Abstract
Background
Approximately 40% of all Enterobacterales (EB) bloodstream infections (BSIs) among solid organ transplant recipients (SOTRs) are due to extended-spectrum β-lactamase (ESBL)–producing organisms, but risk factors for such infections remain ill defined in this population. We sought to determine the risk factors for ESBL-EB BSIs among SOTRs.
Methods
A multicenter case-control study was performed. All SOTRs with an EB BSI at the Hospital of the University of Pennsylvania and University of Maryland Medical Center between 1 January 2007 and 30 June 2018 and at The Johns Hopkins Hospital between 1 January 2005 and 31 December 2015 were included. Cases were those with an ESBL-EB BSI. Controls were those with a non–ESBL-EB BSI. Multivariable logistic regression was performed to determine risk factors for ESBL-EB BSI.
Results
There were 988 episodes of EB BSI, of which 395 (40%) were due to an ESBL-EB. On multivariable analysis, the independent risk factors for ESBL-EB BSI included: ESBL-EB on prior culture (aOR, 12.75; 95% CI, 3.23–50.33; P < .001), a corticosteroid-containing immunosuppression regimen (aOR 1.30; 95% CI 1.03–1.65; P = .030), acute rejection treated with corticosteroids (aOR 1.18; 95% CI 1.16–1.19; P < .001), and exposure to third-generation cephalosporins (aOR 1.95; 95% CI 1.48–2.57; P < .001), echinocandins (aOR 1.61; 95% CI 1.08–2.40; P = .020), and trimethoprim-sulfamethoxazole (aOR 1.35; 95% CI 1.10–1.64; P = .003).
Conclusions
We identified several novel risk factors that are uniquely important to the SOTR population, including exposure to trimethoprim-sulfamethoxazole and corticosteroid-containing immunosuppressive regimens. Further studies exploring these associations and testing interventions aimed at these modifiable risk factors among SOTRs are needed.
Keywords: extended-spectrum beta-lactamase, Enterobacteriaceae, transplant, bloodstream infection
In this multicenter case-control study, we identified risk factors for extended-spectrum B-lactamase–producing Enterobacterales bloodstream infections among solid-organ transplant recipients. Several novel risk factors were identified including exposure to trimethoprim-sulfamethoxazole and corticosteroid-containing immunosuppressive regimens.
Infections due to multidrug-resistant organisms (MDROs) have become ubiquitous among solid organ transplant recipients (SOTRs). In particular, there has been a rapid and sustained emergence of multidrug-resistant gram-negative (MDR-GN) infections. The majority of such MDR-GN infections are due to extended-spectrum β-lactamase (ESBL)–producing Enterobacterales (EB) [1]. Although rates vary by region, reports suggest that approximately 10% of all bloodstream infections (BSIs), and 40% of all EB BSIs, among SOTRs are due to ESBL-producing EB [1, 2].
Solid organ transplantation (SOT) has been previously recognized as an independent risk factor for ESBL-EB infection, with rates up to five times higher than in the general population [3]. This has been explained by the extensive exposure of SOTRs to the typical MDRO risk factors, such as repeated hospitalizations, intensive care unit stays, and antibiotic exposures [4]. However, SOT status remains an independent risk factor for ESBL-EB infection even after accounting for such exposures [3], suggesting there may be unique characteristics of SOTRs that place them at additional risk for MDRO infection. As such, it is crucial that risk factors for MDROs, and particularly ESBL-EB infections, be elucidated specifically among SOTRs.
To our knowledge, there are only three prior studies that have evaluated risk factors for ESBL-EB infections among SOTRs: two were limited to kidney transplant recipients and one was limited to liver transplant recipients [5–7]. In addition to focusing on a single organ type, these prior studies were also limited to single transplant centers with small sample sizes and focused primarily on urinary tract infections or pneumonia. As a result, we sought to perform a large multicenter study to determine risk factors for ESBL-EB BSIs among all SOT types.
METHODS
Study Design and Setting
A multicenter case-control study was performed at three tertiary care transplant centers in the United States: the Hospital of the University of Pennsylvania (HUP) (776 beds), The Johns Hopkins Hospital (JHH) (1154 beds), and the University of Maryland Medical Center (UMMC) (767 beds).
Study Population
The initial source population included all SOTRs with an EB BSI identified at HUP or UMMC between 1 January 2007 and 30 June 2018, and at JHH between 1 January 2005 and 31 December 2015. Case patients were those with an ESBL-producing EB BSI. ESBL production was determined by either confirmatory testing using the double disk method with both cefotaxime and ceftazidime [8], the ESBL ETEST (bioMérieux, Durham, NC), or a ceftriaxone minimum inhibitory concentration (MIC) of 8 µg/mL or greater. This MIC cut point has been previously shown to have a positive predictive value of 100% and a negative predictive value of 99.5% for ESBL-production [9] and is more stringent than including all ceftriaxone-nonsusceptible isolates in this group. Control patients were those with a non–ESBL-EB BSI, as defined by negative confirmatory testing or a ceftriaxone MIC less than 8 µg/mL. This control population was selected in order to focus on the risk factors for ESBL production rather than risk factors for EB BSI. Since the microbiology laboratories at each center process both inpatient and outpatient cultures, the cohort included any SOTR with an EB BSI, regardless of the location from which the culture was drawn.
If the same SOTR developed multiple episodes of EB BSI during the study period, each distinct episode was considered for inclusion. At least 14 days had to elapse between consecutive positive blood cultures, with at least one intercedent negative blood culture, in order to be considered a separate episode. There were no other exclusion criteria applied. The study was approved by the institutional review board at each of the participating transplant centers.
