Copyright by Chengwen Teng 2019

The Dissertation Committee for Chengwen Teng Certifies that this is the approved version of the following dissertation:

ADVERSE DRUG REACTIONS ASSOCIATED WITH ANTIBIOTICS: AN ANALYSIS OF THE FDA ADVERSE EVENT REPORTING SYSTEM

Committee:

Christopher R. Frei, Supervisor

Kelly R. Reveles

James P. Wilson

Elizabeth A. Walter

Carlos A. Alvarez

ADVERSE DRUG REACTIONS ASSOCIATED WITH ANTIBIOTICS: AN ANALYSIS OF THE FDA ADVERSE EVENT REPORTING SYSTEM

by

Chengwen Teng

Dissertation Presented to the Faculty of the Graduate School of

The University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

The University of Texas at Austin December 2019 Dedication

I dedicated this dissertation to my family, who have supported and loved me unconditionally.

Acknowledgements

I would like to thank my supervisor, Dr. Christopher Frei, for his mentoring and support. Dr. Frei has guided me in every aspect of research, including literature review, generating research ideas, research proposal writing, data analysis, interpretation of data, and manuscript writing. He is a great mentor. In addition, I express my appreciation to my dissertation committee members, Dr. Kelly Reveles, Dr. James Wilson, Dr. Elizabeth Walter, and Dr. Carlos Alvarez. Thank you for your guidance and support throughout this dissertation project. Moreover, I express my gratitude to Dr. Kirk Evoy for his guidance on manuscript writing. I would also like to thank Dr. Frei’s research group members Dr. Obiageri Obodozie-Ofoegbu, Xavier Jones, Dr. Daryl Gaspar, Kaitlin Kennedy, Taylor Patek,

Courtney Baus, Dr. Lindsey Groff, Dr. Victor Encarnacion, and Dr. Huda Razzack. I really appreciate your kind help. Furthermore, I would like to thank The University of Texas at Austin Pharmacotherapy Division graduate students Dr. James Shurko and Shiraz Halloush for organizing social events and fun. Lastly, I would like to thank my former classmates at Yangzhou High School, Dr. Guanbo Chen, Dr. Yuanzhen Li, Le Tao, Chunming Li, and Dr. Tian Liu; my former classmates at Wuhan University, Dr. Li Li, Dr. Jiatian Lu, Dr.

Qianfan Zhang, Dr. Hao Yang, Dr. Yijing Liu, and Liangbo Jiang; my former roommates at the University of Southern Mississippi, Yongliang Shi and Chen Xia; my former colleague at the University of Alabama at Birmingham, Dr. Yifeng Gao; my former classmates at the University of Oklahoma College of Pharmacy, Dr. Ankush Rehan, Dr. Jeremy Benson, Dr. Tianyun Gao, and Dr. Joshua Johnston; for friendship and advice. v

Abstract

ADVERSE DRUG REACTIONS ASSOCIATED WITH ANTIBIOTICS: AN ANALYSIS OF THE FDA ADVERSE EVENT REPORTING SYSTEM

Chengwen Teng, Ph.D.

The University of Texas at Austin, 2019

Supervisor: Christopher R. Frei

Antibiotics are associated with adverse drug reactions (ADR), such as

Clostridium difficile infection (CDI), Torsade de pointes/QT prolongation (TdP/QTP), acute kidney injury (AKI), hypoglycemia, and rhabdomyolysis. ADRs lead to significant morbidity and mortality, as well as high health care costs. However, data on the ADR profiles of antibiotics are limited. The FDA Adverse Event Reporting System (FAERS) provides real-world data on ADRs of antibiotics. Using FAERS, this study 1) discovered new ADR associations for FDA-approved antibiotics, 2) identified an antibiotic-ADR association that was worse for patients on two drugs than on either drug alone, 3) determined if a recent FDA warning for hypoglycemia should apply to the entire fluoroquinolone class or just selected members of that class, and 4) detected an antibiotic class-ADR association that was worse in a special population, such as elderly patients. The Medical Dictionary for Regulatory Activities (MedDRA) was used to identify ADR cases. Reporting Odds Ratios (RORs) and corresponding 95% confidence intervals vi (95%CI) for the association between antibiotics and ADRs were calculated. An association was considered statistically significant when the lower limit of the 95%CI was greater than 1.0. This study evaluated the association between antibiotics and certain ADRs, which were CDI, TdP-QTP, AKI, hypoglycemia, and rhabdomyolysis. Several new antibiotic-ADR associations were found, which were associations between amikacin and TdP-QTP, between ertapenem and hypoglycemia, and between meropenem and rhabdomyolysis. Patients on vancomycin and piperacillin-tazobactam had higher association with AKI than those on vancomycin alone and those on piperacillin- tazobactam alone. Moxifloxacin and levofloxacin were associated with hypoglycemia, when patients were also taking sulfonylureas or meglitinides. Ciprofloxacin was not associated with hypoglycemia. The association between penicillin combinations and CDI in patients 65 years or older was higher than that in patients less than 65 years old. The findings of this study will aid clinicians to select antibiotics for patients with bacterial infections.

