Evaluating the Use of Meropenem in Hematologic Patients with Febrile Neutropenia: A Retrospective Observational Single-Cohort Study

To the editor.

Patients with malignant hematologic diseases or after haematopoietic stem cell transplantation (HSCT) have weakened immune systems due to their primary diagnosis and/or treatment.¹ Treatment can induce periods of neutropenia, defined by an absolute neutrophil count (ANC) below <500 cells/μL.^2,3 In addition to often concurrent mucositis, this predisposes to infections, which often occur in this population.⁴ In patients with chemotherapy-induced neutropenia, the prevalence of febrile neutropenia has been suggested to rise up to 80%.⁴ Fever caused by an infection is a dangerous complication and can be lethal; therefore, prompt antibiotic treatment is indicated.^3,4

Meropenem is an ultra broad-spectrum antibiotic in the β-lactam class often used to treat febrile neutropenia.^5,6 Empiric treatment with meropenem reduces mortality risk in patients with infections caused by extended-spectrum beta-lactamase (ESBL)-producing bacteria and other multidrug-resistant (MDR) gram-negative bacteria.² However, meropenem use is associated with an increased risk of clostridium infections and candidemia and an increased risk of acute graft-versus-host disease in patients undergoing allogeneic HSCT.^7–12 For these reasons, as well as to promote antibiotic stewardship, the Dutch Working Party on Antibiotic Policy (SWAB) has labelled carbapenems, including meropenem, as a second-line treatment option.¹³

The nationally recommended primary choice of antimicrobial therapy is dependent on a pre-emptive risk stratification based on the expected duration of neutropenia (≤7 days vs. >7 days).¹³ In high-risk neutropenic patients, antipseudomonal β-lactams such as ceftazidime and piperacillin/tazobactam are the preferred choice of antibiotic therapy.¹³ Standard-risk neutropenic patients with an expected short duration of neutropenia are treated according to their Multinational Association of Supportive Care in Cancer (MASCC) score.

Until 2023, the protocol in our hospital advised treatment of febrile neutropenia in high-risk patients with meropenem for at least 72 hours.⁶ As no data are available on local resistance patterns for antipseudomonal β-lactams, making unguided changes to the treatment protocol to adhere to the national guidelines can be challenging in this vulnerable patient population.

This study aims to determine the frequency of bacteria resistant to ceftazidime, piperacillin/ tazobactam, and meropenem in diagnostic cultures in haematology patients admitted with febrile neutropenia to our hospital. Doing so can provide insight into the appropriateness of meropenem use and possibilities for responsible adjustments to the current empiric febrile neutropenia treatment protocol.

Methods

Study design and outcomes

A retrospective, observational, single-centre study was carried out at Meander MC - a teaching hospital in the Netherlands, using a single cohort design of adult patients admitted with hematologic disease and febrile neutropenia between October 2018 and June 2021. The primary outcome was the frequency of bacteria resistant to the antibiotics of interest in diagnostic blood and urine cultures taken on admission for febrile neutropenia. Our antibiotics of interest were ceftazidime, piperacillin/ tazobactam, and meropenem.

Only the first diagnostic blood and urine samples, taken after the occurrence of fever, were included in the results. Two other relevant diagnostic cultures, namely a line tip and wound culture, were also included in data collection as their results were the base of treatment evaluation.

Although meropenem has an ultra-broad-spectrum coverage, it does not treat infections caused by some gram-positive cocci, such as Staphylococcus epidermis and Enterococcus faecalis.¹⁴ Except for E. faecalis, which is sensitive to piperacillin/tazobactam, the bacteria mentioned above were resistant to both meropenem and either of the alternatives.¹⁴ When these gram-positive cocci are either suspected or found, they are treated with different antibiotics, such as vancomycin.¹³ To avoid reporting results biased as a higher resistance frequency, we displayed bacteria separately based on their resistance status to meropenem. These are not included in the calculations of resistance percentages to piperacillin/tazobactam and ceftazidime.

Data collection and statistical analysis

Data collection was carried out in accordance with the Dutch Medical Treatment Contracts Act (WGBO). The study was approved by the scientific research committee of the hospital. Data were analysed using SPSS (version 24). Categorical variables were reported as frequencies. Continuous variables were defined as mean and standard deviation when normally distributed or as a median and interquartile range when they were not.

