Stenotrophomonas maltophilia Infections in Pediatric Patients – Experience at a European Center for Pediatric Hematology and Oncology

Stenotrophomonas maltophilia is an important nosocomial pathogen in immunocom-promised individuals and characterized by intrinsic resistance to broad-spectrum antibacterial agents. Limited data exists on its clinical relevance in immunocompromised pediatric patients, particularly those with hematological or oncological disorders. In a retrospective single center cohort study in pediatric patients receiving care at a large european pediatric hematology and oncology department, ten cases of invasive S.maltophilia infections (blood stream infections (BSI), 4; BSI and pneumonia, 3, or soft tissue infection, 2; and pneumonia, 1) were identified between 2010 and 2020. Seven patients had lymphoblastic leukemia and/or were post allogeneic hematopoietic cell transplantation. Invasive S.maltophilia infections occurred in a setting of indwelling central venous catheters, granulocytopenia, defective mucocutaneous barriers, treatment with broad-spectrum antibacterial agents, and admission to the intensive care unit. Whole genome sequencing based typing revealed no genetic relationship among four individual S.maltophilia isolates. The case fatality rate and mortality at 100 days post diagnosis were 40 and 50%, respectively, and three patients died from pulmonary hemorrhage. Invasive S.maltophilia infections are an emerging cause of infectious morbidity in patients receiving care at departments of pediatric hematology and oncology and carry a high case fatality rate.


INTRODUCTION
Stenotrophomonas maltophilia (formerly: Pseudomonas or Xanthomonas maltophilia) is an aerobic non-fermenting Gram-negative bacillus (NGNB) that can be found ubiquitously in the environment (1). Next to Pseudomonas aeruginosa and Acinetobacter spp., the organism is considered the third most frequent nosocomial pathogen among nonfermentative bacteria (2,3).
Pneumonia and bloodstream infection (BSI) are the most common clinical manifestations of S.maltophilia infections. Less frequently, S.maltophilia can cause urinary tract infections, cholangitis, peritonitis, wound infections, eye infections, arthritis, meningitis, and endocarditis (4,5). Patients with hematologic malignancies are at high risk for S.maltophilia infection because of chemotherapy-induced neutropenia and immunodeficiency. Frequent exposure to broad-spectrum antibiotics and the presence of central venous catheters further enhance the risk of S.maltophilia infection (6,7). The rate of S.maltophilia BSI among BSIs in this patient population has been reported to be as high as 60% (8)(9)(10)(11).
Treatment of S.maltophilia infection can be difficult because of the organisms inherent resistance to a variety of antibiotics (12,13). Trimethoprim-sulfamethoxazol (TMP-SMX) is the drug of choice, and fluoroquinolones are the proposed alternative. Similar to the treatment of Pneumocystis jirovecii pneumonia, up to five-fold higher than regular doses of TMP-SMX are recommended for severe infections (5,14). Thus, the therapeutic options for S.maltophilia infections are quite different from those available for other NGNB, and appropriate antimicrobial therapy is often delayed through ineffective treatment during initial empirical therapy (15). Accordingly, mortality rates are high in immunocompromised and critically ill patients (11,16), with 30-day mortality rates of S.maltophilia BSIs ranging from 11% to 53% (8,(11)(12)(13)(17)(18)(19).
While series of adult cancer patients with invasive S.maltophilia infections have been published in regular intervals, few reports exist for pediatric patients with cancer and/or allogeneic hematopoietic cell transplantation (HCT) (20)(21)(22). We therefore analyzed the incidence, genetic relatedness, clinical course and outcomes of invasive S.maltophilia infections observed during the past ten years at our institution, a high volume European pediatric cancer center with an active allogeneic HCT program.

Study Design and Setting
The study was a retrospective observational single center cohort study of children and adolescents with oncological or hematological disease including patients with autologous or allogeneic HCT receiving care at the Department of Pediatric Hematology and Oncology of the University Children's Hospital of Münster between January 2010 and July 2020 with the last follow-up in October 2020. The Department's referral patterns and admission data at the time of the study have been reported recently (23). Patients with S.maltophilia infection or colonization were identified through the Hospital's central electronic medical information system. Inclusion criteria were medical care at the Department of Pediatric Hematology and Oncology; a diagnosis of either solid tumor, hematological malignancy, a non-neoplastic hematological disorder, or status post allogeneic HCT; and microbiology confirmation of S.maltophilia in blood, usually sterile body sites or respiratory secretions in the presence of pneumonia. Patient demographics, disease related parameters, clinical course and outcome data were retrieved from the medical information system and analyzed. The primary endpoint of outcome was survival at day +100 post diagnosis. Written informed consent for data collection and analysis was obtained within the consent procedure for cancer treatment, HCT, and specialized medical care approved by the local institutional review board. Data collection was accomplished by a pseudonymized standardized case report form.