vii Table of Contents

List of Tables ...... xii

List of Figures ...... xiii

Chapter One: Overview of the FDA Adverse Event Reporting System ...... 1

Structure of the FDA Adverse Event Reporting System ...... 1

Disproportionality Analysis in the FDA Adverse Event Reporting System ...... 2

Limitations of the FDA Adverse Event Reporting System ...... 4

Chapter Two: Specific Aims ...... 5

Chapter Three: Clostridium difficile Infection Associations with Important Antibiotic Classes ...... 7

Introduction ...... 7

Methods ...... 8

Data Source ...... 8

Study Design ...... 8

Drug Exposure Definition ...... 9

Adverse Drug Reaction Definition ...... 9

Statistical Analysis ...... 10

Results ...... 11

Discussion ...... 15

Limitations ...... 18

Conclusions ...... 18

viii Chapter Four: Torsades de pointes and QT Prolongation Associations with Antibiotics ...... 19

Introduction ...... 19

Methods ...... 20

Data Source ...... 20

Study Design ...... 21

Drug Exposure Definition ...... 21

Adverse Drug Reaction Definition ...... 21

Statistical Analysis ...... 21

Results ...... 22

Discussion ...... 24

Limitations ...... 26

Conclusions ...... 26

Chapter Five: Acute Kidney Injury Associations with Antibiotics ...... 27

Introduction ...... 27

Methods ...... 29

Data Source ...... 29

Study Design ...... 29

Drug Exposure Definition ...... 29

Adverse Drug Reaction Definition ...... 30

Statistical Analysis ...... 30

Results ...... 30

Discussion ...... 32

Limitations ...... 35 ix Conclusion ...... 35

Chapter Six: Hypoglycemia Associations with Antibiotics ...... 36

Introduction ...... 36

Methods ...... 38

Data Source ...... 38

Study Design ...... 38

Drug Exposure Definition ...... 39

Adverse Drug Reaction Definition ...... 39

Statistical Analysis ...... 39

Results ...... 40

Discussion ...... 43

Conclusions ...... 45

Chapter Seven: Rhabdomyolysis Associations with Antibiotics ...... 46

Introduction ...... 46

Methods ...... 47

Data Source ...... 47

Study Design ...... 48

Drug Exposure Definition ...... 48

Adverse Drug Reaction Definition ...... 48

Statistical Analysis ...... 49

Results ...... 49

Discussion ...... 53

Limitations ...... 54

x Conclusions ...... 54

Chapter Eight: Conclusions and Future Directions ...... 56

Conclusions ...... 56

Strengths and limitations ...... 56

Translational application...... 57

Future directions ...... 57

Appendix ...... 58

Appendix 1. Study medications ...... 58

Glossary ...... 61

References ...... 62

Vita ...... 75

xi List of Tables

Table 1.1: Content of each table in FAERS...... 2 Table 1.2: A two by two continency table for a drug (X) - adverse event (Y)

combination...... 3

Table 3.1: A two by two contingency table for a drug (A) – ADR (X) combination ...... 10

Table 3.2: Gender and age information for patients on antibiotics...... 12

xii List of Figures

Figure 3.1: Reporting Odds Ratios (RORs) for Clostridium difficile infection with

antibiotics ...... 14 Figure 3.2: Reporting Odds Ratios (RORs) for Clostridium difficile infection with

antibiotics stratified by age ...... 15 Figure 4.1: Reporting Odds Ratios (RORs) for Torsades de pointes/ QT prolongation

with antibiotics...... 23 Figure 4.2: Adjusted Reporting Odds Ratios (RORs) for Torsades de pointes/QT

prolongation with antibiotics...... 24

Figure 5.1: Reporting Odds Ratios (RORs) for AKI with antibiotics...... 32

Figure 6.1: Reporting Odds Ratios (RORs) for hypoglycemia with antibiotics...... 41 Figure 6.2: Adjusted Reporting Odds Ratios (aRORs) for hypoglycemia with

antibiotics in patients not on a sulfonylurea or a meglitinide...... 42

Figure 7.1: Reporting Odds Ratios (RORs) for rhabdomyolysis with antibiotics...... 51 Figure 7.2: Adjusted Reporting Odds Ratios (aRORs) for rhabdomyolysis with

antibiotics...... 52

xiii Chapter One: Overview of the FDA Adverse Event Reporting System

STRUCTURE OF THE FDA ADVERSE EVENT REPORTING SYSTEM

The FDA Adverse Event Reporting System (FAERS) is a publicly available database organized into Quarterly Data Files, which consists of adverse event reports that were submitted to the United States Food and Drug Administration (FDA).1 FAERS is designed to support the FDA's post-marketing safety surveillance program for drug and therapeutic biologic products. Adverse drug reactions (ADR) are coded using terms in the

Medical Dictionary for Regulatory Activities (MedDRA) terminology.1 FAERS consists of voluntary reports from health care professionals (e.g., physicians, pharmacists, nurses, and others), consumers (e.g., patients, family members, lawyers, and others), and pharmaceutical manufacturers. If a manufacturer receives a report from a health care professional or a consumer, the law requires the manufacturer to send the report to the FDA.1 FAERS is a relational database consisting of seven types of tables (files), which are “demographic”, “drug”, “reaction”, “outcome”, “report sources”, “therapy”, and “indication”. “Primaryid” is the unique number for identifying each FAERS report. “Primaryid” is the primary link field (primary key) between data tables. Some FAERS reports have multiple case versions. FDA recommends using the latest case version for data analysis. See Table 1.1 for the content of each table in FAERS.

