Impact of Specific Antimicrobials and MIC Values on the Outcome of Bloodstream Infections Due to ESBL- or Carbapenemase-producing Enterobacterales in Solid Organ Transplantation: an Observational Multinational Study.

METHODS

Study design: multicentre, international retrospective cohort study.

Sites: multiple expert investigators from different countries will be invited. Criteria for participation will include accessibility to a database with the required data or ability to retrospectively collect the data in a timely manner.

Procedure Centers will receive by email a personal key for access to an electronic Case Report File (CRF), on the official INCREMENT-SOT website at http://www.increment-sotproject.org/

The participant centres will be requested to include:

• Previously published cases: all these cases should be included if possible. The fact that the case was previously published should be specified in the database.
• Additionally, participants will be requested to include consecutive episodes detected by reviewing their databases (clinical, infection control or microbiological records) from January 2004 to October 2016, according to the following criteria:
• For ESBL producers: Consecutive cases for which the enzyme was characterised at least to group level by PCR (it is, CTX-M, SHV, TEM) should be prioritised despite the date of diagnosis. If PCR-characterisation was not performed, cases in which ESBL-production was identified using a standard phenotypic method may be included.
• For carbapenemase-producers: cases in which the carbapenemase gene was detected by a molecular method or the isolate showed an identical phenotype to previously characterized carbapenemase-producers detected at the same site.

Overall, to avoid selection biases, consecutive cases according to previous criteria should be included.

Variables

A common, online database has been designed. Access to this database will be restricted to authorized researchers by use of an individual user name and password.

Main outcome variables: Cure rate at day 14 and all-cause mortality until day 30. For safety issues, frequency and types of renal and liver toxicity.

Secondary outcome variables: Mortality at 72 hours, 7, 14 and 90 days, clinical improvement at 72 hours, clinical cure at day 28. Renal and liver toxicity and other serious adverse events.

Explanatory variables:

• Demographics
• Use of antibiotics within 30 days before the BSI.
• Severity of chronic underlying conditions: McCabe and Charlson index
• Acute severity of underlying disease: Pitt score during the day before BSI.
• Type of acquisition
• Source of BSI
• Severity of SIRS at presentation
• Microorganism, MICs, clinical category (Susceptible, Intermediate, Resistant), guideline (CLSI or EUCAST), mechanisms of resistance if studied.
• Empirical therapy
• Definitive therapy
• Bacteremia Source Control: drainage within the first four days, removal of prosthetic material, removal of infected catheter, etc.
• Renal function when bacteremia occurred (Creatinine clearance).
• Selective Intestinal Decontamination (SID) during the previous 12 months and type of SID (drugs used).
• Variables related with transplantation:
• Type of SOT
• Day after transplantation
• Acute rejection within 30 days before the BSI
• Leukopenia (number of lymphocytes, lymphocyte subpopulations, immunoglobulins and complement).
• Basal immunosuppression
• Induction of immunosuppression
• Prophylaxis with TMP/SMX within 30 days before the BSI
• Surgical reoperation.
• Urinary catheter.
• Urinary stenosis (renal)
• Biliary stenosis (liver)
• Traqueal stenosis (lung)
• Post-transplant dialysis within 30 days before the BSI
• CMV replication within 30 days before the BSI
• CMV disease within 30 days before the BSI
• Mechanical ventilation
• Other underlining condition

