Global impact of antibacterial resistance in patients with hematologic malignancies and hematopoietic cell transplant recipients

Abstract

Patients with hematologic malignancies and hematopoetic cell transplant (HCT) recipients are at high risk of developing bacterial infections. These patients may suffer severe consequences from these infections if they do not receive immediate effective therapies, and thus are uniquely threatened by antimicrobial-resistant bacteria. Here, we outline how the emergence of specific resistant bacteria threatens the effectiveness of established approaches to prevent and treat infections in this population. The emergence of fluoroquinolone resistance among Enterobacterales and viridans group streptococci may decrease the effectiveness of fluoroquinolone prophylaxis during neutropenia. The emergence of Enterobacterales that produce extended-spectrum-β-lactamases or carbapenemases and of increasingly resistant Pseudomonas aeruginosa may result in neutropenic patients experiencing delayed time to active antibacterial therapy and consequently worse clinical outcomes. The ability to select targeted antibacterial therapies after the availability of susceptibility data may be limited in patients infected with metallo-β-lactamase-producing Enterobacterales and difficult-to-treat P. aeruginosa. Vancomycin-resistant enterococci and Stenotrophomonas maltophilia can cause breakthrough infections in patients already being treated with broad-spectrum β-lactam antibiotics. Resistance can also limit the ability to provide oral stepdown antibacterial therapy for patients who could otherwise be discharged from hospitalization. We also outline strategies that have the potential to mitigate the negative impact of antimicrobial resistance, including interventions based on active screening for colonization with resistant bacteria and the use of novel rapid diagnostic assays. Additional research is needed to better understand how these strategies can be leveraged to combat the emerging crisis of antimicrobial resistance in patients with hematologic malignancies and HCT recipients.

INTRODUCTION

Antimicrobial resistance is a major threat to public health that challenges our ability to treat and prevent infections. Antimicrobial resistance is globally responsible for over one million deaths annually, as well as considerable economic cost.^1–4 Although significant advances to combat this challenge have occurred, there are concerns that the global toll of antimicrobial resistance will continue to grow.^3,4 Patients with hematologic malignancies and hematopoietic cell transplant (HCT) recipients are uniquely vulnerable to increase in antimicrobial resistance. Firstly, they are more likely to encounter resistant organisms due to frequent contacts with healthcare settings. Secondly, their frequent exposures to antimicrobial agents increases their risk of acquiring and becoming infected with resistant organisms because antimicrobial agents select for resistant pathogens. Thirdly, periods of prolonged neutropenia and chemotherapy-induced gastrointestinal mucositis place them at high risk of translocating oral and gut bacteria into their bloodstream. Lastly, their significantly immunocompromised status makes them more reliant on active antimicrobial therapy to combat these infections than immunocompetent patients.

This review aims to highlight the specific ways in which antimicrobial resistance negatively impacts established approaches to prevention and treatment of infections in patients with hematologic malignancies and HCT recipients. There are several distinct time periods in which antimicrobial resistance creates problems for these patients: 1) decreased effectiveness of antibacterial prophylaxis during neutropenia; 2) decreased effectiveness of empirical antibacterial therapy for initial fever and neutropenia; 3) fewer options for selection of targeted antibacterial therapy after the availability of antimicrobial susceptibility testing (AST) results; 4) increased risk of breakthrough infections during neutropenia in patients already receiving a broad spectrum β-lactam agent; 5) lack of oral options for stepdown therapy after a neutropenic patient has improved during treatment of their infection (Figure 1). Each of these time periods is impacted by specific antimicrobial-resistant pathogens (Table 1), including antimicrobial-resistant Enterobacterales, Pseudomonas aeruginosa, Stenotrophomonas maltophilia, viridans streptococci, and enterococci. Other resistant bacteria, such as methicillin-resistant Staphylococcus aureus and Acinetobacter baumannii, may also cause infections in this patient population, but are beyond the scope of this review. Additionally, resistance in fungi and viruses, and investigational non-antibiotic treatments (e.g., phage therapy), will instead be covered in other manuscripts in this issue.

Figure 1 –

Timeline of issues caused by antibiotic resistance in patients with hematologic malignancies and HCT recepients and potential interventions

Table 1.

