Acute myeloid leukemia (AML) is an aggressive hematological malignancy characterized by uncontrolled proliferation of immature myeloid precursor cells within the bone marrow, leading to progressive bone marrow failure and severe impairment of normal hematopoiesis. Patients frequently present with anemia, thrombocytopenia, and profound neutropenia, which markedly increase susceptibility to life-threatening infections even before initiation of chemotherapy.¹ Despite significant advances in leukemia treatment, infectious complications continue to represent one of the principal causes of treatment-related morbidity and mortality worldwide, particularly during induction chemotherapy when prolonged neutropenia and disruption of mucosal barriers facilitate microbial invasion into the bloodstream. Infections remain responsible for a substantial proportion of early deaths among AML patients despite improvements in supportive care and antimicrobial therapy.²

Bloodstream infections (BSIs) represent the most severe form of infectious complication in patients with AML because they frequently progress to septic shock, multiple organ dysfunction, prolonged hospitalization, interruption of chemotherapy, and increased mortality.³The epidemiology of bloodstream infections has changed considerably over the past two decades, with Gram-negative organisms again emerging as the predominant pathogens in many regions after a period during which Gram-positive bacteria were more frequently isolated.³ The widespread use of broad-spectrum antibiotics, increasing healthcare-associated infections, invasive procedures, and prolonged hospitalization have accelerated the emergence of antimicrobial-resistant organisms, making empirical treatment increasingly challenging.⁴

Among resistant pathogens, multidrug-resistant (MDR) bacteria have become a major global public health concern. According to the international expert proposal by Magiorakos and colleagues, MDR organisms are defined as isolates demonstrating acquired non-susceptibility to at least one antimicrobial agent in three or more antimicrobial categories^.5 Infections caused by MDR bacteria are associated with delayed administration of effective antimicrobial therapy, prolonged intensive care unit admission, increased healthcare costs, and significantly higher mortality. Hematological malignancy patients are particularly vulnerable because repeated hospital admissions, frequent exposure to broad-spectrum antibiotics, chemotherapy-induced neutropenia, and invasive medical devices create an ideal environment for the selection and transmission of resistant microorganisms.⁶

The burden of antimicrobial resistance is particularly alarming in low- and middle-income countries such as Pakistan, where unrestricted antibiotic availability, inadequate infection prevention practices, overcrowded hospitals, and limited antimicrobial stewardship programs contribute to the rapid emergence of resistant pathogens.⁷ Several Pakistani surveillance studies have demonstrated increasing rates of extended-spectrum β-lactamase (ESBL)-producing Enterobacterales, carbapenem-resistant Acinetobacter baumannii, multidrug-resistant Pseudomonas aeruginosa, and methicillin-resistant Staphylococcus aureus within tertiary healthcare facilities. These organisms frequently complicate the management of immunocompromised patients, including those receiving treatment for hematological malignancies.⁸

Patients with newly diagnosed AML are especially vulnerable during the initial weeks following diagnosis because disease-related immune dysfunction often coexists with chemotherapy-induced neutropenia, severe mucosal injury, frequent blood transfusions, prolonged hospitalization, and the need for central venous catheterization.⁹ These factors substantially increase the risk of bloodstream infection before hematological recovery occurs. Previous studies have identified prolonged neutropenia, prior antibiotic exposure, colonization with resistant organisms, intensive care admission, and invasive procedures as potential predictors of MDR bloodstream infections; however, the magnitude of these associations varies considerably between healthcare settings because local antimicrobial resistance patterns differ substantially across countries and institutions.¹⁰ Consequently, region-specific epidemiological data are essential for developing appropriate empirical antibiotic policies and infection prevention strategies.

