Managing Septic Shock in Patients With Cirrhosis: The Role of Adrenal Insufficency

Abstract

Fernández J, Escorsell A, Zabalza M, et al. (Liver Unit, Hospital Clinic, Barcelona, Spain). Adrenal insufficiency in patients with cirrhosis and septic shock: effect of treatment with hydrocortisone on survival. Hepatology 2006;44:1288–1295.Bacterial infection continues to be a major cause of death among patients with cirrhosis. The demonstrated survival benefit of prophylactic antibiotic therapy in cirrhotic patients with gastrointestinal hemorrhage underscores the significant impact that bacterial sepsis has in this patient population (Hepatology 1999;29:1655–1661). Currently, management of patients with sepsis centers on supportive care. Many areas of uncertainty remain and strategies for management of sepsis need to be prospectively evaluated in patients with advanced liver disease.Relative adrenal insufficiency (RAI) has been linked to refractory shock, resistance to vasoconstrictor medications, and increased mortality in patients with sepsis (N Engl J Med 2003;348:727–734). A recent meta-analysis of data from multiple clinical trials demonstrates a significant survival benefit from stress-dose steroid replacement in patients with septic shock not necessarily associated with cirrhosis (BMJ 2004;329:480–489). In this setting, Fernández et al prospectively assessed the impact of hydrocortisone treatment in cirrhotic patients with septic shock and abnormal short corticotropin test responses (Hepatology 2006;44:1288–1295).Twenty-five consecutive cirrhotic patients with a diagnosis of septic shock admitted to a specialized liver intensive care unit (ICU) from June 2004 to January 2006 were assessed for adrenal insufficiency with a short corticotropin test within 24 hours of ICU admission. Immunocompromised patients, those with advanced hepatocellular carcinoma, and those with a history of steroid therapy within the past 6 months were excluded. Patients with RAI—defined as a baseline cortisol concentration <15 μg/dL and/or increase of plasma cortisol concentration of <9 μg/dL 1 hour after corticotrophin administration in patients with a baseline cortisol concentration <35 μg/dL—were treated with 50 mg of intravenous hydrocortisone every 6 hours. Hydrocortisone was tapered once vasopressor medications were no longer required. Patients in the prospective cohort were compared with a historical control group of the last 50 consecutive patients with cirrhosis and sepsis in whom adrenal function was not assessed and no steroid therapy was given. All patients were treated with a standard protocol that differed only in the assessment of adrenal function and included antibiotic therapy, central venous and urinary catheterization, goal-directed plasma volume expansion and vasopressor therapy, albumin administration in cases of spontaneous bacterial peritonitis, and continuous venovenous hemofiltration and mechanical ventilation as indicated.Within the prospective cohort, 17 of 25 patients had RAI. A trend toward higher rates of RAI in patients with Child–Pugh class C cirrhosis was noted (P = .08), although no statistically significant differences in clinical characteristics were identified between patients with and without adrenal insufficiency. Compared with historical controls, patients in the prospective cohort had similar baseline clinical characteristics, except for a statistically significant lower serum albumin concentration (2.3 ± 5 g/dL vs 2.7 ± 5 g/dL; P < .05). Clinically important outcomes were improved in the prospective cohort compared with controls, including resolution of shock (96% vs 58%; P = .001), time to shock resolution (48% vs 18% within 48 hours of ICU admission; P =.01), ICU and hospital survival (68% vs 38% [P = .03]; 64% vs 32% [P = .003], respectively). Two patients treated with hydrocortisone died of fungal infection; no other serious adverse events were reported in the hydrocortisone treatment group.CommentThe sepsis syndrome is a major precipitant for acute-on-chronic liver failure, thereby leading to hypotension, worsening hepatic function, hepatorenal syndrome, multisystem organ failure, and death (Gut 2005;54:718–725). Broad-spectrum antibiotic therapy and early goal-directed therapy (with target values for targets for central venous oxygen saturation, central venous pressure, and mean arterial pressure) have been demonstrated to improve short-term survival in patients with sepsis not necessarily associated with cirrhosis (N Engl J Med 2001;345:1368–1377). Strategies for the management of septic shock need to be prospectively evaluated in patients with liver disease, among whom infection remains a