1 Table 1.1: Content of each table in FAERS. Table name Content Demographic age, sex, weight, reporter’s type of occupation, and reporter country Drug drug name, product active ingredient, drug’s reported role, route of administration, and dosing Reaction “Preferred Terms” from MedDRA to define an ADR Outcome “DE” (death), “LT” (life-threatening), “HO” (hospitalization), “DS” (disability), “CA” (congenital anomaly), “RI” (required intervention to prevent permanent impairment/damage), and “OT” (other serious important medical event) Report source the source of the report Therapy start date, end date, and duration of the therapy Indication “Preferred Terms” from MedDRA to describe the indication of the drug

Each FAERS report contains demographic information for one patient (one row in the demographic table), who took one or more drugs (one or more rows in the drug table), and had one or more ADRs (one or more rows in the reaction table).

DISPROPORTIONALITY ANALYSIS IN THE FDA ADVERSE EVENT REPORTING SYSTEM

In the literature, FAERS has been used to compare ADR profiles of various drugs and to identify drug-drug interactions. Disproportionality analyses were performed by calculating Reporting Odds Ratios (ROR), Proportional Reporting Ratios (PRR), Relative

Reporting Ratios (RRR), and Information Component (IC) to measure the association between a drug and an ADR.2 Below are the formulas to calculate ROR, PRR, RRR, and

IC:

2 Table 1.2: A two by two continency table for a drug (X) - adverse event (Y) combination Adverse event (Y) Not adverse event (Y) Total Using drug (X) a b a+b Not using drug (X) c d c+d Total a+c b+d a+b+c+d

Reporting Odds Ratio (ROR) of drug (X) – adverse event (Y) combination:2 ROR = (a/b)/(c/d), 95% Confidence Interval (CI) = eln(ROR)±1.96√(1/a+1/b+1/c+1/d)

Proportional Reporting Ratio (PRR) of drug (X) – adverse event (Y) combination:2

PRR = (a/(a+b))/(c/(c+d)), 95% CI = eln(PRR)±1.96√(1/a-1/(a+b)+1/c-1/(c+d)) Relative Reporting Ratio (RRR) of drug (X) – adverse event (Y) combination:3, 4 RRR = (a/(a+b))/((a+c)/(a+c+c+d)), 95% CI = eln(RRR)±1.96√(b/a/(a+b)+(b+d)/(a+c)/(a+b+c+d))

3 IC = log2(RRR) ROR is more sensitive and provides more signals than PRR, RRR, and IC.

RORs have been used to compare ADR profiles of various drugs in FAERS. A research article calculated RORs of “palmarplantar erythrodysesthesia (PPE)” associated with conventional doxorubicin, PEGylated-liposome doxorubicin, and non-PEGylated- liposome doxorubicin and found the highest PPE ROR in PEGylated-liposome doxorubicin.5 Another research article computed RORs for cardiotoxicity associated with targeted-therapy drugs for breast cancer and demonstrated that trastuzumab had the highest ROR.6 RORs have also been used to identify drug-drug interactions. A FAERS study discovered a much higher ROR for acute kidney injury associated with concomitant use of acyclovir or valacyclovir with nonsteroidal anti-inflammatory drugs compared with individual drug use.7 In addition, RORs were calculated for patients stratified by age groups. The ROR of sexual dysfunction associated with dutasteride was much higher in patients 76-92 years of age than in other age groups.8 3 The PRR has also been used to compare ADR profiles of drugs in FAERS. A research study calculated the PRR of death associated with warfarin versus all non- vitamin K antagonist oral anticoagulants and demonstrated mortality benefit for non- vitamin K antagonist oral anticoagulants over warfarin.9

LIMITATIONS OF THE FDA ADVERSE EVENT REPORTING SYSTEM

A causal relationship between a drug and an ADR cannot be established by FAERS. Adverse events are usually underreported in spontaneous reporting systems. Different adverse events had different reporting rates, but the average is only 6%.10 FAERS is not appropriate to estimate incidence rate, because of lack of a denominator. Weber effect, notoriety effect, ripple effect, and masking effect are limitations of FAERS. Weber effect is that the reporting of ADRs increases over the first two years after the launching of the drug and then starts decreasing. Notoriety effect is that the reporting of an ADR increases after it is highlighted, such as an FDA warning. Ripple effect is that the notoriety for a drug increases the reporting of other drugs in the same class. Masking effect is that the signal can be suppressed by a large number of reports in which the same ADR is associated with other drugs.11 The association between a drug and an ADR is confounded by comorbid diseases and concomitant drugs. FAERS has missing information and not all comorbid diseases and concomitant drugs are reported.

4 Chapter Two: Specific Aims

Specific Aim 1: Discover a new ADR association for an FDA-approved antibiotic. Hypothesis 1A: Amikacin is associated with TdP-QTP. Hypothesis 1B: Ertapenem is associated with hypoglycemia. Hypothesis 1C: Meropenem is associated with rhabdomyolysis.

