Hematopoietic stem cell transplantation (HSCT) is one of the greatest advances in modern medicine, which can help patients recover normal hematopoietic and immune function. Therefore, it is an important treatment for hematological malignancies, solid tumors, genetic metabolic diseases, primary immunodeficiency diseases, etc.(Bazinet and Popradi, 2019; Carreras et al., 2019; Castagnoli et al., 2019). Although in the past few decades, with the advent of a series of new drugs and the progress of transplantation technology, the success rate of HSCT has been significantly improved. However, infection remains one of the most common complications and early causes of death after transplantation (Esquirol et al., 2021).

Due to the use of high-dose radiotherapy, chemotherapy and immunosuppressive agents, almost all transplant patients will suffer from agranulocytosis, impaired mucosal barrier and weakened immune function. In addition, factors such as central venous catheter implantation, parenteral nutrition and long-term hospitalization greatly increase the chance of bacterial invasion (Ferreira et al., 2018; Cao et al., 2021). In a review of literature, we found that the distribution pattern of bacterial infections in HSCT recipients reported in different countries varied widely: the ratio of gram-positive bacteria (GPB) to gram-negative bacteria (GNB) ranged from 85% vs.15% to 26% vs.74% (Balletto and Mikulska, 2015). Even in the same country, the data reported by different transplant centers are quite different. For example, in China, Cao WJ et al. found that GNB accounted for an absolute dominant position (71.67%) in bloodstream infections (BSI) in HSCT recipients (Cao et al., 2021). However, the statistical results of Wang D et al. showed that the proportion of GPB and GNB infections was close (49.06% vs. 50.94%) (Wang et al., 2019). Although some epidemiological and clinical data related to bacterial infections in HSCT populations have been previously published, most of them were concentrated in a single category (such as carbapenem-resistant Klebsiella pneumoniae and methicillin-resistant Staphylococcus aureus), few studies specifically focused on mixed bacterial infections (MBI), which hinders the effective management and prognosis improvement of transplant patients.

In this study, we analyzed the distribution, antibiotic resistance and prognosis of MBI in HSCT recipients in the hematology ward of a large general hospital in China over the past 10 years, and paid special attention to the risk factors of MBI acquisition and death, which provided valuable information for a comprehensively understanding of the characteristics of post-transplant infection and adjusting anti-infection treatment strategies.

HSCT recipients are at high risk of infection (Styczynski et al., 2020). MBI often leads to poor prognosis due to the variety of strains involved and the complex mechanisms of drug resistance. Our results indicated that compared with single bacterial infection, HSCT recipients with MBI had a longer platelet engraftment time and a higher risk of adverse outcomes such as sepsis shock and death. The time interval from diagnosis to transplantation > 180 days and ICU admission after transplantation were independent risk factors for MBI acquisition.

In previous reports, we have repeatedly emphasized the disadvantages of prolonged transplantation interval: disease recurrence, repeated chemotherapy, bone marrow suppression, immunosuppression, mucosal damage, impaired organ function, excessive psychological pressure, delayed donor cell implantation, etc. (Liu et al., 2022; Jia et al., 2023). Kikuchi M et al. also confirmed that the interval between diagnosis and transplantation was a risk factor for BSI in the pre-engraftment period of HSCT patients (Kikuchi et al., 2015). That’s why the guidelines recommend early HSCT for patients with hematological malignancies who meet physical conditions (Blackburn et al., 2019). However, limited by age, financial burden, family support, medical resources/level and other factors, there are still a lot of patients who cannot receive transplantation at the most appropriate time. In particular, as the birth rate declines and the population ages, the number of HLA-matched sibling donors is decreasing. Fortunately, with the improvement of haplotype HSCT technology in recent years, the problem of donor source has been greatly solved, resulting in a rapid increase in the number of HSCT cases (Xu et al., 2018).

