Liver Test Patterns Reveal Hidden Risk in Hospitalized Heart Failure Patients

When a patient is rushed to the hospital with acute decompensated heart failure, clinicians order a battery of routine blood tests, including the standard panel of liver enzymes. Those liver values have long been treated as incidental findings, noted in passing and rarely interpreted as a coherent pattern. A new retrospective cohort study published in Clinical Research in Cardiology suggests that this casual approach may be discarding some of the most actionable prognostic information available at the bedside. By classifying patients according to the pattern of their liver test abnormalities rather than looking at each marker in isolation, researchers found strikingly different short-term mortality risks that could reshape how early risk stratification is performed in this high-acuity population.

The study, led by Gil Marcus and Avishay Grupper of the Shamir Medical Center in Israel, together with colleagues, analyzed 8,332 hospital admissions for acute decompensated heart failure occurring between December 2006 and 2017. Of these, 6,792 admissions, or 81.5 percent, had sufficient laboratory data to be classified into one of four cardiohepatic injury phenotypes. The classification rested on peak in-hospital values of four routine liver tests: aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, and total bilirubin. Rather than treating each enzyme as a separate signal, the investigators combined them into clinically meaningful syndromes that hepatologists have long used to characterize liver disease, applying that framework to the failing heart’s impact on the liver.

The four phenotypes capture fundamentally different mechanisms of hepatic injury. Patients with no hepatic injury had normal values across the panel. Hepatocellular injury, dominated by elevations of the transaminases AST and ALT, reflects damage to the liver cells themselves, a pattern classically associated with hypoxic hepatitis, in which reduced oxygen delivery to the liver causes hepatocyte death. Cholestatic injury, marked by elevations of alkaline phosphatase and bilirubin, points to impaired bile flow, a phenomenon often linked in heart failure to congestion, the backward accumulation of pressure and fluid when the failing heart cannot adequately empty the venous system. Combined injury, in which both patterns coexist, suggests simultaneous cellular damage and biliary stasis, and it turned out to be the most ominous category of all.

The distribution of these phenotypes across the cohort was itself revealing. A majority of patients, 58.1 percent, showed no hepatic injury, but the remaining four in ten carried some form of liver abnormality. Hepatocellular injury was the most common pattern at 22.1 percent, followed by cholestatic injury at 13.1 percent, while combined injury affected 6.7 percent of admissions. This means that in more than four out of ten hospitalizations for acute decompensated heart failure, the liver was sending a measurable distress signal, one that the researchers argue deserves systematic interpretation rather than the per-marker glance it typically receives.

The mortality figures associated with each phenotype were dramatic. In-hospital death occurred in just 3.5 percent of patients with no hepatic injury, but rose to 16.3 percent in those with hepatocellular injury, 9.0 percent in those with cholestatic injury, and 24.3 percent in those with combined injury, a difference that was highly statistically significant. In other words, a patient whose routine liver panel showed the combined pattern faced roughly a one-in-four chance of dying during the same hospitalization, nearly seven times the risk of a patient with a clean panel. After adjustment for other clinical variables using multivariable logistic regression, all three injury phenotypes independently predicted in-hospital death compared with no injury, confirming that the association was not simply explained by patients being sicker in more obvious ways.

The post-discharge story proved more nuanced and, in some respects, more surprising. Among 6,128 patients included in adjusted analyses of outcomes after leaving the hospital, the researchers examined 30-day mortality, 30-day readmission, and a composite of the two, using Cox proportional hazards models and Fine–Gray models to account for competing risks. Here a clear divide emerged: hepatocellular injury and combined injury were independently associated with higher 30-day mortality and with the composite outcome, whereas isolated cholestatic injury was not. Both hepatocellular and combined injury also independently predicted 30-day readmission. The implication is that the specific mechanism of liver injury, not merely its presence, determines which patients remain at risk after they walk out of the hospital door.