Data Collection
Data on SOTRs were abstracted from the electronic medical records at each study site by a combination of electronic data extraction, with validation of key fields, and manual chart review. Information was collected on demographics, comorbidities, medications, and details of the EB BSI episode (see Supplementary Information A1 and A2 for a complete list of data collected).
Susceptibility Testing of Enterobacterales Isolates
All EB isolates identified from study subjects were tested as part of routine care for susceptibility to antibiotics at each of the centers’ clinical microbiology laboratories. At HUP, the semiautomated Vitek 2 identification and susceptibility system (bioMérieux, Durham, NC) was utilized; at JHH, the BD Phoenix Automated System (BD Diagnostics, Sparks, MD) was used; and at UMMC, disk diffusion was used prior to 2010 and the Vitek 2 (bioMérieux, Durham, NC) was used after 2010.
Statistical Analysis
Continuous variables were compared using the Student’s t test or Wilcoxon rank-sum test, and categorical variables were compared using the χ 2 or Fisher’s exact test. For the adjusted analyses, mixed-effects multivariable logistic regression was performed with a random effect for study site, in order to adjust for the relatedness of outcomes at each center. Bivariable mixed-effects logistic regression, using a complete case analysis approach, was used to examine the relationship between each potential risk factor and ESBL-EB BSI. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to evaluate the strength of any associations. Variables from bivariable analyses with P values less than .20 were considered for inclusion in the final multivariable model. Manual forward selection was performed to build the multivariable model. Variables were retained in the final model if they were significantly associated with the outcome (P < .05).
Subgroup and sensitivity analyses were then performed evaluating (1) risk factors among all organ types after adjusting for year of BSI; (2) risk factors among all organ types after adjusting for time since SOT (measured as the number of days between transplant and the first positive blood culture for EB); (3) risk factors stratified by organ transplant type; (4) risk factors among all organ types in the HUP subpopulation, where trimethoprim-sulfamethoxazole (TMP-SMX) exposure was limited to prophylactic dosing (defined as 560 g of TMP per week or less); and (5) risk factors when only the first EB BSI for each SOTR was included. The same statistical approach was used for these subgroups and sensitivity analyses as is described for the primary analysis. All analyses were performed using STATA version 15.0 (StataCorp, College Station, TX).
RESULTS
Study Population
A total of 988 EB BSI episodes occurred among 897 SOTRs, of which 395 (40%) were due to ESBL-EB (case contributions per year are shown in Supplementary Figure 1). The median age in the cohort was 57 years (interquartile range [IQR], 48–64), and 421 (42%) were women (Table 1). In total, 585 (59%) received a kidney transplant, 310 (31%) received a liver transplant, 77 (8%) received a heart transplant, 65 (7%) received a lung transplant, and 27 (3%) received a pancreas transplant (not mutually exclusive). In the 6 months prior to the EB BSI, 717 (72%) had been hospitalized and 764 (77%) had received at least one antibiotic course.
Table 1.
Baseline Characteristics of Cases and Controls
| Baseline Characteristics | ESBL-EB BSIs (n = 395) | Non–ESBL-EB BSIs (n = 593) | P |
|---|---|---|---|
| Study site,a n (%) | |||
| Hospital of the University of Pennsylvania | 78 (25) | 231 (75) | <.01 |
| University of Maryland Medical Center | 96 (30) | 225 (70) | |
| The Johns Hopkins Hospital | 221 (62) | 137 (38) | |
| Demographics | |||
| Age (median, IQR), years | 57 (48–65) | 57 (47–64) | .42 |
| Female gender, n (%) | 166 (42) | 252 (43) | .87 |
| Hispanic ethnicity, n (%) | 10 (3) | 13 (2) | .73 |
| Black or African American race, n (%) | 62 (16) | 210 (35) | <.01 |
| White or Caucasian race, n (%) | 108 (27) | 221 (37) | <.01 |
| Comorbidities,b n (%) | |||
| Diabetes mellitus | 157 (40) | 313 (53) | <.01 |
| CKD | 33 (8) | 71 (12) | .07 |
| CKD requiring dialysis | 105 (27) | 236 (40) | <.01 |
| Cirrhosis | 64 (16) | 117 (20) | .16 |
| Prior chemotherapy | 26 (7) | 45 (8) | .55 |
| Cardiomyopathy | 36 (9) | 61 (10) | .54 |
| Structural lung disease | 36 (9) | 41 (7) | .02 |
| HIV | 15 (3.8) | 43 (7) | .02 |
| Organ transplant type, n (%) | |||
| Kidney | 210 (53) | 370 (62) | <.01 |
| Liver | 142 (36) | 162 (27) | <.01 |
| Heart | 28 (7) | 49 (8) | .50 |
| Lung | 35 (9) | 29 (5) | .01 |
| Pancreas | 10 (3) | 17 (3) | .75 |
