Introduction: Hepatorenal syndrome (HRS) is a common complication of advanced cirrhosis with a high mortality rate and limited treatment options. Central to its pathogenesis is severe, but potentially reversible, renal vasoconstriction leading to functional renal failure. Current pharmacological treatment using splanchnic vasoconstrictors is suboptimal and prognosis without liver transplantation is dismal. The peptide hormone relaxin (RLN) mediates haemodynamic adaptations to pregnancy including increased renal blood flow (RBF) and glomerular filtration rate (GFR). I hypothesised that exogenous RLN could be used therapeutically to improve RBF and renal function in the context of experimental cirrhosis and HRS. Methods: To address this I generated pathologically distinct rat models of liver cirrhosis with features of human HRS including renal vasoconstriction and renal failure. Compensated cirrhosis was induced in male rats by 16 weeks of i.p. carbon tetrachloride (CCl4) and decompensated cirrhosis by bile duct ligation (BDL). I studied the effects of acute i.v. or sustained (72 hr) s.c. infusion of RLN compared with vehicle on systemic haemodynamics, RBF, GFR and kidney histology. I used blood oxygen dependent-magnetic resonance imaging (BOLD-MRI) to detect changes in kidney parenchymal oxygenation and Doppler ultrasound to monitor changes in RBF (velocity time integral, VTI) and renal arterial resistance (resistive index, RI). Hepatic and renal expression of the relaxin receptor RXFP1 was determined by quantitative polymerase chain reaction (qPCR) and immunohistochemistry (IHC). Vascular functional responses in isolated renal arteries were assessed by wire myography. Relaxin mediated changes in key vaso-regulatory signalling pathways in the kidney and renal vessels were analysed by qPCR, IHC and ELISA. Results: I showed using in vitro myography that the pathophysiological mechanism that underlies renal vasoconstriction in experimental cirrhosis models is an impairment of endothelium-dependent vasodilatation. Selective targeting of renal vasoconstriction using relaxin improved renal blood flow, tissue oxygenation, and normalized glomerular filtration rate in both compensated and decompensated rat cirrhosis. Furthermore, relaxin treatment restored endothelium-dependent vasodilation in isolated renal vessels from CCl4 cirrhotic rats. Relaxin-induced effects on renal blood flow and glomerular filtration rate were mediated though activation of the AKT/eNOS/nitric oxide signalling pathway in kidney, though systemic nitric oxide levels were unaffected. Crucially for human translation, relaxin did not reduce mean arterial blood pressure even in advanced cirrhosis. Conclusion: My findings identify relaxin as the first potential targeted treatment reversing the vascular dysfunction which causes HRS and directly improving renal function in HRS. Clinical translation in carefully selected populations is warranted.

Background: It has been reported that transjugular intrahepatic portosystemic shunting (TIPS) might be utilized as a salvage option for hepatorenal syndrome (HRS), while randomized controlled trials are pending and real-world contemporary data on inpatient mortality is lacking. Methods: We conducted an observational retrospective cohort study from the National Inpatient Sample from 2005 to 2014. We included all adult patients admitted with HRS and cirrhosis, using ICD 9-CM codes. We excluded cases with variceal bleeding, Budd–Chiari, end-stage renal disease, liver transplant and transfers to acute-care facilities. TIPS’ association with inpatient mortality was assessed using multivariable mixed-effects logistic regression, as well as exact-matching, thus mitigating for TIPS selection bias. The exact-matched analysis was repeated among TIPS-only versus dialysis-only patients. Results: A total of 79,354 patients were included. Nine hundred eighteen (1.2%) underwent TIPS. Between TIPS and non-TIPS groups, mean age (58 years) and gender (65% males) were similar. Overall mortality was 18% in TIPS and 48% in dialysis-only cases (n = 10,379; 13.1%). Ninety six (10.5%) TIPS patients underwent dialysis. In-hospital mortality in TIPS patients was twice less likely than in non-TIPS patients (adjusted odds ratio [aOR] = 0.43, 95% CI 0.30–0.62; p < 0.001), with similar results in matched analysis [exact-matched (em) OR = 0.39, 95% CI 0.17–0.89; p < 0.024; groups = 96; unweighted n = 463]. Head-to-head comparison showed that TIPS-only patients were 3.3 times less likely to succumb inpatient versus dialysis-only patients (contrast aOR = 0.31, 95% CI 0.20–0.46; p < 0.001), with similar findings post-matching (emOR = 0.22, 95% CI 0.15–0.33; p < 0.001; groups = 54, unweighted n = 1457). Conclusions: Contemporary, real-world data reveal that TIPS on its own, and when compared to dialysis, is associated with decreased inpatient mortality when utilized in non-bleeders-HRS patients. Further randomized studies are needed to establish the long-term benefit of TIPS in these patients.