Notable antibiotic-resistant bacteria that threaten our approach to prevention and treatment of bacterial infections in patients with hematologic malignancies and HCT recipients

Bacteria	Antibiotic resistance	Timepoint at which resistance is most problematic	Notable considerations
Gram-negative bacteria
Enterobacterales	Fluoroquinolone (FQ) resistance Extended-spectrum β-lactamase (ESBL) production and carbapenem resistance Metallo-β-lactamase (MBL) production	Prophylaxis Empirical therapy Targeted therapy	FQ prophylaxis may be less effective if high background rates of resistance Piperacillin-tazobactam and cefepime are not optimal therapies for ESBL-producing or carbapenem-resistant Enterobacterales (CRE) MBL-producing CRE are resistant to new β-lactam/ β-lactamase inhibitors
Pseudomonas aeruginosa	Intrinsic resistance Difficult-to-Treat Resistance	Empirical therapy Targeted therapy	Multiple resistance mechanisms (AmpC β-lactamases, outer membrane porin changes, and efflux) Defined as resistance to all first-line anti-pseudomonal β-lactam agents and FQs
Stenotrophomonas maltophilia	Intrinsic resistance	Breakthrough infection	Intrinsic restistance to β-lactams, including carbapenems, due to L1 and L2 β-lactamases
Gram-positive bacteria
Viridans group streptococci	Fluoroquinolone resistance	Prophylaxis	Common cause of breakthrough Gram-positive bacteremia during FQ prophylaxis
Enterococcus faecium	Vancomycin resistance	Breakthrough infection	Infections follow intestinal domination after antibiotic exposure

Resistant bacteria that decrease the effectiveness of fluoroquinolone prophylaxis

Current guidelines recommend consideration of fluoroquinolones for prophylaxis in high-risk patients with hematologic malignancies and HCT recipients who are expected to have profound and protracted neutropenia.^{5, 6} These recommendations were established primarily from two randomized, placebo-controlled trials conducted over 20 years ago that demonstrated that levofloxacin prophylaxis reduced the risk of fever and neutropenia and bloodstream infections (BSIs).^7,8 A subsequent meta-analysis identified a mortality benefit with FQ prophylaxis in neutropenic patients.⁹ A more recent randomized trial conducted in children also identified that levofloxacin prophylaxis may decrease the risk of bacteremia.¹⁰ The effectiveness of FQ prophylaxis to prevent bacterial infections in neutropenic patients is predicated on the activity of FQs against common pathogens in neutropenic patients. The emergence of FQ-R bacteria would threaten the effectiveness of this approach.

Fluoroquinolone-resistant Enterobacterales (FQ-R E)

Enterobacterales comprise the majority of Gram-negative BSIs during chemotherapy-induced neutropenia.^11–13 Unfortunately, FQ resistance has become increasingly common in Enterobacterales in the United States.¹⁴ In a recent multicenter study, 22% of urinary E. coli isolates from ambulatory patients were FQ-R and 28% of isolates from hospitalized patients were FQ-R.¹⁵ A high prevalence of FQ-R E has also been observed among patients with hematologic malignancies and HCT recipients. A multicenter U.S. study of BSIs during fever and neutropenia found that 53% of Enterobacterales bloodstream isolates were FQ-R.¹³ FQ-R E are also common in other countries. A multicenter Italian study of 342 BSIs due to E. coli in patients with hematologic malignancies found that 70% of isolates were resistant to FQs.¹⁶ A study of HCT recipients in 25 countries that included 474 BSIs due to Enterobacterales found that 57% of isolates were FQ-R. Increases in the prevalence of FQ-R E have also been reported after the introduction of FQ prophylaxis.¹⁷ Moreover, another study found that BSIs due to FQ-R Gram-negative bacteria in allogeneic HCT recipients were associated with increased mortality compared to BSIs due to FQ-susceptible Gram-negative bacteria.¹⁸ Emerging data suggest that FQ prophylaxis may be ineffective at preventing Gram-negative BSI in patients colonized with FQ-R E prior to neutropenia. A single-center study in New York City found that 23% of patients receiving HCT had FQ-R E colonization prior to their transplant. Nearly one-third of these colonized patients developed Gram-negative BSI during neutropenia, most commonly with their colonizing strain.¹⁹ The emergence of FQ-R E has led some guidelines to emphasize a nuanced approach to FQ prophylaxis, where local patterns of resistance and balancing of risks and benefits are taken into consideration.^{6, 20}

Fluoroquinolone-resistant viridans group streptococci (FQ-R VGS)

Patients with hematologic malignancies who develop neutropenia and oral mucositis from cytotoxic chemotherapy are at high risk for translocation of VGS into the bloodstream.^{21, 22} VGS BSIs in neutropenic patients can potentially lead to a shock or acute respiratory distress syndrome, with mortality ranging from 2 to 21%.²³ Multiple studies have reported the emergence of breakthrough BSIs due to FQ-R VGS in neutropenic patients receiving levofloxacin prophylaxis.^{24, 25} In a U.S. multicenter study of BSIs that occurred during fever and neutropenia, 72% of VGS bloodstream isolates were resistant to levofloxacin.¹³ Moreover, the prevalence of oropharyngeal colonization with FQ-R VGS may increase after the initiation of FQ prophylaxis.²⁶