major cause of death.The hemodynamic derangements of cirrhosis mirror those of septic shock and adrenal insufficiency. Recent data have emerged regarding the association of RAI in patients with liver dysfunction and sepsis. In patients with acute liver failure, RAI was present in 62% of patients and was associated with lower transplant-free survival (Hepatology 2002;36:395–402). A series of cirrhotic patients with septic shock revealed that RAI was present in 51.5% of patients and was associated with significantly higher ICU and hospital mortality rates. Cumulative 90-day survival rates were 15.3% and 63.2% in adrenal-insufficient and normal adrenal function groups, respectively (Hepatology 2006;43:673–681).The rates of RAI reported by Fernández et al were similar to, if not higher than, previously reported rates. In concordance with prior reports, RAI was more common in patients with more advanced liver disease. Whether this is directly related to impaired hepatic cholesterol synthesis and steroid metabolism, the proinflammatory cytokine milieu, or whether adrenal insufficiency is simply a marker of multisystem organ failure remains unclear. That RAI is frequently present in patients with acute liver failure without evidence of sepsis (Hepatology 2002;36:395–402) suggests that hepatocellular dysfunction may play a significant role. Notably, a specific bacterial etiology for the sepsis syndrome was not identified in 36% and 24% of patients in the prospective and control groups, respectively. The role of RAI and steroid therapy in cirrhotic patients with noninfectious precipitants for the systemic inflammatory response syndrome (such as gastrointestinal hemorrhage or surgery) remains to be determined.Although the overall mortality rate was improved in patients receiving steroid therapy, 2 of 17 patients receiving hydrocortisone therapy developed fatal fungal infections. The site of infection and portal of entry was not reported by the authors. A prior study utilizing higher doses of steroid replacement (hydrocortisone 300 mg/d) reported an increased rate of resistant bacterial infections (Liver Int 2003;23:71–77). No resistant bacterial infections were noted in the current study.The major limitation of the current study is the absence of a contemporary, blinded, placebo-controlled arm. Although the use of clinical management protocols in a single, specialized liver ICU minimized the variability in care between the prospective and historical cohorts, it is possible that the improved survival in the contemporary (prospective) cohort may be due to an unidentified factor other than adrenal replacement therapy. Although this study provides valuable insight into the care of the septic cirrhotic patient, the results need to be validated in a larger, multicenter, placebo-controlled trial before routine use of hydrocortisone for RAI in cirrhotic patients with sepsis can be recommended. Fernández J, Escorsell A, Zabalza M, et al. (Liver Unit, Hospital Clinic, Barcelona, Spain). Adrenal insufficiency in patients with cirrhosis and septic shock: effect of treatment with hydrocortisone on survival. Hepatology 2006;44:1288–1295. Bacterial infection continues to be a major cause of death among patients with cirrhosis. The demonstrated survival benefit of prophylactic antibiotic therapy in cirrhotic patients with gastrointestinal hemorrhage underscores the significant impact that bacterial sepsis has in this patient population (Hepatology 1999;29:1655–1661). Currently, management of patients with sepsis centers on supportive care. Many areas of uncertainty remain and strategies for management of sepsis need to be prospectively evaluated in patients with advanced liver disease. Relative adrenal insufficiency (RAI) has been linked to refractory shock, resistance to vasoconstrictor medications, and increased mortality in patients with sepsis (N Engl J Med 2003;348:727–734). A recent meta-analysis of data from multiple clinical trials demonstrates a significant survival benefit from stress-dose steroid replacement in patients with septic shock not necessarily associated with cirrhosis (BMJ 2004;329:480–489). In this setting, Fernández et al prospectively assessed the impact of hydrocortisone treatment in cirrhotic patients with septic shock and abnormal short corticotropin test responses (Hepatology 2006;44:1288–1295). Twenty-five consecutive cirrhotic patients with a diagnosis of septic shock admitted to a specialized liver intensive care unit (ICU) from June 2004 to January 2006 were assessed for adrenal insufficiency with a short corticotropin test within 24 hours of ICU admission. Immunocompromised patients, those with advanced