Des que es va definir la síndrome hepatorenal (SHR), als anys 60 del segle passat, s’han fet grans progressos en el coneixement d’aquesta entitat. La SHR és una insuficiència renal característica dels pacients amb cirrosi hepàtica (CH) amb molt mal pronòstic a curt plaç. El tractament curatiu és el trasplantament hepàtic perquè elimina la CH que és l’origen del problema. Però no sempre està indicat el trasplantament o no s’arriba a temps de realitzar-lo. Els estudis basats en la fisiopatologia de la SHR han estat la base per investigar sobre tractaments que puguin revertir la insuficiència renal i serveixin de pont cap al trasplantament. Fins al moment de la publicació dels treballs que formen part d’aquesta tesi, s’havien fet alguns estudis amb vasoconstrictors esplàcnics que suggerien que aquests fàrmacs eren els més efectius en la reversió de la SHR. Però no s’havien publicat estudis prospectius i aleatoritzats al respecte i aquesta és l’aportació d’un dels estudis: avaluar l’efecte de la terlipressina junt amb l’albúmina sobre la funció renal i la supervivència en pacients amb CH i SHR. Es van aleatoritzar 46 pacients, 23 en el grup de terlipressina i albúmina i 23 en el grup d’albúmina sola. La milloria de la funció renal es va donar en 10 (43,5%) dels pacients tractats amb terlipressina i albúmina i en 3 (8,7%) dels pacients tractats amb albúmina sola (p=0,017). En l’anàlisi multivariat, els factors independents predictius de milloria de la funció renal van ser el volum urinari basal, la creatinina sèrica, el recompte de leucocits i el tractament amb terlipressina i albúmina. La supervivència a 3 mesos no va ser significativament diferent entre els dos grups de tractament (terlipressina i albúmina 27% vs albúmina sola 19%, p=0,7). En l’anàlisi multivariat, els factors independents predictius de supervivència als 3 mesos van ser el MELD basal i la resposta al tractament. A falta d’altres opcions, la terlipressina junt amb albúmina és l’opció terapèutica més efectiva per revertir la SHR i és el tractament que s’ha d’utilitzar, sobretot en els pacients en lista per a trasplantament hepàtic, com a pont fins a l’arribada de l’organ necessari. D’altra banda, malgrat els progressos realitzats en la SHR, els pacients amb CH poden presentar altres tipus d’insuficiència renal i aquest és un camp poc estudiat. En el segon estudi que forma part d’aquesta tesi, basant-nos en un gran número de pacients hospitalitzats, es van avaluar els diferents tipus d’insuficiència renal que poden presentar els pacients amb CH des del punt de vista de la repercusió en el pronòstic. Es van incloure de forma prospectiva, 562 pacients amb CH i insuficiència renal a l’ingrés o que la van desenvolupar durant el mateix. La causa de la insuficiència renal es va classificar en 4 grups: associada a infeccions, a deplecció de volumen, SHR i nefropatia parenquimatosa. La insuficiència renal associada a infeccions va ser la més freqüent (46%), seguida de l’associada a hipovolèmia (32%), SHR (13%) i nefropatia parenquimatosa (9%). Un 17.6% dels casos presentava combinació de causes. La probabilitat de supervivència als 3 mesos va ser del 73% en la nefropatia parenquimatosa, 46% en la insuficiència renal associada a hipovolèmia, 31% en la insuficiència renal associada a infecciones i 15% en la SHR (p<0.0005). En l’anàlisi multivariat ajustat per factors potencialment confusius, la causa de la insuficiència renal estava independentement associada al pronòstic, junto amb el MELD, el sodi sèric i l’encefalopatia hepàtica en el moment del diagnòstic de la insuficiència renal. Aquesta informació és molt important de cara a avaluar els pacients per a trasplantament hepàtic. / Hepatorenal syndrome (HRS) is a type of renal failure common and specific in patients with advanced cirrhosis. The prognosis of HRS is poor and there are no randomized studies with effective treatment for HRS. In one