The Bacteremia Caused by Non-Lactose Fermenting Gram-Negative Bacilli in Solid Organ Transplant Recipients

Definitions for MDR

In the present review, MDR was defined as acquired non-susceptibility to at least one agent in three or more antimicrobial categories and XDR was defined as non-susceptibility to at least one agent in all but two or fewer antimicrobial categories (i.e., bacterial isolates remain susceptible to only one or two categories) according to the criteria introduced by Magiorakos et al. [38]. According to this definition, NLF GNB producing a carbapenem-hydrolyzing β-lactamase are considered to be MDR isolates given they are not only resistant to all β-lactam antimicrobials, but also frequently to other classes of antimicrobials, such as aminoglycosides and fluoroquinolones.

Relevant morbidity and microbiological spectrum of NLF GNB in SOT with bacteremia

The incidence rates of NLF GNB depend on the type of organisms and transplantation. Previous studies reported that Pseudomonas spp. [1–5,7–10,12–30,32–34,39], Acinetobacter spp. [2,5,6,8–12,14–16,19,20,25,30,33,39], S. maltophilia [3,5,7,8,10,13,14,16,19–21,30,33], B. cepacia [5,10,14,24], and Chryseobacterium spp. [2,11,20] were responsible for 2.3%–38.9%, 0.8%–15.9%, 0.3%–14.9%, 0.4%–8.5% and 1.6%–4% of all pathogens causing BSIs respectively, in SOT recipients, as shown in Table 1. Other investigators found that Pseudomonas and Acinetobacter spp. accounted for 21.1% of pathogens among BSI SOT recipients with septic shock [40] and P. aeruginosa for 13% of gram-negative bacteria causing BSI following SOT [41].

As also shown in Table 1, as far as the type of transplantation is concerned, the incidence rates in recipients with NLF GNB bacteremia have been reported to range from 1.6% to 38.9% of kidney [11,22,27–30], 1.6% to 21.8% of heart [20,22–24,30], 17.6% to 37.3% of lung [5,14,30], 9.1%–21.1% of solitary pancreas or simultaneous kidney-pancreas [6,30], 2.5%–8.3% of intestinal [4,32], 7.3% to 35% of liver [1–3,7–10,12,13,16,17,19,22,26,30,33], and 5.7% to 18.9% of SOT [15,18,21,25,39] recipients with BSIs, respectively.

Relevant mortality rates of NLF GNB bacteremia in SOT

The mortality rates of SOT recipients with NLF GNB bacteremia were also associated with the types of organisms and transplantation. As shown in Table 2, previous studies reported that the mortality rates of SOT recipients with Pseudomonas spp. [10,14,18,20,22,26,34,42], Acinetobacter spp. [20], S.maltophilia [20], B.cepacia [14], and Chryseobacterium spp. [20] bacteremia were 0%–57%, 13%, 33%, 20% and 0%, respectively. It is notable that the highest mortality rate 85.7% caused by P. aeruginosa bacteremia is from a 1971 study [28] where P. aeruginosa is almost invariably isolated from complicated cases in bacteremic kidney recipients.

Also, as shown in Table 2, the mortality rates in recipients with NLF GNB bacteremia have been reported to be 85.7% in kidney [28], 27% in heart [20], 30% in liver [26] and 0% to 47% in SOT [18,22] recipients with BSIs, respectively. Korvick et al. [42] and Bert et al. [10] documented that among liver transplant recipients with P. aeruginosa bacteremia, the 14-d and 15-d mortality rates after BSI were 30% and 28%, respectively. Husain et al. [14] found that the 28-d mortality rate after BSI was 33% in P. aeruginosa bacteremia and 20% in B. cepacia bacteremia in lung transplant recipients. In a 2009 study from the United States, the mortality rate of recipients with P. aeruginosa bacteremia was 42% among SOT and hematopoietic stem cell transplant (HSCT) recipients. ⁴³ A study conducted by Moreno et al. ³⁰ showed that in Spain, 13% of SOT and HSCT recipients with NLF GNB bacteremia died.

Antimicrobial resistance of organisms among SOT recipients with NLF GNB bacteremia

Previous studies [9,30] found that MDR NLF GNB ranged from 9.8% to 12.5% of all NLF GNB among SOT recipients with BSIs. Previous studies suggested that 33.3% of NLF GNB [1], 72.7%–92.8% of A. baumannii [8,30], and 16% of P. aeruginosa [10] were carbapenem-resistant. Bodro et al. [30,34] also claimed that 74.3% of P. aeruginosa were carbapenem- and fluoroquinolone-resistant and 63.2% were XDR in SOT recipients with bacteremia. Others [15,43] reported that from 40% to 43% of Pseudomonas spp. were MDR. MDR NLF GNB causing bacteremia among SOT were shown in Table 3.

