Effectiveness and Safety of Tigecycline Compared with Other Broad-Spectrum Antimicrobials in

Transcription

1 Received Date : 06-Jun-2016 Revised Date : 20-Sep-2016 Accepted Date : 29-Sep-2016 Article type : Original Research Article Effectiveness and Safety of Tigecycline Compared with Other Broad-Spectrum Antimicrobials in Abdominal Solid Organ Transplant Recipients with Polymicrobial Intraabdominal Infections Tyler Liebenstein 1,2, Lucas T Schulz 1,2, Chris Viesselmann 1, 2, Emma Bingen 1,2, Jackson Musuuza 3, Nasia Safdar 3, Warren E. Rose 1,2* Funding 1 Department of Pharmacy, University of Wisconsin Hospital and Clinics, Madison WI 2 School of Pharmacy, University of Wisconsin-Madison, Madison WI 3 School of Medicine and Public Health, Division of Infectious Diseases, Department of Medicine, University of Wisconsin-Madison, Madison, WI This study was supported by internal funding and conducted as part of our routine work. Keywords: polymicrobial infections, transplant, outcomes, broad-spectrum therapy, glycylcycline This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: /phar.1883

2 Running head: Tigecycline outcomes in abdominal transplant recipients *Corresponding Author: Warren Rose, 2 Pharmacy Practice Division, University of Wisconsin-Madison School of Pharmacy, Madison WI USA777 Highland Avenue, Madison, WI 53705; Tel: ; Fax ; warren.rose@wisc.edu Abstract STUDY OBJECTIVE: As patients with abdominal solid organ transplants (SOT) are at increased risk of polymicrobial intraabdominal infections (IAIs) following transplantation, the objective of this study was to compare the effectiveness and adverse event profile of tigecycline with those of other broadspectrum therapies for polymicrobial IAIs in this population. DESIGN: Retrospective cohort study. SETTING: Large academic medical center with multiple outpatient clinics. PATIENTS: A total of 81 adult SOT recipients were included who were treated for confirmed or suspected polymicrobial IAIs from ; of these patients, 27 received tigecycline and 54 received comparator therapy with a broad-spectrum -lactam (e.g., piperacillin-tazobactam, cefepime, or meropenem) with or without glycopeptide or lipopeptide gram-positive therapy (vancomycin or daptomycin) (comparator group). Patients in the comparator group were matched to tigecycline-treated patients based on transplant type (kidney, combined kidney-pancreas, combined kidney-liver, or solitary pancreas) in a 1:2 ratio (tigecycline:other broad-spectrum antibiotics). MEASUREMENTS AND MAIN RESULTS: Data on patient demographics, comorbidities, and clinical variables were collected and compared by using bivariate analyses. Clinical outcomes clinical cure,

3 improvement, or failure, and disease recurrence as well as death within 1 year were analyzed by bivariate analyses and logistic regression. Clinical cure was lower in the tigecycline group versus the comparator group (40.7% vs 72.2%, p=0.008), but cure combined with improvement was similar between the two groups (85.2% vs 88.9%, p=0.724). Multiple logistic regression analysis showed that treatment with comparator antibiotics increased the odds of cure (odds ratio [OR] 1.37, 95% confidence interval [CI] ) and reduced the odds of treatment failure (OR 0.59, 95% CI ) and death within 1 year (OR 0.79, 95% CI ); however, patients receiving comparator antibiotics were more likely to have disease recurrence (OR 1.45, 95% CI ). Patients receiving tigecycline experienced a higher rate of adverse events than those receiving comparator antibiotics (29.6% vs 9.3%, p=0.026). CONCLUSION: Patients receiving tigecycline were less likely to achieve optimal clinical outcomes and had more adverse events. Alternative regimens should be selected over tigecycline for treatment of polymicrobial IAIs in abdominal SOT recipients until additional studies are completed to examine its role in this population. Recipients of abdominal solid organ transplants (SOT),which include liver, pancreas, and/or kidney transplants, are at a particularly high risk of infectious complications, especially in the early posttransplantation period. One primary infectious concern during this period is the occurrence of intraabdominal infections (IAIs). 1 These infections are often complicated because they are typically polymicrobial and may be caused by multidrug-resistant strains, including extended-spectrum - lactamase producing Enterobacteriaceae, methicillin-resistant Staphylococcus aureus (MRSA), and vancomycin-resistant Enterococcus (VRE). 2-4 Antimicrobial therapy prior to SOT is common and increases the risk of selection for antibiotic-resistant organisms. Other factors that contribute to this high risk in SOT recipients include considerable immunosuppression, health care setting exposure, and iatrogenic causes during prolonged hospitalization. In immunocompromised patients, it is imperative to select antibiotics with high potency against the likely or identified causative pathogen(s) since patients are