study of this thesis, 46 patients with cirrhosis and HRS were randomly assigned to receive either terlipressin and albumin (n=23) or albumin alone (n=23) for a maximum of 15 days. Primary outcomes were improvement of renal function and survival at 3 months. Improvement of renal function ocurred in 10 patients (43.5%) treated with terlipressin and albumin compared with 2 patients (8.7%) treated with albumin (P=.017). Independent predictive factors of improvement of renal function were baseline urine volume, serum creatinine, leukocyte count, and treatment with terlipressin and albumin. Survival at 3 months was not significantly different between the 2 groups (terlipressin and albumin 27% vs albumin 19%, P=.7). Independent predictive factors of 3-month survival were baseline MELD score and improvement in renal function. In conclusion, the administration of terlipressin and albumin should be considered for the management of patients with cirrhosis and HRS, particularly in patients who are candidates to liver transplantation. The prognostic value of the different causes of renal failure (RF) in cirrhosis is not well stablished. In the other study of this thesis, 562 consecutive patients with cirrhosis and RF (serum creatinine >1.5mg/dl) were prospectively included. The cause of RF was classified into 4 groups: RF associated with bacterial infections, RF associated with volume deplection, HRS and parenchymal nephropathy. The primary end point was survival at 3 months. The frequency of RF was: RF associated with infections: 213 cases ( 46%),hypovolemia-associated RF: 149 (32%), HRS: 60 (13%), and parenchymal nephropathy: 41 (9%). Three-month probability of survival was73% for parenchymal nephropathy, 46% for hypovolemia-associated RF, 31% for RF associated with infections, and 15% for HRS (P<.0005). In the multivariate analysis, cause of RF was independently associated with prognosis, together with MELD score, serum sodium and hepatic encephalopathy. This information may help in decision making in liver transplantation.

INTRODUÇÃO: Lesão renal aguda (LRA) é uma complicação comum da cirrose frequentemente desencadeada por infecções bacterianas. A mortalidade de pacientes com cirrose e LRA varia de 10 a 100% a depender do estádio da cirrose, etiologia e progressão da LRA e tratamento recebido. Pacientes com LRA que progride possuem mortalidade intra-hospitalar consideravelmente maior do que aqueles que não progridem. Os critérios diagnósticos da LRA baseiam-se na creatinina sérica. Porém, esse biomarcador tem acurácia diagnóstica limitada, já que demora até 48 horas para se alterar e não distingue a etiologia da LRA. Essa última limitação é particularmente importante em pacientes com suspeita de síndrome hepatorrenal (SHR), já que o seu tratamento envolve o uso de medicações de alto custo (albumina e terlipressina) e eficácia limitada. O NGAL (neutrophil gelatinase-associated lipocalin) é um biomarcador de necrose dos túbulos renais e sua dosagem tem sido proposta como marcador diagnóstico mais acurado da LRA na cirrose, pois se altera mais precocemente e seus níveis urinários variam conforme a etiologia e gravidade da LRA. OBJETIVOS: Os objetivos primários do trabalho foram avaliar a acurácia do NGAL urinário para predizer: a) desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; b) resposta ao tratamento combinado com albumina e terlipressina em pacientes com diagnóstico estabelecido de SHR. Os objetivos secundários foram: a) acurácia de marcadores de disfunção hemodinâmica em cirrose (atividade plasmática de renina e noradrenalina sérica) na predição do desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; b) acurácia de marcadores de função hepática, parâmetros hemodinâmicos, marcadores inflamatórios e testes laboratoriais de lesão renal tradicionalmente utilizados na prática clínica na predição do desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; c) acurácia