Specific Aim 2: Identify an antibiotic-ADR association that is worse for patients on two drugs than on either drug alone. Hypothesis 2A: Patients on vancomycin and piperacillin-tazobactam have higher association with AKI than those on vancomycin alone. Hypothesis 2B: Patients on vancomycin and piperacillin-tazobactam have higher association with AKI than those on piperacillin-tazobactam alone.

Specific Aim 3: Determine if a recent FDA warning for hypoglycemia should apply to the entire fluoroquinolone class or just selected members of that class. Hypothesis 3A: The fluoroquinolone class is not associated with hypoglycemia. Hypothesis 3B: Levofloxacin is not associated with hypoglycemia. Hypothesis 3C: Ciprofloxacin is not associated with hypoglycemia. Hypothesis 3D: Moxifloxacin is not associated with hypoglycemia.

Specific Aim 4: Detect an antibiotic class-ADR association that is worse in a special population, such as elderly patients. Hypothesis 4: The association between penicillin combinations and CDI in patients 65 years or older is higher than that in patients less than 65 years old. 5 The Institutional Review Board (IRB) of the University of Texas Health Science Center at San Antonio determined that this project does not require IRB approval because it is not human research. The IRB protocol number is HSC20190309N.

6 Chapter Three: Clostridium difficile Infection Associations with Important Antibiotic Classes

The study described in this chapter has been published as a research article:

Teng C, Reveles KR, Obodozie-Ofoegbu OO, Frei CR. Clostridium difficile infection risk with important antibiotic classes: an analysis of the FDA Adverse Event

Reporting System. Int J Med Sci. 2019;16(5):630-5.

Authors’ contributions:

Study concept and design: Teng and Frei. Statistical analysis: Teng. Interpretation of data: Teng, Reveles, and Frei. Drafting of the manuscript: Teng. Critical revision of the manuscript for important intellectual content: All authors. Study supervision: Frei.

INTRODUCTION

Clostridium difficile infection (CDI) is a great public health concern in hospital and community settings. In the first decade of the twenty-first century, United States hospitals noted a profound increase in CDI incidence.12 Since then, national standards required hospitals to implement effective infection control interventions and antimicrobial stewardship programs to prevent CDI. Nationally-representative studies now indicate that CDI rates among hospitalized patients might be declining.13 With the decline in CDI incidence in hospitals, there appears to have been a concurrent shift to community-onset CDI.14 A rich and diverse intestinal microbiota prevents CDI; disruption of microbiota, especially due to antibiotic use, can lead to loss of colonization resistance and

7 proliferation of Clostridium difficile.15, 16 Antibiotic exposure is the most important risk factor in both hospital and community-onset CDI.17-19 In previous meta-analyses conducted between 1988 and 2009, clindamycin, fluoroquinolones, and cephalosporins had the highest CDI risks.17-19 Given the change in CDI epidemiology in recent years, more recent data are needed to evaluate the current CDI associations with various antibiotics. The FDA

Adverse Event Reporting System (FAERS) provides recent data on CDI and antibiotics.1 The objective of this study is to evaluate CDI associations with antibiotics using FAERS data from 2015 to 2017.

METHODS

Data Source

FAERS is a publicly available database organized into Quarterly Data Files, which contain adverse event reports that were submitted to United States Food and Drug Administration (FDA).1 FAERS data include patient demographic information (age and sex), drug information (drug name, active ingredient, route of administration, and drug’s reported role in the event), and reaction information. Each report lists a primary suspected drug with one or more adverse reactions and may include other drugs. Clinical outcomes, such as death and hospitalization, may also be reported.

Study Design

FAERS data from January 1, 2015 to December 31, 2017 were obtained from the FDA. Some adverse event reports were submitted multiple times with updated

8 information. Therefore, duplicate reports were removed by case number, with the most recent submission included in the study. Reports containing drugs which were administered in oral, subcutaneous, intramuscular, intravenous, and parenteral routes were included in the study, while other routes of administration were excluded.

Drug Exposure Definition

Each antibiotic was identified in the FAERS drug files by generic and brand names listed in the Drugs@FDA Database.20, 21 Only drugs with a reported role coded as “PS” (Primary Suspect Drug) or “SS” (Secondary Suspect Drug) were included in this study.22 Antibiotics with less than three CDI reports were excluded from the data analysis.23

Adverse Drug Reaction Definition

FAERS defines adverse drug reactions using Preferred Terms from the Medical Dictionary for Regulatory Activities (MedDRA). MedDRA includes a hierarchy of terms, which are (from the highest to the lowest) System Organ Classes (SOC), High Level Group Term (HLGT), High Level Term (HLT), Preferred Term (PT), and Lowest Level

Term (LLT). Standardised MedDRA Queries (SMQs) are groupings of MedDRA terms, usually at the PT level, which relate to an adverse drug reaction. Pseudomembranous colitis (SMQ), including Preferred Terms “Clostridial infection”, “Clostridial sepsis”, “Clostridium bacteraemia”, “Clostridium colitis”, “Clostridium difficile colitis”, “Clostridium difficile infection”, “Clostridium test positive”, “Gastroenteritis clostridial”,

9 and “Pseudomembranous colitis” were used to identify CDI cases.24 “Clostridium difficile sepsis”, which is a Lowest Level Term, was also used in the study.