Definitions

• Clinically significant bacteremia: bacteremia that occurs in a patient who fulfils criteria for systemic inflammatory response (see below, sepsis criteria).
• Charlson index: punctuation is automatically calculated by filling the data in the database. Alternatively, if Pitt score was previously calculated, it may be added directly to the database. For all diseases, a medical diagnosis in chart is enough. Additionally, the following criteria should be used: 1,Diabetes mellitus: antidiabetic therapy (oral or insulin); 2, Chronic pulmonary disease: any disease leading to chronic respiratory insufficiency; 3,Myocardial infarction: EKG evidence; 4, Congestive heart failure: NYHA grade II or higher; 5, Peripheral arterial disease: when causing skin ulcer or the need for revascularization or amputation; 6. Dementia: if significantly limiting independence for basic activities; 7,Connective tissue disease: if requiring immunosuppressive therapy, 8.Liver disease: chronic hepatitis, significant liver fibrosis or cirrhosis; 9,Kidney disease: creatinine clearance <30 ml/min or the need for chromic dialysis; 10 Any tumour: any malignancy requiring chemotherapy and/or radiotherapy or palliative care.
• McCabe classification (modified). This is a classification for the chronic underlying condition (not the acute condition): non-fatal underlying disease (no underlying disease or related death is expected to occur in the next 5 years), ultimately fatal underlying disease (related death is expected to occur in the next 5 years), or rapidly fatal disease (related death is expected to occur in the next year).
• Pitt score: punctuation is automatically calculated by filling the data in the database, which should be retrospectively collected in the 24 hours prior to diagnosis of bacteremia. Alternatively, if Pitt score had been calculated previously, it can be directly added.
• SIRS severity
• Sepsis: at least 2 of the following: temperature >38ºC or <36ºC, respiratory rate >20 or PaCO2 <32 mmHg, heart rate >90, altered mental status, systolic blood pressure <90 mmHg, leukocyte count >12.000/mm3 or <4,000/mm3 or immature forms >10%.
• Severe sepsis: sepsis plus one of the following: hypotension (systolic BP <90 mmHg, median BP <70 mmHg, decrease in median BO >40 mmGh), organ dysfunction (respiratory, renal, liver, neurologic, hematologic), or hyperlactatemia (> 3 mmol/L)
• Septic shock: sustained hypotension not responding to fluid support therapy and requiring inotropic support.
• Acquisition. Nosocomial if infection signs/symptoms started >48 hours after hospital admission or in less than 48 hours after hospital discharge. Otherwise, the case should be considered community-onset.
• If community-onset, the episode is considered healthcare-associated if fulfilling any of the following criteria in the previous 3 months: hospitalization in acute care center, any kind of dialysis, surgery, specialized home care, attention at day-hospital, any kind of invasive procedure (endoscopy, urinary or vascular catheterization, etc.) or long-term care facility resident.
• Source: CDC definitions for nosocomial infections will be used as a reference; however, clinical and microbiological criteria as evaluated by the investigators may be used for interpretation. A source does not need to be microbiologically confirmed if enough clinical criteria are present.
• Empirical therapy: administered before susceptibility report is available.
• Definitive therapy: administered once the susceptibility report is available. If empirical therapy was continued, it is not necessary to fill in the definitive therapy data.
• Outcome definitions:
• Improvement: partial control or resolution of signs and symptoms related to the infection, or resolution but antibiotic therapy is still necessary.
• Non-improvement or deterioration: clinical situation qualified as similar or worse in comparison to that at the point of diagnosis of bacteremia.
• Cure: resolution of all signs and symptoms related to the infection, and antibiotic therapy is no longer necessary.
• Dead: death of the patient for whatever the reason.
• Renal toxicity: development of acute kidney injury according to RIFLE criteria (Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, Acute Dialysis Quality Initiative workgroup: Acute renal failure - definition, outcome measures, animal models, fluid therapy and information technology needs: The second international consensus conference of the acute dialysis quality initiative (ADQI) group. Crit Care 2004, 8:R204 - R212.
• Liver toxicity: Patients will be considered to have hepatotoxicity when they presented alanine transaminase, aspartate aminotransferase or bilirubin elevations of more than two times the upper limit of the normal range. Toxicity is considered severe when symptomatic elevations are more than three times or asymptomatic elevations are more than five times the normal levels.
• Severe adverse event (SAD): Any event that becomes fatal or life-threatening, produces a significant or persistent alteration (according to the Principal Investigator), requires patient hospitalization or prolongation of an existing hospitalization, produces a congenital anomaly or results in overdose of the drugs included in this stud. Any toxicity of grade 4 will be considered a SAD (National Cancer Institute Common Toxicity Criteria Version 3.0).

Quality of data

Data will be approved and signed by the responsible investigator in each center. It is recommended that data collectors designated by PIs in each Institution are trained in Good Clinical Practice. All data will be centrally reviewed; queries will be sent for missing data as well as data showing inconsistencies or discrepancies. Data will be analysed per center; those centers showing significant differences with the average will be requested to review their data.

Statistical Analysis Plan

• Objective 1: To evaluate the efficacy and safety of different antibiotics used for the treatment of infections caused by ESBL- and carbapenemase-producing Enterobacterales in the SOT population.

Separate statistical analyses will be performed by grouping the different antibiotics to be evaluated in the two subcohorts of empirical and definitive therapy, which will be analysed separately; active/inactive empirical therapy will be a potential confounder when analysing definitive therapy and viceversa. To be assigned to specific treatment arms, patients must have received the drugs according to the following specific criteria:

• Empirical therapy: patients received the drug in monotherapy (except if combination is considered) within 24 hours of the blood culture being drawn and prior to availability of antibiogram results. The antibiotic must have been administered for at least 48 hours, with the single exception of patients who died before 48 hours, who will be included if death occurred after 1 complete day of therapy with the assigned regimen (and will be otherwise excluded).
• Definitive therapy: the drug was administered in monotherapy (except if combination is considered) once susceptibility data was available. When more than one drug was used, only drugs that were administered for at least half the duration of the definitive therapy will be included in this study.