de marcadores de disfunção hemodinâmica em cirrose (atividade plasmática de renina e noradrenalina sérica) na predição de resposta ao tratamento combinado com albumina e terlipressina em pacientes com diagnóstico estabelecido de SHR; d) acurácia de marcadores de função hepática, parâmetros hemodinâmicos, marcadores inflamatórios e testes laboratoriais de lesão renal tradicionalmente utilizados na prática clínica na predição de resposta ao tratamento combinado com albumina e terlipressina em pacientes com diagnóstico estabelecido de SHR; e) comparar o conceito tradicional de LRA em cirrose com a nova classificação ICA-AKI para predizer mortalidade intra-hospitalar, em 30 e 90 dias em pacientes com cirrose e infecção bacteriana; f) acurácia do NGAL para predizer mortalidade intra-hospitalar em pacientes com cirrose e infecção bacteriana. PACIENTES E MÉTODOS: Critérios de inclusão: a) cirrose; b) ascite e/ou hidrotórax hepático; c) idade maior que 18 anos; d) concordância em participar no estudo; e) LRA e/ou infecção bacteriana. Critérios de exclusão: a) comorbidades graves; b) choque; c) nefropatia intrínseca; d) uso de drogas nefrotóxicas; e) diálise prévia; f) transplante hepático. Foram coletadas amostras de urina para dosagem de NGAL e sangue para dosagem de atividade plasmática de renina e noradrenalina sérica no momento da inclusão do paciente no estudo. Os pacientes com SHR receberam o tratamento padrão atual com albumina e terlipressina. RESULTADOS: Foram incluídos 199 pacientes: 179 com infecção bacteriana para avaliação de desenvolvimento ou progressão da LRA e 58 com SHR para avaliação da resposta ao tratamento (38 pacientes foram avaliados nas duas partes do estudo). O NGAL urinário apresentou associação com a progressão da LRA (AUC: 0,67; p=0,002), mas não com o seu desenvolvimento (p=0,973). Outras variáveis associadas à progressão da LRA foram INR (p=0,033), MELD (p=0,012), FEUr (p=0,026) e relação proteinúria/creatinina urinária (p=0,023). Houve associação entre NGAL urinário e a resposta ao tratamento combinado com albumina e terlipressina em pacientes com SHR (AUC: 0,70; p=0,007). Outras variáveis associadas à resposta ao tratamento da SHR foram INR (p=0,028), MELD (p=0,004) e relação proteinúria/creatinina urinária (p=0,003). Combinando o MELD com o NGAL urinário foi possível identificar subgrupos de pacientes com taxas de resposta ao tratamento com albumina e terlipressina distintas (9,1% x 48,1% x 80,0%, p < 0,001). Tanto o conceito tradicional de LRA na cirrose quanto a classificação ICA-AKI foram capazes de predizer mortalidade intra-hospitalar, em 30 e 90 dias (p < 0,05). O NGAL urinário também foi capaz de predizer mortalidade intra-hospitalar (AUC: 0,71; p < 0,001). CONCLUSÕES: a) NGAL urinário aumentado foi associado à progressão da LRA em pacientes com cirrose e infecção bacteriana; b) atividade plasmática de renina e noradrenalina sérica não se correlacionaram ao desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; c) NGAL urinário foi preditor de resposta ao tratamento combinado com albumina e terlipressina em pacientes com SHR; d) atividade plasmática de renina e noradrenalina sérica não se correlacionaram à resposta ao tratamento combinado com albumina e terlipressina em pacientes com SHR; e) tanto o conceito tradicional de LRA em cirrose quanto a classificação ICA-AKI mostraram-se adequados para predizer mortalidade intra-hospitalar, em 30 e 90 dias em pacientes com cirrose e infecção bacteriana; f) NGAL urinário aumentado foi associado à maior mortalidade intra-hospitalar em pacientes com cirrose e infecção bacteriana / INTRODUCTION: Acute kidney injury (AKI) is a frequent complication of cirrhosis, and is often triggered by bacterial infection. The mortality of patients with cirrhosis and AKI varies from 10 to 100%, depending on the stage of cirrhosis, etiology and progression of AKI, and