Statistical Analysis

A disproportionality analysis was performed by calculating Reporting Odds Ratios (RORs) and corresponding 95% confidence intervals (95%CI) for the association between CDI and each antibiotic class or individual antibiotic.2 ROR was calculated as the ratio of the odds of reporting CDI versus all other events for a given drug, compared with this reporting odds for other drugs present in FAERS.2 An association was considered to be statistically significant if the 95%CI did not include 1.0 (see Table 3.1 for the calculation of ROR and CI).2 A higher ROR suggests a stronger association between the antibiotic and CDI. A subgroup analysis was performed on patients who were 65 years or older and patients less than 65 years old. The Cochran-Armitage Trend

Test was used to assess a change in the trend of CDI reports in patients who took fluoroquinolones from 2004 to 2017. Data analysis was performed using Microsoft Access 2016, Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA), SAS 9.4, and JMP Pro 13.2.1 (SAS Institute, Cary, NC).

Table 3.1: A two by two contingency table for a drug (A) – ADR (X) combination ADR (X) Other ADRs Total Drug (A) a b a+b Other drugs c d c+d Total a+c b+d a+b+c+d

† ADR = adverse drug reaction; ROR = (a/b)/(c/d); 95% Confidence Interval (CI) = eln(ROR)±1.96√(1/a+1/b+1/c+1/d)

10

RESULTS

After inclusion and exclusion criteria were applied and duplicate reports were removed, FAERS contained a total of 2,042,801 reports from January 1, 2015 to December 31, 2017. There were 5,187 CDI reports from 2015 to 2017, which were included in the data analysis. Female patients represented 61% of CDI patients who had gender information. CDI patients who had age information had a median age (IQR, interquartile range) of 62 (27) years. Please see Table 3.2 for the gender and age information of patients who were taking various antibiotics.

11 Table 3.2: Gender and age information for patients on antibiotics Antibiotic Class/Antibiotic % Female Median age (IQR) Lincosamides (clindamycin) 58 58 (28)

Monobactams (aztreonam) 56 55 (35)

Penicillin combinations 48 62 (28) Piperacillin-tazobactam 39 64 (25) Amoxicillin-clavulanate 54 60 (32) Ampicillin-sulbactam 46 67 (33)

Carbapenems 44 63 (29) Meropenem 44 61 (32) Ertapenem 44 69 (24) Imipenem-cilastatin 46 63 (27)

Cephalosporins, monobactams, and carbapenems 47 63 (34)

Third/fourth-generation cephalosporins 49 62 (39) Cefepime 44 64 (20) Cefotaxime 39 44 (61) Ceftriaxone 51 63 (41)

Tetracyclines 60 51 (35) Tetracycline 60 26 (32) Doxycycline 60 51 (35)

Macrolides 61 54 (35) Erythromycin 63 55 (26) Clarithromycin 61 55 (33) Azithromycin 59 48 (41)

Fluoroquinolones 58 58 (27) Ofloxacin 43 68 (24) Ciprofloxacin 57 57 (28) Levofloxacin 61 59 (24) Moxifloxacin 55 55 (26)

Trimethoprim-sulfamethoxazole 45 60 (28)

12 † IQR = interquartile range

The lincosamide class had the highest CDI ROR (46.95, 95%CI: 39.49-55.82) among all antibiotic classes included in the study (Figure 3.1). Clindamycin was the only antibiotic in the lincosamide class which met the inclusion criteria. The monobactam class (including aztreonam only) demonstrated the second highest CDI ROR (29.97,

95%CI: 14.60-61.54). The CDI ROR of the trimethoprim-sulfonamides class was the lowest (3.32, 95%CI: 2.03-5.43).

13 Figure 3.1: Reporting Odds Ratios (RORs) for Clostridium difficile infection with antibiotics

† CI = confidence interval; CDI = Clostridium difficile infection

Among patients who took penicillin combinations, carbapenems, cephalosporins, tetracyclines, macrolides, fluoroquinolones, and trimethoprim-sulfamethoxazole, patients who were 65 years or older had a higher CDI ROR than those less than 65 years old (Figure 3.2). Among patients who took lincosamides, patients who were 65 years or older had a lower CDI ROR than those less than 65 years old.

14 Figure 3.2: Reporting Odds Ratios (RORs) for Clostridium difficile infection with antibiotics stratified by age

† CI = confidence interval; CDI = Clostridium difficile infection; yrs = years

The Cochran-Armitage Trend Test demonstrated that there was a significant relationship between the proportion of CDI reports in patients who took fluoroquinolones and the year of reporting (p<0.0001). From 2004 to 2010, 2.3% of fluoroquinolone reports had CDI. From 2011 to 2017, 1.7% of fluoroquinolone reports had CDI.