hepatocellular carcinoma, and those with a history of steroid therapy within the past 6 months were excluded. Patients with RAI—defined as a baseline cortisol concentration <15 μg/dL and/or increase of plasma cortisol concentration of <9 μg/dL 1 hour after corticotrophin administration in patients with a baseline cortisol concentration <35 μg/dL—were treated with 50 mg of intravenous hydrocortisone every 6 hours. Hydrocortisone was tapered once vasopressor medications were no longer required. Patients in the prospective cohort were compared with a historical control group of the last 50 consecutive patients with cirrhosis and sepsis in whom adrenal function was not assessed and no steroid therapy was given. All patients were treated with a standard protocol that differed only in the assessment of adrenal function and included antibiotic therapy, central venous and urinary catheterization, goal-directed plasma volume expansion and vasopressor therapy, albumin administration in cases of spontaneous bacterial peritonitis, and continuous venovenous hemofiltration and mechanical ventilation as indicated. Within the prospective cohort, 17 of 25 patients had RAI. A trend toward higher rates of RAI in patients with Child–Pugh class C cirrhosis was noted (P = .08), although no statistically significant differences in clinical characteristics were identified between patients with and without adrenal insufficiency. Compared with historical controls, patients in the prospective cohort had similar baseline clinical characteristics, except for a statistically significant lower serum albumin concentration (2.3 ± 5 g/dL vs 2.7 ± 5 g/dL; P < .05). Clinically important outcomes were improved in the prospective cohort compared with controls, including resolution of shock (96% vs 58%; P = .001), time to shock resolution (48% vs 18% within 48 hours of ICU admission; P =.01), ICU and hospital survival (68% vs 38% [P = .03]; 64% vs 32% [P = .003], respectively). Two patients treated with hydrocortisone died of fungal infection; no other serious adverse events were reported in the hydrocortisone treatment group. CommentThe sepsis syndrome is a major precipitant for acute-on-chronic liver failure, thereby leading to hypotension, worsening hepatic function, hepatorenal syndrome, multisystem organ failure, and death (Gut 2005;54:718–725). Broad-spectrum antibiotic therapy and early goal-directed therapy (with target values for targets for central venous oxygen saturation, central venous pressure, and mean arterial pressure) have been demonstrated to improve short-term survival in patients with sepsis not necessarily associated with cirrhosis (N Engl J Med 2001;345:1368–1377). Strategies for the management of septic shock need to be prospectively evaluated in patients with liver disease, among whom infection remains a major cause of death.The hemodynamic derangements of cirrhosis mirror those of septic shock and adrenal insufficiency. Recent data have emerged regarding the association of RAI in patients with liver dysfunction and sepsis. In patients with acute liver failure, RAI was present in 62% of patients and was associated with lower transplant-free survival (Hepatology 2002;36:395–402). A series of cirrhotic patients with septic shock revealed that RAI was present in 51.5% of patients and was associated with significantly higher ICU and hospital mortality rates. Cumulative 90-day survival rates were 15.3% and 63.2% in adrenal-insufficient and normal adrenal function groups, respectively (Hepatology 2006;43:673–681).The rates of RAI reported by Fernández et al were similar to, if not higher than, previously reported rates. In concordance with prior reports, RAI was more common in patients with more advanced liver disease. Whether this is directly related to impaired hepatic cholesterol synthesis and steroid metabolism, the proinflammatory cytokine milieu, or whether adrenal insufficiency is simply a marker of multisystem organ failure remains unclear. That RAI is frequently present in patients with acute liver failure without evidence of sepsis (Hepatology 2002;36:395–402) suggests that hepatocellular dysfunction may play a significant role. Notably, a specific bacterial etiology for the sepsis syndrome was not identified in 36% and 24% of patients in the prospective and control groups, respectively. The role of RAI and steroid therapy in cirrhotic patients with noninfectious precipitants for the systemic inflammatory response syndrome (such as gastrointestinal hemorrhage or surgery) remains to be determined.Although the overall mortality rate was improved in patients receiving steroid therapy, 2 of 17 patients receiving hydrocortisone therapy