treatment. Patients with AKI that progresses have a higher in-hospital mortality than those with AKI that does not progress. AKI diagnostic criteria are based on serum creatinine. However, this biomarker has limited diagnostic accuracy, taking up to 48 hours to rise, and does not distinguish the etiology of AKI. This latter limitation is particularly important in patients with suspected hepatorenal syndrome (HRS), because treatment involves high cost medication (albumin and terlipressin) with limited efficacy. NGAL (neutrophil gelatinase-associated lipocalin) is a biomarker of renal tubular necrosis which has been proposed as a more accurate AKI biomarker in cirrhosis because it rises earlier than creatinine and its urinary levels vary according to the etiology of AKI. OBJECTIVES: The primary endpoints were the establishment of the accuracy of urinary NGAL to predict: a) development or progression of AKI in patients with cirrhosis and bacterial infection; b) response to treatment with albumin and terlipressin in patients with diagnosed HRS. The secondary endpoints were: a) accuracy of biomarkers of hemodynamic dysfunction in cirrhosis (renin plasma activity and serum noradrenaline) to predict AKI development or progression in patients with cirrhosis and bacterial infection; b) accuracy of biomarkers of hepatic function, hemodynamics, inflammation and kidney injury routinely used in clinical practice to predict AKI development or progression in patients with cirrhosis and bacterial infection; c) accuracy of biomarkers of hemodynamic dysfunction in cirrhosis (renin plasma activity and serum noradrenaline) to predict response to treatment with albumin and terlipressin in patients with diagnosis of HRS; d) accuracy of biomarkers of hepatic function, hemodynamics, inflammation and kidney injury routinely used in clinical practice to predict response to treatment with albumin and terlipressin in patients with diagnosis of HRS; e) compare the traditional definition of AKI in cirrhosis with the new classification ICA-AKI to predict in-hospital, 30-day and 90-day mortality in patients with cirrhosis and bacterial infection. f) accuracy of urinary NGAL in predicting in-hospital mortality in patients with cirrhosis and bacterial infection. PATIENTS AND METHODS: Inclusion criteria: a) diagnosis of cirrhosis; b) presence of ascites and/or hepatic hydrothorax; c) age over 18 years old; d) consent to participate in the study; e) AKI and/or bacterial infection. Exclusion criteria: a) severe systemic comorbidities; b) shock; c) intrinsic nephropathy; d) nephrotoxic drug use; e) previous dialysis; f) liver transplantation recipient. Urine and blood samples were collected for urinary NGAL and renin plasma activity and serum noradrenaline measurement at the inclusion. Patients with HRS received standard treatment with albumin and terlipressin. RESULTS: 199 patients were included: 179 with bacterial infection for the analysis of AKI development or progression and 58 with HRS for the analysis of response to treatment (38 patients were analyzed in both parts of the study). Urinary NGAL was associated with AKI progression (AUC: 0.67, p=0.002), but not with AKI development (p=0.973). Other variables associated with AKI progression were INR (p=0.003), MELD (p=0.012), FEUr (p=0.026) and proteinuria/urinary creatinine ratio (p=0.023). There was also an association of urinary NGAL with response to HRS treatment with albumin and terlipressin (AUC: 0.70, p=0.007). Other variables associated with response to HRS treatment were INR (p=0.028), MELD (p=0.004) and proteinuria/urinary creatinine ratio (p=0.003). Combining MELD and urinary NGAL we could identify subgroups of patients with distinct response rates to treatment with albumin and terlipressin (9.1% x 48.1% x 80.0%, p < 0.001). Both the traditional definition of AKI in cirrhosis and the classification ICA-AKI predicted in-hospital, 