DISCUSSION

Our antibiotic CDI association rank order was similar to previous meta- analyses.17-19 Our results demonstrated significant CDI associations (from strongest to weakest) with lincosamides, monobactams, penicillin combinations, carbapenems,

15 cephalosporins, tetracyclines, macrolides, fluoroquinolones, and trimethoprim- sulfonamides. In a prior meta-analysis of antibiotics and the risk of community-associated CDI (CA-CDI), the risks from the highest to the lowest were: clindamycin, fluoroquinolones, CMCs, macrolides, trimethoprim-sulfonamides, and penicillins, with no effect of tetracycline on CDI risk.17 In another prior meta-analysis of CA-CDI and antibiotics, the risks from the highest to the lowest were: clindamycin, fluoroquinolones, cephalosporins, penicillins, macrolides, and trimethoprim-sulfonamides, while no association was found between tetracyclines and CDIs.18 Regarding hospital-acquired CDI (HA-CDI), a prior meta-analysis indicated that the associations from the strongest to weakest were: third- generation cephalosporins, clindamycin, second-generation cephalosporins, fourth- generation cephalosporins, carbapenems, trimethoprim-sulfonamides, fluoroquinolones, and penicillin combinations.19 FAERS data do not specify whether CDI is community- associated or hospital-acquired; therefore, our results are likely a mixture of CA-CDI and HA-CDI. The higher CDI RORs associated with clindamycin, penicillin combinations, and carbapenems may be due to their activity against anaerobes and disruption of gut flora.25 Clindamycin had the highest CDI ROR in our study, which is consistent with the highest

CDI risks associated with clindamycin in prior meta-analyses.17, 18 Piperacillin- tazobactam had the second highest ROR in our study; the reasons might include the broad-spectrum antimicrobial activity of piperacillin-tazobactam and the great extent of gut flora disruption as a result.26, 27 Trimethoprim-sulfonamides had the lowest CDI ROR among the antibiotic classes included in our study. In previous meta-analyses, trimethoprim-sulfonamides also had one of the lowest CDI risks.17-19

16 Our results demonstrated that fluoroquinolones had a weaker association with CDI compared with most of the antibiotic classes included in the study, except for trimethoprim-sulfonamides. Prior meta-analyses have implicated fluoroquinolones as one of the highest risk antibiotics for CDI;17, 18 however, these studies used data during the CDI epidemic that was associated with the fluoroquinolone-resistant ribotype 027 Clostridium difficile strain.28, 29 A more recent meta-analysis by Vardakas et al. did not implicate fluoroquinolones as one of the highest risk antibiotics, which is consistent with our findings.30 Given that ribotype 027 strains are now endemic in healthcare settings, our data suggest that fluoroquinolones might not be as important of a CDI risk factor as before considering the recent changes in CDI epidemiology.31 A recent article published in 2017 demonstrated that a concomitant decline in inpatient fluoroquinolone use and the NAP1/027 strain may have contributed to the decrease in the incidence rate of long-term- care facility-onset CDI from 2011 to 2015.32 Our results from the Cochran-Armitage

Trend Test also indicated that there was a trend of decrease in CDI risk with fluoroquinolones from 2004 to 2017. In the subgroup analysis, the CDI ROR rank order in both subgroups (< 65 years old and ≥ 65 years old) were similar to that in all patients. Our results showed that older patients had a higher CDI ROR among most of the antibiotic classes analyzed (Figure

3.2). It is known that CDI risk is higher in patients 65 years or older.33 Knowledge of the CDI risk associated with antibiotic classes has important implications for antimicrobial stewardship. Therapeutic interchanges could be identified, especially for those patients who have a high baseline risk for CDI (e.g., elderly, frequent hospitalizations, and comorbid conditions). For example, to treat non-severe purulent skin and skin structure infections in patients with a high risk of CDI, trimethoprim-

17 sulfamethoxazole could be preferred to clindamycin, considering the much lower CDI ROR of trimethoprim-sulfamethoxazole.34

LIMITATIONS

A causal relationship between a drug and an adverse drug reaction (ADR) cannot be established by FAERS. The spontaneous and voluntary reporting of ADRs may lead to significant bias due to underreporting and lack of overall drug use data.35, 36 The association between a drug and an ADR is confounded by concomitant drugs and comorbidities. Media attention and recent drug approval might affect the reporting behaviors. Furthermore, epidemiological shift in the circulating Clostridium difficile strains in the United States might account for the weaker association between fluoroquinolones and CDI in our study; however, the FAERS study design does not permit us to investigate this hypothesis. Therefore, we believe the next step in this line of research will be to confirm these findings in a future case-control or cohort study.

CONCLUSIONS

All antibiotic classes evaluated in the study were significantly associated with

CDI. Lincosamides (e.g., clindamycin) had the highest CDI ROR and trimethoprim- sulfonamides had the lowest CDI ROR of all the antibiotic classes investigated in this study. Results from FAERS should be interpreted with caution in the context of data limitations. Antibiotic stewardship is needed to prevent CDI and to improve health outcomes.

18 Chapter Four: Torsades de pointes and QT Prolongation Associations with Antibiotics

The study described in this chapter has been published as a research article:

Teng C, Walter EA, Gaspar DKS, Obodozie-Ofoegbu OO, Frei CR. Torsades de pointes and QT prolongation associations with antibiotics: a pharmacovigilance study of the FDA Adverse Event Reporting System. Int J Med Sci. 2019;16(7):1018-22.

Authors’ contributions:

Study concept and design: Teng, Walter, and Frei. Statistical analysis: Teng.

Interpretation of data: Teng and Frei. Drafting of the manuscript: Teng and Gaspar.

Critical revision of the manuscript for important intellectual content: All authors. Study supervision: Frei.

This study has also been presented as a poster:

Teng C, Gaspar DKS, Frei CR. Torsade de pointes/QT prolongation risks with antibiotics: A contemporary analysis of the FDA Adverse Event Reporting System.