Some studies also focused on drug-resistance of NLF GNB to commonly used antibiotics rather than MDR among SOT recipients with NLF GNB bacteremia. Previous studies suggested that 11.1% of NLF GNB in liver recipients [1] and 22.2% of Pseudomonas spp. in bacteremic heart recipients [24] were ciprofloxacin-resistant; Bert et al. [10] reported that resistance to ceftazidime was observed in 7.7% of P. aeruginosa, 100% of S. maltophilia, and 50% of A. baumannii, respectively. Among living donor liver transplant recipients with NLF GNB bacteremia, Kim et al [2] found 100% of Chryseobacterium meningosepticum (1/1) to be cephalosporin-resistant.

The risk factors for BSI and mortality caused by NLF GNB among SOT

Despite these high rates of morbidity and mortality, the predictors of NLF GNB bacteremia and its related mortality in SOT recipients remain ill-defined, with most studies being limited by small sample size and a lack of specificity to NLF GNB bacteremia. Korvick et al. [42] demonstrated that among liver transplant recipients with P. aeruginosa bacteremia, survival rates were substantially lower for patients with hypotension, on mechanical ventilators, and increasing severity of illness. Johnson et al. [43] established that among SOT and HSCT recipients, independent risk factors for MDR Pseudomonas BSI included previous transplantation, hospital-acquired BSI, and prior intensive care unit (ICU) admission, and the only independent predictor of mortality related to P. aeruginosa BSI was onset of BSI in the ICU. A recent study identified prior transplantation, nosocomial acquisition, and septic shock at onset as factors independently associated with XDR P. aeruginosa bacteremia among SOT recipients [34]. No more studies focusing on the association of clinical or laboratory parameters and the risk for NLF GNB bacteremia and its related mortality were found so far.

Predominant rods in SOT recipients with NLF GNB bacteremia and the tendency of sensitive NLF GNB toward resistant rods

Among those SOT recipients with NLF GNB bacteremia, P. aeruginosa [1,3,5,10,12,14,15,19,21,24,25,31,37], A. baumannii [8,9,11,20,30] or S. maltophilia [7,16] were reported to be predominant organisms. Some studies claimed that P. aeruginosa [17,18,22,23,26,27,28,29,32] or A. baumannii [6] was the only NLF GNB causing BSI among SOT recipients. Among these investigations, a study [32] reported that P. aeruginosa was the most common organism among all pathogens causing BSI in SOT recipients. Predominant NLF GNB rods in SOT recipients with bacteremia were illustrated in Table 1.

NLF GNB bacteremia has not appeared to be problematic in early studies of infections in liver recipients with only zero to nine episodes of P. aeruginosa bacteremia documented in four series [44–47]. However, the clinical importance of NLF GNB bacteremia has increased over the last two decades. Since 2000, more and more studies [9,14–16,18] reported that MDR or extensive-drug resistant NLF GNB were up to 30% or greater in SOT recipients with BSIs. A study of BSI liver recipients by Bodro et al. [39] and a study of BSI SOT by Kim et al. [8] found that 72.7% and 92.8% of A. baumannii were carbapenem-resistant, respectively. Bodro et al. [39] found that 74.3% of P. aeruginosa were carbapenem- and fluoroquinolone-resistant.

Sources of NLF GNB bacteremia

The most common source of NLF GNB bacteremia was the urinary tract in kidney transplant recipients [11,22,28,29]. The most common portal identified was the abdominal/biliary duct [1,8,9,10,17,22] or intravascular catheters [7,16,26] in liver transplant recipients and the abdominal/biliary duct [2,19] or surgical site [3] in living donor liver transplantation. The most common source of BSIs was the lung in lung transplant recipients [14,31] and lung [22,24] or intravascular catheters [20] in heart recipients. Central venous catheters were the most common origin identified in pediatric small bowel transplantation [4]. However, the most frequent source comprised the lung [40], the urinary tract [18,34,41], abdomen [21], and intravascular catheters [30] in SOT recipients with NLF GNB bacteremia. Table 4 showed the most common primary sources of NLF GNB bacteremia among SOT recipients.

