Good Enough? Terlipressin, Hepatorenal Syndrome, and the Usage of RRT

Hepatorenal syndrome (HRS), now in its most severe iteration referred to as HRS-AKI, is a rapidly progressive condition unique to patients with cirrhosis which, if left untreated, portends a survival of only weeks to months.¹ HRS-AKI is a functional form of AKI, with a pathophysiologic cascade triggered by portal hypertension ultimately leading to vigorous activation of the sympathetic nervous system and renin–angiotensin–aldosterone system, resulting in severe constriction of renal vasculature and a rapid and relentless fall in GFR.² While still associated with a grim prognosis, by their very nature, the derangements leading to HRS-AKI compel hope. If the stimulus for renal vasoconstriction is lifted, GFR may be expected to be restored. The definitive treatment has been and remains liver transplantation, which eliminates the cirrhotic pathophysiologic environment engendering the renal vasoconstriction. However, the supply of potential allografts lags significantly behind demand, and many patients are regardless ineligible for transplant. For patients on the transplant waiting list and potentially for those deemed ineligible, pharmacologic treatment with vasoconstrictive agents may mitigate the activation of sympathetic nervous system and renin–angiotensin–aldosterone system, restore renal perfusion, and reverse HRS-AKI.³ While such an outcome does not address the underlying cirrhosis, reversal of AKI may buy time until a transplant is available or, at a minimum, provide a bridge through a period of acute illness.

Terlipressin, a synthetic long-acting vasopressin analog, acts preferentially in the splanchnic bed to shunt blood systemically, increase effective circulating arterial volume, reduce renal vasoconstriction, and restore GFR.⁴ Terlipressin, paired with albumin, has been the treatment of choice for HRS-AKI in much of the world for nearly two decades.^4–7 Despite this widespread usage, terlipressin until very recently has not been available in North America. Two large, randomized phase III clinical trials, OT-0401 and REVERSE, demonstrated higher rates of HRS reversal, but the results did not reach statistical significance.^8,9 More recently, the CONFIRM trial randomized 300 patients 2:1 to terlipressin verses placebo (both groups also treated with albumin) and found a higher rate of verified HRS reversal, defined as two consecutive serum creatinine measurements <1.5 mg/dl up to day 14 and survival without RRT for at least an additional 10 days, in patients receiving terlipressin, 32 versus 17%, P = 0.006.¹⁰ In pooled data across the three trials, complete reversal of HRS-AKI (defined as creatinine ≤1.5 mg/dl by day 14 or discharge) occurred more frequently in patients receiving terlipressin, 117/352 (33%) versus 42/256 (16%), P < 0.001.^11,12

Despite these robust data on the salutary effect of terlipressin, the US Food and Drug Administration approval process was contentious and support far from universal.¹³ One of the primary concerns was that despite voluminous evidence associating HRS-AKI reversal with improved survival and a clear association between terlipressin use and reversal,¹⁴ none of the North American trials alone or when pooled^8–10,15 demonstrated a reduction in mortality with terlipressin use. In addition to the trials not being powered for this end point, in patients as complex as those with HRS-AKI, who are at high risk of mortality from causes not remediable by renal recovery, tethering proof of efficacy to the demonstration of improved survival is onerous. With short-term changes in creatinine not universally accepted as a viable end point for AKI trials, there is a need to objectively assess the utility of terlipressin by means other than mortality.¹⁶

The use of RRT in patients with HRS-AKI portends an especially poor prognosis, with 6-month mortality rates exceeding 80% in the absence of transplantation.^7,17 Dialysis is technically challenging, with elevated risks of complications, including bleeding associated with access placement and intradialytic hypotension.¹⁷ HRS-AKI patients frequently require continuous RRT with associated increased costs and resource utilization. In patients ultimately receiving an allograft, RRT at the time of transplant is associated with short-term complications and reduced long-term survival.¹⁸ An intervention reducing the need for RRT is of clear clinical significance, and RRT may be a more appropriate hard trial end point than mortality for an agent explicitly designed and approved to treat AKI. The utilization of RRT was a secondary end point in CONFIRM and was significantly lower at 14 days in patients treated with terlipressin but did not reach statistical significance at day 30 (P = 0.07) or day 60 (P = 0.09), although this study was not powered to do so.¹⁰

In this issue of Kidney360, Velez et al. used pooled data from 3 phase III trials (OT-0401, REVERSE, and CONFIRM) to evaluate the effect of terlipressin therapy on RRT and determine baseline predictors of RRT avoidance.¹⁵ Overall, the use of RRT was significantly reduced in patients receiving terlipressin versus placebo by day 30 (28% versus 36%, respectively; P = 0.04) and day 60 (30% versus 38%; P = 0.045) while not differing significantly by day 90 (32% versus 38%; P = 0.08). For those receiving a liver allograft, 166/608 (27%) post-transplant RRT use was significantly lower in patients receiving terlipressin by day 60 (21% versus 40%; P = 0.008) and day 90 (25% versus 43%; P = 0.018). Treatment with terlipressin independently predicted avoidance of RRT, odds ratio (OR), 1.56 (1.02–2.39), while the higher baseline serum creatinine level and Child-Pugh scores were negatively associated with RRT avoidance, OR, 0.52 (0.42–0.66), and OR, 0.89 (0.80–0.99), respectively. Model of end-stage liver disease score was not included in the multivariable model due to colinearity with serum creatinine. Among those receiving terlipressin, RRT-free survival was higher at day 90 in the overall population (37% versus 29%; P = 0.03) and among patients who received a transplant (60% versus 40%; P = 0.01). Importantly, there was no difference at any time point in proportion of patients receiving a liver transplant or in overall mortality.