| Transplant characteristics | |||
| Time from transplant (median, IQR), days | 335 (83–2530) | 886 (160–2967) | <.01 |
| Multiple organ transplants prior to EB BSI, n (%) | 67 (17) | 94 (16) | .63 |
| Induction immunosuppression given at transplant, n (%) | 149 (41) | 260 (53) | <.01 |
| Induction with ATG, n (%) | 76 (19) | 110 (19) | .79 |
| Induction with basiliximab, n (%) | 15 (4) | 11 (2) | .06 |
| Recipient CMV seropositive, n (%) | 128 (42) | 220 (54) | <.01 |
| Donor CMV seropositive, n (%) | 98 (37) | 180 (51) | <.01 |
| Living donor, n (%) | 98 (25) | 109 (18) | .02 |
| Complications of transplant, n (%) | |||
| Primary graft dysfunction | 65 (18) | 75 (15) | .20 |
| Reoperation within four weeks of transplant | 124 (36) | 156 (33) | .31 |
| Allograft failure (prior to EB BSI) | 122 (33) | 128 (22) | <.01 |
| CMV infectionc | 96 (26) | 108 (19) | .01 |
| Acute rejection and immunosuppression,d n (%) | |||
| Any rejection | 40 (10) | 58 (10) | .81 |
| Rejection treated with ATG | 7 (2) | 19 (3) | .17 |
| Rejection treated with corticosteroids | 25 (6) | 36 (6) | .87 |
| Rituximab exposuree | 8 (2) | 15 (3) | .59 |
| Chronic immunosuppression at time of BSI, n (%) | |||
| Corticosteroids | 313 (79) | 353 (60) | <.01 |
| Tacrolimus | 308 (78) | 400 (67) | <.01 |
| Mycophenolate mofetil | 223 (56) | 342 (58) | .71 |
| Sirolimus | 33 (8) | 46 (8) | .74 |
| Azathioprine | 13 (3) | 28 (5) | .27 |
| Cyclosporine | 24 (6) | 29 (5) | .42 |
| Hospital and antibiotic exposures,f n (%) | |||
| Hospitalization | 301 (76) | 411 (69) | .02 |
| Any antibiotic | 346 (88) | 414 (70) | <.01 |
| Aminoglycoside | 81 (21) | 34 (6) | <.01 |
| Amoxicillin or ampicillin | 31 (8) | 43 (7) | .73 |
| Amoxicillin-clavulanate or ampicillin-sulbactam | 53 (13) | 80 (13) | .97 |
| Atovaquone | 51 (13) | 45 (8) | .01 |
| Aztreonam | 18 (5) | 14 (2) | .06 |
| Carbapenem | 152 (38) | 89 (15) | <.01 |
| First-generation cephalosporin | 100 (25) | 126 (21) | .14 |
| Third-generation cephalosporin | 117 (30) | 74 (12) | <.01 |
| Cefepime | 141 (36) | 120 (20) | <.01 |
| Clindamycin | 38 (10) | 45 (8) | .26 |
| Colistin or polymyxin B | 28 (7) | 8 (1) | <.01 |
| Daptomycin | 36 (9) | 18 (3) | <.01 |
| Echinocandin | 23 (6) | 21 (4) | .09 |
| Fluoroquinolone | 152 (38) | 159 (27) | <.01 |
| Fluconazole | 39 (10) | 58 (10) | .96 |
| Linezolid | 102 (26) | 72 (12) | <.01 |
| Macrolide | 53 (13) | 44 (7) | <.01 |
| Metronidazole | 126 (32) | 113 (19) | <.01 |
| Piperacillin-tazobactam | 203 (51) | 195 (33) | <.01 |
| TMP-SMX | 181 (46) | 169 (29) | <.01 |
| Vancomycin (IV) | 235 (59) | 230 (39) | <.01 |
| Voriconazole or posaconazole | 16 (4) | 14 (2) | .13 |
| Microbiological results,g n (%) | |||
| Prior EB (any organism, any anatomical site) | 234 (60) | 265 (49) | <.01 |
| Prior ESBL-EB | 147 (37) | 28 (5) | <.01 |
| Prior carbapenem-resistant EB | 58 (15) | 7 (1) | <.01 |
| Prior EB on genitourinary tract culture | 128 (32) | 195 (33) | .88 |
| Prior EB on blood culture | 70 (18) | 63 (11) | <.01 |
| Prior EB on respiratory culture | 71 (18) | 46 (8) | <.01 |
| Prior EB on skin or soft tissue culture | 21 (5) | 8 (1) | <.01 |
| Prior EB on intra-abdominal culture | 31 (8) | 9 (2) | <.01 |
This table includes data from 988 EB BSIs among 897 solid-organ transplant recipients, identified at the Hospital of the University of Pennsylvania and the University of Maryland Medical Center between 1 January 2007 and 30 June 2018 and at The Johns Hopkins Hospital between 1 January 2005 and 31 December 2015. Data are presented as numbers (percentages) except where noted. The percentages represent the proportion of cases or controls with the specified baseline characteristic, except where noted. Only those variables with a P value < .20 are included in this table, as well as those of notable biological importance.
Abbreviations: ATG, antithymocyte globulin; BSI, bloodstream infection; CKD, chronic kidney disease; CMV, cytomegalovirus; EB, Enterobacterales; ESBL, extended-spectrum β-lactamase; HIV, human immunodeficiency virus; IQR, interquartile range; IV, intravenous; TMP-SMX, trimethoprim-sulfamethoxazole.
aPercentages for study sites represent the proportion of solid-organ transplant recipients at that site with an ESBL-EB BSI or a non–ESBL-EB BSI (rather than the proportion of cases or controls enrolled from that site).
bComorbidities assessed at time of EB BSI.
cCMV infection assessed through 6 months prior to EB BSI.
dAcute rejection assessed through 3 months prior to EB BSI.
eRituximab exposure assessed through 1 year prior to EB BSI.
fHospital and antibiotic exposures assessed through 6 months prior to EB BSI.
gMicrobiological results assessed through 1 year prior to EB BSI.
