Infection and sepsis significantly impact the prognosis of LT patients, even in an era where 5-year survival rates are nearly 80% [2, 3, 15]. Post-LT infections caused by antibiotic-resistant bacteria have become increasingly common, including the lethal carbapenem-resistant gram-negative bacteria. A decade ago, we previously reported on the spectrum of MDR gram-negative bacterial infections post-LT [4]. However, most studies have focused only on specific infectious pathogens and there is a lack of comprehensive analysis on serious post-LT infections such as sepsis, which plays a crucial role in determining prognosis. Therefore, this study aims to provide a comprehensive analysis of infection patterns and resistance profiles while simultaneously investigating risk factors for sepsis in LT patients. Furthermore, building upon our previous study conducted from 2007–2010, we aim to demonstrate the shifting pattern of bacterial infections in recent decades within the field of LT.
In our previous study, the incidence of bacterial infection was as high as 59.9%. Specifically, 29.5% were attributed to gram-positive bacteria and 30.4% to gram-negative bacteria (Fig. 1A) [4]. In this large cohort study, a total of 20.1% of patients were infected by either bacteria or fungi (Fig. 1B). Although there has been a remarkable decrease in the overall bacterial infection rate over recent years, the ratio between gram-negative and gram-positive bacteria remains nearly unchanged at approximately 1:1 (Fig. 1A and B). This reduction in infection rates can be attributed to advancements in infection prevention protocols, immunosuppressive therapies, surgical procedures, and peri-operative management practices. The pathogen spectrums of 2007–2010 (MDR gram-negative bacteria) and 2010–2023 were shown in Fig. 1C and D. Among specific gram-negative pathogens identified, the top three were A. baumannii (36.7%), K. pneumoniae (26.7%), and E. coli (15.6%), which aligns with previous reports from years ago [4]. However, it is noteworthy that both A. baumannii and K. pneumoniae have shown an increase of approximately 6–7%, while E. coli has decreased by 8.3%. Furthermore, carbapenem resistance was observed in up to 31.1% of gram-negative bacteria including A. baumannii (42.4%), K. pneumoniae (29.2%), E. coli (7.1%), P. aeruginosa (33.3%) and S. maltophilia (100%).
Without a doubt, bacterial infection is the most prevalent post-LT infection. The incidence of bacterial infection varies among different transplantation centers and eras. Previous studies have reported incidences of post-LT infection ranging from 18.4% [16], 26.6% [17], 44% [18], to 68.6% [19] in these centers, respectively. In our study, the infection rate was determined to be 20.1%. Notably, the spectrum of infections was significantly associated with geographical regions and environmental factors, while perioperative management further influenced post-LT infections' occurrence and characteristics. Therefore, studying the spectrum of post-LT infections within a single center holds great significance. This large-scale study conducted at a single center in East China successfully revealed the regional spectrum of post-LT infections.
Infections caused by MDR bacteria, particularly carbapenem-resistant gram-negative bacteria, have emerged as a significant public health challenge due to limited antibiotic options and high case-fatality rates [20]. The Centers for Disease Control and Prevention (CDC) has identified CRE and CRAB as two of the five urgent threats to public health (https://www.cdc.gov/drugresistance/pdf/threats-report/2019-ar-threats-report-508.pdf). Additionally, the World Health Organization (WHO) has classified CRAB, CRPA, and CRE as Priority 1 (critical) pathogens requiring research, discovery, and development of new antibiotics [21]. According to the most recent data from CHINET (www.chinets.com), approximately 20% of P. aeruginosa isolates are carbapenem-resistant along with 78% of A. baumannii isolates and 30% of K. pneumoniae isolates. Overall, our study found that 31.1% of GNB were resistant to carbapenems. Furthermore, we observed common resistance among post-LT infections caused by MRSA, MRCNS, VRE, and ESBLs. Importantly, LT patients infected with these drug-resistant pathogens had a higher likelihood of developing sepsis (OR 2.351; P = 0.033). Previous studies have reported that infection with MDR pathogens significantly increases the risk of inappropriate empiric therapy administration and mortality [22–24]. Based on our findings presented in this study, we propose that the association between infection with MDR pathogens and increased mortality may be attributed to a higher incidence rate of subsequent sepsis.
Other independent risk factors for infection and sepsis in our study include advanced age, prolonged mechanical ventilation, extended ICU stay, and elevated bilirubin levels. Advanced age has been widely recognized as a well-known risk factor for infection and has been consistently identified in previous studies [9, 25]. In our previous study [4], we also established prolonged mechanical ventilation as an independent risk factor for both GNB infection and MDR organisms. Here, we further demonstrate that prolonged mechanical ventilation is associated with an increased risk of sepsis development among recipients. LT patients are particularly susceptible to sepsis due to various factors such as cirrhosis-associated immune dysfunction, frequent hospital admissions, multiple antibiotic courses, lengthy stays in the ICU, and immunosuppressive therapies [2, 26]. Yoshizumi T et al. reported that a MELD > 15 was an independent risk factor for bacterial sepsis development [27]. Regarding MELD score, our findings indicate that bilirubin levels exceeding 90 µmol/L constitute a significant risk factor for the development of sepsis. Furthermore, other studies have also reported an association between elevated bilirubin levels, sepsis occurrence, and subsequent outcomes [28, 29]. Notably, bilirubin > 90 µmol/L serves as a more specific and modifiable indicator of liver function impairment. This finding suggests the importance of implementing jaundice-reducing treatments prior to LT to mitigate the occurrence of post-LT sepsis. In our previous study, we identified post-LT without the administration of prednisone as an independent protective factor against GNB infection. In contrast, in this study, the use of post-LT prednisone showed only a non-significant trend. This discrepancy may be attributed to a significantly lower incidence of infection or advancements in perioperative management.
There are several limitations inherent to this retrospective study. Firstly, the retrospective nature of the study inherently imposes certain limitations. Secondly, it is important to note that this study was conducted at a single center. While conducting a multi-center study on infection patterns may be impractical, it would be valuable to validate and compare these findings using data from other transplantation centers. Finally, the correlation between ICI and infection after LT remains uncertain and warrants further investigation; however, it should be noted that this study has a limited sample size.