The lack of documented mortality benefit with terlipressin use continues to be vexing and may initially elicit the question as to whether reduction in RRT is a clinically meaningful outcome if overall survival is not altered. Here, it may be helpful to reason from first principles. Terlipressin was approved by the US Food and Drug Administration as a treatment for a specific form of AKI (indeed, it is the only agent approved in the United States with a formal indication for any form of AKI) and may rightly be evaluated against that standard. In patients with cirrhosis and AKI, the association between AKI progressing to a higher stage and increased mortality is well documented.^19–21 A treatment reducing the need for RRT then would seem prima facie valuable. Were a treatment available for acute tubular necrosis which reduced the need for RRT though not having a documented effect on mortality, it would nonetheless likely be greeted with enthusiastic acclaim. In patients with cirrhosis, peritransplant RRT confers a worse prognosis when receiving a liver allograft,^22,23 with dialysis either preoperatively or postoperatively associated with increased 5-year mortality.²⁴ Pretransplant dialysis is an independent predictor of HRS nonreversal after transplant and the risk of nonreversal increases with duration of dialysis.²⁵ Critically, in patients who achieve reversal of HRS-AKI and do not require RRT, post-transplant outcomes are similar to patients without HRS-AKI.²⁶ Post-transplant, the need for RRT within 90 days is a strong predictor for progression to ESKD, and only 37% of patients who require post-transplant RRT may be alive at 5 years.²⁷

Importantly, in the study by Velez, experiencing a partial reversal, defined as a >30% reduction in serum creatinine but not achieving complete reversal, was significantly associated with avoidance of RRT. Patients with creatinine >5 mg/dl are unlikely to respond to terlipressin.¹¹ In one sense, this is just a mathematical tautology; the higher creatinine is at treatment initiation, the less likely it is to “respond” if response is defined as creatinine falling below 1.5 mg/dl. This begs the question of whether such a stringent definition of treatment success obscures some benefit, for example, a treatment-induced fall in creatinine from 5.2 to 2.6 mg/dl would not qualify as HRS-AKI reversal but would seem likely to lead to avoidance of RRT. Although the data presented here would seem to confirm this, caution remains warranted. In most settings, a study demonstrating reduced usage of dialysis can be assumed to have achieved this outcome by reducing the need for dialysis. Interpreting such data in patients with advanced cirrhosis however is complicated by far from universal opinions regarding the situational appropriateness of RRT utilization. In addition to assessments of clinical need, ethical and even philosophical considerations are frequently brought to bear when contemplating potential dialysis recipients, and opinions vary across and within specialties as to treatment candidacy in the context of AKI etiology and transplant eligibility. An intervention that rendered patients as less appealing RRT candidates might reduce dialysis rates while not actually improving renal outcomes or favorably altering their clinical course. This scenario is especially pertinent with terlipressin, where concerns have been raised regarding increased rates of respiratory adverse events which might be hypothesized in individual patients to contribute to a perception of clinical futility.¹⁰ For this reason, the documented improvement seen in 90-day RRT-free survival is especially reassuring.

This study is not without limitations. In a setting with so much alternatively mediated death, RRT-free survival analysis may have been strengthened by using competing risk and hierarchical analyses, but sufficient data regarding timing of events were unavailable in this post hoc analysis. Decisions regarding dialysis in patients with cirrhosis are often affected by eligibility for transplant. It would be extremely informative to see the effect of terlipressin on RRT rates broken down by transplant listing status. For transplant recipients, indications for RRT preoperatively and postoperatively vary and postoperative RRT can be required due to intraoperative complications unreflective of the efficacy of preoperative AKI treatment. Further analysis must be performed to parse the effect of terlipressin on pretransplant as opposed to post-transplant RRT requirements.

Despite these limitations, this pooled analysis of the effect of terlipressin on RRT utilization in patients with HRS-AKI is a welcome addition to the body of research on this agent as it comes to North American market. Future studies will likely enroll patients at an earlier point in their disease than those included in the trials analyzed here and may thereby produce the heretofore elusive evidence of mortality benefit. Until that time, reduction in the need for RRT is an encouraging next step on the path to improved outcomes for patients with HRS-AKI.

Disclosures

J.M. Belcher reports the following: Consultancy: Mallinckrodt Pharmaceuticals; Honoraria: Mallinckrodt Pharmaceuticals; and Advisory or Leadership Role: Mallinckrodt Pharmaceuticals.

Funding

None.

Author Contributions

Writing – original draft: Justin M. Belcher.

Writing – review & editing: Justin M. Belcher.