we reverse HRS?

Vasoconstrictors may improve renal perfusion by improving effective (central) blood volume and enhancing arterial blood pressure (Velez et al, Nephron 2015). The conventional approach in the US is to use midodrine (an oral alpha-agonist) and octreotide (likely an expensive placebo) (Pomier-Layargues et al, Hepatology 2003). In Europe, Terlipressin is the first-line agent. It is not available in the US. The problem is that whenever Terlipressin went up for registration, it failed to reverse HRS (to creatinine <1.5mg/dL) more than placebo (Sanyal et al, Gastroenterology 2008 and Boyer et al, Gastroenterology 2016).

Enter CONFIRM

The CONFIRM trial enrolled 300 patients with HRS to be randomized 2:1 to either bolus terlipressin (1 mg q6 hours) or placebo. Both groups received albumin. Although verified reversal of HRS was reported in 63 (32%) in the terlipressin group and 17 (17%) in the placebo group (P=0.006), this finding was overshadowed. Terlipressin was associated with no improvements in mortality (51% vs 45%) or transplantation (23% vs 29%) at 90 days. Furthermore, death within 90 days due to respiratory disorders occurred in 22 (11%) receiving terlipressin and 2 (2%) in receiving placebo.

In July 2020, the FDA’s Cardiovascular and Renal Drugs Advisory Committee voted 8-7 to recommend approval. However, in September 2020, the FDA rejected approved in Complete Response Letter requesting additional data.

Does terlipressin work?

The FDA agreed to a primary outcome defined by reversal of HRS. Although it is a surrogate outcome, it is clearly mechanistically appropriate for the broader aims of care in this population. As such, the trial met its prespecified primary endpoint.

Why you might not find the results of CONFIRM persuasive?

Terlipressin did not improve mortality or transplant-free survival. Furthermore, the rate of respiratory failure is alarming.

Why is CONFIRM confusing?

Several features of CONFIRM make for a complicated interpretation.

First, patients required creatinine >2.25 mg/dL for entry. And the average creatinine at enrollment was 3.5! Prior trials used lower cutoffs and we currently define HRS as an increase in the creatinine level of >0.3 mg/dL above baseline that fails to respond to a volume challenge. Most patients were also pre-treated with octreotide/midodrine. What this means is that the inclusion procedures for the trial may have selected a cohort with definite HRS but also were less likely to experience recovery.
Second, CONFIRM used bolus terlipressin. Continuous infusions may be safer (Cavallin et al Hepatology 2016).
Third, while we do not have a sense of how much albumin is too much albumin, it is intuitive that continuous, daily infusions of albumin in patients without improving kidney function could be associated with volume overload. As Dr. Garcia-Tsao notes in her terrific editorial, “Respiratory failure probably resulted from pulmonary edema, which in turn resulted from an increase in afterload (from terlipressin) and an increase in preload (from albumin, which was used in 83% of patients in the terlipressin group).” (Garcia-Tsao, NEJM 2021).
Fourth, it would be incredibly challenging to demonstrate a survival benefit in the patients who were enrolled. A glance at Table 1 should humble any clinical trialist who claims “my patients are too sick for a trial.” The patients in CONFIRM had a creatinine of 3.5, MELD-Na of 33 with acute-on-chronic liver failure, 40% with alcoholic hepatitis, and roughly 45% with SIRS. Any trial powered to improve survival in this population would have be massive.

What are the next steps?

Emotions are raw after CONFIRM. It is neither a miracle drug nor is it without value. While others may have additional ideas, I see the greatest value in 3 strands of future investigation

First, “Early Terli.” Many people suggest that terlipressin should be initiated prior to such severe kidney injury. That might be more effective. We need a trial to prove it.

Second, individualized, judicious use of albumin. The clearest message from CONFIRM as well as the ATTIRE trial published in the same issue of NEJM is that we need to be more careful with albumin (China et al, NEJM 2021). One of the most exciting paths forward is to use point-of-care ultrasound to evaluate central blood volume and determine whether a patient needs any albumin, let alone daily albumin (Velez et al, Am J Nephrol 2019). This strategy is poised to reduce the risk of pulmonary edema and renal venous congestion. However, we need multicenter trials.

Third, norepinephrine has been associated with outcomes equivalent to terlipressin in small studies. Further evaluation of this treatment is warranted.

Elliot B. Tapper MD, @ebtapper

Director, Michigan Cirrhosis Program

Ann Arbor, MI

Conflicts of interest:

Elliot Tapper has consulted for Mallinckrodt on their Ornithine Phenylacetate program. Mallinckrodt makes terlipressin.