Standard Operating Procedures
All patients received treatment for their underlying condition according to standard protocols of the German Society for Oncology and Hematology (GPOH) or individual recommendations of the respective study groups. Up to December 2014, antibacterial prophylaxis was given to patients undergoing HCT and consisted of penicillin, ciprofloxacin and metronidazole in allogeneic and penicillin and ciprofloxacin in autologous HCT recipients, respectively. Antibacterial prophylaxis was discontinued starting 2015. Initial empirical antibacterial therapy for fever and neutropenia consisted of ceftazidime plus gentamycin until December 2016 and was then replaced by piperacillin/tazobactam. Unstable patients were to start with meropenem plus vancomycin and were subsequently deescalated, as feasible. This regimen was also used for escalation in patients with fever persisting for more than 48-72 hours or a new fever after defervescence, with or without additional empirical antifungal therapy at the discretion of the attending physician. Suspected or proven infections were treated according to current management recommendations. All patients received TMP-SMX 8 mg/kg (max. 320 mg) twice weekly as prophylaxis for prevention of Pneumocystis jirovecii pneumonitis, and topical polyenes or azoles for prevention of oropharyngeal candidiasis. Prophylaxis with TMP-SMX was continued until three months after end of therapy in cancer patients and until immunoreconstitution in allogeneic HCT recipients. Standard antifungal prophylaxis consisted of fluconazole for allogenic HCT recipients, and either posaconazole or voriconazole for patients with acute myeloid leukemia or recurrent leukemia (24,25). Blood cultures were drawn in case of fever and daily until defervescence and negative results. Aerobic and anaerobic cultures with age-appropriate blood volumes were obtained from each lumen of an indwelling catheter or from a peripheral vein, if no catheter was present. Respiratory cultures were obtained by tracheal aspiration in intubated patients (n=4) and by sputum induction in non-intubated patients, respectively. Cultures form other body sites were obtained only when infection was clinically or radiologically suspected or on a case-by-case basis to monitor bacterial colonization by swabs from the throat and the perianal region. All patients were routinely screened for colonization with methicillin-resistant Staphylococcus aureus by a combined swab from the throat and the nares at each hospital admission.

Definitions
Blood stream infection was defined as ≥ one positive blood culture for either S.maltophilia or any other bacterial and fungal pathogens obtained in a patient with fever and other signs of infection, where present. Infections at other body sites were defined by clinical and/or radiographic criteria. Pulmonary infection was considered to be radiological evidence of pneumonic infiltrates together with detection of S.maltophilia in respiratory secretions and BSI. In the absence of documented BSI, respiratory evidence of S.maltophilia together with direct detection of S.maltophilia in the intraoperative tissue cultures, as in patient 3 after open abscess surgery ( Table 1), was considered pulmonary infection.

Whole Genome Sequencing-Based Typing
To determine the clonal relationship of S.maltophilia strains isolated from blood cultures, available isolates were subjected to whole genome sequencing (WGS)-based typing using the Illumina MiSeq platform (Illumina Inc., San Diego, USA) as described previously (26). Due to the retrospective character of this study, only four individual samples from four different patients were available for testing. Using SeqSphere+ software version 2.0 beta (Ridom GmbH, Münster, Germany), all coding regions were extracted and compared in a gene-by-gene approach (core genome multilocus sequence typing, cgMLST) using SM K279a strain (GenBank accession number AM743169.1) as a reference sequence. Instead of a published cgMLST scheme, which is not yet available, this ad hoc scheme was used to differentiate the cluster. SeqSphere+ software was used to display the clonal relationship in a minimum spanning tree. For backwards compatibility with classical molecular typing, i. e. MLST, the MLST sequence types were extracted from the WGS data in silico.

Statistical Analysis
Statistical analyses were carried out with SPSS Statistics 26 (IBM Corporation, Armonk, NY, USA) software package. Overall survival (OS) was calculated from primary diagnosis to death or last follow-up. Comparison of the frequency of S.maltophilia infections over time and statistical exploration of associations between patient-and disease related parameters and mortality were performed by the Fisher's Exact test; univariate and multivariate analyses were not performed due to the limited sample size. The level of statistical significance was set at p<0.05 (two-sided).