4 unable to mount a full immune response during the critical infection period. 5 Overcoming this often requires dosing antibiotics to achieve high antibiotic concentrations at the site of infection as well as combination therapy to cover potential multidrug-resistant organisms 5 Even though complicated bacterial infections are a common occurrence during the posttransplantation period, few studies have compared broad-spectrum antimicrobial therapies for outcome and safety in this population. Tigecycline is a glycylcycline antibiotic with a broad spectrum of activity, including many antibiotic-resistant gram-positive, gram-negative, and anaerobic organisms. 6, 7 Tigecycline is indicated for the treatment of community-acquired bacterial pneumonia, complicated skin and skin structure infections, and complicated IAIs. Although tigecycline resistance remains low except for intrinsically resistant Pseudomonas aeruginosa and Proteus species, concerns exist regarding increased risk of mortality when used as initial therapy for serious infections. 8, 9 This poor response has been partially attributed to the bacteriostatic activity and low blood concentrations of tigecycline. 10 The United States Food and Drug Administration (FDA) issued a labeling change warning of increased mortality risk, 13 but it is not clear if this applies to all patient populations and infections. To our knowledge, the effectiveness of tigecycline during the post-transplantation period in the SOT population remains unstudied even though there are many potential advantages to using tigecycline for polymicrobial infections. Thus, the objectives of this study were to evaluate the clinical effectiveness and adverse event profile of tigecycline compared with other broad-spectrum antimicrobials for the treatment of polymicrobial IAIs in abdominal SOT recipients. Methods Study Design, Setting, and Patient Population In this retrospective cohort study, we compared the clinical outcomes and safety of tigecycline with other broad-spectrum antimicrobials in hospitalized abdominal transplant recipients for the treatment of polymicrobial IAIs from at the University of Wisconsin Hospital and Clinics (Madison, WI). The comparator cohort consisted of patients who received therapy with a broad-spectrum -lactam (e.g., piperacillin-tazobactam, cefepime, or meropenem) with or without glycopeptide or lipopeptide gram-

5 positive therapy (vancomycin or daptomycin) for confirmed or suspected polymicrobial infection. Patients in the comparator group were matched to tigecycline-treated patients over the same time period based on transplant type in a 1:2 (tigecycline:other broad-spectrum antibiotics) ratio. Eligible patients included adults aged years with a history of abdominal SOT treated with tigecycline or comparator therapy for at least 48 hours as an inpatient with suspected or confirmed polymicrobial IAI. Patients with a culture of an organism inherently resistant to tigecycline (e.g., Pseudomonas aeruginosa) were included if an additional active antibiotic was used to treat that pathogen. Patients were excluded if they received inappropriate therapy for concomitant invasive fungal or opportunistic viral infection, or were not evaluable by chart review. Approval from the University of Wisconsin Institutional Review Board committee for human subject research was obtained prior to study initiation. Data Collection Patients were identified through query of the institution s electronic medical record. Patient data collection was performed by using a structured data record form. Patient demographic and clinical characteristics, including body weight, hospital length of stay, transplant type, and primary and secondary infectious diagnoses, were collected. Other patient variables collected included inflammatory markers, hemodialysis or continuous renal replacement therapy during hospitalization, vasopressor use, presence of concomitant chronic or opportunistic infections, recent antimicrobial exposure, intensive care unit (ICU) stay 90 days prior to treatment, bacteremia and fungemia 90 days prior to treatment, source control interventions, antimicrobial adverse events, death during hospitalization, and death within 1 year after hospital discharge. Presence of Clostridium difficile infection during treatment and within 90 days of study drug discontinuation was recorded.

