This prospective observational case–control study was conducted at Minia University Hospital between August 2022 and December 2024. The primary objectives of this study were to determine whether serum copeptin (hemodynamic/vasopressin axis) and urinary [TIMP-2]·[IGFBP7] (tubular stress) can distinguish HRS-AKI from HRS-CKD and from non-HRS decompensated cirrhosis, and to evaluate whether these biomarkers, alone and in combination with MELD 3.0, improve short-term mortality risk prediction in patients with HRS-AKI. A secondary objective was to explore clinically actionable cutoff values for copeptin and [TIMP-2]·[IGFBP7] that could support early recognition and risk-guided management in routine practice. The study enrolled 80 patients with HCV-related liver cirrhosis and 20 age- and sex-matched healthy controls. Approval was obtained from the local ethics committee, and the study adhered to the Helsinki Declaration. All participants provided informed written consent.

We prospectively enrolled adult patients with HCV-related cirrhosis who were followed at our tertiary liver unit. All enrolled patients had previously received direct-acting antiviral therapy and had achieved sustained virologic response (SVR12), confirmed by undetectable HCV RNA at enrollment. To minimize etiologic heterogeneity, we restricted inclusion to hepatitis C-related cirrhosis and excluded other causes of chronic liver disease (e.g., metabolic dysfunction-associated steatotic liver disease, alcoholic liver disease, autoimmune or cholestatic liver disease). This restriction was chosen to avoid confounding from differing comorbidity profiles and pathophysiologic mechanisms, particularly the higher prevalence of hypertension, diabetes, obesity, and intrinsic kidney disease in non-HCV etiologies, which could influence baseline copeptin and [TIMP-2]·[IGFBP7] levels and obscure biomarker–outcome relationships.

Patients with HCV-related liver cirrhosis (LC) were stratified into three groups1: the compensated LC group (n=20) with no clinically or radiologically detectable ascites2; the decompensated LC group (n=30) presenting with diuretic-responsive ascites; and3 the HRS group (n=30), which was further subdivided into HRS-AKI (n=15) meeting ADQI/ICA criteria, and HRS-CKD (n=15) with eGFR <60mL/min/1.73m2 for ≥3 months (Fig. 1). Inclusion criteria required HCV-positive liver cirrhosis confirmed by liver biopsy or by clinical, radiographic, and biochemical investigations, with HRS diagnosis following the revised ADQI/ICA criteria. Exclusion criteria comprised1: serious comorbidities (NYHA class IV heart failure, O2-dependent COPD, advanced cancer)2; non-HCV cirrhosis etiologies3; shock4; structural kidney disease (presence of significant proteinuria (e.g., >500mg/day or equivalent spot UACR)), persistent hematuria suggestive of glomerular disease, and abnormal renal imaging (e.g., small echogenic kidneys or obstructive uropathy)5; nephrotoxic drug use within 30 days6; active urinary tract infection7; anuria >12h8; current renal replacement therapy; or9 history of solid organ transplantation.

Fig. 1.

Study enrollment and patient allocation. A total of 80 patients with HCV-related liver cirrhosis were enrolled and stratified into three groups: compensated cirrhosis (n=20), decompensated cirrhosis (n=30), and hepatorenal syndrome (n=30). The flow diagram outlines patient eligibility, exclusions, and final group assignment.

The sample size was determined a priori using G*Power (version 3.1) to detect a clinically meaningful difference in urinary [TIMP-2]·[IGFBP7] between patients with hepatorenal syndrome (HRS) and healthy controls. Based on prior reports of large biomarker gradients between AKI and non-AKI populations, we assumed a standardized effect size (Cohen's d) of 0.80 for the primary comparison, with a two-sided type I error of 0.05 and 80% power. Under these assumptions, a minimum of 26 participants per group was required for a two-sample t test. Because the primary a priori power calculation was based on the main biomarker contrast between HRS and controls, comparisons between the HRS subtypes (HRS-AKI vs HRS-CKD; n=15 each) should be interpreted as exploratory. We have therefore framed subtype analyses as hypothesis-generating and highlighted the need for external validation.

To account for potential exclusions or missing samples and to allow for multivariable modeling, we planned to enroll 80 patients with HCV-related cirrhosis (distributed across compensated, decompensated, and HRS strata) and 20 healthy controls (total n=100). This final sample size exceeded the minimum per-group requirement and provided ≥80% power to detect the anticipated between-group differences in [TIMP-2]·[IGFBP7] and serum copeptin, as well as to evaluate associations between these biomarkers and HRS-AKI using logistic regression models adjusted for key covariates.

The study groups were well-matched for age and sex (p=0.36 and 0.91 respectively; Table 1). HRS patients had lower body mass index compared with compensated cirrhotics (p<0.01) and showed progressive hematologic abnormalities, with declining hemoglobin and pronounced thrombocytopenia relative to controls (both p<0.01).

Liver function tests demonstrated marked synthetic impairment in HRS, with higher bilirubin (p<0.01), lower albumin (p=0.03), and more prolonged INR compared with compensated cirrhosis (p=0.02), whereas aminotransferase levels did not clearly distinguish between cirrhosis stages (p>0.05). Renal dysfunction was most evident in HRS, where serum creatinine was approximately doubled relative to other groups (p<0.01), while urine albumin-to-creatinine ratio remained similar across groups (p=0.24).

