Patients with cancer are at a high risk of infection and often the focus of the infection is not apparent. Bloodstream infections (BSIs) have been the leading complications in such patients [1, 2]. In adult patients with malignancies, the crude mortality rates of BSIs range from 18%-42% [3, 4]. A variety of factors including frequent hospitalization, exposure to invasive procedures, use of broad-spectrum antibiotics and chemotherapy make cancer patients more susceptible to BSIs [1, 5]. Chemotherapy renders cancer patients to be neutropenic making them more susceptible to potentially life-threatening BSIs [6, 7].

Bacteria are the primary causative agents of BSIs [8, 9]. The gram-positive pathogens such as Staphylococcus aureus (prevalence of S. aureus ranging from 18% - 35%) [10, 11] and the gram-negative bacteria such as E. coil (13% - 23.1%), K. pneumoniae (4.6% - 12%) , and/or P.aeruginosa (6% - 12.5%) [10, 12, 13] had been reported as the common pathogens causing BSIs in cancer patients. However, in recent years, saprophytic bacteria have been emerging as potential human pathogens causing life-threatening infections in patients with malignancies. A case report from Delhi State Cancer Institute, for example, indicated that Aeromonas hydrophila, a saprophyte in the soil, caused a serious soft tissue infection in a woman with breast malignancy [14]. A study from Texas Children’s Hospital isolated the plant pathogen Pantoea agglomerans from a blood culture of children [15]. Another study from Egypt had also reported the rare bacterial pathogen Chryseobacterium meningosepticum from blood cultures of patients with cancer [16].

The spectrum and drug resistance profile of the pathogens causing BSIs in patients with cancer have been reported to show significant fluctuations in different geographical areas and time points [8, 17]. Data regarding the emerging saprophytic potential nosocomial pathogens is limited, particularly in sub-Saharan countries, despite the escalating burden of cancer. This study therefore aimed at determining the spectrum and drug resistance profile of the rare bacterial pathogens causing BSIs in febrile cancer patients at a referral hospital in Ethiopia.

This study is the first to describe the rare bacterial isolates as causes of BSIs in Ethiopian patients with cancer. The fermentative gram-negative rod, Aeromonas hydrophilia species, was the most predominant isolate of the rare bacteria isolated. Previous studies from Egypt had also reported this pathogen as an agent of BSIs in patients with hematological malignancy [18, 19]. Aeromonas species are widely distributed in the aquatic environment and patients acquire infection by oral consumption or direct contact with contaminated water or seafood [20]. Aeromonas may cause serious fatal infections such as hepatobiliary infection, invasive skin and soft tissue infections, primary bacteremia, burn infections, pleuropulmonary infection, meningitis and endocarditis in the immunocompromised patient [21, 22]. Though Aeromonas isolates are susceptible to a broad range of antibiotics, some species may produce beta-lactamase which makes them resistant to ampicillin and first-generation cephalosporins [23]. In this study, all the four isolates were resistant to amoxicillin-clavulanic acid, ampicillin, and trimethoprim-sulphamethoxazole suggesting the emergence of this saprophytic pathogen as a drug resistant potential nosocomial pathogen in Ethiopia.

Chryseomonas luteola, a saprophyte found in the soil and water, has only rarely been reported as a human bacterial pathogen [24, 25]. In a case report from a Moroccan University Hospital, this bacterium was indicated causing serious infection in a newborn with a respiratory failure [24]. In the current study, we isolated two Chryseomonas luteola pathogens from neutropenic patients with non-Hodgkin lymphoma. In the previous reports, Chryseomonas luteola isolates showed variable sensitivity to ampicillin, tetracycline, and co-trimoxazole. In this study, both isolates were resistant to five of the nine antibiotics tested including amoxicillin-clavulanic acid, ampicillin, chloramphenicol, ceftriaxone, and trimethoprim-sulphamethoxazole. This also suggests the emergence of this bacterium as a potential nosocomial pathogen in Ethiopia.