The reason for the high rates and wide ranges for NLF GNB bacteremia and its related mortality

Potent desensitization and induction agents were reported to be related to increased risk of certain infections [48]. Some virulence factors intrinsic to NLF GNB had a negative impact on the outcomes of SOT. In addition, bacteremia caused by MDR NLF GNB was associated with greater mortality compared with the ones caused by non-MDR counterparts in SOT recipients [30]. Inappropriate therapy frequently present in NLF GNB bacteremia is also associated with greater mortality [49,50]. A study of the effect of carbapenem resistance on mortality in bacteremia caused by P. aeruginosa revealed that the worst outcomes were mainly because of the greater severity of the underlying disease rather than virulence factors of the organism [51]. Furthermore, NLF GNB bacteremias predispose patients to a substantial risk for septic shock, which is associated with greater mortality among SOT recipients [40]. Bacteremia, per se, may also be a marker for other factors that confer greater mortality, comprising poor graft function, surgical complications, and prolonged ICU stay [52].

The greater number of GNB bacteremia and the higher mortality rate in SOT population can be partly explained by the poor general condition, the disruption of the gastrointestinal barrier because of surgery, reinterventions, prolonged use of central venous catheters, administration of total parenteral nutrition post-transplantation, and the great immunosuppression required because of the immunity of the allograft [4]. The wide ranges for NLF GNB bacteremia and mortality, however, reflect geographic differences in the exposure to community and hospital-acquired pathogens, the variety of diagnostic criteria, the diversity of type, duration, and intensity of immunosuppressive drugs used, the different sample size, study period, and duration of follow-up, and the different prophylactic and therapeutic antimicrobial strategies at different transplant centers.

Treatment of NLF GNB bacteremia

Because the most frequent source was the lung, the urinary tract, the surgical incision, the abdomen/biliary duct, or intravascular catheters [30] in SOT recipients with NLF GNB bacteremia, source control through central venous catheter removal, urinary catheter removal, surgical incision debridement, abscess, the biliary tree and body fluid drainage are crucial factors associated with better outcomes of SOT recipients with NLF GNB bacteremia [53].

The antibiotics against MDR NLF GNB have several limitations including less clinical experience, greater incidence of adverse effects, less knowledge of the pharmacokinetics and pharmacodynamics of these second-line drugs, and, in most cases, are only in parenteral formulations. The incidence of adverse effects may be increased in SOT patients because of the concomitant use of nephrotoxic agents (such as calcineurin inhibitors), decreased glomerular filtration rate, or the need for renal replacement therapies [53].

Combination therapy may be beneficial to enhance antimicrobial activity, provide synergistic interactions, relieve adverse effects and minimize superinfections [54]. Treatment of MDR NLF GNB should comprise combination therapies including two to three classes of antibiotics according to resistance phenotypes. P. aeruginosa is a substantial nosocomial pathogen in all types of SOT recipients, being responsible for early post-transplant bacteremia [53]. For MDR P. aeruginosa infections, combination therapies including beta-lactam, aminoglycoside, and fluoroquinolone are recommended for 10 to 14 d [55,56]. A. baumannii is associated with prolonged hospital stay and invasive procedures in SOT patients. Many MDR A. baumannii have only demonstrated susceptibility to colistin and polymixin B and its congeners. The use of colistin in the transplant population has greater incidence of adverse events including renal and neurological toxicity [57,58]. For MDR B. cepacia infections, triple combinations including meropenem, aminoglycosides, and either ceftazidime or trimethoprim sulfamethoxazole are recommended. MDR Stenotrophomonas infections require high dose trimethoprim sulfamethoxazole combined with ceftazidime and levofloxacin [53].

Although promising new antimicrobial agents are available, P. aeruginosa and A. baumannii are difficult to treat. For example, Ceftolozane is a novel anti-pseudomonal cephalosporin, which has been combined with tazobactam and demonstrated activity against carbapenem-resistant P. aeruginosa [59,60]. Pivmecillinam is a pro-drug of mecillinam, a betalactam with specific activity against NLF GNB [53].

NLF GNB bacteremias, especially because of P. aeruginosa and A. baumannii, have been an increasing emerging problem among SOT recipients in the modern era. The diversity of NLF GNB bacteremia is influenced by several factors comprising medical surrounding, type of donor, transplantation and immunosuppressant, prophylactic antimicrobial strategies, and most particularly, regional bacterial epidemiology. Because of the difficulty of the treatment, MDR pathogens may be especially of great concern among patients with NLF GNB bacteremia after SOT.