Results

Population demographics

100 patients (58 male, 42 female) admitted between October 2018 and June 2021 were enrolled in this study. The median age was 65.0 (54.0–73.8) years, the median BMI was 24.95 (22.3–29.4) kg/m² and the median duration of hospital admission was 21.5 (9.3–31.0) days. Additional population demographics are shown in Table 1.

Table 1.

Population demographics and characteristics (n = 100).

Variable	Frequency
Age (years)
<60 years	35
≥60 years	65
Sex
Male	58
Female	42
Primary diagnosis *
AML	38
NHL	31
MM	11
CLL/HCL	9
MDS	4
ALL	4
RAEB	1
CML	1
Myelofibrosis	1
Vital status
Alive	86
Death ≤30 days after admission	10
Death during admission	4
Lines
None	55
Central	44
Midline	1
Duration of hospital admission
<7 days	25
≥7 days	75
Readmission ≤ 3 months
No	73
Yes	27

^*Acute Myeloid Leukemia (AML); Non-Hodgkin Lymphomas (NHL, all grades), Multiple Myeloma (MM); Chronic Lymphatic Leukaemia (CLL); Hairy Cell Leukaemia (HCL); Myelodysplastic Syndrome (MDS); Acute Lymphatic Leukaemia (ALL); Refractory Anemia with Excess Blasts (RAEB); Chronic Myeloid Leukaemia (CML)

Resistance frequencies

Blood and urine cultures were taken in 100 and 62 patients, respectively. Two other diagnostic cultures originated from a line tip and wound. Resistance to ceftazidime was found in seven (7%) patients, divided over seven blood cultures and one wound culture (Figure 1). Resistance to piperacillin/tazobactam was confirmed in only a single urine culture from one (1%) patient.

Figure 1.

Microbiologic data on diagnostic cultures (blood/urine/other). Other: a line tip and wound culture. Pip/tazo: piperacillin/tazobactam. Only the first cultures taken after diagnosis of febrile neutropenia are included.

Discussion

The aim of this study was to determine the frequency of bacteria resistant to ceftazidime, piperacillin/tazobactam and meropenem in diagnostic cultures of hematologic patients admitted to our hospital with febrile neutropenia. Retrospective analysis of diagnostic cultures showed a resistance frequency of 7% to ceftazidime and 1% to piperacillin/tazobactam. Furthermore, three diagnostic cultures showed E. faecalis, which is susceptible to the latter but a poor target for meropenem. These frequencies support that the hospital’s empiric treatment protocol for haematology patients admitted with febrile neutropenia can be safely adjusted from meropenem to piperacillin/tazobactam.

It is important to keep in mind that Enterobacterales with intrinsic, chromosomally encoded AmpC beta-lactamase (“AmpC producers”), such as S. marcescens, C. freundii and E. cloacae complex, can develop resistance to penicillins and cephalosporins due to the selection of de-repressed mutants during treatment.¹⁵ The risk of resistance selection is especially high during therapy with third-generation cephalosporins such as ceftazidime, whereas piperacillin/tazobactam is only a weak inducer of AmpC derepression. Data from observational studies suggest that piperacillin/ tazobactam may be a treatment option for bloodstream infections with AmpC producers, but no clinical trials are available.^16–19 Therefore, current guidelines for antibiotic use in our hospital do not recommend the use of penicillins (including piperacillin/tazobactam) and cephalosporins for the treatment of infections with AmpC producers and susceptibility results for penicillins, and cephalosporins are not reported to the clinicians. For critically ill patients with febrile neutropenia admitted to the intensive care unit, the first choice of treatment remains a carbapenem.

This study provides data from a relatively large sample readily applicable to the hospital’s clinical practice. However, a prospective follow-up study comparing clinical outcomes before and after the suggested treatment adjustments can strengthen the recommendations made. These outcomes should involve mortality risk and resistance patterns at a minimum to confirm the expected benefits, including antibiotic stewardship, without impairing clinical outcomes.

Patient characteristics available at admission, such as age, BMI, and recent hospital admissions, hold predictive value and allow for more precise risk stratification.^1,2 Including prospectively validated MASCC scores or other alternatives would allow for more accurate assessments, thus further guiding clinicians to the most appropriate antibiotic therapy.^13,20

Conclusions

Based on the results of this study, we have changed our protocol of empiric antibiotic therapy of chemotherapy-induced neutropenia from meropenem to piperacillin/tazobactam. Making this carefully considered change helps us promote antibiotic stewardship while preserving our patients’ safety.