Demographic and Clinical Characteristics
Between January 2010 and July 2020, a total of 502 distinct BSIs were identified in children with oncological or hematological disease including patients with autologous or allogeneic HCT receiving care at the Department of Pediatric Hematology and Oncology of the University Children's Hospital of Münster. Of these, nine BSIs were due to S.maltophilia, accounting for a rate of S.maltophilia BSIs of 1.8% among all BSIs and of 7% among all Gram-negative BSIs, respectively. Considering one additional patient with documented pulmonary infection and positive cultures from all other sources, there were a total of ten invasive S.maltophilia infections in ten patients. Over time, there was a numerical, but not statistically significant increase in infected patients in the second half (2016-2020) of the study ( Figure 1).
The demographic and clinical characteristics of the ten patients with invasive infections are listed in Table 1. Five patients each were male and female, and the median age was 10.4 years (range, 0.8 to 17.9 years). Five patients had acute lymphoblastic leukemia, and five patients had received allogeneic HCT and were between 13 and 523 days (median: 63) posttransplant. Seven patients were receiving antineoplastic or immunosuppressive therapy, and all had an indwelling central venous catheter at the time of diagnosis (Broviac-type, n=6; percutaneous transient catheter, n=3; port-a-cath-type, n=1). Among the 10 patients with invasive S.maltophilia infections, four had isolated BSIs, three a BSI and concomitant pneumonia, two a BSI and concomitant soft tissue infection, and one patient had pneumonia with an intrapulmonary abscess without positive blood cultures. In seven patients, superficial colonization by S.maltophilia was detected. Most affected patients (n=8) were receiving broad-spectrum antibacterial agents at the time of diagnosis, most frequently carbapenems (n=8), glycopeptides (n=7), and quinolones (n=6). All had an increased C-reactive protein level, and seven patients were profoundly granulocytopenic with an absolute neutrophil count < 500/uL. Six patients required admission to the intensive care unit at presentation, and four of these patients received mechanical ventilation because of pneumonia (n=3) and respiratory failure not related to pneumonia but to multiorgan failure (n=1) ( Table 1).  Figure 3 for details); **** patient 1 had a catheter exit-site infection, and patient 8 had necrotizing fasciitis involving the lower extremities and buttocks. ***** Four patients died in direct causal relationship to the infection from pulmonary hemorrhage (patients 3,4,5) and necrotizing fasciitis (patient 7) with multiorgan failure, and one patient (patient 10) died two months after completion of treatment from unrelated causes in hospice care. 1

Antimicrobial Susceptibilities and Genotyping
Using disk diffusion methodology in accordance with the standards of European Committee on Antimicrobial Susceptibility Testing (EUCAST) and an agar diffusion diameter of > 16 mm assumed as susceptible (increased exposure), 70% of all ten initial isolates were susceptible to TMP-SMX. However, in one of the seven patients with a TMP-SMX-susceptible initial isolate, a follow-up blood stream isolate obtained three days after the initial one was tested nonsusceptible. WGS-based typing and gene by gene comparison of four initial S.maltophilia blood culture isolates obtained from four different patients showed allelic differences between strains of at least 1604 alleles, thereby excluding any genetic relatedness of subjected S.maltophilia isolates ( Figure 2).

Antimicrobial Management and Outcome
The indwelling central venous catheter was removed shortly after diagnosis in seven of the nine patients with positive blood cultures. One patient (patient 7, Table 1) received repeated granulocyte transfusions. Antimicrobial treatment of S.maltophilia infection was highly heterogeneous and included combinations of meropenem (7), fluoroquinolones (7), tigecyclin (5), colistin (3), TMP-SMX (2), ceftazidime (1), fosfomycin (1), and tobramycin (1) administered for a total treatment duration of 1 to 45 days (median: 14.5 days). Of note, in retrospect, it is difficult to distinguish precisely between therapy directed at S.maltophilia, empiric treatment for suspected infections or directed treatment of confirmed concomitant infections, but the agents TMP-SMX and moxifloxacin were added only when S.maltophilia was detected. The 30-day mortality rate and the overall mortality rate were 30% and 50%, respectively, after a median follow-up time of 123 days (range, 2 to 2446 days). Four patients died in direct causal relationship to the infection after 2, 3, 10 and 45 days after diagnosis from pulmonary hemorrhage (patients 3,4,5, Figure 3) and necrotizing fasciitis (patient 7) with multiorgan failure ( Table 1). Explorative statistical analysis of factors associated with overall mortality in patients with invasive S.maltophilia