Resistant bacteria that decrease the effectiveness of empirical therapy for initial fever and neutropenia

Fever and neutropenia in patients with hematologic malignancies is a medical emergency because delays in active antimicrobial therapy for Gram-negative BSIs can lead to a mortality rate as high as 70%.²⁷ Thus, neutropenic patients rely on receiving empirical antimicrobial therapies that are active against the infecting pathogen to prevent clinical decompensation. Unfortunately, the etiology of infection, if present, is unknown at the time of fever and neutropenia and it typically takes 2–4 days for a clinical microbiology laboratory to report the causative pathogen and its susceptibility profiles.²⁸ Thus, the emergence of Gram-negative bacteria that are resistant to commonly used empirical antimicrobial therapies leads to delays in effective therapy and increased mortality. Infectious Diseases Society of America (IDSA) guidelines recommend cefepime, piperacillin-tazobactam, or a carbapenem as empirical therapy for fever in high-risk neutropenic patients.²⁹ A recent survey of 29 U.S. institutions found that the most common empirical Gram-negative agent recommended for fever and neutropenia was cefepime, followed by piperacillin-tazobactam.³⁰ Another study of 14 U.S. centers demonstrated that 62% of patients with fever and neutropenia were treated empirically with cefepime, 23% were treated with piperacillin-tazobactam, and 8% were treated with meropenem.¹³ However, due to the emergence of resistant Gram-negative bacteria, 13% of these patients did not receive empirical therapy that had in vitro activity against their Gram-negative pathogen.¹³ The most concerning organisms that are frequently resistant to first-line empirical agents for fever and neutropenia are ESBL-E, CRE, and Pseudomonas aeruginosa.

Extended-spectrum β-lactamase-producing Enterobacterales (ESBL-E)

ESBL-E possess β-lactamases capable of hydrolyzing penicillins, extended-spectrum cephalosporins, and aztreonam. Infections with ESBL-E are not adequately treated with cefepime or piperacillin-tazobactam. A recent multicenter U.S. surveillance study showed that less than 10% of ESBL-E isolates from 2016–2020 were susceptible to cefepime.³¹ Moreover, an observational study of ESBL-E BSI in the general population found that cefepime therapy had inferior clinical outcomes compared to carbapenem therapy, even when the bloodstream isolates were cefepime susceptible.³² Although many ESBL-E isolates test susceptible to piperacillin-tazobactam, a multicenter randomized trial in the general patient population found that patients with ESBL-E BSI who were treated with piperacillin-tazobactam had a 3-fold increase in 30-day mortality compared to patients treated with carbapenems.³³

ESBL-E have become increasingly common causes of BSI in patients with hematologic malignancies and HCT recipients. Reports from different countries indicate that ESBL-E represent 21–46% of BSIs due to Enterobacterales in patients with hematologic malignancies (Table 2).^34–40 A study from Lebanon reported a rate as high as 78%, highlighting the importance of geographic variation and local epidemiology.⁴¹ A single-center longitudinal study from Spain noted that the proportion of ESBL production among Enterobacterales bloodstream isolates in HCT recipients has increased over time, from 0% in 1998–2002, to 8% in 1998–2002, to 17% 2003–2007, to 34% in 2008–2012 and 21% from 2013–2017.⁴² Mortality after ESBL-E BSI in patients with hematologic malignancies and HCT recipients is high,^{34–38, 43–45} and some studies have noted increased mortality after ESBL-E BSIs compared to non-ESBL-E BSIs.^{34, 46, 47} Although routine empirical therapy with a carbapenem would optimally treat patients infected with ESBL-E BSIs, this is not an optimal strategy because routine empirical carbapenem use would likely exacerbate the emerging problem of CRE. Furthermore, routine use of carbapenems and other agents active against anaerobes in allogeneic HCT recipients has been associated with increased risk of graft-versus-host disease, possibly related to effects on the gut microbiome.^48–51

Table 2.