30-day and 90-day mortality (p < 0.05). Urinary NGAL was also associated with in-hospital mortality (AUC: 0.71, p < 0.001). CONCLUSIONS: a) High urinary levels of NGAL were associated with AKI progression in patients with cirrhosis and bacterial infection; b) renin plasma activity and serum noradrenaline were not associated with AKI development or progression in patients with cirrhosis and bacterial infection; c) urinary NGAL was associated with response to combined treatment with albumin and terlipressin in patients with HRS; d) renin plasma activity and serum noradrenaline were not associated with response to combined treatment with albumin and terlipressin in patients with HRS; e) both traditional definition ok AKI in cirrhosis and ICA-AKI classification were able to predict in-hospital, 30-day and 90-day mortality in patients with cirrhosis and bacterial infection; f) high urinary levels of NGAL were associated with in-hospital mortality in patients with cirrhosis and bacterial infection

INTRODUÇÃO: Lesão renal aguda (LRA) é uma complicação comum da cirrose frequentemente desencadeada por infecções bacterianas. A mortalidade de pacientes com cirrose e LRA varia de 10 a 100% a depender do estádio da cirrose, etiologia e progressão da LRA e tratamento recebido. Pacientes com LRA que progride possuem mortalidade intra-hospitalar consideravelmente maior do que aqueles que não progridem. Os critérios diagnósticos da LRA baseiam-se na creatinina sérica. Porém, esse biomarcador tem acurácia diagnóstica limitada, já que demora até 48 horas para se alterar e não distingue a etiologia da LRA. Essa última limitação é particularmente importante em pacientes com suspeita de síndrome hepatorrenal (SHR), já que o seu tratamento envolve o uso de medicações de alto custo (albumina e terlipressina) e eficácia limitada. O NGAL (neutrophil gelatinase-associated lipocalin) é um biomarcador de necrose dos túbulos renais e sua dosagem tem sido proposta como marcador diagnóstico mais acurado da LRA na cirrose, pois se altera mais precocemente e seus níveis urinários variam conforme a etiologia e gravidade da LRA. OBJETIVOS: Os objetivos primários do trabalho foram avaliar a acurácia do NGAL urinário para predizer: a) desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; b) resposta ao tratamento combinado com albumina e terlipressina em pacientes com diagnóstico estabelecido de SHR. Os objetivos secundários foram: a) acurácia de marcadores de disfunção hemodinâmica em cirrose (atividade plasmática de renina e noradrenalina sérica) na predição do desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; b) acurácia de marcadores de função hepática, parâmetros hemodinâmicos, marcadores inflamatórios e testes laboratoriais de lesão renal tradicionalmente utilizados na prática clínica na predição do desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; c) acurácia de marcadores de disfunção hemodinâmica em cirrose (atividade plasmática de renina e noradrenalina sérica) na predição de resposta ao tratamento combinado com albumina e terlipressina em pacientes com diagnóstico estabelecido de SHR; d) acurácia de marcadores de função hepática, parâmetros hemodinâmicos, marcadores inflamatórios e testes laboratoriais de lesão renal tradicionalmente utilizados na prática clínica na predição de resposta ao tratamento combinado com albumina e terlipressina em pacientes com diagnóstico estabelecido de SHR; e) comparar o conceito tradicional de LRA em cirrose com a nova classificação ICA-AKI para predizer mortalidade intra-hospitalar, em 30 e 90 dias em pacientes com cirrose e infecção bacteriana; f) acurácia do NGAL para predizer mortalidade intra-hospitalar em pacientes com cirrose e infecção bacteriana. PACIENTES E MÉTODOS: Critérios de inclusão: a) cirrose; b) ascite e/ou hidrotórax hepático; c) idade maior que 18 anos; d) concordância em participar no estudo; e) LRA