6 Outcomes The indication for use of tigecycline or comparator broad-spectrum antimicrobials was identified as either empiric, primary, or consolidation therapy; escalation in antibiotic coverage spectrum secondary to observed antimicrobial resistance; or failure or treatment-limiting adverse effects of previous antimicrobial regimen. Patients were assessed for clinical cure, improvement, or failure at hospital discharge. Clinical cure was defined as the absence of all pretreatment signs and symptoms of infection, and no continued antibiotic treatment deemed necessary. Clinical improvement was defined as improvement of pretreatment signs and symptoms of infection (i.e., fever, white blood cell count, abdominal pain), but continued antibiotic therapy without change was required against the suspected or isolated organism until completion. Clinical failure was defined as persistent, worsening, or new or recurrent signs and symptoms of infection, the need to change therapy based on lack of response (e.g., fever, elevated inflammatory markers, persistent cultures), or discontinuation of study antibiotic due to adverse events according to medical team documentation in the electronic health record of at least a 11, 12 potential association. A change in therapy was defined as a change to or addition of tigecycline or comparator agents for the same intended organism(s). Two unblinded study investigators determined the outcome based on the collected clinical data. Patients were evaluated for recurrence of infection within 1 year after hospital discharge, defined by culture of the same organism(s) at the original body site, death during hospitalization, and death within 1 year after hospital discharge. Statistical Analysis Bivariate analyses were conducted to compare demographic and clinical characteristics of patients treated with tigecycline versus those treated with comparator antibiotics. The Fisher exact test was used to compare categorical data. A two-sample t test was used to compare normally distributed data, and the nonparametric Wilcoxon Mann-Whitney test was used if the dependent variable was not normally distributed.

7 A logistic regression model predicting the odds of clinical cure or improvement comparing patients treated with tigecycline and those treated with comparator antibiotics was fitted. Potential confounders adjusted for in the final model were selected using the forward stepwise selection process. Model building started with a simple model that only included treatment type (therapy), age, and sex as covariates. During model building, all potential confounders that were marginally associated with the outcome (P < 0.10) from the univariate analysis were subsequently and sequentially added to the model. Only those variables with P < 0.10 in the logistic regression models were kept in the final models. We also used subject area expertise to assess if the variables we selected for inclusion in the final model were risk factors for the outcomes examined. A similar procedure for covariate selection as above was used for the outcome variables clinical failure, death during hospitalization, death within 1 year after hospital discharge and disease recurrence. Separate final logistic regression models were performed for each these outcomes adjusted for the selected covariates. For all analyses, a P value of less than 0.05 was considered to indicate a statistically significant difference. Statistical analysis was performed by using SAS software, version 9.2 (SAS Institute Inc., Cary, NC). Results Baseline Characteristics One-hundred thirty-eight patients received tigecycline at our institution during the study period. Of these, 36 patients received an abdominal SOT. Nine patients were excluded because they received tigecycline for indications other than IAI (7 respiratory tract infections, 1 pyelonephritis, 1 osteomyelitis); therefore, 27 patients were evaluated in the tigecycline group. The comparator group included 54 matched SOT recipients who received comparator broad-spectrum antimicrobials. The majority of patients were liver transplant recipients (63%). Other transplant types were kidney (22.2%), combined kidney-pancreas (7.4%), combined kidney-liver (3.7%), and solitary pancreas

