Frequency of microbial isolates and pattern of antimicrobial resistance in patients with hematological malignancies: a cross-sectional study from Palestine

In recent decades, major developments have been made in the care of cancer patients that have significantly improved patient survival. However, despite these developments, patients with hematologic malignancies remain at an extraordinarily high risk of infections. This is the result of a complex interaction between basic immunodeficiency and therapeutic practices such as surgery, radiation, and chemotherapy [1‐4].

Previous studies have reported that the prevalence of bacterial bloodstream infections among patients with hematologic malignancies ranges from 11 to 38%, and the rough mortality rate reaches up to 40% [5‐7]. Furthermore, studies have shown gram-negative bacteria (GNB) to be the most frequently isolated organisms during the 1960s and 1970s [8]. However, over the following two decades, the proportion of gram-positive bacteria (GPB) has increased.

A study in India showed that E. coli was the most common isolated organism, followed by coagulase-negative staphylococci (CoNS) [9]. Another study showed that GNB were the most common isolated organisms, and the best empirical treatment for them was non-carbapenem-based anti-pseudomonal antibiotics [5]. However, another study conducted in 2018 in Ethiopia showed a change in prevalence from GNB to GPB, mainly as a result of increased use of urinary catheters, and multi-drug resistance was detected in 46.3% of bacterial isolates [10].

In Palestine, cancer is the third leading cause of death after heart disease and cerebrovascular disease, accounting for 10.3% of total deaths [11]. However, Palestinian cancer services started to spring up in the early 2000s, and it took until 2008 to organize cancer care at Augusta Victoria Hospital in East Jerusalem [12]. In the Gaza Strip, a study showed a high prevalence of resistance to amoxicillin (73.2%) in isolates of Staphylococcus aureus, and high resistance to penicillin (40.4%) in Streptococcus pneumonia [13]. However, no published reports have shown the epidemiology and antimicrobial resistance patterns of potential etiological agents among patients with hematologic malignancies in Palestine.

Antibiotic resistance is a growing concern in global health [14‐20]. Overuse and constant consumption of antibiotics, due to lack of control programs in hospitals and over-the-counter antibiotics, lead to multi-drug resistant pathogens. These then lead to increased mortality, length of hospitalizations, and health care costs [21, 22].

This study provides information on the spectrum of microbial isolates and their antimicrobial resistance patterns in patients with hematologic cancer at An-Najah National University Hospital. This study is the first to evaluate bacterial and fungal resistance patterns among hematological malignancies in Palestine. This information will help decrease morbidity and mortality by helping to establish empirical treatment guidelines and antibiotic stewardship programs. These will reduce antibiotic overuse and, subsequently, decrease hospitalizations. This study also highlights the immense effect and burden of multi-drug resistant organisms.

Infections are the most common cause of death in cancer patients, especially among those with hematologic malignancies, with studies reporting that approximately 60% of deaths are infection-related [26]. This increased risk of infections can be due to host or treatment-related causes. Host-related factors consist of immunodeficiency, comorbid illnesses, mucosal ulcerations, previous infections, nutritional deficiency, and stress [26], while treatment-related factors include invasive procedures, surgery, radiation, immunosuppressive drugs, and use of antimicrobials [27]. These infections can be caused by various pathogens such as viruses, bacteria, fungi, etc. Bacteria are the leading cause of infections in cancer patients, followed by fungi [27].

In our study, Pseudomonas aeruginosa (27, 43.6%) was the predominant bacterium among GNB, followed by E. coli (20, 32.3%) that can be divided into non-ESBL (10, 50%) and ESBL-E. coli (10, 50%). These were followed by Acinetobacter baumannii and Klebsiella pneumonia, with six isolates each (9.7%). These results are in conjunction with other studies conducted in India and Pakistan. In the former, they reconfirmed the predominance of GNB in patients with hematologic cancers, with E. coli, Pseudomonas, and Klebsiella having the largest shares [28] In the latter study, which evaluated GNB isolated from bloodstream infections of patients on chemotherapy, Pseudomonas aeruginosa was the most frequent bacteria, followed by E. coli, Klebsiella, Proteus, and Shigella [29]. These results are also similar to a study conducted in Italy, where E. coli was the most frequent organism, followed by Pseudomonas aeruginosa, Klebsiella pneumoniae, and Enterobacter cloacae [30]. In another study carried out in Sudan, E. coli represented the most frequently isolated bacterium among GNB, followed by Pseudomonas aeruginosa [31]. Meanwhile, a study conducted in Egypt found that the GNB most frequently isolated from all samples was Klebsiella pneumoniae followed by E. coli [32].