Prevalence and characteristics of bloodstream infections due to ESBL-E and associated mortality in patients with hematologic malignancies and HCT recipients

First author	Location (No. of centers)	Years	Organisms	% of isolates that are ESBL producers	Mortality of ESBL-E BSI vs non-ESBL-E BSI	Mortality type
Kang30	Korea (18)	2008–2009	E. coli, K. pneumoniae	33% (37/156)	45% vs. 14%	30-day
Mechergui36	Tunisia (1)	2002–2011	E. coli, K. pneumoniae	21% (74/360)	NR	NR
Metan31	Turkey (1)	2006–2011	Enterobacterales	33% (40/120)	13 vs. 15%	7-day
Trecarichi32	Italy (9)	2009–2012	Enterobacterales	37% (98/265)a	26% vs. 5%	21-day
Haddad37	Lebanon (1)	2007–2017	E. coli, K. pneumoniae	78% (86/110)b	NR	NR
Islas-Munoz35	Mexico (1)c	2016–2017	E. coli, K. pneumoniae, E. cloacae	38% (42/110)	NR	NR
Liang33	China (3)	2010–2018	E. coli, K. pneumoniae	46% (204/449)	17% vs. 17%	30-day
Chumbita34	Spain (14)	2019	E. coli, K. pneumoniae	21% (42/197)	14% vs. 13%	30-day

Abbreviations: ESBL-E, extended-spectrum beta-lactamase-producing Enterobacterales; HCT, hematopoietic cell transplant; No, number; NR, not reported.

^aResistance to third-generation cephalosporins was used as a surrogate for ESBL production

^bMethod of ESBL determination not specified

^cAll enrolled patients in this study underwent HCT

Carbapenem-resistant Enterobacterales (CRE)

CRE often harbor carbapenemases that hydrolyze not only extended-spectrum cephalosporins and aztreonam, but also hydrolyze carbapenems. In the U.S., China, and South America, K. pneumoniae carbapenemase (KPC) is the most common carbapenemase, although other enzymes may be more prominent in other areas.^52–57 Multicenter studies in New York and Italy have found that CRE represent 5–6% of Gram-negative BSIs in patients with hematologic malignancies.^{36, 58} Among Enterobacterales in the U.S., Klebsiella pneumoniae is most likely to be carbapenem resistant because it is mostly likely to harbor KPC.^{52, 53} Studies from New York, Italy, Israel, and China have found that 18–35% of Klebsiella bloodstream isolates from patients with hematologic malignancies are carbapenem resistant.^{36, 58–61} Unsurprisingly, mortality related to CRE infections in patients with hematologic malignancies is high, ranging from 29% to 58% (Table 3).^{58, 61–67} Although new agents, such as novel β-lactam/β-lactamase inhibitors combinations, have become available, these agents are not active against all CRE, such as metallo-β-lactamase (MBL)-producing CRE.^68–72 Moreover, they are unlikely to be used empirically in neutropenic patients, which may lead to delays in active therapy for neutropenic patients infected with CRE.^68–72

Table 3.

Characteristics of CRE infections and associated mortality in patients with hematologic malignancies and HCT recipients

First author	Location (No. of centers)	Years	Pts	Infection type	Organisms	Mortality (type)	Infection-related mortality
Satlin44	New York, USA (2)	2008–2012	43	Bacteremia	K. pneumoniae (n=30), E. cloacae (n=8), others (n=5)	53% (30-day)	51%
Girmenia54	Italy (45)	2010–2013	112	Bacteremia (n=99), Pneumonia (n=12), Skin (n=1)	K. pneumoniae	58% (90-day)	54%
Tofas55	Greece (4)	2010–2014	50	Bacteremia	K. pneumoniae	50% (14-day)	50%
Trecarichi56	Italy (13)	2010–2014	161	Bacteremia	K. pneumoniae	52% (21-day)	NR
Castón57	Spain, Israel (4)	2012–2016	31	Bacteremia	Enterobacterales	45% (30-day)	NR
Liu49	China (1)	2014–2018	20	Bacteremia	K. pneumoniae	55% (30-day	NR
Tumbarello58	Italy (22)	2018–2020	46	Bacteremia (n=40), cUTI (n=4), LRTI (n=2)	K. pneumoniae	44% (30-day)	NR
Zhang59	China (1)	2012–2021	94	Bacteremia	Enterobacterales	29% (30-day)	NR

Abbreviations: CRE, carbapenem-resistant Enterobacterales; cUTI; complicated urinary tractin infection; HCT, hematopoietic cell transplant; LRTI, lower respiratory tract infection; No, number; NR, not reported; Pts, patients.