Why should hepatocellular involvement carry this extended shadow of risk while cholestasis does not? The authors’ framework suggests a plausible physiological explanation. Hepatocellular enzyme elevations in acute heart failure are thought to reflect forward failure, a drop in cardiac output that deprives the liver of oxygen, essentially a shock-like state affecting the hepatic parenchyma. That kind of circulatory compromise signals profound hemodynamic instability, which plausibly persists beyond discharge and drives both early death and early readmission. Cholestatic patterns, by contrast, may track congestion that responds to decongestive therapy during the index admission; once the patient is euvolemic, the cholestatic signal may lose its independent prognostic weight. The study did not directly test these mechanisms, but the differential outcomes are consistent with this hemodynamic interpretation of the two injury types.

The practical implications are considerable. The four liver tests used in the study are cheap, universally available, and already drawn in virtually every patient admitted with acute heart failure, meaning that pattern-based interpretation requires no new testing, no new equipment, and no additional cost. A clinician reviewing an admission panel could, in principle, immediately place the patient into one of the four phenotypes and adjust the intensity of monitoring, counseling about prognosis, and post-discharge follow-up accordingly. Patients flagged with hepatocellular or combined injury might merit closer early follow-up, tighter diuretic titration schedules, or enrollment in post-discharge transition-of-care programs, while the phenotype could also serve as a stratification variable in future clinical trials of acute heart failure therapies.

The findings also fit into a growing body of literature on cardiohepatic interactions. Previous studies, including analyses from the PROTECT, ASCEND-HF, EVEREST, CHARM, and TOPCAT programs, have documented that abnormal liver function tests are common in heart failure and correlate with worse outcomes, and recent work from the STRONG-HF trial has linked changes in liver tests after discharge to prognosis. What distinguishes the new study is its explicit phenotypic framing: instead of asking whether any individual enzyme is abnormal, it asks what constellation of abnormalities is present, borrowing the hepatological taxonomy of hepatocellular, cholestatic, and mixed injury. This approach mirrors a broader movement in cardiology toward phenotype-driven medicine, in which heterogeneous syndromes like acute heart failure are decomposed into mechanistically coherent subgroups with distinct prognoses and, potentially, distinct treatment responses.

As with any retrospective single-cohort analysis, certain caveats apply. The study relied on peak in-hospital laboratory values, and the cohort spans admissions from 2006 to 2017, an era that predates the widespread adoption of sodium-glucose cotransporter 2 inhibitors, which have since transformed heart failure therapy and may themselves interact with liver outcomes. Residual confounding cannot be excluded, and the classification thresholds and their generalizability to other health systems will require validation in independent cohorts. Nevertheless, the central message stands on solid ground: the pattern of routine liver test abnormalities in acute decompensated heart failure is not noise but signal, and phenotypes with hepatocellular involvement identify patients at elevated risk both during hospitalization and in the vulnerable first month afterward. For a condition that sends millions of patients to hospitals worldwide each year and carries substantial early mortality, a free risk-stratification tool hiding in plain sight on the standard chemistry panel is a finding worth taking seriously.

Subject of Research: Cardiohepatic injury phenotypes and short-term outcomes in acute decompensated heart failure

Article Title: Cardio-hepatic injury phenotypes and short-term outcomes in acute decompensated heart failure

Article References: Marcus, G., Maymon, S. L., Minha, I., Kalmanovich, E., Moravsky, G., Fuchs, S., Minha, S., & Grupper, A. (2026). Cardio-hepatic injury phenotypes and short-term outcomes in acute decompensated heart failure. Clinical Research in Cardiology. https://doi.org/10.1007/s00392-026-03043-w

Image Credits: AI Generated

DOI: 10.1007/s00392-026-03043-w

Keywords: acute decompensated heart failure, cardiohepatic syndrome, hepatocellular injury, cholestatic injury, liver function tests, in-hospital mortality, 30-day readmission, risk stratification, prognosis, retrospective cohort study, hypoxic hepatitis, congestive hepatopathy