e/ou infecção bacteriana. Critérios de exclusão: a) comorbidades graves; b) choque; c) nefropatia intrínseca; d) uso de drogas nefrotóxicas; e) diálise prévia; f) transplante hepático. Foram coletadas amostras de urina para dosagem de NGAL e sangue para dosagem de atividade plasmática de renina e noradrenalina sérica no momento da inclusão do paciente no estudo. Os pacientes com SHR receberam o tratamento padrão atual com albumina e terlipressina. RESULTADOS: Foram incluídos 199 pacientes: 179 com infecção bacteriana para avaliação de desenvolvimento ou progressão da LRA e 58 com SHR para avaliação da resposta ao tratamento (38 pacientes foram avaliados nas duas partes do estudo). O NGAL urinário apresentou associação com a progressão da LRA (AUC: 0,67; p=0,002), mas não com o seu desenvolvimento (p=0,973). Outras variáveis associadas à progressão da LRA foram INR (p=0,033), MELD (p=0,012), FEUr (p=0,026) e relação proteinúria/creatinina urinária (p=0,023). Houve associação entre NGAL urinário e a resposta ao tratamento combinado com albumina e terlipressina em pacientes com SHR (AUC: 0,70; p=0,007). Outras variáveis associadas à resposta ao tratamento da SHR foram INR (p=0,028), MELD (p=0,004) e relação proteinúria/creatinina urinária (p=0,003). Combinando o MELD com o NGAL urinário foi possível identificar subgrupos de pacientes com taxas de resposta ao tratamento com albumina e terlipressina distintas (9,1% x 48,1% x 80,0%, p < 0,001). Tanto o conceito tradicional de LRA na cirrose quanto a classificação ICA-AKI foram capazes de predizer mortalidade intra-hospitalar, em 30 e 90 dias (p < 0,05). O NGAL urinário também foi capaz de predizer mortalidade intra-hospitalar (AUC: 0,71; p < 0,001). CONCLUSÕES: a) NGAL urinário aumentado foi associado à progressão da LRA em pacientes com cirrose e infecção bacteriana; b) atividade plasmática de renina e noradrenalina sérica não se correlacionaram ao desenvolvimento ou progressão da LRA em pacientes com cirrose e infecção bacteriana; c) NGAL urinário foi preditor de resposta ao tratamento combinado com albumina e terlipressina em pacientes com SHR; d) atividade plasmática de renina e noradrenalina sérica não se correlacionaram à resposta ao tratamento combinado com albumina e terlipressina em pacientes com SHR; e) tanto o conceito tradicional de LRA em cirrose quanto a classificação ICA-AKI mostraram-se adequados para predizer mortalidade intra-hospitalar, em 30 e 90 dias em pacientes com cirrose e infecção bacteriana; f) NGAL urinário aumentado foi associado à maior mortalidade intra-hospitalar em pacientes com cirrose e infecção bacteriana / INTRODUCTION: Acute kidney injury (AKI) is a frequent complication of cirrhosis, and is often triggered by bacterial infection. The mortality of patients with cirrhosis and AKI varies from 10 to 100%, depending on the stage of cirrhosis, etiology and progression of AKI, and treatment. Patients with AKI that progresses have a higher in-hospital mortality than those with AKI that does not progress. AKI diagnostic criteria are based on serum creatinine. However, this biomarker has limited diagnostic accuracy, taking up to 48 hours to rise, and does not distinguish the etiology of AKI. This latter limitation is particularly important in patients with suspected hepatorenal syndrome (HRS), because treatment involves high cost medication (albumin and terlipressin) with limited efficacy. NGAL (neutrophil gelatinase-associated lipocalin) is a biomarker of renal tubular necrosis which has been proposed as a more accurate AKI biomarker in cirrhosis because it rises earlier than creatinine and its urinary levels vary according to the etiology of AKI. OBJECTIVES: The primary endpoints were the establishment of the accuracy of urinary NGAL to predict: a) development or progression of AKI in patients with cirrhosis and bacterial infection; b) response to treatment with albumin and