Translational Science 2019, Washington, DC, March 5-8, 2019.

INTRODUCTION

Drug-induced QT interval prolongation (QTP) is able to cause Torsades de pointes (TdP), a potentially fatal ventricular arrhythmia.37 The risk of TdP/QTP must be considered when selecting antibiotic therapy. In 2010, a study evaluated the risk of TdP with antibiotics using the United States FDA Adverse Event Reporting System (FAERS) and identified macrolides, fluoroquinolones, and linezolid as TdP agents.38 Macrolides 19 and fluoroquinolones are known to cause QTP via blockade of the rapidly activating delayed rectifier potassium channel (hERG/IKr channel).39-43 Linezolid has been associated with TdP;38 however, a double-blind placebo-controlled four-way crossover study with 40 healthy subjects found that linezolid had no effect on the QT interval itself.44 An observational cohort study of 1,270 patients indicated that beta-lactamase inhibitors were associated with QTP.45 Ceftriaxone, when used with lansoprazole, was significantly associated with QTP in a study of FAERS and electronic health records.46 A case report stated that imipenem-cilastatin and piperacillin-tazobactam caused hypokalemia leading to TdP in a patient.47 In this study, we investigated FAERS to analyze the association between TdP/QTP and common antibiotic agents, including macrolides, fluoroquinolones, oxazolidinones, penicillins, carbapenems, cephalosporins, aminoglycosides, metronidazole, and glycopeptide antibiotics.

METHODS

Data Source

FAERS is a publicly available database, which is composed of adverse event reports that were submitted to United States Food and Drug Administration (FDA).1 FAERS data contain drug information (drug name, active ingredient, route of administration, the drug’s reported role in the event) and reaction information. Each report has a primary suspected drug with one or more adverse drug reactions (ADR) and may include other drugs taken by the patient.

20 Study Design

FAERS data from January 1, 2015 to December 31, 2017 were included in the study. Some reports were submitted to FDA multiple times with updated information. Therefore, duplicate reports were removed by case number, with only the most recently submitted version included in the study. Reports containing drugs which were administered in oral, intramuscular, subcutaneous, intravenous, and parenteral routes were included in the study, while other routes of administration were excluded.

Drug Exposure Definition

Each antibiotic was identified in FAERS by generic and brand names listed in the Drugs@FDA Database.20, 21 Drugs with a reported role coded as “PS” (Primary Suspect Drug) or “SS” (Secondary Suspect Drug) were evaluated for inclusion.22 Antibiotics with less than three TdP/QTP reports were excluded from data analysis.23

Adverse Drug Reaction Definition

FAERS defines ADRs using Preferred Terms (PT) from the Medical Dictionary for Regulatory Activities (MedDRA).24 Preferred Terms “Electrocardiogram QT prolonged”, “Long QT syndrome”, and “Torsade de pointes” were used to identify TdP/QTP cases.

Statistical Analysis

A disproportionality analysis was conducted by computing Reporting Odds Ratios (ROR) and corresponding 95% confidence intervals (95%CI) for the association between

21 TdP/QTP and each antibiotic class or individual antibiotic.2 ROR was calculated as the ratio of the odds of reporting TdP/QTP versus all other ADRs for a given drug, compared with this reporting odds for all other drugs present in FAERS.2 An association was considered to be statistically significant if the lower limit of 95%CI was above 1.0.2 An adjusted ROR was calculated after removing reports of potentially confounding antiarrhythmic drugs from the data analysis. These drugs include amiodarone, azimilide, disopyramide, dofetilide, flecainide, ibutilide, mexiletine, propafenone, propranolol, quinidine, and sotalol. Data analysis was performed using Microsoft Access 2016,

Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA), and JMP Pro 13.2.1 (SAS Institute, Cary, NC).

RESULTS

After applying inclusion and exclusion criteria and removing duplicate reports,

FAERS contained 2,042,801 reports from January 1, 2015 to December 31, 2017. There were 3,960 TdP/QTP reports from the study period, which were included in the data analysis. Females accounted for 60% of TdP/QTP reports. TdP/QTP patients had a median age (IQR, interquartile range) of 55 (34) years. Macrolides had the highest TdP/QTP ROR among all antibiotics in the study. Of the 4,092 reports associated with macrolides, 108 reports were associated with TdP/QTP.

The RORs for agents significantly associated with TdP/QTP were: macrolides 14.32 (11.80-17.38), linezolid 12.41 (8.52-18.08), amikacin 11.80 (5.57-24.97), imipenem- cilastatin 6.61 (3.13-13.94), fluoroquinolones 5.68 (4.78-6.76), penicillin combinations 3.42 (2.35-4.96), and ceftriaxone 2.55 (1.41-4.62) (Figure 4.1).

22 Figure 4.1: Reporting Odds Ratios (RORs) for Torsades de pointes/ QT prolongation with antibiotics.

† CI = confidence interval; TdP/QTP = Torsades de pointes/QT prolongation

An adjusted ROR was performed to exclude reports among patients who were taking concomitant antiarrhythmic agents. This was done to reduce confounding variables that may also contribute to TdP/QTP. The adjusted RORs for agents significantly associated with TdP/QTP were: macrolides 13.02 (10.63-15.95), linezolid 12.57 (8.63-18.32), amikacin 12.24 (5.78-25.91), imipenem-cilastatin 6.77 (3.21-14.28), fluoroquinolones 5.50 (4.60-6.56), penicillin combinations 2.99 (2.00-4.48), and ceftriaxone 2.60 (1.44-4.71) (Figure 4.2). The ROR association rank did not differ when adjusted to exclude antiarrhythmic agents.