Overview of Enterobacterales Bloodstream Infection Microbiology
The most common EB organisms isolated on blood culture were Klebsiella pneumoniae (379, 38%) and Escherichia coli (368, 37%) (Table 2). In the ESBL group, Klebsiella species were most common (180, 46%), while E. coli was most common (244, 41%) in the non-ESBL group. There were 259 (26%) EB isolates demonstrating fluoroquinolone resistance, 174 (17%) with piperacillin-tazobactam resistance, 77 (8%) with carbapenem resistance (all of which were included in the ESBL group), and 513 (52%) with TMP-SMX resistance.
Table 2.
Overview of Enterobacterales Bloodstream Infection Microbiology
| Microbiologic Characteristics | All EB BSIs (n = 988) | ESBL-EB BSIs (n = 395) | Non–ESBL-EB BSIs (n = 593) | P |
|---|---|---|---|---|
| Organism(s) on blood culture,a n (%) | ||||
| Escherichia coli | 366 (37) | 122 (31) | 244 (41) | <.01 |
| Klebsiella spp. | 412 (42) | 180 (46) | 232 (39) | .04 |
| Enterobacter spp. | 131 (13) | 67 (17) | 64 (11) | .01 |
| Serratia spp. | 36 (4) | 11 (3) | 25 (4) | .24 |
| Proteus spp. | 24 (2) | 7 (2) | 17 (3) | .27 |
| Citrobacter spp. | 16 (2) | 6 (2) | 10 (2) | .84 |
| Polymicrobial BSI per index blood culture, n (%) | 36 (4) | 8 (2) | 28 (5) | .03 |
| Antimicrobial resistance | ||||
| Piperacillin-tazobactam resistance | 170 (17) | 143 (36) | 27 (5) | <.01 |
| Fluoroquinolone resistance | 256 (26) | 168 (43) | 88 (15) | <.01 |
| TMP-SMX resistance | 513 (52) | 258 (44) | 255 (65) | <.01 |
| Carbapenem resistance | 77 (8) | 77 (8) | 0 (0) | <.01 |
This table includes data from 988 EB BSIs among 897 solid-organ transplant recipients, identified at the Hospital of the University of Pennsylvania and the University of Maryland Medical Center between 1 January 2007 and 30 June 2018 and at The Johns Hopkins Hospital between 1 January 2005 and 31 December 2015. Data are presented as numbers (percentages) except where noted.
Abbreviations: BSI, bloodstream infection; EB, Enterobacterales; ESBL, extended-spectrum β-lactamase; TMP-SMX, trimethoprim-sulfamethoxazole.
aThese events are not mutually exclusive; each subject may have had multiple BSIs and/or multiple organisms on each blood culture.
Risk Factors for Extended-spectrum β-lactamase–Producing Enterobacterales Bloodstream Infections Among All Solid Organ Transplant Recipients
On multivariable analysis, six independent risk factors were associated with an increased odds of ESBL-EB BSI compared with non–ESBL EB BSI (Table 3): isolation of an ESBL-EB organism from any anatomical site in the prior 1 year (adjusted OR [aOR], 12.75; 95% CI, 3.23–50.33; P < .001), a corticosteroid-containing chronic immunosuppression regimen (aOR, 1.30; 95% CI, 1.03–1.65; P = .030), an acute rejection episode treated with corticosteroids in the prior 3 months (aOR, 1.18; 95% CI, 1.16–1.19; P < .001), exposure to third-generation cephalosporins in the prior 6 months (aOR, 1.95; 95% CI, 1.48–2.57; P < .001), exposure to echinocandins in the prior 6 months (aOR, 1.61; 95% CI, 1.08–2.40; P = .020), and exposure to TMP-SMX in the prior 6 months (aOR, 1.35; 95% CI, 1.10–1.64; P = .003). Kidney transplant recipients had a significantly lower odds of ESBL-EB BSI compared with the other organ transplant types (aOR, .71; 95% CI, .52–.98; P = .036).
Table 3.
Mixed-Effects Multivariable Logistic Regression of Risk Factors for Extended-Spectrum β-lactamase–Producing Enterobacterales Bloodstream Infection Among All Solid-Organ Transplant Recipient Types
| Risk Factor | aOR | 95% CI | P |
|---|---|---|---|
| Prior ESBL-EB on culturea | 12.75 | 3.23–50.33 | <.001 |
| TMP-SMX exposureb | 1.35 | 1.10–1.64 | .003 |
| Echinocandin exposureb | 1.61 | 1.08–2.40 | .020 |
| Third-generation cephalosporin exposureb | 1.95 | 1.48–2.57 | <.001 |
| Corticosteroid-containing chronic immunosuppression regimenc | 1.30 | 1.03–1.65 | .030 |
| Acute rejection treated with corticosteroidsd | 1.18 | 1.16–1.19 | <.001 |
| Kidney transplant recipient | 0.71 | .52–.98 | .036 |
This table includes data from 988 EB BSIs among 897 solid organ transplant recipients, identified at the Hospital of the University of Pennsylvania and the University of Maryland Medical Center between 1 January 2007 and 30 June 2018 and at The Johns Hopkins Hospital between 1 January 2005 and 31 December 2015.
Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; EB, Enterobacterales; ESBL, extended-spectrum β-lactamase; TMP-SMX, trimethoprim-sulfamethoxazole.
aESBL-EB organism isolated on microbiological culture from any anatomical site in the 1 year prior to the EB BSI.
bExposure within the 6 months prior to the EB BSI.
cImmunosuppression at the time of the EB BSI.
dRejection within the 3 months prior to the EB BSI.