terlipressin in patients with diagnosed HRS. The secondary endpoints were: a) accuracy of biomarkers of hemodynamic dysfunction in cirrhosis (renin plasma activity and serum noradrenaline) to predict AKI development or progression in patients with cirrhosis and bacterial infection; b) accuracy of biomarkers of hepatic function, hemodynamics, inflammation and kidney injury routinely used in clinical practice to predict AKI development or progression in patients with cirrhosis and bacterial infection; c) accuracy of biomarkers of hemodynamic dysfunction in cirrhosis (renin plasma activity and serum noradrenaline) to predict response to treatment with albumin and terlipressin in patients with diagnosis of HRS; d) accuracy of biomarkers of hepatic function, hemodynamics, inflammation and kidney injury routinely used in clinical practice to predict response to treatment with albumin and terlipressin in patients with diagnosis of HRS; e) compare the traditional definition of AKI in cirrhosis with the new classification ICA-AKI to predict in-hospital, 30-day and 90-day mortality in patients with cirrhosis and bacterial infection. f) accuracy of urinary NGAL in predicting in-hospital mortality in patients with cirrhosis and bacterial infection. PATIENTS AND METHODS: Inclusion criteria: a) diagnosis of cirrhosis; b) presence of ascites and/or hepatic hydrothorax; c) age over 18 years old; d) consent to participate in the study; e) AKI and/or bacterial infection. Exclusion criteria: a) severe systemic comorbidities; b) shock; c) intrinsic nephropathy; d) nephrotoxic drug use; e) previous dialysis; f) liver transplantation recipient. Urine and blood samples were collected for urinary NGAL and renin plasma activity and serum noradrenaline measurement at the inclusion. Patients with HRS received standard treatment with albumin and terlipressin. RESULTS: 199 patients were included: 179 with bacterial infection for the analysis of AKI development or progression and 58 with HRS for the analysis of response to treatment (38 patients were analyzed in both parts of the study). Urinary NGAL was associated with AKI progression (AUC: 0.67, p=0.002), but not with AKI development (p=0.973). Other variables associated with AKI progression were INR (p=0.003), MELD (p=0.012), FEUr (p=0.026) and proteinuria/urinary creatinine ratio (p=0.023). There was also an association of urinary NGAL with response to HRS treatment with albumin and terlipressin (AUC: 0.70, p=0.007). Other variables associated with response to HRS treatment were INR (p=0.028), MELD (p=0.004) and proteinuria/urinary creatinine ratio (p=0.003). Combining MELD and urinary NGAL we could identify subgroups of patients with distinct response rates to treatment with albumin and terlipressin (9.1% x 48.1% x 80.0%, p < 0.001). Both the traditional definition of AKI in cirrhosis and the classification ICA-AKI predicted in-hospital, 30-day and 90-day mortality (p < 0.05). Urinary NGAL was also associated with in-hospital mortality (AUC: 0.71, p < 0.001). CONCLUSIONS: a) High urinary levels of NGAL were associated with AKI progression in patients with cirrhosis and bacterial infection; b) renin plasma activity and serum noradrenaline were not associated with AKI development or progression in patients with cirrhosis and bacterial infection; c) urinary NGAL was associated with response to combined treatment with albumin and terlipressin in patients with HRS; d) renin plasma activity and serum noradrenaline were not associated with response to combined treatment with albumin and terlipressin in patients with HRS; e) both traditional definition ok AKI in cirrhosis and ICA-AKI classification were able to predict in-hospital, 30-day and 90-day mortality in patients with cirrhosis and bacterial infection; f) high urinary levels of NGAL were associated with in-hospital mortality in patients with cirrhosis and bacterial infection