8 (3.7%). Baseline demographic and clinical characteristics were similar between groups (Table 1). Duration of hospital stay (38.3 vs 33.1 days, p = 0.491) and Charlson comorbidity index score (4.8 vs 4.3, p = 0.152) were similar between the tigecycline and comparator groups, respectively. Patients in the tigecycline group were significantly more likely to have received antibiotics within the past 90 days (51.9% vs 20.4%, p = 0.005) and to have concomitant opportunistic viral infections (25.9% vs 5.6%, p = 0.014). The proportion of patients with fungemia within 90 days prior to treatment was comparable between the two groups (3.7 vs 5.6%, p = >0.99, although the proportion of those with bacteremia within 90 days prior to treatment (51.9% vs 18.5%, p = 0.004) and combined fungemia and bacteremia (51.9% vs 22.2%, p = 0.011) were higher among patients treated with tigecycline. The primary infection type was from an intraabdominal source (Table 2). All patients in the tigecycline group had primary IAIs, as did all but one patient in the comparator group. Many patients also had infections at secondary sites. Tigecycline was used as empiric or primary therapy in five patients (18.5%). Other reasons for use of tigecycline were consolidation therapy (11 patients [40.7%]), escalation in antibiotic coverage spectrum (6 patients [22.2%]), failure of previous antimicrobials (2 patients [7.4%]), and treatment-limiting adverse effects of previous antimicrobials (3 patients [11.1%]). Antibiotics received in the comparator group were highly variable and often changed during the hospitalization. Comparator antibiotics received at least once during hospitalization included piperacillintazobactam (59%), cefepime (21%), meropenem (33%), vancomycin (59%), and daptomycin (76%). Microbiology Table 3 summarizes the patient microbiologic data for the IAIs. Twenty-four patients in the tigecycline group (88.9%) and 53 patients in the comparator group (98.1%) had culture-confirmed polymicrobial IAI. The most common organism isolated was Enterococcus species, including a high rate of VRE in both the tigecycline (74.1%) and comparator (66.7%) groups. Pseudomonas aeruginosa was identified as a pathogen in 29.6% of patients in the tigecycline group. When this occurred, additional therapy targeted at this organism was added.

9 Clinical Outcomes Source control interventions were performed in 24 patients (89%) in the tigecycline group and in 100% of patients in the comparator group, all of which were deemed as adequate source control. Methods of source control in the tigecycline group were percutaneous aspiration (58.3%), abdominal washout (37.5%), and retransplantation (4.2%). Methods of source control in the comparator group were percutaneous aspiration (55.6%) and abdominal washout (44.4%). The clinical outcome comparison between groups is displayed in Figure 1. Clinical cure was achieved in 11 and 39 patients in the tigecycline and comparator groups, respectively (40.7% vs 72.2%, P=0.008). When the three patients in the tigecycline group who did not have source control interventions attempted were removed, clinical cure was achieved in 9 and 39 patients in the tigecycline and comparator groups, respectively (37.5% vs 72.2%, P=0.005). The outcomes of combined clinical cure and clinical improvement, treatment failure, and recurrence were similar between the two groups. Death during hospitalization occurred in five patients in the tigecycline group and eight patients in the comparator group (18.5% vs 14.8%, P=0.751). A total of nine deaths occurred within 1 year of receipt of antibiotics in the tigecycline group compared with 15 in the comparator group (33.3% vs 27.8%, P=0.615). Logistic regression results are presented in Table 4. Patients treated with comparator antibiotics were approximately 1.4-fold more likely to achieve clinical cure or improvement than those treated with tigecycline. The odds of treatment failure decreased by 41% for patients treated with comparator antibiotics versus those treated with tigecycline. Odds of death within 1 year after hospital discharge decreased by 21% for patients treated with comparator antibiotics compared to those treated with tigecycline. However, patients treated with comparator antibiotics were approximately 1.5-fold more likely to have disease recurrence after their treatment.