Regarding GPB, CoNS represented the most frequent species isolated in our study (32, 56.1%), followed by Enterococcus faecium (10, 17.5%), Enterococcus faecalis (5, 8.7%) and Staphylococcus aureus (4, 7.0%). These results are comparable to the aforementioned Italian study, where CoNS were the most common species, followed by Enterococcus spp., viridans group streptococci (VGS) and Staphylococcus aureus (11). In the Indian study, the most frequent GPB isolates were CoNS, then Staphylococcus aureus, Streptococcus spp., and Enterococcus spp. (14).

In our study, the bacteria most commonly isolated were Pseudomonas aeruginosa (22%), E. coli (16.3%), and Staphylococcus epidermidis (11.4%), followed by Enterococcus faecium and Staphylococcus haemolyticus (8.1% each), and then Klebsiella pneumonia, Acinetobacter baumanii, and Staphylococcus hominis (4.9% each). In comparison, when looking at patients with hematologic malignancies in Japan, E. coli was the most commonly seen bacterium, followed by Klebsiella spp., Pseudomonas aeruginosa, Staphylococcus aureus, Enterobacter spp. Citrobacter spp., and Acinetobacter spp. [33].

Hard to spot but lethal if missed, invasive fungal infections—predominantly caused by Aspergillus and Candida—are the leading infectious cause of mortality in patients with myelosuppression due to chemotherapy [34]. In our study, Candida had the highest share of fungal infections, in contrast to a study in Italy where most infections were caused by Aspergillus spp., followed by Candida [35].

In our study, Pseudomonas aeruginosa exhibited high resistance to ciprofloxacin (60%), in concordance with numbers found in similar Italian studies [30, 36], and with a Spanish study that observed resistance to ciprofloxacin among cancer patients in general [37]. Pseudomonas aeruginosa isolates in our study also had high resistance to carbapenems, including imipenem (59.3%), meropenem (48%), and gentamicin (48%). These numbers resemble those found in another study where the resistance rate to carbapenems was 60% [36], and in an Italian study where the resistance rate to meropenem was 71.2% [30]. However, this is in contrast to an American study that found the resistance to imipenem seen among solid and hematological cancer patients was only 6% [38]. Also among our Pseudomonas aeruginosa isolates, piperacillin resistance was found to be 54.2%, while in a previously mentioned study it was found to be 24% [36]. Meanwhile, among cephalosporins, cefepime, and ceftazidime, resistance rates were 16% and 24% respectively. The reasons behind these low rates of resistance to cephalosporins are the infrequent use of these agents, as the prescription of cefepime is highly restricted, and piperacillin-tazobactam is the most commonly used initial therapy for neutropenic fever instead. This highlights the importance of diversification of antibiotic use, such as prescribing third-generation cephalosporins (ceftazidime) for neutropenic fever [39], to avoid selection of carbapenem resistance by extensive carbapenem use. However, the selection of empiric antimicrobial therapy should be based on multiple factors, including but not limited to the clinical status of the patient, previous cultures and colonization, and institutional antibiograms [40, 41].

Among GNB, 21 CRE (31.8%) were detected, more than that seen in febrile neutropenic patients with hematological cancer in Japan [42]. In our study, the resistance of E. coli isolates to amikacin was only 5%, similar to the results of another study where 85.2% of E. coli isolates were found to be sensitive to amikacin [30]. On the other hand, ESBL-E. coli exhibited 100% resistance to both cephalosporins and ampicillin, similar to previous research, where the vast majority of ESBL-producing isolates were resistant to all generations of cephalosporins [42]. E. coli in our study also exhibited high resistance to levofloxacin and TMP-SMX (75% and 63.2%, respectively), similar to the results found in a previous study [30]. This could be due to the frequent use of fluoroquinolones (especially levofloxacin) for prophylaxis in patients with prolonged neutropenia [39].