Pseudomonas aeruginosa

Pseudomonas aeruginosa is a significant pathogen in neutropenic patients and an anti-pseudomonal β-lactam agent is an essential component of empirical therapy.^{27, 29} Pseudomonas aeruginosa is responsible for 9–28% of Gram-negative BSIs in patients undergoing HCT.^{13, 73–77} It is notorious for its ability to develop antimicrobial resistance via a variety of mechanisms.⁷⁸ A multicenter report by the Centers for Disease Control and Prevention found that 20% of central line-associated BSIs due to P. aeruginosa in oncology units from 2009–2012 were carbapenem-resistant and 14% were resistant to anti-pseudomonal cephalosporins.⁷⁹ A recent multicenter study of P. aeruginosa BSIs in 280 patients with hematologic malignancies in Spain found that 36% of isolates were resistant to cefepime, piperacillin-tazobactam, or a carbapenem.⁸⁰ Moreover, receipt of empirical therapy that was not active against P. aeruginosa bloodstream isolates in this study was associated with a 2.7-fold increase in odds of 30-day mortality compared to receipt of active empirical therapy. A multicenter Italian study of P. aeruginosa bloodstream isolates from hematology patients demonstrated even more extensive resistance profiles, with only 58% of isolates being susceptible to piperacillin-tazobactam and 29% to meropenem.³⁶ Patients with hematologic malignancies who have BSIs due to multidrug-resistant (MDR) P. aeruginosa (isolates resistant to at least 3 antibiotic classes) are more likely to receive inactive empirical antibacterial therapy and have greater mortality compared to patients infected with non-MDR strains.^{36, 80, 81} Combination empirical therapy with an anti-pseudomonal β-lactam and an aminoglycoside may increase the chance of providing an active antibiotic more quickly while awaiting AST results. However, the use of aminoglycosides carries risks of additional adverse events and older studies did not find that combination empirical therapy improves outcomes in neutropenic patients.⁸²

Resistance that limits selection of targeted therapies

Targeted therapy refers to antimicrobial agents administered after the culprit organism and its antimicrobial susceptibilities are identified. Prior to the availability of new β-lactam/β-lactamase inhibitors, treatment options for many patients infected with CRE and MDR P. aeruginosa were limited to polymyxins, tigecycline, and aminoglycosides, drugs that are relatively ineffective for BSIs or have unfavorable toxicity profiles.^83–85 The availability of new β-lactam/β-lactamase inhibitors is a major advance for patients with MDR infections because treatments with these agents are associated with improved outcomes compared to prior therapies.^{70, 71, 86, 87} However, patients with hematologic malignancies and HCT recipients can still become infected with pathogens resistant to these newer agents, which limits the selection of adequate targeted therapies. Additionally, access to these newer agents may be limited in certain countries, particularly in low- and middle-income countries that often have high rates of antimicrobial resistance.⁸⁸ The pathogens that are most commonly encountered that have limited treatment options are MBL-producing CRE and difficult-to-treat resistance (DTR) P. aeruginosa.

Metallo-β-lactamase-producing Enterobacterales (MBL-producing CRE)

CRE that produce MBLs pose a significant challenge for targeted therapy because these carbapenemases are not inhibited by new β-lactamase inhibitors avibactam, vaborbactam, and relebactam. MBLs include NDM, VIM, and IMP enzymes. NDM has become a predominant carbapenemase in India, Pakistan, and the UK, and is becoming more common in other regions, being identified in 11–12% of CRE isolates in the U.S. and Western Europe, and 23–24% of CRE isolates from Eastern Europe and Latin America.^{54, 56, 57, 89} Infections due to NDM-producing CRE have also been reported in patients with hematologic malignancies. A U.S. cancer center reported 5 cases of BSI due to NDM-producing Enterobacterales in patients with hematologic malignancies over an 11-month period, and three of these patients died.⁹⁰ Another report from China identified BSIs due to NDM-producing K. pneumoniae in patients with hematologic malignancies that were resistant to not only new β-lactam/β-lactamase inhibitors, but also to cefiderocol.⁹¹ Treatment of MBL-producing CRE is often limited to the combination of aztreonam with ceftazidime-avibactam concurrently, or consideration of other regimens such as cefiderocol, aminoglycosides, or polymyxins.⁹² As these highly resistant pathogens become increasingly common, more patients with hematologic malignancies and HCT recipients will face limited options for targeted therapies.

Difficult-to-treat resistance (DTR) Pseudomonas aeruginosa

DTR P. aeruginosa isolates are resistant to all first-line anti-pseudomonal β-lactam agents and fluoroquinolones.⁹³ Multicenter studies in the general population have shown that P. aeruginosa is the most common DTR pathogen, with a U.S. study demonstrating it accounted for 38% of DTR infection episodes, and a Korean study showing that 18% of their P. aeruginosa bloodstream isolates were DTR.^{94, 95} The latter study also demonstrated higher mortality among patients with DTR Gram-negative BSI compared to non-DTR Gram-negative BSI.⁹⁵ Novel agents such as ceftolozane-tazobactam, which often retain activity against DTR P. aeruginosa isolates, have demonstrated effectiveness in neutropenic patients with hematologic malignancies who have BSIs due to these organisms.^{96, 97} However, carbapenemase-producing P. aeruginosa isolates have emerged in many global regions that are resistant to ceftolozane-tazobactam and ceftazidime-avibactam.^{98, 99} Moreover, non-carbapenemase-producing P. aeruginosa isolates have emerged that are resistant to ceftolozane-tazobactam and ceftazidime-avibactam because of mutations in intrinsic P. aeruginosa cephalosporinases, and this resistance can emerge on therapy.^100–102 These mutations may also confer cross-resistance to cefiderocol, which limits options for antimicrobial therapy even further.¹⁰³