10 Adverse Events Eight patients in the tigecycline group (29.6%) experienced at least one adverse event compared to five patients (9.3%) in the comparator group (P = 0.026). Nausea was frequent in the tigecycline group (six patients). Other adverse events in the tigecycline group included thrombocytopenia in one patient, hepatotoxicity in two patients, and pancreatitis in one patient. Tigecycline was discontinued in four patients with nausea, one patient with hepatotoxicity, and both patients with thrombocytopenia and pancreatitis. Nausea and pancreatitis resolved in all five patients on discontinuation of tigecycline. Neither the thrombocytopenia nor hepatotoxicity resolved with discontinuation of tigecycline, perhaps reflective of worsening overall liver disease rather than adverse events of tigecycline. Of the five adverse events in the comparator group, two patients experienced nausea, two patients had creatine kinase level elevations attributed to daptomycin, and one patient had myelosuppression attributed to piperacillintazobactam. Discussion Tigecycline represents a therapeutic option for patients with skin and soft tissue infections, community-acquired pneumonia, and IAIs. Its broad spectrum of activity against many multidrugresistant isolates and high tissue penetration (volume of distribution 7 L/kg) would seem to lend itself well to treat these infections. However, a substantial amount of uncertainty exists regarding the clinical outcomes of patients receiving tigecycline. The FDA added a black box warning to the tigecycline drug label in This was the result of a meta-analysis of clinical trials showing an increase in all-cause mortality for tigecycline versus comparator treatments. 8 The increase in mortality was primarily seen in patients with ventilator-associated pneumonia, for which tigecycline is not FDA approved. However, the black box warning states that tigecycline should be reserved for use in situations when alternative treatments are not suitable. 13 Furthermore, tigecycline has not been widely studied in SOT recipients, a population for which tigecycline could represent an important therapeutic option due to antimicrobial

11 resistance concerns and the frequency of IAIs arising from surgical manipulation and immunosuppression. In our study, only 18.5% of patients received tigecycline as empiric or primary therapy. All other patients received tigecycline either for resistant bacteria on culture, failure or adverse effects of other antimicrobials, or consolidation of several broad-spectrum antimicrobials into a single agent. Patients in this study had polymicrobial IAIs. The most commonly isolated pathogen was Enterococcus species, including a substantial number of VRE isolates in both groups (Table 3). Although tigecycline is not approved for VRE treatment, it has demonstrated reliable activity against this organism. 14 Rates of VRE colonization have been shown to increase after receipt of SOT, and VRE colonization has been strongly correlated with VRE-related infections. 15 No patients in the tigecycline group and only one in the comparator group were infected with Staphylococcus aureus, likely reflective of the infrequent prevalence of this pathogen in IAIs. We were interested in the outcome of mortality in patients, including those with clinical failure, since mortality has been reported to be increased with tigecycline treatment for serious infections. 16 In the univariate analysis, a higher proportion of deaths during hospitalization occurred in the tigecycline group compared to the comparator group (18.5% vs 14.8%, P=0.751), although the small sample size limited any statistical significance. These results are similar to those reported in the meta-analyses of 14 randomized trials with tigecycline where all-cause mortality was higher with tigecycline (odds ratio 1.28), but this difference also was not statistically significant (95% confidence interval ). 9 Another meta-analysis with these same trial data concluded that tigecycline significantly increased the risk of mortality. 8 A review of these differences suggests that a general statement about tigecycline-associated mortality is difficult to articulate and may depend on the underlying mortality in the study population. 17 Results of the multiple logistic regression with death during hospitalization were not presented due to lack of robustness of the model. Recent evidence suggests that tigecycline may be a useful treatment for Clostridium difficile infection (CDI) We were interested to see if tigecycline may have a protective effect against the