DISCUSSION
Stenotrophomonas maltophilia is a non-fermentative, Gramnegative bacillus that has emerged as important nosocomial pathogen in immunocompromised and critically ill patients (16,27). Published experience in pediatric patients with cancer and/or allogeneic HCT is limited to two separate studies reporting on a total of 24 S.maltophilia BSIs (22,28) and several larger pediatric series that include a relevant proportion of patients with hematological malignancies or solid tumors (24,25,29,30) ( Table 2). In the study presented here, S.maltophilia accounted for 1.8% of all BSIs and for 7% of those caused by Gram-negative rods. Invasive S.maltophilia infection was associated with a diagnosis of acute leukemia and/or allogeneic HCT, or immunodeficiency, and occurred in a setting of impaired host defences, defective mucocutaneous barriers, indwelling central venous catheters, treatment with broad-spectrum antibacterial agents, and admission to the intensive care unit. Four patients died in direct relationship to the infection, including three patients with pneumonia and pulmonary hemorrhage (Figure 3), which has been reported to be associated with S.maltophilia infection and status post allogeneic HCT (29,44,45). Similar to others (28), we found a numerical increase in S. maltophilia infections over time. Molecular typing of a limited number of blood culture isolates, however, confirmed that isolates were genetically not related and suggests the absence of a nosocomial outbreak (46). The exact route of acquisition of S.maltophilia often remains unknown. Nevertheless, isolation of the organism from mucosal surfaces of the respiratory and/or the lower gastrointestinal tract may herald later infection as many patients with S.maltophilia BSIs were reported to be colonized prior to infection (5,17). Indeed, the oral microbiome has recently been described as a potential reservoir, and real-time monitoring of the oral S.maltophilia relative abundance has been suggested to identify patients at risk for invasive infection (30). In our limited cohort, concomitant colonization was detected in the majority of cases with invasive S. maltophilia infection, but overall, there was no apparent relationship between pharyngeal or respiratory colonization and invasive infection.
Similar to previous reports (20), the majority of invasive S. maltophilia infections in our cohort was associated with indwelling central venous catheters. Six of the ten patients were treated at the intensive care unit and four were on invasive ventilation. Intensive care, mechanical ventilation, and/or central venous catheterization have been identified as risk factors for S. maltophilia BSI and/or dismal outcome ( Table 2). Several studies suggest a survival benefit for removal of indwelling central venous catheters (8,9,17,(47)(48)(49), and international guidelines strongly recommend prompt catheter removal in S. maltophilia associated BSIs (50), independent on whether the catheter is considered the source of the infection or being colonized secondary to ongoing bacteremia. Patients with S.maltophilia BSIs often have polymicrobial infections (5), and their relative frequency in children seems to be higher as observed with Pseudomonas aeruginosa (35). In the cohort presented here, concomitant BSI occurred in 30% of patients with S.maltophilia infection, which is below the rate in previous series of pediatric patients (31,33). Bacteria most commonly recovered in temporal context with S.maltophilia were coagulase-negative Staphylococcus and Enterococcus spp (8,45). It remains unclear whether the detection of S.maltophilia is a consequence of appropriate antimicrobial therapy for other BSIs or whether the concurrent invasive infections simply reflect the sum of immunodeficiency in the affected patients.
In eight of the ten cases, S.maltophilia infection occurred as breakthrough infection in patients receiving broad-spectrum antibacterial agents. Prior use of carbapenems has been repeatedly described as a risk factor for S.maltophilia infection (12,17,44,(51)(52)(53), and cumulative carbapenem use has been identified to be associated with S.maltophilia in leukemia patients with altered oral microbiome (30). Similarly, in the majority of studies in pediatric patients investigating factors related with outcome, prior use of carbapenems was associated with dismal outcome of S.maltophilia BSI (22, 35, 39) ( Table 2). As a consequence, clinicians should be aware that breakthrough infection with S.maltophilia may occur in severely ill patients being treated with carbapenems.