Resistant bacteria that cause breakthrough infections in patients receiving broad-spectrum β-lactam therapy

Many patients who receive broad-spectrum β-lactam therapies for initial fever and neutropenia can subsequently develop breakthrough infections due to bacteria that are resistant to the antibiotic that they are receiving. These breakthrough pathogens are typically resistant to empirical therapies for fever and neutropenia, leading to delays in receipt of effective therapy. Recurrent BSIs in patients receiving empirical therapies for fever and neutropenia may occur in 10% of these patients by day 30.¹³ Two of the most prominent pathogens that arise in this setting are VRE and Stenotrophomonas maltophilia.

Vancomycin-resistant enterococci (VRE)

Although VRE are uncommon causes of BSI during initial fever and neutropenia, VRE are common bloodstream pathogens in patients already being treated with first-line agents for fever and neutropenia. One study of HCT recipients found that VRE did not cause any BSIs in neutropenic patients not receiving antibacterial agents, but caused 55% of BSIs in neutropenic patients receiving a broad-spectrum β-lactam agent.¹⁰⁴ The cumulative incidence of VRE BSI after an allogeneic HCT has been reported as 6 to 16% and some studies have reported that VRE is the most common cause of BSI early after allogeneic HCT.^104–108 VRE BSI is thought to occur frequently in this population due to exposure to broad-spectrum antimicrobial therapies that disrupt normal enteric flora and lead to intestinal domination by VRE.¹⁰⁹ Although VRE was traditionally thought of as a pathogen of limited virulence, VRE BSI can have significant clinical consequences in neutropenic patients, with rates of septic shock and mortality that are similar to those associated with Gram-negative BSIs.¹⁰⁴

There are few treatment options for VRE BSIs in neutropenic patients and the treatments have significant limitations. Linezolid can cause myelosuppression, which is of potential concern for allogeneic HCT recipients,¹¹⁰ although some studies support the safety of linezolid in this population.^{111, 112} Daptomycin requires dosages that are greater than what is listed in the U.S. Food and Drug Administration (FDA) label to achieve drug exposures that are associated with efficacy in animal and clinical models.^{113, 114}

Stenotrophomonas maltophilia

Stenotrophomonas maltophilia is an environmental Gram-negative bacillus that is an emerging pathogen in immunocompromised patients who have been exposed to antimicrobial agents. Stenotrophomonas maltophilia is intrinsically resistant to carbapenems and other β-lactam agents used empirically for fever and neutropenia due to a chromosomally-encoded L1 metallo-β-lactamase and an L2 cephalosporinase.¹¹⁵ One surveillance study found that S. maltophilia is the most common carbapenem-resistant Gram-negative bloodstream pathogen in the U.S.¹¹⁶ Stenotrophomonas maltophilia can account for 2–7% of Gram-negative BSIs in patients with hematologic malignancies.^{35, 36, 117, 118} The most important risk factor for S. maltophilia infections in patients with hematologic malignancies is prior use of broad-spectrum therapies, particularly carbapenems.^119–121 Other risk factors in this population include acute leukemia, degree of neutropenia, mucositis, and presence of a central venous catheter. Risk scores based on some of these variables have been proposed to predict which patients may have S. maltophilia infections.¹²¹ Mortality rates for S. maltophilia BSI in patients with hematologic malignancies range from 24 to 38%.^{122, 123} Combination therapy with two agents with in vitro activity is recommended by IDSA for the treatment of S. maltophilia BSI,⁹² which can be challenging in neutropenic patients. Trimethoprim-sulfamethoxazole (TMP-SMX) is one of the most commonly used agents for S. maltophilia, but can cause myelosuppression, which is of particular concern in HCT recepients at time of engraftment.^{124, 125}

Resistance that limits use of oral stepdown agents, leading to continued parenteral therapy