12 development of CDI in our study population. IAIs are often polymicrobial, and broad-spectrum antibiotics are often required for treatment. Many antibiotic options for IAIs, such as the broad-spectrum -lactams used in our comparator study population, increase the risk of CDI in hospitalized patients. 21 Only one patient (3.7%) in the tigecycline group developed CDI within 90 days of completion of antimicrobial therapy compared to five patients (9.3%) in the comparator group (P=0.658). Although this difference was not statistically significant in our study, it remains a potential area of research in larger populations. Our study has limitations. Patients were not randomized to either tigecycline or comparator antibiotics. Although it is difficult to adjust for all confounding factors, we conducted multiple logistic regression analyses and adjusted for potential known confounders such as comorbidities, duration of hospital stay, and history of hemodialysis. The absence of randomization increases the likelihood of confounding by indication in this study. This would be a plausible explanation for the results if only the very sick patients, who would be more likely to have poor outcomes, were treated with tigecycline rather than with comparator antibiotics. This is a low probability in our study since only 7.4% of the patients received tigecycline because of failure of previous antimicrobials. Most of the tigecycline group (40.7%) was given as consolidation therapy, whereas only 5% of the patients received it as empiric therapy. Another limitation of the study was the failure of all logistic regression models to reach statistical significance. A post hoc power analysis for our sample size of 27 patients and effect size of 0.31 indicates that we had ~11% power to detect differences between the tigecycline versus comparator groups. Finally, this study shows the experience of a single medical center, limiting the external validity of the results. Despite these limitations, the findings from this study have important clinical significance regarding tigecycline use in SOT recipients since, to our knowledge, no other studies have reported tigecycline outcomes for polymicrobial infections in this population.

13 Conclusion The patients in our study were less likely to achieve clinical cure with receipt of tigecycline and had poorer outcomes compared to other antimicrobials. Tigecycline was also associated with a higher frequency of adverse events, particularly nausea. Our results suggest that, as in other patient populations, a careful consideration of risks and benefits should be undertaken before using tigecycline as first-line therapy for IAIs in abdominal SOT recipients. There is need for larger, multicenter, randomized trials to assess the efficacy of tigecycline among patients with SOT. Acknowledgments We thank Kurt Osterby and the staff of the UW Health Center for Clinical Knowledge Management for their assistant with data acquisition and management in this study. References 1. Cervera C, Fernandez-Ruiz M, Valledor A, et al. Epidemiology and risk factors for late infection in solid organ transplant recipients. Transpl Infect Dis 2011;6: Lee SO, Kang SH, Abdel-Massih RC, Brown RA, Razonable RR. Spectrum of early-onset and lateonset bacteremias after liver transplantation: implications for management. Liver Transpl 2011;6: Kawecki D, Kwiatkowski A, Michalak G, et al. Etiologic agents of bacteremia in the early period after simultaneous pancreas-kidney transplantation. Transplant Proc 2009;8: Chabros L, Mlynarczyk G, Sawicka-Grzelak A, et al. Comparison of bacteria isolated from urinary tract infections in patients on transplant versus urologic wards. Transplant Proc 2011;8: Solomkin JS, Mazuski JE, Bradley JS, et al. Diagnosis and management of complicated intraabdominal infection in adults and children: guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis 2010;2:

