Delisting of liver transplant candidates because of bacterial sepsis

It is estimated that bacterial infections are present in 30% of patients with cirrhosis upon hospital admission, which is at least 4- to 5-fold higher than in the general population.1 Spontaneous bacterial peritonitis (SBP) and urinary infections are the most frequent infections found in patients with cirrhosis; others include pneumonia, cellulitis, and bacteremia.1, 2 Approximately 30% of bacterial infections are community acquired, 30% are health care associated (HCA), and 35% to 40% are nosocomial.2, 3 Risk factors include poor liver function, variceal bleeding, low-protein ascites, and prior episodes of SBP.1-3 Historically, gram-negative bacteria were implicated in close to 80% of patients of SBP. The epidemiologic pattern of SBP has experienced drastic changes over the past 2 decades. The increasing presence of gram-positive bacteria (mostly Streptococcus viridans, Staphylococcus aureus, and Enterococcus fecalis) reflects invasive procedures during hospitalization.3 There is also an alarming increase in infections caused by multi-resistant bacteria in cirrhosis. The most common are extended-spectrum β-lactamase–producing Enterobacteriaceae, nonfermentable gram-negative bacilli (ie, Pseudomonas aeruginosa), methicillin-resistant Staphylococcus aureus, vancomycin-susceptible Enterococcus, and vancomycin-resistant Enterococcus. Although resistance varies among different geographical areas, these multiresistant bacteria are more frequently isolated in nosocomial infections (35%-39%) compared with HCA (14%-20%) or community-acquired episodes (0%-4%).2 Risk factors for multiresistant bacterial infections include nosocomial acquisition of infection, β-lactams administration within the last 3 months, and infection by multiresistant bacteria in the last 6 months, all common events in patients with cirrhosis listed for liver transplantation.2 These infections by multiresistant bacteria not only diminish the efficacy of current first-line regimens but also increase the risk of treatment failures of second-line therapies.4-7

Patients with cirrhosis have an increased risk of developing bacterial infections, sepsis, sepsis-induced organ failure, and death. Therefore, it is not surprising that mortality rates in this patient population reach 30% in the first month and reach more than 60% after 1 year.8, 9 In Western countries, bacterial infection is the main precipitating event of acute-on-chronic liver failure (ACLF), an important entity characterized by the rapid deterioration in a patient with cirrhosis, frequent requirement for pressor and other organ support, and high short-term mortality.10 Early liver transplantation seems to be an adequate therapeutic option for ACLF in selected patients11; however, if these patients are listed for liver transplantation, many will be delisted because of uncontrolled infection or deterioration of their condition that precludes the surgery even if the patient is appropriately treated with antibiotics.

In this issue of Liver Transplantation, Reddy et al.12 from the North American Consortium for the Study of End-Stage Liver Diseases delve into this matter. In this study, the authors analyzed patients with cirrhosis with infections. Of the 413 patients included, the study focuses on 136 who were listed for liver transplantation. Of these, 42% were delisted or died, 35% underwent transplantation, and only 23% achieved transplant-free survival 6 months after the hospital admission for an infection. In the delisted patients, the number of organ system failures at index hospitalization was highest (19% had 3 or 4 individual organ failures). In fact, an increasing number of organ failures, particularly respiratory failure and brain failure, was associated with an increased risk of delisting/death. Of the 45 patients transplanted (of 47) with survival data, 39 had <2 organ system failures, and 6 had 3 to 4 organ system failures. Survival following transplantation was statistically higher compared to those who were delisted or died. Patients who underwent transplantation had fewer organ system failures compared to those who were delisted and/or died. Thus, at least 3 organ failures had occurred in 9% of those ultimately transplanted, in 19% of those who died or were delisted, but in only 3% of those who achieved transplant-free 6-month survival.

The results of this analysis are unfortunately representative of what occurs in most transplant centers, namely, infected patients with cirrhosis and extrahepatic organ failure are at a significantly higher risk of dropping off the transplant list or dying before liver transplantation compared to those without infection. A limitation of this analysis is the absence of predefined criteria for delisting. This may vary between centers, in particular when the decision is related to infections that are treatable. The authors were unable to determine the therapy of these infections, and thus, it is difficult to know if these patients, albeit treated with antibiotics, had a community-acquired, a HCA, or nosocomial infection. However, the information on organ system failure is highly relevant as these criteria could potentially be used as an objective delisting rule in infected patients. In the study, if there were 2 associated organ failures, patients were still transplanted whereas if there were 3 or 4, most programs considered them too sick for transplantation.

What should we learn from these findings? First, the type of infection (aside from urinary tract infection, which had a low impact on delisting) does not seem to play a major role on the decision to delist, although listed candidates were more likely to be on SBP prophylaxis; second is that more than 2 organ failures are adverse prognostic factors with both respiratory failure and brain failure being independent predictors of delisting; and third is that a high proportion of patients will end up delisted after an infection. Information on the use of norfloxacin as a prophylactic agent in these patients was not available. In patients with advanced liver disease, long-term administration of norfloxacin orally (400 mg/day) prevents the development of hepatorenal syndrome and improves survival,13 reflecting its ability to prevent bacterial translocation, suppress proinflammatory cytokines, and improve circulatory function.

In summary, the authors have demonstrated that a large proportion of listed patients with cirrhosis drop off the liver transplant list because of infections and that this is not necessarily driven by the type of infection per se but by the number of organ failures related to the infection. Although current empiric regimens are a good option for community-acquired bacterial infections, in nosocomial infections empiric treatment should be adapted according to local patterns of resistance. Therefore, all efforts should be focused toward prevention of these infections and use of appropriate antibiotic therapy in those admitted to the hospital with bacterial infections.