Terlipressin: Hopes Fulfilled or Dashed?

One of the most dreaded complications of end stage liver disease is hepatorenal syndrome AKI (HRS-AKI; formally known as hepatorenal syndrome 1 [HRS-1]). HRS-AKI carries a high mortality, and treatment options are limited. Guidelines recommend terlipressin as the first-line treatment of HRS-AKI; however, terlipressin has not been approved by the Food and Drug Administration (FDA) for use in the United States.

The recently published CONFIRM study is the largest randomized, placebo-controlled study to evaluate the efficacy and safety of terlipressin (¹). CONFIRM included 300 patients with HRS-AKI. The primary endpoint of CONFIRM was HRS reversal and survival without KRT for at least 10 days, which occurred in 32% versus 17% (P=0.006) in terlipressin and albumin (T+A)–treated patients versus placebo and albumin (P+A)–treated patients, respectively. The effects of T+A versus P+A on the four prespecified secondary endpoints were HRS reversal: 39% versus 18%, respectively (P<0.001); HRS reversal without need for KRT within 30 days: 34% versus 17%, respectively (P=0.001); HRS reversal in patients with systemic inflammatory response syndrome: 37% versus 6%, respectively (P<0.001); and reversal of HRS without recurrence of HRS within 30 days: 26% versus 17%, respectively (P=0.08). Although HRS reversal is an important clinical endpoint, it should be noted that T+A versus P+A administration was associated with neither a reduction in death within 90 days (51% versus 45%) nor an increase in lifesaving liver transplantation (23% versus 29%).

Of equal importance were the adverse findings. While on study drug, the incidence of death from respiratory failure was 3% in the T+A group as compared with 0% in the P+A group. Respiratory disorders as a cause of death within 90 days were seen in 11% versus 2% in T+A versus P+A, respectively, and respiratory failure as an adverse event leading to death by day 14 was seen in 19% versus 0%, respectively, of patients. The findings of respiratory failure were unexpected and concerning. So, how do these data compare with previous terlipressin trials?

There are three relevant studies and one meta-analysis that warrant review. In the study by Sanyal et al. (²), the incidence of serious adverse events related to study treatment was 9% in participants treated with T+A versus 2% in participants treated with P+A. Similar to CONFIRM, 40% of T+A–treated patients with serious adverse events had respiratory distress (²). The TAHRS study reported an incidence of “circulatory overload” of 30% in the T+A arm as compared with 17% in the P+A arm (P=0.87) (³). Finally, in the REVERSE trial, treatment-related deaths were seen in 3% of patients in the T+A group, and no treatment-related deaths were reported in the P+A arm. Also, multiorgan failure and pulmonary edema were both more commonly seen with T+A (11% versus 5% and 11% versus 7%, respectively) (⁴). None of the aforementioned trials showed an improvement in transplant-free survival.

A meta-analysis of 18 trials by Wang et al. (⁵) suggested that T+A–treated patients had a higher rate of HRS reversal (odds ratio [OR], 4.96; 95% confidence interval [CI], 2.23 to 11.0, P=0 .001) and a lower incidence of death compared with placebo (relative risk, 0.63; 95% CI, 0.44 to 0.91; P=0.01). The relative risk for more adverse events with T+A was 1.57. Wang et al. (⁵) also compared terlipressin with the other vasoconstrictor most commonly used in HRS treatment, norepinephrine. In a meta-analysis of eight studies comparing T+A and norepinephrine plus albumin, both medications were equally effective in reversing HRS (OR, 1.01; 95% CI, 0.65 to 1.57). Notably, adverse events were more common with T+A as compared with norepinephrine plus albumin (OR, 2.72; 95% CI, 1.33 to 5.55). As such, some have considered norepinephrine as first-line treatment for HRS for patients in the intensive care unit. Nevertheless, in a recent study in patients with acute on chronic liver failure, terlipressin was associated with a higher HRS reversal rate (40% versus 17%; P=0.004), less need for KRT (57% versus 80%; P=0.006), and an improved 28-day mortality (48% versus 20%; P=0.001), but also a higher incidence of adverse events (23% versus 8%; P=0.02) when compared with norepinephrine (⁶).

This raises important questions. What caused the adverse safety signals suggested in most of these trials? Is there a biologic rationale to explain why terlipressin use could be associated with respiratory failure? Terlipressin is a prodrug that is converted by endopeptidases into lysine-vasopressin. Lysine-vasopressin is a nonselective vasopressin analog that binds to the V1 (V1a) and V3 (V1b) receptors. Vasopressin and terlipressin both mediate systemic vasoconstriction by binding to the V1 receptor on vascular smooth muscle cells, a mechanism by which terlipressin causes splanchnic vasoconstriction and a reduction in portal venous pressure. Nevertheless, vasopressin V1a and V1b receptor expression has also been demonstrated in the lung with a high affinity of lysine-vasopressin for lung V1b receptors (⁷).

