Decreased protein C function predicts mortality in patients with cirrhosis

INTRODUCTION

Protein C (PrC), a physiological anticoagulant, regulates inflammation and cell death and has known predictive/therapeutic roles in sepsis. Accumulating evidences suggest plasma hypercoagulability results in progression of fibrosis and formation of microclots causing end-organ dysfunction. We investigated a possible association between natural anticoagulants-PrC, protein S (PrS) and antithrombin III (AT)-and clinical outcomes in cirrhotics.

METHODS

Functional PrC, PrS and AT were analysed in 515 cirrhotic patients and compared with 229 noncirrhotics. Among those with cirrhosis, we conducted multivariable predictive model on 3-month survival to assess the prognostic ability of anticoagulants.

RESULTS

Protein C (P < .001), PrS (P < .001) and AT (P < .001) levels were lower in cirrhotics compared with noncirrhotics. In addition, patients with Child-Pugh (CP)-C had significantly lower (P < .05) functional PrC, PrS and AT levels than CP-B, CP-A and noncirrhotic patients. Low PrC function correlated with markers of liver dysfunction and inflammation: INR(r = -.72, P < .001), bilirubin (r = -.620, P < .001), albumin (r = .539, P < .001), creatinine (r = -.417, P < .001), ferritin (r = -.68, P = .035), procalcitonin (r = -.79, P = .01), raised ESR (r = .56, P < .001) and liver fibrosis (r = -.840, P < .001). Patients who died (n = 160) had significantly lower median PrC function (23.8%, 16.3-33.0]) compared with those who remained alive (74.9%, [59.7-92.5]); P < .001. In a multivariable predictive model using PrC, and MELD score, we found a significant impact of low PrC levels on survival (P < .001, IRR = 0.97, 95% CI = 0.96-0.98). Receiver operating characteristic (ROC) curve analysis revealed that functional PrC levels <52% were associated with increased mortality (P < .001).

CONCLUSION

Low functional protein C level correlated with markers of liver dysfunction, inflammation and sepsis and independently predicted mortality at 3 months in cirrhotics, especially if functional levels were <52%.

Microvesicles in hepatic and peripheral vein can predict nonresponse to corticosteroid therapy in severe alcoholic hepatitis

BACKGROUND

Severe alcoholic hepatitis patients have high mortality and limited response to corticosteroids. Microvesicles reflect cellular stress and disease conditions.

AIMS

To investigate whether microvesicles are associated with severity, response to steroid therapy and inflammation in severe alcoholic hepatitis.

METHODS

Microvesicles originating from different cells were studied pre-therapy in 101 patients; (71 responder to corticosteroid therapy and 30 nonresponders) and 20 healthy controls. Microvesicles and cells were determined in peripheral and hepatic vein samples using flow cytometry and correlated with outcomes. Inflammatory signalling pathways and functional alterations of immune cells after stimulation with microvesicles were also investigated.

RESULTS

Microvesicles mean levels were higher in nonresponders for T cells (CD3⁺ CD4⁺ ; 10.1 MV/μL vs 5.4; P = 0.06), macrophages (CD68⁺ CD11b⁺ ; 136.5 vs 121.9 MV/μL; P = 0.01), haematopoietic stem-cells (CD45⁺ CD34⁺ ; 116.8 vs 13.4 MV/μL; P = 0.0001) and hepatocytes (ASGPR⁺ ; 470 vs 361 MV/μL; P = 0.01); the latter two predicting steroid nonresponse in 94% patients at baseline in peripheral plasma. Microvesicle levels correlated with histological and liver disease severity indices. Whereas, in non-responders hepatic vein CD34+ cells were lower (P = 0.02), the CD34+ microvesicles there from were higher (P = 0.04), thus suggesting impaired regeneration. Also, microvesicles of 0.2-0.4 μm size were higher in nonresponders (P < 0.03) at baseline. Microvesicles from patients trigger more (P = 0.04) ROS generation, TNF-α production (P = 0.04) and up-regulate pro-inflammatory cytokine related genes in neutrophils in vitro.