Standard prognostic scores increased in parallel with disease severity, with Child–Pugh and MELD 3.0 scores lowest in compensated cirrhosis and highest in HRS (p<0.01), consistent with HRS as a terminal complication of portal hypertension.

Urinary [TIMP-2]·[IGFBP7] and serum copeptin levels showed a significant stepwise rise from healthy controls through compensated and decompensated cirrhosis to HRS (both p<0.01; Table 1 and Fig. 2). Biomarker levels were lowest in controls, increased modestly in compensated cirrhosis, and were highest in HRS. Copeptin showed a modest decrease in decompensated versus compensated cirrhosis, before increasing again in HRS.

Fig. 2.

Biomarker levels across cirrhosis stages. Panel A illustrates urinary [TIMP-2]·[IGFBP7] (ng/mL)2/1000 levels, which progressively increase from healthy controls to HRS-AKI. Panel B shows serum copeptin levels (pmol/L), demonstrating a similar rising trend with disease severity. The p-values displayed on the figure represent pairwise comparisons between the HRS group (reference) and each of the other study groups (healthy controls, compensated cirrhosis, decompensated cirrhosis, and the alternate HRS subtype). Pairwise p-values were derived from one-way ANOVA followed by Duncan's multiple range post hoc testing (two-sided), with p<0.05 considered statistically significant.

Within the HRS subgroups, HRS-AKI patients had substantially 41% higher [TIMP-2]·[IGFBP7] and 28% higher copeptin levels than HRS-CKD patients (both p<0.01; Table 1), indicating more pronounced acute tubular stress and circulatory disturbances in HRS-AKI.

Serum copeptin and urinary [TIMP-2]·[IGFBP7] were strongly correlated (Pearson r=0.72, 95% CI 0.61–0.80; p<0.001; Fig. 3).

Fig. 3.

Correlation between serum copeptin and urinary [TIMP-2]·[IGFBP7]. This scatter plot illustrates the strong positive correlation between serum copeptin and urinary [TIMP-2]·[IGFBP7] levels (r=0.72, p<0.001), supporting their combined use as mechanistically complementary biomarkers for HRS-AKI.

The strong correlation between serum copeptin and urinary [TIMP-2]·[IGFBP7] (r=0.72, p<0.001) suggests a tight coupling between hemodynamic stress and tubular injury in advanced cirrhosis. Both biomarkers showed progressive elevation from compensated cirrhosis to HRS-AKI, reflecting the underlying pathophysiology of cirrhosis-associated renal injury. However, the modest decrease in copeptin observed in decompensated versus compensated cirrhosis was small in magnitude and may reflect clinical heterogeneity (e.g., differences in diuretic exposure, plasma osmolality/sodium, and timing relative to acute stress) rather than a true reduction in vasopressin-system activation; copeptin increased again in HRS, consistent with progressive effective arterial underfilling and neurohormonal activation.

[TIMP-2] and IGFBP7 are expressed by tubular epithelial cells and upregulated in response to ischemic or inflammatory stress, where they induce G1 cell-cycle arrest to prevent cell division under injurious conditions. Their product, [TIMP-2]·[IGFBP7], therefore reflects early tubular stress before overt decline in glomerular filtration.18 This highly statistically significant relationship indicates that as copeptin levels rise, [TIMP-2]·[IGFBP7] levels increase proportionally, with the correlation coefficient falling within the “strong” range (0.50–0.74). This is stronger than typically seen between traditional renal markers (e.g., creatinine/cystatin C correlations average r=0.60–0.65).

Copeptin is a stable surrogate for arginine vasopressin release and integrates non-osmotic neurohormonal activation driven by splanchnic vasodilation and effective arterial underfilling in advanced cirrhosis. Rising copeptin levels across disease stages and their further increase in HRS-AKI mirror the hemodynamic crisis that characterizes this syndrome, with activation of the vasopressin, sympathetic, and renin–angiotensin–aldosterone systems and progressive renal vasoconstriction.13

Taken together, these patterns support a unified pathophysiologic model of HRS-AKI in which circulatory dysfunction (captured by copeptin) and structural tubular stress (captured by [TIMP-2]·[IGFBP7]) act synergistically. The higher [TIMP-2]·[IGFBP7] levels in HRS-AKI versus HRS-CKD are consistent with more intense acute tubular injury in the AKI phenotype, while copeptin quantifies the severity of the hemodynamic insult. This dual-marker framework aligns with contemporary ICA-AKI concepts emphasizing the interaction between splanchnic vasodilation, neurohormonal activation, and ischemic tubular damage, and provides a biologic rationale for combining hemodynamic and tubular stress biomarkers in risk stratification models.8 This complements prior work by Zhang et al., 14 though our larger cohort and inclusion of HRS-CKD provide novel stratification. Notably, preserved UACR across groups supports the functional nature of HRS.