Regarding the six isolates of Acinetobacter baumannii, the highest resistance rates were observed to carbapenems (80% to meropenem and 83.3% to imipenem) and piperacillin-tazobactam (83.3%) similar to a related study held in Turkey [43]. Isolates also exhibited high resistance to gentamicin (66.7%). Four of these isolates were tested for resistance to colistin and all were sensitive, in agreement with prior research where all isolates of Acinetobacter baumannii were susceptible to colistin [44]. Finally, the six isolates of Klebsiella pneumonia were 100% susceptible to piperacillin/tazobactam, carbapenems, fluoroquinolones, ceftazidime, and cefepime. In other studies, 55.8% of Klebsiella isolates were resistant to piperacillin/tazobactam [30], 44.9% were resistant to meropenem while 1% were resistant to imipenem [38], 69.8% were resistant to ciprofloxacin, 58.1% were resistant to ceftazidime [30], and 20% were resistant to cefepime [42].

Among the 10 Enterococcus faecium isolates and the 5 Enterococcus faecalis isolates, 90% of Enterococcus faecium isolates were VRE while none of the Enterococcus faecalis isolates were VRE. Regarding Enterococcus faecium, 40% of isolates were resistant to streptomycin, 30% were resistant to gentamicin, and 11.1% were resistant to tigecycline. Meanwhile, Enterococcus faecalis species had 80% resistance to streptomycin, 50% resistance to gentamicin, and 33.3% resistance to tigecycline. In particular, none of the Enterococcus faecalis or Enterococcus faecium isolates was resistant to linezolid, in agreement with prior research [30].

Among the CoNS (Staph. epidermidis, hominis and haemolyticus), no isolates were resistant to vancomycin or linezolid, while 93.3% were resistant to oxacillin, similar to the results of a previous study [30]. Regarding the 14 isolates of Staphylococcus epidermidis, all were resistant to penicillin and cephalosporins, and 54.5% were resistant to trimethoprim-sulfamethoxazole. Regarding the four isolates of Staphylococcus aureus, 75% were resistant to oxacillin, a high percentage compared to patients in Italy (36.4%) [30]. Additionally, 66.6% were resistant to cefuroxime and 50% were resistant to ceftriaxone. However, all were sensitive to both vancomycin and linezolid, similar to those in the former Italian study [30].

Regarding antifungal resistance rates, all were sensitive to caspofungin, comparable to a similar study in which caspofungin resistance rates were 5% [45]. All were sensitive to fluconazole, voriconazole, flucytosine, and micafungin. When reviewing the literature on Candida infections in patients with hematologic malignancies, a study showed that 27.6% [37] were resistant to fluconazole. Meanwhile, in another study, 8% of Candida were resistant to voriconazole and 5% were resistant to caspofungin [45].

51.5% of GNB and 68.4% of GPB in this study were MDRO. Among GNB, Acinetobacter baumanni had the highest rate of MDRO (83.3%), whereas among GPB, CoNS had the highest rate (81.3%). Meanwhile, in a similar study in which MDROs were isolated in 13% of patients, the most frequently isolated MDRO was Klebsiella pneumoniae, followed by MRSA, Acinetobacter baumanni, Pseudomonas, E. coli, and VRE [46].

This study is the first in Palestine to determine the microbial profile of infections in patients with hematological malignancies. However, there were some limitations to our study. First, not all data were written in the patient’s medical reports such as white blood cell counts, absolute neutrophil counts, and patient temperatures at the time of culture, so we could not assess neutropenic fever and its relationship with other variables. Furthermore, some data were not collected, such as the last time a patient received a chemotherapy session. Second, our data were collected from only one center that may not be representative of other centers. Third, some patients died or left the hospital before the culture results were ready, so they did not receive any treatment other than empirical antibiotics. Finally, the study did not assess increases in antibiotic resistance year over year.

Patients with hematologic malignancies are at risk for a variety of serious infections that cause significant morbidity and mortality. The most common bacterial isolates among GNB were Pseudomonas aeruginosa and E. coli, while coagulase-negative staphylococci and Enterococcus faecium were the most common among GPB. Our study showed alarming rates of resistance to the most widely used antibiotics, thus highlighting the need to develop local guidelines for antimicrobial use based on local resistance patterns of these organisms. This study also emphasizes the need to develop antimicrobial stewardship programs in local hospitals. Enforcing the implementation of infection control policies would help curb the spread of these MDROs and reduce the morbidity, mortality, and economic burden of these serious infections.