Transition from parenteral to oral therapy is appropriate for most patients with Gram-negative BSI after clinical improvement and source control, particularly when highly bioavailable oral agents such as FQs and TMP-SMX are used.¹²⁶ However, antimicrobial resistance can limit the ability to switch to an active oral agent for patients with hematologic maligiancies and HCT recipients. The need to continue parenteral therapy leads to longer hospital stays or necessitates the use of a peripherally inserted central catheter, which in turn causes increased morbidity and healthcare costs.^{127, 128} For example, ESBL-E and CRE are frequently also resistant to FQs and TMP-SMX, preventing transition to these oral options.^129–131 Pseudomonas aeruginosa is intrinsically resistant to all available oral agents except for FQs,¹³² and thus the emergence of FQ-R P. aeruginosa precludes the potential for oral therapy. An international study of HCT recipients across 3 continents reported that only 30% of P. aeruginosa bloodstream isolates were susceptible to FQs.¹³³

Interventions to curb the impact of antimicrobial resistance

Screening for colonization with resistant bacteria:

Screening patients with hematologic malignancies for colonization with resistant bacteria prior to intensive chemotherapy has many potential uses to curb the impact of antimicrobial resistance. First, identifying patients colonized with resistant bacteria can lead to the implementation of isolation precautions that may decrease transmission of these organisms among patients on a hematology or transplant unit. Active surveillance has been shown to be an important component of strategies to control the spread of resistant organisms such as CRE,^134–136 and has been implemented effectively in high-risk cancer patients and patients undergoing HCT.^{137, 138}

Screening for colonization with FQ-R E may identify patients for whom FQ prophylaxis is likely to be ineffective. A study of 234 HCT recipients receiving levofloxacin prophylaxis during neutropenia found that 23% were colonized with FQ-R E prior to transplant.¹⁹ Thirty-one percent of FQ-R E-colonized patients developed Gram-negative bacteremia, compared to only 1% of patients without FQ-R E colonization. Moreover, FQ-R E BSIs were typically caused by colonizing strains. These data suggest that screening for FQ-R E colonization may identify patients for whom an alternate approach to FQ prophylaxis is needed, such as prophylaxis with a different antibiotic (e.g., oral cephalosporin), omission of prophylaxis, or decolonization strategies. However, further studies are needed to understand how to optimize these screening-based strategies.

Active screening for targeted decolonization of resistant organisms is an area of active investigation. Theoretically, identifying colonization with a resistant gut organism could allow for decolonization to prevent translocation of bacteria into the bloodstream during neutropenia. Studies of selective digestive decontamination have shown limited short-term effectiveness in suppression of resistant organisms.^139–143 However, using oral antibiotics for decolonization is not routinely recommended because of the transient nature of these effects and concerns related to the emergence of resistance to the agents used for decolonization and detrimental consequences to the gut microbiome.¹⁴⁴ Fecal microbiota transplantation is another emerging strategy that could suppress colonization with resistant bacteria. However, there are currently concerns regarding the risk of this intervention causing transmission of dangerous pathogens to immunocompromised patients.¹⁴⁵

Screening for colonization with resistant bacteria can also be used to guide empirical antimicrobial therapy in patients who subsequently have fever and neutropenia. A study of 312 HCT recipients found that 32% of patients colonized with ESBL-E prior to their transplant developed ESBL-E bacteremia during their transplant admission, compared to 0.4% of patients not colonized with ESBL-E.¹²⁹ Moreover, the ESBL-E BSIs typically occurred during initial fever and neutropenia and were caused by the colonizing strains. These data suggest that screening for colonization with ESBL-E could be used to identify patients at high risk of ESBL-E BSI, for whom empirical carbapenem therapy is appropriate for fever and neutropenia. Guidelines of the American Society of Transplantation and Cellular Therapy suggest that an individualized antibiotic plan can be formulated for empirical therapy based on screening for colonization with resistant bacteria.¹⁴⁴ However, additional research is needed to define the optimal frequency of screening and the optimal screening method (e.g., molecular vs. selective culture) for these individualized antibiotic plans.

Use of rapid diagnostic assays to detect etiologies of infection:

Emerging rapid diagnostics tools can identify the etiologies of infection more quickly than culture-based methods, which allows clinicians to expeditiously initiate appropriate therapy for BSIs. An overview of FDA-cleared rapid diagnostics for BSIs is shown in Table 4. Multiple molecular diagnostic assays are available that can be applied directly to positive blood culture broths to identify bloodstream pathogens and the presence of resistance genes, such as bla_KPC for Gram-negative bacteria and vanA for Gram-positive bacteria, within hours of positive blood culture results.^146–150 This effectively decreases the time to determination of crucial information about bloodstream pathogens from 24–72 hours after blood culture positivity to about 2 hours, allowing for earlier optimization of antibacterial therapy.¹⁵¹ Despite these advantages, clinical data demonstrating improved outcomes with use of these rapid diagnostic assays are limited, in part because most of these studies evaluated a limited number of patients infected with resistant organisms. In contrast, a multicenter observational study of 137 patients with CRE bacteremia found that use of a rapid molecular diagnostic blood culture assay was associated with decreased time to appropriate therapy and decreased mortality for these infections.¹⁵² Data for other resistant organisms and in patients with hematologic malignancies and HCT recipients are limited.