Although several international studies have investigated MDR bloodstream infections among patients with hematological malignancies, evidence from Pakistan remains limited and is largely restricted to single-center reports with relatively small sample sizes. Differences in healthcare infrastructure, antimicrobial prescribing practices, microbiological diagnostic capacity, and local resistance profiles limit the applicability of findings from high-income countries to resource-constrained settings. Identification of locally relevant risk factors is therefore essential to improve empirical antimicrobial therapy, strengthen infection control measures, and optimize antimicrobial stewardship programs within Pakistani hospitals. The present study was conducted to determine the frequency of multidrug-resistant bloodstream infections and identify independent risk factors associated with their occurrence among adults with newly diagnosed acute myeloid leukemia receiving care at tertiary healthcare hospitals in Pakistan.

This multicenter analytical cross-sectional study was conducted over a period of one year, from January 2025 to December 2025, in the Departments of Hematology and Microbiology of three tertiary care teaching hospitals in Pakistan. These hospitals serve as major referral centers for patients with hematological malignancies and provide specialized diagnostic, microbiological, and supportive care services. The study was designed to determine the frequency of multidrug-resistant (MDR) bloodstream infections (BSIs) and identify factors associated with their occurrence among adults with newly diagnosed acute myeloid leukemia (AML). Ethical approval was obtained from the Institutional Review Board (IRB) or Ethical Review Committee of each participating institution before commencement of the study. Written informed consent was obtained from all participants or their legally authorized attendants prior to enrollment. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The sample size was calculated using the WHO sample size calculator, assuming a 35% prevalence of multidrug-resistant bloodstream infections, a 95% confidence level, and a 7% margin of error, yielding a minimum required sample of 179 patients. After accounting for incomplete data, the final sample size was increased to 198 participants. Eligible patients were recruited using consecutive non-probability sampling. Adult patients aged 18 years or older with newly diagnosed acute myeloid leukemia, confirmed according to the World Health Organization (WHO) classification based on bone marrow morphology, immunophenotyping, and cytogenetic or molecular investigations where available, were eligible for inclusion. Patients who developed clinically suspected bloodstream infection with at least one positive blood culture before or during induction chemotherapy were enrolled. Bloodstream infection was defined as the isolation of a pathogenic microorganism from one or more blood cultures in the presence of compatible clinical signs and symptoms, including fever (≥38.3°C once or ≥38.0°C sustained for more than one hour), chills, hypotension, or other manifestations of systemic infection. Patients with relapsed or refractory AML, acute promyelocytic leukemia (AML-M3), prior chemotherapy for any hematological malignancy, hematopoietic stem cell transplantation, polymicrobial culture contamination, incomplete microbiological records, fungal bloodstream infections, viral infections without bacterial isolation, or refusal to provide informed consent were excluded. Patients transferred from other hospitals after more than 72 hours of hospitalization were also excluded to minimize variability in prior antimicrobial exposure and infection control practices. Demographic, clinical, laboratory, microbiological, and treatment-related data were collected prospectively using a standardized data collection form. Variables included age, sex, body mass index, residence, smoking status, diabetes mellitus, chronic kidney disease, liver disease, baseline Eastern Cooperative Oncology Group (ECOG) performance status, AML subtype, complete blood count, absolute neutrophil count, serum creatinine, liver function tests, serum albumin, and inflammatory markers where available. Clinical variables included duration of neutropenia, presence of febrile neutropenia, central venous catheter insertion, urinary catheterization, previous hospitalization within the preceding three months, intensive care unit (ICU) admission, blood transfusion requirements, mucositis, duration of hospital stay before onset of bloodstream infection, previous exposure to broad-spectrum antibiotics within the last 30 days, and timing of induction chemotherapy. Peripheral blood cultures were obtained under strict aseptic precautions before initiation or modification of antimicrobial therapy whenever possible. Two sets of blood cultures were collected from separate venipuncture sites, with each set consisting of aerobic and anaerobic culture bottles. For patients with central venous catheters, paired blood cultures from both the catheter and peripheral vein were collected when clinically indicated. Blood culture bottles were incubated using an automated blood culture monitoring system where available; in centers lacking automated facilities, conventional incubation and manual monitoring techniques were employed according to standard microbiological protocols. Positive cultures were subcultured onto appropriate media including blood agar, chocolate agar, and MacConkey agar for bacterial isolation and identification. Microorganisms were identified using standard microbiological methods, including colony morphology, Gram staining, biochemical testing, and automated identification systems where available. Antimicrobial susceptibility testing was performed using the Kirby–Bauer disk diffusion method and interpreted according to the latest Clinical and Laboratory Standards Institute (CLSI) guidelines. Minimum inhibitory concentration (MIC) testing was performed for selected antimicrobial agents whenever clinically indicated or when facilities were available. Quality control procedures were maintained throughout the study using standard American Type Culture Collection (ATCC) reference strains recommended by CLSI. Multidrug resistance was defined according to the international consensus criteria proposed by Magiorakos et al., whereby bacterial isolates demonstrating acquired non-susceptibility to at least one antimicrobial agent in three or more antimicrobial classes were classified as multidrug-resistant. Extended-spectrum β-lactamase (ESBL)-producing Enterobacterales, carbapenem-resistant Gram-negative bacilli, methicillin-resistant Staphylococcus aureus (MRSA), and vancomycin-resistant Enterococcus species were categorized according to CLSI interpretive criteria and international recommendations. The primary outcome was the occurrence of bloodstream infection caused by multidrug-resistant organisms. Secondary outcomes included in-hospital mortality, duration of hospitalization, intensive care unit admission, septic shock, and microbiological spectrum of bloodstream isolates. Potential risk factors evaluated included age, sex, baseline comorbidities, prolonged neutropenia (absolute neutrophil count <500 cells/µL for more than 10 days), prior broad-spectrum antibiotic exposure, duration of hospitalization before bloodstream infection, invasive procedures, central venous catheter use, urinary catheterization, mechanical ventilation, blood transfusion frequency, serum albumin concentration, and ICU admission. Data were entered into Microsoft Excel and analyzed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Continuous variables were tested for normality using the Shapiro–Wilk test and expressed as mean ± standard deviation or median (interquartile range), as appropriate, while categorical variables were presented as frequencies and percentages. Comparisons between patients with MDR and non-MDR bloodstream infections were performed using the independent-samples t-test or Mann–Whitney U test for continuous variables and the Chi-square or Fisher's exact test for categorical variables. Variables with a p-value <0.20 on univariate analysis were included in a multivariable logistic regression model to identify independent risk factors for MDR bloodstream infection, with adjusted odds ratios (AORs) and 95% confidence intervals (CIs) reported. A two-tailed p-value <0.05 was considered statistically significant. Data quality was ensured through standardized data collection, periodic record verification, and adherence to CLSI guidelines for microbiological procedures and antimicrobial susceptibility testing.