14 6. Rose WE, Rybak MJ. Tigecycline: first of a new class of antimicrobial agents. Pharmacotherapy 2006;8: Heizmann WR, Loschmann PA, Eckmann C, von Eiff C, Bodmann KF, Petrik C. Clinical efficacy of tigecycline used as monotherapy or in combination regimens for complicated infections with documented involvement of multiresistant bacteria. Infection 2015;43: Yahav D, Lador A, Paul M, Leibovici L. Efficacy and safety of tigecycline: a systematic review and meta-analysis. J Antimicrob Chemother 2011;9: Tasina E, Haidich AB, Kokkali S, Arvanitidou M. Efficacy and safety of tigecycline for the treatment of infectious diseases: a meta-analysis. Lancet Infect Dis 2011;11: Nemeth J, Oesch G, Kuster SP. Bacteriostatic versus bactericidal antibiotics for patients with serious bacterial infections: systematic review and meta-analysis. J Antimicrob Chemother 2015;70: Dubrovskaya Y, Chen TY, Scipione MR, et al. Risk factors for treatment failure of polymyxin B monotherapy for carbapenem-resistant Klebsiella pneumoniae infections. Antimicrob Agents Chemother 2013;11: Sakoulas G, Brown J, Lamp KC, Friedrich LV, Lindfield KC. Clinical outcomes of patients receiving daptomycin for the treatment of Staphylococcus aureus infections and assessment of clinical factors for daptomycin failure: a retrospective cohort study utilizing the Cubicin Outcomes Registry and Experience. Clin Ther 2009;9: Tygacil [package insert]. Philadelphia, PA: Wyeth Pharmaceuticals Inc; Sader HS, Flamm RK, Jones RN. Tigecycline activity tested against antimicrobial resistant surveillance subsets of clinical bacteria collected worldwide (2011). Diagn Microbiol Infect Dis 2013;2: Ziakas PD, Pliakos EE, Zervou FN, Knoll BM, Rice LB, Mylonakis E. MRSA and VRE colonization in solid organ transplantation: a meta-analysis of published studies. Am J Transplant 2014;8: Prasad P, Sun J, Danner RL, Natanson C. Excess deaths associated with tigecycline after approval based on noninferiority trials. Clin Infect Dis 2012;12:

15 17. Verde PE, Curcio D. Imbalanced mortality evidence for tigecycline: 2011, the year of the metaanalysis. Clin Infect Dis 2012;3: Aldape MJ, Heeney DD, Bryant AE, Stevens DL. Tigecycline suppresses toxin A and B production and sporulation in Clostridium difficile. J Antimicrob Chemother 2015;70: Larson KC, Belliveau PP, Spooner LM. Tigecycline for the treatment of severe Clostridium difficile infection. Ann Pharmacother 2011;7-8: Herpers BL, Vlaminckx B, Burkhardt O, et al. Intravenous tigecycline as adjunctive or alternative therapy for severe refractory Clostridium difficile infection. Clin Infect Dis 2009;12: Slimings C, Riley TV. Antibiotics and hospital-acquired Clostridium difficile infection: update of systematic review and meta-analysis. J Antimicrob Chemother 2014;4: TABLES Table 1. Baseline Demographic and Clinical Characteristics of the Study Patients Characteristic Tigecycline Comparator Group p Value Group (n=54) (n=27) Age (yrs) 52.6 ± ± Male 17 (63.0) 34 (63.0) >0.999 Height (in.) 66.7 ± ± Actual body weight (kg) 78.7 ± ± Ideal body weight (kg) 63.8 ± ± Duration of hospitalization (days) 38.3 ± ± Charlson comorbidity index score 4.8 ± ±