Table 4.

FDA-cleared diagnostic tests that may rapidly identify bloodstream infections due to problematic antibiotic-resistant bacteria in patients with hematologic malignancies and HCT recipients

Test	Methodology	No. of bacteria identified	Notable bacterial resistance identified	Turnaround Time (hrs)
Genotypic methods applied to positive blood culture bottlesa
BioFire® BCID2	Nested multiplex PCR with melt curve analysis	26	CTX-M-producing and carbapenemase-producing Enterobacterales VRE	1–2
Verigene®	Microarray-based detection using capture and detection probes	9 (Gram-negative panel) 13 (Gram-positive panel)	CTX-M-producing and carbapenemase-producing Enterobacterales VRE	2
Genmark® ePlex®	Multiplex nucleic acid amplification assay based on competitive DNA hybridization and electrochemical detection	21 (Gram-negative panel) 20 (Gram-positive panel)	CTX-M-producing and carbapenemase-producing Enterobacterales VRE	1.5
Phenotypic methods applied to positive blood culture bottlesa
Accelerate Pheno®	Fluorescence in situ hybridization (ID), time-lapse imaging of bacteria under dark-field microscopy (AST)	14	Enterobacterales and P. aeruginosa resistant to piperacillin-tazobactam, cephalosporins, carbapenems, ciprofloxacin, and/or aminoglycosides VRE	2 (ID) 7 (AST)
Method applied directly to blood without the need for growth in blood culture bottles
T2Bacteria® Panel	Detection of bacterial cells via T2 Magnetic resonance on whole blood	E. coli E. faecium K. pneumoniae P. aeruginosa S. aureus	N/A	3.5–7b

Abbreviations: AST, antimicrobial susceptibility testing; BCID, Blood Culture Identification; FDA, U.S. Food and Drug Administration; ID, identification; N/A, not applicable; PCR, polymerase chain reaction; VRE, vancomycin-resistant enterococci.

^aAll listed methods applied to positive blood culture bottles can additionally detect the presence of methicillin-resistant Staphylococcus aureus

^bTurnaround time for direct-from-blood methods refers to time from obtaining blood samples, as opposed to blood culture-based methods for which turnaround time represents time from blood culture positivity

A primary limitation of these rapid molecular diagnostic assays is that detecting the presence or absence of a small number of resistance genes does not provide complete antimicrobial susceptibility information. This is particularly true for P. aeruginosa, which is frequently resistant to β-lactams due to changes in outer membrane porins and upregulation of efflux pumps, not necessarily the β-lactamases targeted by these panels.⁷⁸ Thus, the ability to perform rapid phenotypic AST can offer advantages to rapid molecular diagnostic assays. One FDA-cleared assay has the ability to report phenotypic AST results within 7 hours of blood culture positivity.^{153, 154} In a randomized trial, this assay demonstrated decreased time to antimicrobial escalation for antimicrobial-resistant BSIs, but no differences in clinical outcomes compared to a standard AST method.¹⁵⁵

A limitation of all these assays is that they can only be utilized after a blood culture is flagged as positive, a process that takes 12–24 hours for the most common target pathogens.¹⁵⁶ A direct-from-blood non-culture-based diagnostic assay is now available that can successfully identify E. coli, K. pneumoniae, P. aeruginosa, E. faecium, and Staphylococcus aureus within 4–8 hours from blood culture collection.¹⁵⁷ However, this assay does not provide information on antimicrobial susceptibilities and only detects five organisms, limiting its clinical utility. Studies that characterize the utility of rapid diagnostic assays in patients with hematologic malignancies and HCT recipients are needed to better understand their impact on clinical outcomes in this population.

CONCLUSIONS

Antimicrobial resistance has an ever-increasing impact on patients with hematologic malignancies and HCT recipients. Resistance poses unique challenges to this vulnerable patient population, threatening the effectiveness of antibacterial prophylaxis, empirical therapy for fever and neutropenia, selection of targeted therapy, treatment of breakthrough infections, and transitions to oral therapy. Screening to identify patients colonized with resistant organisms and the implementation of novel rapid diagnostic assays are important strategies to mitigate these emerging challenges.