A total of 198 adults with newly diagnosed acute myeloid leukemia (AML) who developed culture-confirmed bloodstream infections during the study period were included in the final analysis. The mean age of the study population was 44.8 ± 13.9 years (range: 18–76 years), and 118 (59.6%) were males. Multidrug-resistant (MDR) bloodstream infections were identified in 76 (38.4%) patients, while 122 (61.6%) had bloodstream infections caused by non-MDR organisms.

Table I. Baseline demographic and clinical characteristics of the study population (n=198)

The majority of bloodstream infections were caused by Gram-negative bacteria (72.2%), followed by Gram-positive bacteria (23.2%), while polymicrobial infections accounted for only a small proportion of isolates. Klebsiella pneumoniae was the most frequently isolated pathogen, followed by Escherichia coli, Acinetobacter baumannii, and Pseudomonas aeruginosa. Among Gram-positive organisms, Staphylococcus aureus and Enterococcus faecium were the predominant isolates.

‘

Table II. Distribution of bloodstream pathogens

Patients with MDR bloodstream infections had significantly higher rates of previous broad-spectrum antibiotic exposure, prolonged neutropenia, central venous catheterization, previous hospitalization, and intensive care unit admission compared with patients infected with susceptible organisms. Diabetes mellitus and chronic kidney disease were also more common among the MDR group but did not reach statistical significance.