The primary endpoint variable, namely "mortality from any cause until day 30", of patients treated in both cohorts will be plotted using Kaplan-Meier curves and compared using the log-rank test. Multivariate analyses will be performed using Cox regression. The other primary endpoint variable, "Cure rate at day 14", will be analysed as a dichotomous outcome. Regression analyses will use the logistic regression model. Variables with p value <0.2 in the bivariate analysis will be introduced into models (including those related to transplant). A propensity score (the probability of receiving each of the two treatment types under comparison) will be calculated using a non-parsimonious multivariate logistic regression model in which the treatment type will be the outcome variable and propensity scores will be introduced into the models. Interactions between treatment and other variables will be explored and included in the model if they cause a significant modifying effect. Variables will be selected using a backward stepwise process. Goodness of fit will be assessed all throughout. The Akaike Information Criterion (AIC) will be used to select the final logistic models. The models chosen will be those that minimize the Kullback-Leibler divergence between the model and the actual data. The intraclass correlation coefficient will be calculated in order to assess the need for controlling the "Center effect" through a multilevel analysis.

If the sample size is sufficient, an additional analysis restricted to patients matched on the basis of propensity score, will also be performed. Each patient who received one type of treatment will be matched with one who received treatment with the other type of antibiotic evaluated, using calipers of width equal to 0.2 of the standard deviation of the logit of the propensity score.

• Objective 2. To compare the efficacy of different antimicrobials between SOT and non-SOT patients (using matched controls from the "non-transplant" INCREMENT cohort).
• Cases that occurred in transplant patients (Transplant-INCREMENT cohort) will be controlled with cases in non-transplant patients selected from the INCREMENT cohort. Two control groups will be defined according to the following criteria:
• Clinical variables: age, source of BSI, severity of SIRS (septic shock), acquisition.
• Propensity score to receive a particular treatment.
• Subcohorts with patients treated with the treatments to be compared will be selected.
• The same criteria about empirical and definitive treatments from objective 1 will be followed.
• Primary endpoints will also be the same than for objective 1, i.e. "mortality from any cause until day 30" and "cure rate at day 14".
• A propensity score to receive the two treatment types under comparison will be calculated by obtaining a non-parsimonious multivariate model by logistic regression in which the outcome variable will be the treatment type. Explanatory variables will include age, gender, center, type of ward, acquisition, Charlson index, Pitt score, severity of SIRS and source.
• After univariate analysis, multivariate analysis to investigate the adjusted association of treatment type and transplantation with the main and secondary outcome variables will be performed by using logistic regression (for clinical response at day 14) and by Cox regression for mortality. If time until death is unavailable, logistic regression will be used for 30-day mortality. Logistic regression will also be used for 72-hour and 30-day clinical response. The propensity score will be added in all cases; also, Charlson score, Pitt score, severity of SIRS, source and transplant-related variables will be added. Interactions between treatment and other variables will be explored and included in the model if they cause a significant modifying effect. Variables will be selected using a backward stepwise process. Goodness of fit will be assessed throughout. The Akaike Information Criterion (AIC) will be used to select the final logistic models. The models chosen will be those that minimize the Kullback-Leibler divergence between the model and the actual data. The intraclass correlation coefficient will be calculated in order to assess the necessity of controlling the "Center effect" through a multilevel analysis.

The analyses will be performed using R software (version 3.0.1), SPSS 18.0 software and multilevel analysis (MLwiN Version 2.1, University of Bristol).

Microbiological studies

Feasibility tests provided by the different institutions confirmed that most participating hospitals have a stock collection of ESBL and/or carbapenemase producers associated with the clinical cases that will be included in the INCREMENT-SOT database during the study period. Participating centers will be invited to keep these multiresistant isolates until funding is available to ship these bacterial samples from the different participating institutions to a central laboratory in Spain, in order to generate a central collection of microorganisms. This central collection of multidrug-resistant Enterobacterales will be used to conduct phenotypic (antimicrobial susceptibility by microdilution) and genotypic studies (resistance determinants, molecular typing, plasmids study, phylogenetic studies) and the relation of these parameters with clinical data will be evaluated. Also, once the collection is generated, it is expected that research groups in the consortium will be in a strong position to apply for funding in competitive calls to conduct this type of studies.