Undoubtedly, ICU gathered severe patients from different departments in the whole hospital, including trauma, post-surgery, multiple organ failure, etc., so the source of pathogens is very extensive. In addition, invasive procedures such as intravenous catheterization, urethral catheterization, tracheal intubation, incision and drainage further increase the risk of MBI. Reports indicate that 5%-55% of patients will be admitted to ICU after HSCT (Afessa and Azoulay, 2010). Sepsis/septic shock, acute respiratory failure, acute renal failure, mind change, GVHD and veno-occlusive disease are the most common causes. Studies from Barlas T et al. also showed that the incidence of BSI/catheter-related BSI in HSCT patients admitted to ICU was significantly increased (Barlas et al., 2021).

Engraftment period > 20 days, CRRT and septic shock were 3 independent risk factors for death in patients with MBI. Kikuchi M et al. reported that the cumulative incidences of pre- and post- engraftment BSI were 38.9% and 17.2%, respectively. Pre-engraftment BSI was associated with an increased risk of death. The OS at day 180 for patients with or without pre-engraftment BSI was 70.0% and 82.7% (Kikuchi et al., 2015). It proves that delayed engraftment will increases the probability of infection-related death in transplant patients. Not only that, but the extension of engraftment time means patients need to experience a longer period of agranulocytosis. During this period, their immunity was extremely low, unable to resist the invasion of various pathogens, and some opportunistic infections also break out at this time (Omrani and Almaghrabi, 2017; Suwabe et al., 2022). Moreover, infection may induce or aggravate GVHD, leading to transplant failure and even patient death (Kolb et al., 2018; Mori et al., 2018).

CRRT is a continuous, slow, isotonic extracorporeal circulation technique which can help the body remove excess water and certain solutes (endotoxin, inflammatory cytokines, middle molecular substances, etc.) (Tandukar and Palevsky, 2019). It is an important part of blood purification technology. After 2000, the clinical application of CRRT was no longer limited to renal replacement, but expanded to multiple organ support therapy (Ricci et al., 2018). However, a series of complications may occur during CRRT, such as bleeding, thrombosis, infection, nutritional loss, hypotension, hypothermia, electrolyte and acid-base imbalance (Elbahlawan et al., 2021; Kovvuru and Velez, 2021). Especially in the vulnerable population after transplantation, these risks are further amplified. For example, low platelet counts will increase the probability of bleeding or even lead to hematoma formation during subcutaneous puncture and Seldinger catheterization. Vascular connection device, rehydration connection device, extracorporeal circulation and prolonged indwelling of venous catheters also greatly increased the risk of infection. Studies have shown that pipeline joints and exposed parts of pipelines are the most common invasion sites of bacteria (Zhang et al., 2016; Fisher and Mokrzycki, 2023). In addition, excessive ultrafiltration volume and replacement fluid composition deviation can also affect the patient’s vital signs, dealing a fatal blow to the already serious illness. Neofytos D et al. found in a multicenter prospective analysis that hemodialysis was associated with reduced survival due to infection in adult HSCT recipients (Neofytos et al., 2009). Elbahlawan L et al. also found that the mortality rate of children undergoing CRRT after HSCT was as high as 52%-65% (Elbahlawan et al., 2021). It suggests that doctors should strictly control the indications of CRRT and closely monitor the possible complications in the process to ensure patient safety.

Sepsis shock is a systemic infection with acute onset and critical condition (Cecconi et al., 2018). Because of a large amount of inflammatory cytokines entering the blood and insufficient tissue perfusion, resulting in severe circulatory disorders and abnormal cell metabolism. The patient presents with persistent hypotension and hyperlactatemia that are difficult to correct. As the most serious complication of infection, septic shock has become one of the major global healthcare problems, affecting millions of people each year and causing one-third to one-sixth of deaths (Evans et al., 2021). Our data showed that the incidence of septic shock in HSCT recipients with MBI was 31.2%. Since the current research on post-transplantation infection is mostly focused on a single strain or a certain category, such as GNB or GPB, the available literature on MBI is limited, and we cannot fully compare this data with other transplantation centers for the time being. A recent report by Facchin G et al. pointed out that although the incidence of septic shock caused by polymicrobial bloodstream infections in HSCT patients was only 8%, the associated mortality rate was as high as 75% (Facchin et al., 2022).