The development of pulmonary edema in the CONFIRM study has been hypothesized to result from increased systemic afterload caused by terlipressin and increased venous preload from administration of albumin and its associated intravascular and extracellular fluid expansion. However, if this explanation was sufficient, then one should expect respiratory failure with any systemic vasoconstrictor administered in conjunction with albumin in HRS. Instead, respiratory complications have not been reported with norepinephrine and albumin. Although studies of norepinephrine are fewer in number and typically conducted in an intensive care unit setting (where enhanced monitoring may confound results), it bears asking the following question: are there fundamental and relevant differences between these two vasoactive medications that may account for discordant adverse event profiles?

Although both terlipressin and norepinephrine cause vasoconstriction in the systemic circulation and, thereby, increase afterload on the left heart, norepinephrine also causes vasoconstriction in the precapillary pulmonary vasculature and may have a weak inotropic effect on the heart. Terlipressin not only lacks the inotropic effect to potentially counter increased systemic vascular resistance but may also cause pulmonary arterial vasodilation (⁸) and pulmonary venoconstriction (⁹). Stated in another way, although speculative, the combination of T+A in patients with cirrhosis may lead to a unique perfect storm of hydrostatic stress in the lung microvasculature, including (1) a high cardiac output state of liver disease, (2) volume expansion aggravated by administration of large doses of albumin, (3) increased postpulmonary capillary pressure from systemic vasoconstriction and pulmonary venoconstriction, and (4) a lack of precapillary pulmonary vasoconstriction and potentially even pulmonary vasodilation (conceptual diagram in Figure 1A). Alongside the possibility that vasopressin increases alveolar-capillary leak by hydrostatic mechanisms, vasopressin may also negatively alter the intrinsic permeability of the alveolar capillary barrier. In combination with the aforementioned hydrostatic forces, this may conspire to cause pulmonary edema and respiratory failure (¹⁰). In contrast, the precapillary pulmonary vasoconstrictive effect of norepinephrine, stimulation of active alveolar sodium, and fluid reabsorption by norepinephrine may, in fact, protect the lung microvasculature and limit the development of pulmonary edema by decreasing the pressure differential across the pulmonary capillaries and stimulating fluid reabsorption (conceptual diagram in Figure 1B).

Figure 1.: Differential effects of terlipressin versus norepinephrine on the pulmonary vasculature. (A) Terlipressin causes dilation of the pulmonary arteries and constriction of the pulmonary veins, and it also likely increases the permeability of the alveolar capillary barrier, thereby contributing to pulmonary edema. (B) Norepinephrine has a positive inotropic effect but also causes constriction of the pulmonary arteries, and it may also stimulate alveolar sodium and fluid reabsorption, thereby possibly protecting from pulmonary edema. RA, right atrium; RV, right ventricle; LA, left atrium; LV, left ventricle.

With regard to albumin use in the CONFIRM study, albumin was administered to 91% of patients in the P+A group as compared with only 83% in the T+A group (¹). The total amount of albumin given per patient was 240 g in the P+A group versus 199 g in the T+A group (¹). There was no clear relationship between the dose of albumin administered and the development of respiratory failure in the placebo arm; however, there was a trend toward more respiratory failure at higher doses in the T+A arm, with an 11% incidence in the lowest quintile of albumin administration rising to 17% in the highest quartile (FDA briefing document at the Cardiovascular and Renal Drugs Advisory Committee meeting on July 15, 2020). Although speculative, this may suggest that volume expansion is detrimental but that this relationship is most pronounced in the setting of hydrostatic vulnerability.

What does this mean for the treating physician? For now, terlipressin remains unavailable, and many providers in the United States will likely continue to use “off-label” norepinephrine and albumin in patients with HRS-AKI in the intensive care unit. This is our practice, and it is unlikely to change in the wake of CONFIRM. The major barrier to this approach is the availability of intensive care unit beds, which has been exacerbated by the coronavirus disease 2019 pandemic. That having been said, one possible solution was recently pioneered at Stanford, where they suggest that norepinephrine may be used outside the intensive care unit setting (¹¹). More studies are needed to explore the safety and feasibility of such an approach. For those clinicians around the world who have the option of using terlipressin and for regulatory agencies in the United States and abroad, we would advise careful review of the safety data from the CONFIRM trial. When terlipressin is used, it should be done so with caution and with careful attention to respiratory distress, pulse oximetry, and frequent chest radiographs or point of care ultrasound monitoring. More judicious use of albumin might help to mitigate the respiratory risks of terlipressin, but this requires further study.

Disclosures

P.J. Leary reports consultancy agreements with Bayer; receiving research funding from the American Heart Association, the National Heart, Lung, and Blood Institute, and the Pulmonary Hypertension Association; and receiving honoraria from Bayer. E.R. Swenson reports employment with the Veterans Affairs Puget Sound Health Care System. All remaining authors have nothing to disclose.

Funding

None.