Pneumocystis jirovecci pneumonia (PCP) is an opportunistic fungal infection that can be life-threatening in immunodeficient patients (1). Most often affected are patients with HIV infection (2). However, in recent years, there has been an increase in the incidence of PCP in non-HIV patients (3, 4). Thus, various risk factors other than HIV have been identified including hematologic malignancies, immunosuppressive therapy, solid organ transplantations, and autoimmune diseases (3, 5).

Liver cirrhosis (LC) leads to an impairment of the immune system called cirrhosis-associated immune dysfunction (CAID) (6–9). The term refers to the main syndromic abnormalities of immune function, immunodeficiency and systemic inflammation that are present in LC. The natural course of advanced LC, regardless of its etiology, is often complicated by CAID with this constituting the pathophysiological hallmark of an increased susceptibility to bacterial infections (10).

Up to now, liver cirrhosis is not considered to be a typical risk factor for PCP and little is known about the mortality of PCP in patients with liver cirrhosis.

In the current study, we analyzed predictors for Pneumocystis jirovecii associated mortality in patients with Pneumocystis jirovecci infection treated at intensive care units (ICU) of a German university hospital.

To our knowledge, this is the largest study analyzing Pneumocystis jirovecci infection in patients with advanced LC. In our study, patients with LC and Pneumocystis jirovecci infection had a significantly higher mortality compared to infected patients without LC. It has to be noted that up to now only a few case series described Pneumocystis jirovecci infection in patients with LC (13, 14). In those case series, it has been suggested that LC might be an independent risk factor for PCP. Our data clearly indicate a heightened susceptibility to fungal infections. In our study, the majority of patients with LC presented with advanced LC with a mean Meld Score of 32.3. Furthermore, our data show a significant correlation of MELD score and mortality (r = 0.612, p < 0.001). We were also able to show that a cut-off of a Meld Score above 25 can identify patients with the highest mortality risk.

Advanced liver disease itself, as in our cohort can be considered as a cause of acquired immunodeficiency and a higher mortality (7, 15). There are up to now only a few case reports that underline the possible relevance of immunodeficiency in decompensated LC and the risk for opportunistic infections. In case of an acute alcoholic hepatitis, especially under treatment with steroids the threat of PCP is better accepted, although with a low evidence (16). A weakness of the immune system in patients with LC is well accepted and described as CAID (17). But this has been shown mainly for bacterial infections and those infections are the main triggers for an acute on chronic liver failure (ACLF). In our cohort patients with liver cirrhosis had high MELD score, Child-Pugh score and SOFA score values which suggest high grade systemic inflammation in patients with decompensated cirrhosis. Patients with end stage liver disease and sepsis may develop acute-on-chronic liver failure (ACLF), that results in an immune exhaustion, characterized by a dysregulation of immune effector cells (6, 8). This includes a reduced potential of CD4+ lymphocytes and NK-cells in cirrhosis (18, 19). Risk for Pneumocystis jirovecciinfection is associated with low lymphocyte count and mortality of Pneumocystis jirovecci infected patients is associated with both, low lymphocyte and low NK cell counts (20, 21). Therefore, the immune-dysfunctional pattern of patients with sepsis in advanced liver disease may support Pneumocystis jirovecci infection. Even a low colonization with Pneumocystis jirovecci, as in some patients of our study especially in context with invasive ventilation represents a serious risk for PCP. Our results clearly show that advanced LC is a risk factor for Pneumocystis jirovecciin the airways associated with a high mortality. As a clinical and therapeutic consequence, prophylactic therapy as for other immunodeficient patients should be considered at an early stage of the infection.

Our patients were extremely fragile, and the high mortality rate in our study could be related not only to Pneumocystis jirovecci, but to the severe co-morbidities and organ failures. This is underlined by the fact that patients with a high SOFA score independently of the diagnosis LC showed the highest mortality rate. Other reasons might be concomitant infections and/or liver transplant delay. Of note, the mean MELD score of the patients with LC in our study was 32 points, which correspondents with a 90-day mortality of about 60% (22, 23).

Our study has several limitations, mainly due to its retrospective character. Thus, we cannot provide data about the immune status of our patients. Furthermore, there was no specific therapy protocol at which point and for what criteria specific therapy for Pneumocystis jirovecci had to be started. In our study therapeutic decision was based on clinical decisions and correlated with microbiological diagnostics and existing clinical risk scores. Furthermore, in our study patients from two different ICU wards were included. Therefore, the composition of the patient population is heterogeneous with a potential bias for patient outcome.

The strength of our study lies in the substantial number of patients with advanced liver cirrhosis (LC) and Pneumocystis jirovecci infection. Our data clearly demonstrate that Pneumocystis jirovecciinfection is a serious opportunistic infection in patients with LC. Additionally, our findings support the concept of cirrhosis-associated immune dysfunction (CAID) and show the increased susceptibility of patients with liver cirrhosis to fungal infections. This susceptibility was correlated with the Child-Pugh and MELD scores. We established a MELD score cut-off above 25 to identify patients with the highest risk of mortality.

In conclusion, our study demonstrates that Pneumocystis jirovecci infection in patients with advanced liver cirrhosis is associated with high mortality. Therefore, Pneumocystis jirovecciinfection should be included in the differential diagnosis for patients with advanced liver cirrhosis and acute-on-chronic liver failure (ACLF). We recommend initiating early antifungal Pneumocystis jirovecci treatment for this vulnerable patient cohort.

Furthermore, we emphasize the importance of early recognition of immune dysregulation as a consequence of cirrhosis and the early identification of Pneumocystis jirovecci infection. We also underscore the role of early treatment of Pneumocystis jirovecci infections in cirrhosis to improve patient outcomes in both transplant and non-transplant settings.

Our data contribute a new phenotypic manifestation, specifically Pneumocystis jirovecciinfection, to the mechanisms of cirrhosis-associated immunosuppression and may lead to immune-targeted therapies to reduce susceptibility to infection and mortality in patients with cirrhosis.