In this study, CoNS + E. coli/A. baumannii/P. aeruginosa and K. pneumoniae + A. baumannii were the main combination of mixed infections. Consistent with previous reports, CoNS and E. coli were the most frequent isolates of GPB and GNB, respectively (Macesic et al., 2014; Forcina et al., 2017). What is different is that Trifilio S et al. found that the combination of clostridium difficile + gram-positive/negative bacteria was the most common pattern of polymicrobial bacterial infection in HSCT patients, and was thought to be associated with an increased risk of acute gastrointestinal graft-versus-host disease (Trifilio et al., 2015). Sy A et al. showed that Staphylococcus species (22.0%), Pseudomonas (12.1%) and Escherichia Coli (8.9%) were the main components of late polymicrobial bacterial infection in transplant recipients (Sy et al., 2021). We speculate that this may be related to a variety of factors such as primary disease, pre-treatment regimen/intensity, antibiotic prevention, HSCT type and so on. Besides, the gradual transformation of major pathogenic bacteria worldwide may also contribute to this difference (Sahitya et al., 2021).

CoNS is derived from the inherent symbiotic flora of skin and gut, and can be migrated or disseminated through exogenous channels such as catheters (Argemi et al., 2019). Our antibiotic susceptibility tests showed that CoNS was completely sensitive to vancomycin and linezolid, and slightly resistant to penicillin and aminoglycosides. This was somewhat different from what Niyazi D previously reported (Niyazi et al., 2022): They found CoNS resistance rates to penicillin, erythromycin, gentamicin, and ciprofloxacin all exceeding 60%, significantly higher than our statistical results. We speculate that this may be caused by the high proportion of methicillin-resistant CoNS in their study cohort. In contrast, the drug resistance of GNB was much more severe: only tigecycline and amikacin maintained relatively high sensitivity as a whole, but the anti-infection effect of the former against P. aeruginosa and the latter against S. maltophilia was also very poor. The resistance rates of P. aeruginosa and A. baumannii to carbapenems were more than 58% and 85%, respectively, higher than 37.9% and 63.6% reported by Averbuch D (Averbuch et al., 2017). It may be related to the type of primary disease, inappropriate empirical antimicrobial therapy and more multidrug-resistant strains. The above data show that GNB has a greater impact on MBI. Fortunately, in recent years, some new antibiotics, such as Ceftazidime avibactam, Ceftolozane/Tazobactam, Meropenem-vaborbactam, Imipenem/Cilastatin/Relebactam, etc., have been proved to have good efficacy and high safety against GNB, especially carbapenem resistant GNB (Pogue et al., 2020; Alosaimy et al., 2021; Clerici et al., 2021; Heo, 2021). Unfortunately, due to the large time span of this study, many antimicrobial susceptibility tests could not be performed at that time. Additionally, factors such as time to market, drug prices, and patients’ economic affordability limited the use of these novel antibiotics, which increased the difficulty of collecting drug sensitivity results, and ultimately made these data unable to be included in statistical analysis. Different from previous reports that the degree of HLA matching, aGVHD, mechanical ventilation and MDR increase the infection-related mortality of HSCT recipients, here, we did not find statistical differences in the above factors between the survival group and the death group (Forcina et al., 2017; Chueansuwan et al., 2018; Liu et al., 2018; Styczynski et al., 2020).

Admittedly, our study had some limitations. First, it is a single-center retrospective design, and the results may not be well generalized to other transplant centers due to population heterogeneity. Secondly, the number of patients included in this study was relatively small. Thirdly, all patients received antibiotic prophylaxis, but this practice was not completely uniform across the transplant center. Finally, although the prognostic model was adjusted for many key clinical covariates, we cannot exclude the influence of other unmeasured or temporarily unknown factors.

In summary, HSCT recipients with MBI had severe antibiotic resistance and high mortality. Engraftment period > 20 days, CRRT and septic shock were independent risk factors associated with mortality. Therefore, shortening transplant waiting time, promoting donor stem cell implantation as soon as possible, strictly controlling CRRT indications, and early identification and timely correction of shock may help improve the prognosis of HSCT patients.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by the Institutional Review Committee of Xiangya Hospital. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

YJ and YFL were involved in the study conception and design. YL, YFL, and XFC participated in the data management. All authors participated in the interpretation and statistical analysis of data. YFL wrote the draft and YJ revised it. All authors contributed to the article and approved the submitted version.