Chryseobacterium meningosepticum was isolated from a blood culture of a leukemic patient. This bacterium was already defined as etiology of bloodstream infection among cancer patients by the Infectious Diseases Society of America [26]. A study from Egypt had also reported this bacterium from blood cultures of patients with cancer [16]. Chryseobacterium meningosepticum is found in soil and can survive in chlorine-treated municipal water supplies [27] and could colonize patients via contaminated medical devices involving fluids (respirators, intubation tubes, incubators for newborns, ice chests, syringes, intravascular catheters, etc) [28, 29] and cause severe infections in immunocompromised people and in children [16, 28].

Two isolates of Sphignomonas poucimobilis were also observed in this study; one from a 40-year-old female patient with acute leukemia and the other from a 24-year-old female patient with chronic lymphocytic leukemia. These non-fermenting gram-negative bacilli create a significant problem in clinical settings, being the most widespread cause of nosocomial infections. They are opportunistic pathogens that take advantage of underlying conditions and diseases [30]. According to a review of the case reports by Lin et al, about 42 isolates of S.poucimobilis were indicated as agents of BSIs in cancer patients (mostly in patients with hematological malignancy) [30]. The existence of this bacterium in our setting could pose a challenge in managing and treating patients appropriately as standard methods to undertake antibiotic susceptibility testing are not on hand for this pathogen [30].

Pantoea agglomerans, a very a rare fermentative gram-negative pathogen, was also isolated in our study as an agent of BSIs in patients with acute myeloid leukemia. Previously, this pathogen was isolated from patients with hematological malignancy in Belgium [31]. Pantoea agglomerans is a plant pathogen widely distributed in the soil [32]. A review of case reports indicated that P.agglomerans was mostly associated with penetrating trauma by vegetative material and catheter-related bacteremia in immunocompromised patients and in children in the hospital [15].

Serratia ficaria has been isolated from human clinical samples in rare instances [33]. Although its pathogenicity was always questionable, S. ficaria was later found to be able to cause severe infections (septicemia) or deep suppurations such as gallbladder empyema [34]. Thus, S. ficaria is an opportunistic pathogen responsible for colonization or serious infections in compromised patients. In this study, we isolated this organism from a 24-year-old female patient with acute lymphocytic leukemia. Another study in France also isolated Serratia ficaria from a blood of 83-year-old adenocarcinoma patient [34]. Other case reports [33, 35] also reported this bacterium from a blood culture of patients with underlying disease other than neoplasm. In previous reports, S. ficaria was usually susceptible to numerous antibiotics but always resistant to cephalothin (cefazolin was the prophylactic antibiotic in the present case) [34]. In our study, the S. ficaria isolate was susceptible to seven of the nine antibiotics tested while it was resistant to teteracycline and nalidixic acid. We didn't test it against cephalothin (cefazolin).

We also recovered two isolates of Pseudomonas fluorescens from blood cultures of male febrile neutropenic patients with hematological malignancy. Previously, P. fluorescens was reported from a bone marrow transplant patients with hematological malignancies, who became colonized with the organism from a contaminated water dispenser that supplied bottled natural spring water in a hospital [36]. The case holds also true in our setting where the patients didn't have adequate access to clean water supply. A previous study showed that P. fluorescens strains isolated from cancer patients were susceptible to gentamicin, neomycin, tetracyclines, polymyxin B, and colistin and resistant to chloramphenicol, ampicillin, and narrow-spectrum cephalosporin [37, 38]. In the current study, both of our isolates were resistant to amoxicillin-clavulanic acid, ampicillin, and trimethoprim-sulphamethoxazole and one of the strains were resistant to teteracycline, ciprofloxacillin and nalidixic acid. This is of some concern because these classes of antibiotics are commonly recommended for the management of neutropenic sepsis in hematological malignancies.

To mention some of the limitations: first we used only aerobic culture techniques; anaerobic microorganisms that might have been contributing for BSIs were not defined. The fact that we took only blood sample might also have limited the range of bacterial pathogens to be isolated.

This study revealed the emergence of saprophytic bacteria as potential drug-resistant nosocomial pathogens in Ethiopian patients with cancer. The majority of the isolates were resistant to the commonly used chemotherapic antibiotics imposing a challenge to the health care program. As these pathogens are ubiquitous in the environment, infection prevention actions should be strengthened in the hospital and early diagnosis and treatment with appropriate antibiotics are warranted for those already infected.