Antibacterial therapy for S.maltophilia infections is challenging because most clinical isolates are resistant to agents commonly used for empirical treatment of febrile neutropenia or documented infections by Gram-negative organisms, including extendedspectrum penicillins, third-generation cephalosporins, carbapenems, and aminoglycosides (53). In addition, current recommendations for treatment are based on historical evidence, case series, and in vitro susceptibility data rather than pharmacokinetic/pharmacodynamic considerations and results of controlled clinical trials (5). In general, TMP-SMX is the drug of choice for infections by susceptible S.maltophilia isolates based on its potent in vitro activity and documented clinical efficacy (10,46). Nevertheless, susceptibility varies between geographic regions and resistance is an emerging threat (2,5,10,16,54,55). Alternatives to treatment with TMP-SMX include fluoroquinolones and tigecycline (12,13). However, in contrast to TMP-SMX, clinical breakpoints for these agents have not been defined, which makes a valid interpretation of in vitro susceptibility testing results with regards to the prediction of clinical efficacy difficult. In our study, seven of ten S.maltophilia initial isolates from patients with invasive infections were susceptible in vitro to TMP-SMX, and in one of these patients, isolates became resistant during treatment. Apart from primary or secondary resistance, further concerns in immunocompromised patients with cancer and/or allogeneic HCT include the myelotoxicity of therapeutic doses of TMP-SMX (18) and the widespread use of low and intermittent doses of the agent for antibacterial or anti-Pneumocystis prophylaxis that may result in the selection of resistant S.maltophilia strains (17). Indeed, based on emerging resistance, it has been suggested by individual experts to consider escalating therapies in immunocompromised or critically ill patients (49,56). Previous observations on the use of fluoroquinolones against invasive S.maltophilia infections have demonstrated comparable patient survival relative to TMP-SMX (12,13,57), and quinolone prophylaxis in adult cancer patients has been associated with a reduced incidence of invasive S.maltophilia infections (58). Nevertheless, quinolone monotherapy for S.maltophilia BSIs should be critically reflected (19), as rapid emergence of resistance to these agents has been observed both in vitro and in vivo (5,59).
Considering the small number of patients, the 30-day mortality rate of 30% in patients with invasive S.maltophilia infections in our study is in line with 30-day mortality rates of S.maltophilia BSIs of 33% and 38% reported by others (17,60). In pediatric studies, the reported all-cause mortality rates in patients with S.maltophilia infections range from 12.5% to 61% with an attributable mortality of 0% to 18%, respectively (22,31,33,(37)(38)(39)41) (Table 2). Many studies across all age groups have reported risk factors for mortality associated with S.maltophilia BSIs including prolonged hospitalization prior to BSI onset, previous exposure to antimicrobial agents, use of indwelling medical devices, a compromised health status, complex medical care, granulocytopenia and/or transplantation, and inappropriate therapy (5, 8-10, 15, 17, 18, 36, 48, 61-64). While the limited number of patients included precluded robust statistical assessments, the presence of pneumonia and admission to the intensive care unit were both significantly associated in explorative analyses with mortality in the present study (Supplementary Table 1). Nevertheless, as it is often difficult to distinguish between colonization and infection, identification of risk factors for mortality is ultimately limited to BSIs and may not consider the full spectrum of diseases caused by the organism (7).
To conclude, as reflected in this limited series of heterogenous patients, defined therapeutic strategies for invasive S.maltophilia infections in immunocompromised pediatric patients, including those with cancer and/or allogeneic HCT, so far lack uniformity but remain an important goal. Clinicians should be aware that breakthrough infections by S.maltophilia may occur during the administration of broad-spectrum antibiotics, especially following carbapenem use, and that these infections may be associated with fulminant and fatal pulmonary hemorrhage, in particular in allogeneic HCT patients (29). Detection of BSI by S.maltophilia should prompt the removal of indwelling central venous catheters and the immediate initiation of therapeutic doses of TMP-SMX. Initial combination with second generation fluoroquinolones and tigecycline until return of resistance testing  and achievement of a stable clinical response may be considered in view of the high case fatality rates.

DATA AVAILABILITY STATEMENT
The data analyzed in this study is subject to the following licenses/restrictions: Clinical data. Requests to access these datasets should be directed to (andreas.groll@ukmuenster.de).

ETHICS STATEMENT
Written informed consent was obtained from the individual(s), and minor(s)' legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

AUTHOR CONTRIBUTIONS
Retrospective analysis of single-center data and literature review was conducted by SKZ, supported by HH and with clinical input by CR, AH, KM, AR, and AHG. Identification and susceptibility testing by NJF, whole genome sequencing based typing was performed and analyzed by SK. Statistical analysis was performed by AR, SKZ and AHG. Manuscript was written by SKZ and was edited by SK and AHG. All authors contributed to the article and approved the submitted version.

FUNDING
The study was funded by internal resources.