Both urinary [TIMP-2]·[IGFBP7] and serum copeptin showed excellent discrimination for HRS-AKI with AUCs of 0.91 and 0.90, respectively. At their optimal cutoffs, [TIMP-2]·[IGFBP7] provided higher sensitivity, whereas copeptin offered slightly greater specificity, suggesting complementary diagnostic roles.

Importantly, these diagnostic associations remained strong after adjusting for age and sex, and persisted after further adjustment for MELD 3.0. In multivariable models, both biomarkers independently contributed to HRS-AKI prediction: every 1-unit increase in copeptin (pmol/L) was associated with a 2.78-fold increase in the odds of HRS-AKI, and each 1-unit rise in [TIMP-2]·[IGFBP7] ((ng/mL)2/1000) was associated with a 2.58-fold higher odds. These effect sizes surpass many conventional AKI biomarkers (e.g., NGAL typically shows aOR 1.5–2.0).19 These associations persisted after MELD 3.0 score adjustment, indicating that the biomarkers capture pathophysiologic information not fully reflected in liver disease severity scores. Clinically, this supports using [TIMP-2]·[IGFBP7] as an early, sensitive screen for HRS-AKI and copeptin as a confirmatory marker that strengthens rule-in decisions when elevated. However, the observation that copeptin and [TIMP-2]·[IGFBP7] may facilitate earlier detection of HRS-AKI should be interpreted with caution, as it is based on a relatively small HRS-AKI subgroup (n=15) and remains promising but preliminary. These early-detection signals require confirmation in larger, multicenter cohorts before they can be translated into firm clinical decision thresholds or routine screening strategies.

Short-term mortality in HRS-AKI was high in our cohort (about two-thirds of patients died within 30 days), and clearly exceeding reported averages (50–60%).20,21 This elevated mortality may reflect1 delayed diagnosis in our cohort,2 HCV-specific disease progression, and3 limited terlipressin availability. In our cohort, short-term mortality clustered in HRS-AKI, whereas compensated and decompensated non-HRS groups did not experience deaths over 30 days. However, we cannot definitively conclude that renal failure is the dominant mortality driver without longer follow-up and cause-of-death data. The striking 40% absolute mortality difference between HRS subtypes (AKI vs CKD) highlights the time-sensitive nature of AKI management. These findings, consistent with ICA 2024 guidelines but showing higher-than-expected HRS-AKI mortality, strongly support the clinical utility of early biomarkers such as [TIMP-2]·[IGFBP7] and copeptin to identify high-risk patients when interventions may still alter outcomes, particularly given the 3.5-fold greater mortality risk for HRS-AKI versus HRS-CKD patients in this cohort.8

In practical terms, copeptin and [TIMP-2]·[IGFBP7] could be incorporated as add-on tests in patients with decompensated cirrhosis who present with AKI and already undergo routine blood and urine sampling. A single time-point measurement of urinary [TIMP-2]·[IGFBP7] and serum copeptin at presentation would complement, rather than replace, serum creatinine and MELD 3.0: [TIMP-2]·[IGFBP7] could serve as a sensitive screen for HRS-AKI and early tubular stress, while elevated copeptin would help confirm a hemodynamic HRS phenotype and flag patients who may benefit from earlier vasoconstrictor therapy, closer monitoring, and expedited transplant evaluation.

Within HRS-AKI, non-survivors displayed significantly higher copeptin and urinary [TIMP-2]·[IGFBP7] levels than survivors, and both biomarkers showed excellent prognostic performance for 30-day mortality. Urinary [TIMP-2]·[IGFBP7] (AUC 0.92) slightly outperformed copeptin (AUC 0.87), aligning with its role in detecting subclinical tubular injury before functional decline.4 At an exploratory cutoff of >0.52 (ng/mL)2/1000, [TIMP-2]·[IGFBP7] achieved high sensitivity and specificity for mortality, while copeptin >14.6pmol/L provided similar sensitivity with lower specificity.

These findings suggest a pragmatic approach in which combined assessment of copeptin and [TIMP-2]·[IGFBP7] helps identify HRS-AKI patients at highest short-term risk who may benefit from early vasoconstrictor therapy, intensive monitoring, and timely consideration of liver transplantation. The relatively narrow copeptin range among non-survivors (15.6–19.4pmol/L) suggests a possible “critical threshold” above 15pmol/L beyond which mortality escalates dramatically. Urinary [TIMP-2]·[IGFBP7] >0.65 (ng/mL)2/1000 captured all non-survivors, potentially serving as a marker of irreversible injury. The strong performance of both markers in predicting HRS-AKI mortality supports their combined use in prognostic models. Given the limited number of fatal events, these proposed cutoffs should be considered hypothesis-generating and require external validation in larger, prospective cohorts before routine clinical use.

Our prognostic analyses were based on MELD 3.0, which incorporates sodium, albumin, and sex in addition to bilirubin, creatinine, and INR. While this reflects contemporary practice in our center, the incremental prognostic contribution of copeptin and [TIMP-2]·[IGFBP7] may differ in settings that still use earlier MELD formulations (MELD or MELD-Na). Future studies should therefore evaluate these biomarkers alongside the specific MELD-based model used in each transplant program.