Given the prolonged period over which this cohort was accrued (2005–2018), we secondarily adjusted for the year of BSI diagnosis, in order to evaluate whether any of the identified risk factors were markers of temporal trends. In doing so, we found that the primary risk factors remained unchanged, but year of BSI was also significantly associated with the odds of ESBL-EB BSI (aOR, 1.08; 95% CI, 1.03–1.14; P = .002 per year) (Supplementary Table 1). We also evaluated time since SOT as a potential confounder; we similarly found that it did not appreciably change the identified risk factors, although it was significantly associated with the outcome as well (aOR, .99; 95% CI, .99–.99; P = .04 per day) (Supplementary Table 2).
Risk Factors for Extended-Spectrum β-lactamase–Producing Enterobacterales Bloodstream Infection Among Each Organ Transplant Type
After stratifying by organ transplant type, several independent risk factors were identified that were unique for each organ type (Table 4): (1) for liver transplant recipients, exposure to voriconazole or posaconazole in the prior 6 months was a significant risk factor; (2) for kidney transplant recipients, obstructive uropathy, increasing age, exposure to daptomycin in the prior 6 months, and skin or soft tissue infection with an EB organism in the prior year were risk factors, while acute rejection treated with total plasma exchange in the prior 3 months was protective; (3) for heart transplant recipients, African American or Black race was a risk factor, while induction immunosuppression with basiliximab was protective; (4) for lung transplant recipients, idiopathic pulmonary fibrosis, need for reoperation within 4 weeks of transplant, exposure to cefepime in the prior 6 months, and exposure to an aminoglycoside in the prior 6 months were also risk factors. Due to the small number of pancreas transplant recipients, only bivariable analyses were performed (Supplementary Table 3).
Table 4.
Mixed-Effects Multivariable Logistic Regression of Risk Factors for Extended-Spectrum β-lactamase–Producing Enterobacterales Bloodstream Infection Stratified by Organ Transplant Type
| Risk Factor | aOR | 95% CI | P |
|---|---|---|---|
| Liver transplant recipients | |||
| Prior ESBL-EB on culturea | 10.50 | 3.00–36.93 | <.001 |
| TMP-SMX exposureb | 1.75 | 1.19–2.58 | .005 |
| Third-generation cephalosporin exposureb | 1.86 | 1.39–2.50 | <.001 |
| Voriconazole or posaconazole exposureb | 2.69 | 2.63–2.75 | <.001 |
| Corticosteroid-containing chronic immunosuppression regimenc | 1.62 | 1.34–1.95 | <.001 |
| Kidney transplant recipients | |||
| Prior skin/soft-tissue infection due to EBd | 5.83 | 2.26–15.02 | <.001 |
| Third-generation cephalosporin exposureb | 1.58 | 1.04–2.41 | .034 |
| Daptomycin exposureb | 3.24 | 1.98–5.28 | <.001 |
| Prior ESBL-EB on culturea | 9.92 | 1.53–64.21 | .016 |
| Obstructive uropathy as indication for transplant | 3.03 | 1.20–7.66 | .019 |
| Age (per year) | 1.09 | 1.00–1.02 | .012 |
| Acute rejection treated with therapeutic plasma exchangee | 0.50 | .41–.63 | <.001 |
| Heart transplant recipients | |||
| Prior ESBL-EB on culturea | 29.10 | 11.50–73.40 | <.001 |
| Third-generation cephalosporin exposureb | 5.26 | 1.57–17.63 | .007 |
| African American or Black race | 1.80 | 1.55–2.08 | <.001 |
| TMP-SMX exposureb | 1.80 | 1.23–2.63 | .002 |
| Basiliximab induction | 0.22 | .19–.26 | <.001 |
| Corticosteroid-containing chronic immunosuppression regimenb | 2.63 | 1.03–6.74 | .044 |
| Lung transplant recipients | |||
| Prior ESBL-EB on culturea | 23.10 | 1.53–346.75 | .023 |
| Idiopathic pulmonary fibrosis as indication for transplant | 7.13 | 1.99–25.55 | .003 |
| Cefepime exposureb | 8.52 | 2.95–24.66 | <.001 |
| Aminoglycoside exposureb | 2.22 | 1.01–4.89 | .047 |
| Need for reoperation within 4 weeks posttransplant | 1.64 | 1.37–1.97 | <.001 |
This table includes data from 988 EB BSIs among 897 solid-organ transplant recipients, identified at the Hospital of the University of Pennsylvania and the University of Maryland Medical Center between 1 January 2007 and 30 June 2018 and at The Johns Hopkins Hospital between 1 January 2005 and 31 December 2015.
Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; EB, Enterobacterales; ESBL, extended-spectrum β-lactamase; TMP-SMX, trimethoprim-sulfamethoxazole.
aESBL-EB organism isolated on microbiological culture from any anatomical site in the 1 year prior to the EB BSI.
bExposure within the 6 months prior to the EB BSI.
cImmunosuppression at the time of the EB BSI.
dInfection within 1 year prior to the EB BSI.
eRejection within 3 months prior to the EB BSI.
Association Between Trimethoprim-sulfamethoxazole Exposure and Extended-Spectrum β-lactamase–Producing Enterobacterales Bloodstream Infection
Given the critical role of TMP-SMX in SOT prophylaxis, we performed two secondary analyses to further explore the association between TMP-SMX exposure and risk of ESBL-EB BSI. First, we examined whether exposure to TMP-SMX might be a marker for time from transplant, since SOTRs receive the majority of their TMP-SMX exposure in the first-year after transplantation. After adding time since transplant to the final multivariable model, the risk factors were not substantively altered, including the association between TMP-SMX exposure and ESBL-EB BSI (aOR, 1.30; 95% CI, 1.04–1.63; P = .021) (Supplementary Table 2).