CONCLUSIONS

Pre-therapy peripheral plasma levels of CD34⁺ and ASGPR⁺ microvesicles are reliable non-invasive markers of steroid nonresponse and mortality in patients with severe alcoholic hepatitis.

Hepatic Sarcoidosis: Clinico-pathological characterization of symptomatic cases

AIM

The aim of this study was to investigate the clinical and pathological features of hepatic sarcoidosis in symptomatic cases.

METHODS

Twenty-two symptomatic hepatic sarcoidosis cases were included in the study. Hepatic sarcoidosis was determined by typical imaging, histopathology, and high angiotensin-converting enzyme levels. Demographic data, laboratory data, imaging findings, liver biopsies, and clinical findings were analyzed. Portal hypertension (PH) was defined by the presence of ascites and/or varices; imaging findings suggestive of PH-splenomegaly (> 12 cm on longest axis); portal vein dilation (> 13 mm); collateral vessel formation; and hepatic venous pressure gradient ≥ 6 mmHg.

RESULTS

Mean age was 49.63 ± 10.7 years. Liver tests showed elevated serum alkaline phosphatase and gamma-glutamyl transpeptidase levels (95%). Serum albumin levels were low (< 3 g/dl) in 32% of the patients. Histologically, hepatic granulomas were located in the portal/periportal areas, with or without parenchymal involvement (77%). Duct damage (27%), absent portal veins (32%), and hepatomegaly (41%) were also observed. Clinically, chronic cholestatic symptoms and PH features were observed in 41% and 50% of the patients, respectively. Three-quarters of patients with PH features were non-cirrhotic. Cirrhosis and bleeding varices were observed in 14%. Hepatic sarcoidosis overlaps with primary biliary cirrhosis (PBC) and primary sclerosing cholangitis (PSC) was observed in two cases.

CONCLUSIONS

Sarcoidosis causes significant hepatic disease. PH and jaundice are main clinical presentations in liver sarcoidosis patients. Imaging findings of PH should be carefully reviewed, as it can occur even before the establishment of cirrhosis. Hepatic sarcoidosis mimics and overlaps with PBC and PSC.

Serum ferritin predicts early mortality in patients with decompensated cirrhosis

BACKGROUND & AIMS

Serum ferritin is a known marker of hepatic necro-inflammation and has been studied to predict 1 year mortality and post-transplant survival in decompensated cirrhotics. However, there are no studies evaluating ferritin as a predictor of early mortality. We investigated whether serum ferritin levels could predict 15 day and 30 day mortality in patients with decompensated cirrhosis.

METHODS

318 patients with decompensated cirrhosis were included.

RESULTS

Patients of decompensated cirrhosis [257 males, mean age of 51 [±13]years, were followed for a median of 31 days. Serum ferritin levels were significantly different between survivors and non-survivors [p<0.001] and showed significant correlation with MELD score [p<0.001], CTP score [p<0.001], leucocyte counts [TLC] [p<0.001], serum sodium [p<0.001], ACLF grades [p=0.005], spontaneous bacterial peritonitis [SBP] [p=0.02], hepatic encephalopathy [HE] [p<0.001] and hepatorenal syndrome [HRS] [p=0.012]. Serum ferritin, etiology, MELD, HE, CTP score, sodium, TLC, and ACLF grades were significant predictors of mortality on univariate analysis. Ferritin [p=0.04, HR 1.66 95% CI (1.02-2.73)] was a significant predictor of early mortality on multivariate analysis along with HE [p=0.006, HR 3.47 95% CI (2.13-8.41)] (Model 1), TLC [p=0.02, HR 1.81 95% CI (1.06-3.07)] (Model 2), ACLF grades [p=0.018, HR 2.013,95% CI (1.126-3.60)], and CTP score [p<0.0001, HR 1.36 95% CI (1.17-1.59)] (Model 3).

CONCLUSION

Serum ferritin levels correlate with severity of hepatic decompensation and are associated with early liver related death independent of the MELD score in hospitalized patients with decompensated cirrhosis. This could also have a potential therapeutic implication.