16 Hemodialysis 11 (40.7) 20 (37.0) Continuous renal replacement therapy 6 (22.2) 16 (29.6) SIRS criteria 11 (40.7) 15 (27.8) ICU care within 30 days prior to treatment 10 (37.0) 10 (18.5) Recent antibiotics within 90 days prior to treatment 14 (51.9) 11 (20.4) WBC change in first 48 hrs (cells/mm 3 ) ± ± ICU-level care for abdominal infection 10 (37.0) 28 (51.) Vasopressor use 7 (25.9) 18 (33.3) Time from transplantation to infection (yrs) 4.9 ± ± 2.3 <0.001 Immunosuppression Corticosteroid 24 (88.9) 49 (90.7) >0.999 Antiproliferative inhibitor 13 (48.1) 31 (57.4) Calcineurin inhibitor 16 (59.3) 21 (59.3) >0.999 mtor inhibitor 1 (3.7) Intravenous immunoglobulin 2 (7.4) 3 (5.6) >0.999 Receiving 3 immunosuppressants 8 (29.6) 24 (38.9) Invasive fungal infection 13 (48.1) 29 (53.7) Opportunistic viral infection 7 (25.9) 3 (5.6) Bacteremia within 90 days prior to treatment 14 (51.9) 10 (18.5) Fungemia within 90 days prior to treatment 1 (3.7) 3 (5.6) >0.999 Combined bacteremia and fungemia 14 (51.9) 12 (22.2) Data are mean ± SD values or no. (%) of patients. SIRS = systemic inflammatory response syndrome; ICU = intensive care unit; mtor = mammalian target of rapamycin.

17 Table 2. Primary and Secondary Infection Types Infection Type Infection Source Tigecycline Group (n=27) Comparator Group (n=54) Primary Intraabdominal 27 (100) 53 (98.1) Urinary tract 0 (0) 1 (1.9) a Secondary Urinary tract 8 (30.0) 15 (27.8) Data are no. (%) of patients. Bacteremia 14 (51.9) 10 (18.5) Candidemia 1 (3.7) 3 (5.6) Bacterial pneumonia 2 (7.4) 3 (5.6) Osteomyelitis 0 (0) 1 (1.9) a Strong clinical suspicion of intraabdominal infection, although no cultures were obtained.

18 Table 3. Pathogens Collected in Abdominal Fluid Cultures in Patients with Intraabdominal Infection a Pathogen Gram-positive Tigecycline Group Comparator Group (n=27) (n=54) Enterococcus, vancomycin resistant 20 (74.1) 36 (66.7) Enterococcus, vancomycin sensitive 9 (33.3) 14 (25.9) Staphylococcus, coagulase negative 5 (18.5) 19 (35.2) Streptococcus species 4 (14.8) 7 (13.0) Gram-negative Escherichia coli 3 (11.1) 14 (25.9) Klebsiella pneumonia 5 (18.5) 5 (9.3) Pseudomonas aeruginosa 8 (29.6) 6 (11.1) Stenotrophomonas maltophila 4 (14.8) 3 (5.6) Fungi Candida albicans 5 (18.5) 17 (31.5) Candida glabrata 5 (18.5) 9 (16.7) Data are no. (%) of patients. a Isolates recovered in less than 10% of patients in both groups included methicillin-resistant Staphylococcus aureus, methicillin-sensitive Staphylococcus aureus, Enterococcus gallinarium, Aeromonas species, Lactobacillus, Corynebacterium species, Peptostreptococcus species, Klebsiella oxytoca, Enterobacter cloacae, Enterobacter aerogenes, Citrobacter freundii, Acinetobacter baumannii, Proteus mirabilis, Prevotella species, Morganella morganii, Veillonella species, Haemophilus parainfluenzae, Bacteroides species, Candida krusei, Candida tropicalis, Candida guillermundii, and Saccharomyces cervesiae.

19 Table 4. Logistic Regression Analysis of Patients Treated with Comparator Agents versus Tigecycline Outcome Odds ratio 95% Confidence Interval Clinical cure or improvement a Treatment failure b Death within 1 year c Disease recurrence d All models were adjusted for age and sex. a Also adjusted for history of hemodialysis, prior intensive care unit care, Charlson comorbidity index score, and duration of hospital stay. b Also adjusted for prior intensive care unit care, abdominal infection treatment in the intensive care unit, and Charlson comorbidity index score. c Also adjusted for Charlson comorbidity index score. d Also adjusted for history of hemodialysis.

20 FIGURE LEGEND Figure 1. Clinical outcomes of the tigecycline group (n=27) versus the comparator antibiotics group (n=54).