In addition, adequate empiric antibacterial coverage during the period of the initial empirical antibacterial therapy while one is waiting for the offending organism identity and its antimicrobial susceptibility pattern is essential to obtaining a better prognosis for SOT recipients with NLF GNB bacteremia. Detailed knowledge of the local epidemiology of bacteremia in the transplant patients is, therefore, important during this critical period.

Control and prevention of NLF GNB bacteremia

NLF GNB mainly lead to hospital infections. It has been shown that restricted use of fluoroquinolones and antibiotic schedule rotation are important measures to prevent the emergence of MDR strains [61]. The measures to prevent NLF GNB bacteremia in SOT recipients are the same as those in non-immunosuppressed hospitalized patients. Preservation of kidney function, taking effective isolation measures, appropriate hand washing, restriction of the use of invasive devices and equipment, removing unnecessary catheter at the earliest possible time, and minimizing post-operative length of stay in ICU can reduce the incidence of NLF GNB bacteremia. Strict adherence to the principles of intravascular catheter management recommended by the U.S. Centers for Disease Control and Prevention (CDC) [62] can sharply reduce the incidence of catheter-related bacteremia [63].

Contact isolation is indicated for transplant recipients harboring MDR P. aeruginosa and B. cepacia. As nebulizers can potentially transmit B. cepacia, the use of these devices should include strict hygiene measures [53]. The duration of broad-spectrum antibiotics should be kept as short as possible and when results of blood cultures are available, antibiotics should be adjusted according to susceptibility patterns to avoid the emergence of antibacterial resistance. Given the role of immunosuppressive agents in lowering host reactivity to certain ambient organisms in SOT is undoubted, reduction or removal of immunosuppressive agents can prevent the onset of NLF GNB bacteremia when serious NLF GNB infections occur. Because corticosteroids can block transcription of cytokine genes and non-specific inhibition of T lymphocytes and macrophages, maintenance treatment without prednisone decreases the development of NLF GNB bacteremia.

Donor-derived MDR gram-negative infections have recently been reported [64]. Although organ donors with bacteremia should be ideally cured before organ procurement has been recommended, documentation of clearance of donor bacteremia is often not feasible in clinical practice attributable to the limited time window for procurement and incubation period required for cultures. The use of organs from donors with a history of colonization or infections with NLF GNB remains controversial and should be individualized [65].

The effects of infection control measures on reducing infection rates and mortality

Various strategies, such as effective isolation and antimicrobial prophylactic strategies, early removal of the urinary catheter, or applying CDC recommendations for the care of venous catheters, are concrete and efficacious to decrease the rate of nosocomial infections and improve outcomes of SOT. Singh et al. [66] reported an impact of an aggressive infection control strategy on infection in liver transplant recipients and the rate of bacteremia decreased from 26% to 4%. Thus, efforts to control all the risk factors for NLF GNB bacteremias, especially MDR ones, and to curb the spread of these dangerous pathogens, is the most important.

Selective bowel decontamination regimens and rifaximin have been studied to determine whether it could reduce early post-transplant enteric GNB infections and some reported reduced infection [67–69]. Early studies also showed an added benefit from low-dose trimethoprim sulfamethoxazole prophylaxis for reducing urinary tract infection [70,71]. Cytomegalovirus prophylaxis also lowered the risk of bacterial infections including NLF GNB bacteremia. Other studies have studied the impact of infectious disease service consultation on infection and found it decreases all-cause mortality, likely through increased adherence to standards of care in management of infection, including removal of intravascular foci of infection, obtaining follow-up blood cultures, use of beta-lactam therapy when appropriate, and administration of prolonged treatment of complicated infections [72,73].

Future research

Given the considerable mortality and morbidity, the exploration of more intensive or alternative prophylactic measures and effective treatments for NLF GNB bacteremia is warranted for SOT recipients. However, little is known about the risk factors for the development of NLF GNB bacteremia and its associated mortality after SOT transplantation. Furthermore, most of the literature provided a general comment about preventive measures, not centered specifically for SOT. Additional research, therefore, is needed to better identify these risk factors for NLF GNB bacteremia and its associated mortality and specifically preventive strategies for SOT population.

Treatment data specific to SOT recipients with NLF GNB bacteremia are also lacking. Because inadequate empiric therapy has been shown to be an independent predictor of mortality in the setting of bacteremia, multi-center studies providing more robust and generalizable data regarding knowledge of local microbiology are needed. In addition, more funding and clinical trials are urgently needed to explore therapeutic options for MDR NLF GNB.