Second, we sought to determine whether the association between TMP-SMX exposure and ESBL-EB BSI was being driven by prophylactic or treatment dosing of TMP-SMX. We restricted this analysis to only those subjects enrolled at HUP, as this is the only site at which detailed dosing data were available for the complete cohort. In this subgroup, there were 309 episodes of EB BSI, of which 78 (25%) were due to ESBL-EB. Among the HUP SOTRs, 133 (43%) had received any dose of TMP-SMX in the prior 6 months, and 123 (92% of those who had received TMP-SMX) had been on prophylactic dosing. In the multivariable analysis, we found that any TMP-SMX exposure was only marginally associated with an increased odds of ESBL-EB BSI (aOR, 1.98; 95% CI, .97–4.01; P = .062), but TMP-SMX prophylaxis was associated with a significantly increased odds of ESBL-EB BSI (aOR, 2.10; 95% CI, 1.03–4.28; P = .042) (Supplementary Table 4).
Risk Factors for Extended-Spectrum β-lactamase–Producing Enterobacterales Bloodstream Infection in the Incident Bloodstream Infection Cohort
In this sensitivity analysis, there were 897 unique SOTRs with an EB BSI, of which 356 (40%) were due to ESBL-EB (Supplementary Information A2). The risk factors identified were minimally altered in this analysis (Supplementary Tables 5 and 6), except for the following: (1) for all SOTRs, an acute rejection episode treated with corticosteroids was no longer a significant risk factor; (2) for kidney transplant recipients, prior daptomycin exposure was no longer a significant risk factor; and (3) for heart transplant recipients, prior TMP-SMX exposure was no longer a significant risk factor.
DISCUSSION
In this multicenter case-control study, we identified several important risk factors for ESBL-EB BSI among SOTRs, including prior colonization or infection with an ESBL-EB organism, corticosteroid-containing immunosuppression, and prior exposure to antimicrobials, most notably TMP-SMX. To our knowledge, this is the first determination of risk factors for ESBL-EB BSI among all organ transplant types and represents the largest cohort of SOTRs with ESBL-EB BSI documented to date.
Perhaps the most notable risk factor identified in our study is prior exposure to TMP-SMX; we observed a 35% increase in the odds of ESBL-EB BSI with any TMP-SMX exposure in the prior 6 months. Interestingly, in subgroup analyses, we found that (1) TMP-SMX prophylaxis was driving this association and (2) it was particularly problematic among liver transplant recipients.
Three prior studies of the general population have identified exposure to TMP-SMX as a risk factor for extended-spectrum cephalosporin-resistant EB infections [10–12], although other studies have not [13–16]. Trimethoprim-sulfamethoxazole resistance has been found to frequently co-localize with resistance to extended-spectrum cephalosporins, and prior studies have found both the sul and dfr genes that encode for TMP-SMX resistance frequently circulate on plasmids also harboring ESBL genes [11, 17, 18]. It is thus mechanistically plausible that exposure to TMP-SMX could select for plasmids that also confer resistance to extended-spectrum cephalosporins. The association between TMP-SMX and ESBL-EB BSI is likely more evident in this cohort of SOTRs, in whom a large proportion were exposed to TMP-SMX, and in whom the chief exposure was both long term and low dose. It is not immediately evident why this was most notable among liver transplant recipients, although it may possibly relate to changes in drug metabolism following liver transplantation [19] or their increased susceptibility to infections with enteric organisms [20]. This finding has significant implications for the transplant community, since TMP-SMX plays a critical role in posttransplant prophylaxis against Pneumocystis and Toxoplasma, and yet there is an urgent need to curtail the rapid rise of MDRO infections among SOTRs.
In addition to TMP-SMX exposure, we identified several other risk factors for ESBL-EB BSI that relate to immunosuppression regimens and are thus uniquely important to the SOTR population. First, we found that corticosteroid-containing immunosuppression regimens, as well as rejection episodes treated with corticosteroids, were associated with an increased odds of ESBL-EB BSI. This is likely related to the impaired cellular immunity and hyperglycemia associated with corticosteroid use [21–23]. Second, we found that, among kidney transplant recipients, acute rejection treated with therapeutic plasma exchange was protective; most plasma exchange used for rejection is given in conjunction with intravenous immunoglobulin (IVIg), which may confer protective antibody profiles and reduce the risk of bacterial infection [24, 25]. Third, we found that, among heart transplant recipients, basiliximab induction was protective. This is consistent with prior literature that suggests basiliximab has a more modest infection signal than other induction agents, particularly antithymocyte globulin [26, 27]. Using basiliximab in place of alternative induction immunosuppression and in an effort to reduce corticosteroid exposure may result in fewer bacterial complications.
In addition to these findings, antifungal exposure was also identified as a potential risk factor. This may be explained by the recent descriptions of mycobiome-microbiome interactions in the gut: it has been shown that mice exposed to antifungals develop more severe colitis syndromes than those that are not [28], and in patients with Crohn disease, the gut microbiome has been found to be enriched with non–albicans Candida, which is positively correlated with levels of Serratia and E. coli species [29]. It is thus plausible that antifungal exposures result in microbiome disruption that increases the risk of EB infection, including ESBL-EB.
Finally, we found that in every organ transplant type prior infection or colonization with an ESBL-EB was a significant risk factor for ESBL-EB BSI. This is consistent with several prior studies that have shown that colonization with MDR-GNs is associated with high rates of subsequent infection [30–32]; this finding raises the question of whether SOTRs should be proactively screened for ESBL-EB prior to or following transplantation in order to guide empiric therapy in case of EB BSI.