Table III. Comparison of patients with MDR and non-MDR bloodstream infections

Multivariable logistic regression analysis demonstrated that previous exposure to broad-spectrum antibiotics, prolonged neutropenia, central venous catheter use, ICU admission, and prolonged hospitalization remained independent predictors of MDR bloodstream infection after adjustment for potential confounding variables.

Table IV. Multivariable logistic regression analysis for predictors of MDR bloodstream infection

Clinical outcomes were significantly poorer among patients with MDR bloodstream infections. The median duration of hospitalization was considerably longer in the MDR group (24 vs. 15 days; p<0.001). Septic shock developed in 27 (35.5%) patients with MDR infection compared with 18 (14.8%) patients in the non-MDR group (p=0.001). In-hospital mortality was also significantly higher among patients with MDR bloodstream infections (31.6% vs. 14.8%; p=0.004).

Table V. Clinical outcomes according to MDR status

These findings demonstrate that multidrug-resistant bloodstream infections affected more than one-third of adults with newly diagnosed AML and were predominantly caused by Gram-negative organisms. Previous antibiotic exposure, prolonged neutropenia, central venous catheterization, ICU admission, and prolonged hospitalization were independently associated with MDR bloodstream infections, which were also linked to substantially worse clinical outcomes, including longer hospital stay and increased in-hospital mortality.

A total of 198 adults with newly diagnosed acute myeloid leukemia (AML) who developed culture-confirmed bloodstream infections during the study period were included in the final analysis. The mean age of the study population was 44.8 ± 13.9 years (range: 18–76 years), and 118 (59.6%) were males. Multidrug-resistant (MDR) bloodstream infections were identified in 76 (38.4%) patients, while 122 (61.6%) had bloodstream infections caused by non-MDR organisms.

Table I. Baseline demographic and clinical characteristics of the study population (n=198)

The majority of bloodstream infections were caused by Gram-negative bacteria (72.2%), followed by Gram-positive bacteria (23.2%), while polymicrobial infections accounted for only a small proportion of isolates. Klebsiella pneumoniae was the most frequently isolated pathogen, followed by Escherichia coli, Acinetobacter baumannii, and Pseudomonas aeruginosa. Among Gram-positive organisms, Staphylococcus aureus and Enterococcus faecium were the predominant isolates.

‘

Table II. Distribution of bloodstream pathogens

Patients with MDR bloodstream infections had significantly higher rates of previous broad-spectrum antibiotic exposure, prolonged neutropenia, central venous catheterization, previous hospitalization, and intensive care unit admission compared with patients infected with susceptible organisms. Diabetes mellitus and chronic kidney disease were also more common among the MDR group but did not reach statistical significance.

Table III. Comparison of patients with MDR and non-MDR bloodstream infections

Multivariable logistic regression analysis demonstrated that previous exposure to broad-spectrum antibiotics, prolonged neutropenia, central venous catheter use, ICU admission, and prolonged hospitalization remained independent predictors of MDR bloodstream infection after adjustment for potential confounding variables.

Table IV. Multivariable logistic regression analysis for predictors of MDR bloodstream infection

Clinical outcomes were significantly poorer among patients with MDR bloodstream infections. The median duration of hospitalization was considerably longer in the MDR group (24 vs. 15 days; p<0.001). Septic shock developed in 27 (35.5%) patients with MDR infection compared with 18 (14.8%) patients in the non-MDR group (p=0.001). In-hospital mortality was also significantly higher among patients with MDR bloodstream infections (31.6% vs. 14.8%; p=0.004).

Table V. Clinical outcomes according to MDR status

These findings demonstrate that multidrug-resistant bloodstream infections affected more than one-third of adults with newly diagnosed AML and were predominantly caused by Gram-negative organisms. Previous antibiotic exposure, prolonged neutropenia, central venous catheterization, ICU admission, and prolonged hospitalization were independently associated with MDR bloodstream infections, which were also linked to substantially worse clinical outcomes, including longer hospital stay and increased in-hospital mortality.