developed fatal fungal infections. The site of infection and portal of entry was not reported by the authors. A prior study utilizing higher doses of steroid replacement (hydrocortisone 300 mg/d) reported an increased rate of resistant bacterial infections (Liver Int 2003;23:71–77). No resistant bacterial infections were noted in the current study.The major limitation of the current study is the absence of a contemporary, blinded, placebo-controlled arm. Although the use of clinical management protocols in a single, specialized liver ICU minimized the variability in care between the prospective and historical cohorts, it is possible that the improved survival in the contemporary (prospective) cohort may be due to an unidentified factor other than adrenal replacement therapy. Although this study provides valuable insight into the care of the septic cirrhotic patient, the results need to be validated in a larger, multicenter, placebo-controlled trial before routine use of hydrocortisone for RAI in cirrhotic patients with sepsis can be recommended. The sepsis syndrome is a major precipitant for acute-on-chronic liver failure, thereby leading to hypotension, worsening hepatic function, hepatorenal syndrome, multisystem organ failure, and death (Gut 2005;54:718–725). Broad-spectrum antibiotic therapy and early goal-directed therapy (with target values for targets for central venous oxygen saturation, central venous pressure, and mean arterial pressure) have been demonstrated to improve short-term survival in patients with sepsis not necessarily associated with cirrhosis (N Engl J Med 2001;345:1368–1377). Strategies for the management of septic shock need to be prospectively evaluated in patients with liver disease, among whom infection remains a major cause of death. The hemodynamic derangements of cirrhosis mirror those of septic shock and adrenal insufficiency. Recent data have emerged regarding the association of RAI in patients with liver dysfunction and sepsis. In patients with acute liver failure, RAI was present in 62% of patients and was associated with lower transplant-free survival (Hepatology 2002;36:395–402). A series of cirrhotic patients with septic shock revealed that RAI was present in 51.5% of patients and was associated with significantly higher ICU and hospital mortality rates. Cumulative 90-day survival rates were 15.3% and 63.2% in adrenal-insufficient and normal adrenal function groups, respectively (Hepatology 2006;43:673–681). The rates of RAI reported by Fernández et al were similar to, if not higher than, previously reported rates. In concordance with prior reports, RAI was more common in patients with more advanced liver disease. Whether this is directly related to impaired hepatic cholesterol synthesis and steroid metabolism, the proinflammatory cytokine milieu, or whether adrenal insufficiency is simply a marker of multisystem organ failure remains unclear. That RAI is frequently present in patients with acute liver failure without evidence of sepsis (Hepatology 2002;36:395–402) suggests that hepatocellular dysfunction may play a significant role. Notably, a specific bacterial etiology for the sepsis syndrome was not identified in 36% and 24% of patients in the prospective and control groups, respectively. The role of RAI and steroid therapy in cirrhotic patients with noninfectious precipitants for the systemic inflammatory response syndrome (such as gastrointestinal hemorrhage or surgery) remains to be determined. Although the overall mortality rate was improved in patients receiving steroid therapy, 2 of 17 patients receiving hydrocortisone therapy developed fatal fungal infections. The site of infection and portal of entry was not reported by the authors. A prior study utilizing higher doses of steroid replacement (hydrocortisone 300 mg/d) reported an increased rate of resistant bacterial infections (Liver Int 2003;23:71–77). No resistant bacterial infections were noted in the current study. The major limitation of the current study is the absence of a contemporary, blinded, placebo-controlled arm. Although the use of clinical management protocols in a single, specialized liver ICU minimized the variability in care between the prospective and historical cohorts, it is possible that the improved survival in the contemporary (prospective) cohort may be due to an unidentified factor other than adrenal replacement therapy. Although this study provides valuable insight into the care of the septic cirrhotic patient, the results need to be validated in a larger, multicenter, placebo-controlled trial before routine use of hydrocortisone for RAI in cirrhotic patients with sepsis can be recommended.