There are several potential limitations of this study. First, because the study was performed retrospectively at three different centers and spanned more than 10 years, the testing methods for ESBL detection evolved over time. Thus, misclassification bias is possible, although we expect it would be nondifferential and bias the results toward the null. Further, there is a validated phenotypic approach for identifying ESBL producers among E. coli, Klebsiella, and Proteus species, but we do not have such an approach for the other EB organisms. Thus, it is possible that the non–E. coli/Klebsiella/Proteus organisms included in this study were harboring non-ESBL resistance mechanisms, such as AmpC β-lactamases. Reassuringly, these organisms accounted for the minority of the EB BSIs included in this study. Next, we chose to employ a control group with susceptible EB BSI, rather than including all SOTRs who were hospitalized, which can result in an overestimation of the risk of antibiotics [33]. Finally, we saw notably different rates of ESBL-EB among the three study centers; there are several potential explanations for this, including differences in patient population (ie, demographics, organ transplant types, severity of underlying disease at the time of transplantation) or SOTR management (such as accepting organ donors with ESBL-EB colonization or higher rates of antimicrobial use posttransplant); we attempted to account for these differences through our mixed-effects modeling approach and through adjustment for all likely confounders, but it remains possible that additional unmeasured confounders were present.
In summary, we identified several novel risk factors for ESBL-EB BSIs among SOTRs, many of which are unique to the transplant population, including prophylactic antibiotics and specific immunosuppressive regimens. Given the exceptionally high rate of MDROs and ESBL-EB infections in the transplant population, there is a pressing need for further studies evaluating interventions targeting the modifiable risk factors.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Financial support. This work was supported by the Antibacterial Resistance Leadership Group (grant number 5 UM 1AI104681-05 with a subaward fellowship grant to J. A. A.); the Transplant Foundation’s Innovative Research Grant Program, an affiliate of the Gift of Life Donor Program (Donation and Transplantation grant to J. A. A.); the National Institutes of Health (grant numbers K24-AI080942 [to E. L.] and K01-AI137317 [to J. A. A.]); and by a Centers for Disease Control and Prevention Cooperative Agreement FOA#CK16-004-Epicenters for the Prevention of Healthcare Associated Infections (to E. L.).
Potential conflicts of interest. E. A. B. receives research support from Shire and Merck, is a member of Data and Safety Monitoring Boards (DSMBs) for Bristol-Myers Squibb and GlaxoSmithKline, is a member of the Scientific Advisory Committee for Merck, and is a consultant for Shionogi. J. H. H. was affiliated with the University of Pennsylvania during the conduct of this research and is now employed by GlaxoSmithKline. E. L. is a member of a DSMB for Merck and a member of a scientific advisory committee for Paratek and Shionogi and reports travel reimbursement from the Infectious Diseases Society of America, Society for Healthcare Epidemiology of America, the National Institutes of Health, Centers for Disease Control and Prevention, and International Conference of Prevention and Infection Control. K. A. serves on a scientific advisory board for Becton Dickinson. None of these conflicts are relevant to this article. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1. Oriol I, Sabé N, Simonetti AF, et al. Changing trends in the aetiology, treatment and outcomes of bloodstream infection occurring in the first year after solid organ transplantation: a single-centre prospective cohort study. Transpl Int 2017; 30:903–13. [DOI] [PubMed] [Google Scholar]
- 2. Linares L, García-Goez JF, Cervera C, et al. Early bacteremia after solid organ transplantation. Transplant Proc 2009; 41:2262–4. [DOI] [PubMed] [Google Scholar]
- 3. Rodríguez-Baño J, Picón E, Gijón P, et al. Risk factors and prognosis of nosocomial bloodstream infections caused by extended-spectrum-beta-lactamase-producing Escherichia coli. J Clin Microbiol 2010; 48:1726–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Pouch SM, Patel G; AST Infectious Diseases Community of Practice . Multidrug-resistant gram-negative bacterial infections in solid organ transplant recipients-guidelines from the American Society of Transplantation Infectious Diseases community of practice. Clin Transplant 2019; 33:e13594. [DOI] [PubMed] [Google Scholar]
- 5. Linares L, Cervera C, Cofán F, et al. Risk factors for infection with extended-spectrum and AmpC beta-lactamase-producing gram-negative rods in renal transplantation. Am J Transplant 2008; 8:1000–5. [DOI] [PubMed] [Google Scholar]
- 6. Espinar MJ, Miranda IM, Costa-de-Oliveira S, Rocha R, Rodrigues AG, Pina-Vaz C. Urinary tract infections in kidney transplant patients due to Escherichia coli and Klebsiella pneumoniae-producing extended-spectrum β-lactamases: risk factors and molecular epidemiology. PLoS One 2015; 10:e0134737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Zhong L, Men TY, Li H, et al. Multidrug-resistant gram-negative bacterial infections after liver transplantation—spectrum and risk factors. J Infect 2012; 64:299–310. [DOI] [PubMed] [Google Scholar]
- 8. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing M100-S20. Wayne, PA: Clinical and Laboratory Standards Institute, 2010. [Google Scholar]
- 9. Huang Y, Carroll KC, Cosgrove SE, Tamma PD. Determining the optimal ceftriaxone MIC for triggering extended-spectrum β-lactamase confirmatory testing. J Clin Microbiol 2014; 52:2228–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Anesi JA, Lautenbach E, Nachamkin I, et al. Clinical and molecular characterization of community-onset urinary tract infections due to extended-spectrum cephalosporin-resistant Enterobacteriaceae. Infect Control Hosp Epidemiol 2016; 37:1433–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Wiener J, Quinn JP, Bradford PA, et al. Multiple antibiotic-resistant Klebsiella and Escherichia coli in nursing homes. JAMA 1999; 281:517–23. [DOI] [PubMed] [Google Scholar]
- 12. Nakai H, Hagihara M, Kato H, et al. Prevalence and risk factors of infections caused by extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae. J Infect Chemother 2016; 22:319–26. [DOI] [PubMed] [Google Scholar]
- 13. Colodner R, Rock W, Chazan B, et al. Risk factors for the development of extended-spectrum beta-lactamase-producing bacteria in nonhospitalized patients. Eur J Clin Microbiol Infect Dis 2004; 23:163–7. [DOI] [PubMed] [Google Scholar]
- 14. Calbo E, Romaní V, Xercavins M, et al. Risk factors for community-onset urinary tract infections due to Escherichia coli harbouring extended-spectrum beta-lactamases. J Antimicrob Chemother 2006; 57:780–3. [DOI] [PubMed] [Google Scholar]
- 15. Rodríguez-Baño J, Navarro MD, Romero L, et al. Risk-factors for emerging bloodstream infections caused by extended-spectrum beta-lactamase-producing Escherichia coli. Clin Microbiol Infect 2008; 14:180–3. [DOI] [PubMed] [Google Scholar]
- 16. Lautenbach E, Patel JB, Bilker WB, Edelstein PH, Fishman NO. Extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae: risk factors for infection and impact of resistance on outcomes. Clin Infect Dis 2001; 32:1162–71. [DOI] [PubMed] [Google Scholar]
- 17. Pitout JD, Thomson KS, Hanson ND, Ehrhardt AF, Coudron P, Sanders CC. Plasmid-mediated resistance to expanded-spectrum cephalosporins among Enterobacter aerogenes strains. Antimicrob Agents Chemother 1998; 42:596–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Amador P, Fernandes R, Prudencio C, Duarte I. Prevalence of antibiotic resistance genes in multidrug-resistant Enterobacteriaceae on portuguese livestock manure. Antibiotics (Basel) 2019; 8:23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Liu S, Frye RF, Branch RA, Venkataramanan R, Fung JJ, Burckart GJ. Effect of age and postoperative time on cytochrome p450 enzyme activity following liver transplantation. J Clin Pharmacol 2005; 45:666–73. [DOI] [PubMed] [Google Scholar]
- 20. Singh N, Wagener MM, Obman A, Cacciarelli TV, de Vera ME, Gayowski T. Bacteremias in liver transplant recipients: shift toward gram-negative bacteria as predominant pathogens. Liver Transpl 2004; 10:844–9. [DOI] [PubMed] [Google Scholar]
- 21. Cain DW, Cidlowski JA. Immune regulation by glucocorticoids. Nat Rev Immunol 2017; 17:233–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Tamez-Pérez HE, Quintanilla-Flores DL, Rodríguez-Gutiérrez R, González-González JG, Tamez-Peña AL. Steroid hyperglycemia: prevalence, early detection and therapeutic recommendations: a narrative review. World J Diabetes 2015; 6:1073–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Thaiss CA, Levy M, Grosheva I, et al. Hyperglycemia drives intestinal barrier dysfunction and risk for enteric infection. Science 2018; 359:1376–83. [DOI] [PubMed] [Google Scholar]
- 24. Guptill JT, Juel VC, Massey JM, et al. Effect of therapeutic plasma exchange on immunoglobulins in myasthenia gravis. Autoimmunity 2016; 49:472–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Xie P, Tao M, Peng K, et al. Plasmapheresis therapy in kidney transplant rejection. Blood Purif 2019; 47:73–84. [DOI] [PubMed] [Google Scholar]
- 26. Mattei MF, Redonnet M, Gandjbakhch I, et al. Lower risk of infectious deaths in cardiac transplant patients receiving basiliximab versus anti-thymocyte globulin as induction therapy. J Heart Lung Transplant 2007; 26:693–9. [DOI] [PubMed] [Google Scholar]
- 27. Spada M, Petz W, Bertani A, et al. Randomized trial of basiliximab induction versus steroid therapy in pediatric liver allograft recipients under tacrolimus immunosuppression. Am J Transplant 2006; 6:1913–21. [DOI] [PubMed] [Google Scholar]
- 28. Qiu X, Zhang F, Yang X, et al. Changes in the composition of intestinal fungi and their role in mice with dextran sulfate sodium-induced colitis. Sci Rep 2015; 5:10416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Hoarau G, Mukherjee PK, Gower-Rousseau C, et al. Bacteriome and mycobiome interactions underscore microbial dysbiosis in familial Crohn’s disease. MBio 2016; 7:e01250–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Lübbert C, Becker-Rux D, Rodloff AC, et al. Colonization of liver transplant recipients with KPC-producing Klebsiella pneumoniae is associated with high infection rates and excess mortality: a case-control analysis. Infection 2014; 42:309–16. [DOI] [PubMed] [Google Scholar]
- 31. Bert F, Larroque B, Paugam-Burtz C, et al. Pretransplant fecal carriage of extended-spectrum beta-lactamase-producing Enterobacteriaceae and infection after liver transplant, France. Emerg Infect Dis 2012; 18:908–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Giannella M, Bartoletti M, Morelli MC, et al. Risk factors for infection with carbapenem-resistant Klebsiella pneumoniae after liver transplantation: the importance of pre- and posttransplant colonization. Am J Transplant 2015; 15:1708–15. [DOI] [PubMed] [Google Scholar]
- 33. Harris AD, Karchmer TB, Carmeli Y, Samore MH. Methodological principles of case-control studies that analyzed risk factors for antibiotic resistance: a systematic review. Clin Infect Dis 2001; 32:1055–61. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.