23 Figure 4.2: Adjusted Reporting Odds Ratios (RORs) for Torsades de pointes/QT prolongation with antibiotics.

† CI = confidence interval; TdP/QTP = Torsades de pointes/QT prolongation

Amikacin was associated with a total of seven TdP/QTP reports. In these reports, amikacin was the secondary suspect drug of TdP/QTP, while the primary suspect drugs were bedaquiline, clofazimine, linezolid, and ciprofloxacin. Piperacillin-tazobactam was associated with a total of seventeen TdP/QTP reports. In these reports, three reports had Clostridium difficile colitis and seven reports had electrolyte abnormalities.

DISCUSSION

Our study found a significantly higher ROR for TdP/QTP as compared to all other adverse events for these antibiotics, which were (ROR from highest to lowest) 24 azithromycin, erythromycin, linezolid, amikacin, moxifloxacin, clarithromycin, ofloxacin, imipenem-cilastatin, piperacillin-tazobactam, ciprofloxacin, levofloxacin, ceftriaxone, and amoxicillin-clavulanate. An FAERS study published in 2010 indicated significant TdP associations (from strongest to weakest) with moxifloxacin, levofloxacin, erythromycin, ciprofloxacin, gatifloxacin, clarithromycin, azithromycin, and linezolid.38 Both studies showed TdP associations with macrolides, fluoroquinolones, and linezolid.

The 2010 FAERS study only included TdP in their data analysis while our study included not only TdP but also QTP. Since QTP is a precursor of TdP, including QTP increases sensitivity of signal detection. The 2010 FAERS study included drugs administered through all routes, including topical routes, which may have limited systemic absorption and are less likely to cause TdP. Our study only included drugs administered in oral, subcutaneous, intramuscular, intravenous, and parenteral routes and excluded other routes of administration, such as topical routes. Therefore, TdP/QTP events in our study are more likely to be caused by a drug than those in the 2010 FAERS study.38 Our study confirmed previously known TdP/QTP associations with macrolides, linezolid, imipenem-cilastatin, fluoroquinolones, penicillin combinations, and ceftriaxone.38, 45-47 Penicillin combinations have a high incidence of diarrhea, and diarrhea may lead to electrolyte abnormalities, which are significant risk factors for

TdP/QTP. In our study, out of seventeen piperacillin-tazobactam TdP/QTP reports, three reports had Clostridium difficile colitis and seven reports had electrolyte abnormalities.

Amikacin was found to be associated with TdP/QTP in our study, which was not reported in the literature. However, amikacin was the secondary suspect drug in all TdP/QTP reports, while the primary suspect drugs were known to be associated with TdP/QTP. Amikacin might play a role in TdP/QTP, but the causal relationship is not warranted.

25 LIMITATIONS

A causal relationship between a drug and an ADR cannot be determined by FAERS. Significant bias may occur because of the spontaneous and voluntary reporting of ADRs.35, 36 Media attention for a particular ADR might affect the reporting behaviors. The association between a drug and an ADR is confounded by comorbid diseases and concomitant drugs. For example, concomitant QT-prolonging drugs, such as ondansetron, antidepressants, antipsychotics, methadone, arsenic, and azole antifungals, are confounders when studying the associations between TdP/QTP and antibiotics. Diarrhea is also a potential confounder because diarrhea may lead to electrolyte abnormalities, which may cause TdP/QTP. Antibiotics, such as penicillin combinations and fluoroquinolones, cause diarrhea in many patients. The higher TdP/QTP ROR for penicillin combinations might be due to their ability of causing diarrhea. The higher TdP/QTP ROR for fluoroquinolones might be due to a combination of their ability of causing diarrhea and their blockade of hERG/IKr channel.

CONCLUSIONS

This study confirms prior evidence for significant TdP/QTP associations with macrolides, linezolid, imipenem-cilastatin, fluoroquinolones, penicillin combinations, and ceftriaxone. This study also discovers a new association between amikacin and TdP/QTP. Results obtained from FAERS should be interpreted with caution in the context of data limitations. Antibiotic stewardship is needed to prevent TdP/QTP and to improve health outcomes.

26 Chapter Five: Acute Kidney Injury Associations with Antibiotics

The study described in this chapter has been submitted as a manuscript to a journal and is currently under review:

Patek TM, Teng C, Kennedy KE, Alvarez CA, Frei CR. Comparing acute kidney injury associations among antibiotics: a pharmacovigilance study of the FDA Adverse

Event Reporting System (FAERS).

Authors’ contributions:

Study concept and design: Teng, Frei. Statistical analysis: Teng. Interpretation of data: All authors. Drafting of the manuscript: Patek, Teng. Critical revision of the manuscript for important intellectual content: All authors. Study supervision: Frei.

This study will also be presented as a poster:

Patek TM, Teng C, Kennedy KE, Frei CR. Comparing acute kidney injury risk among antibiotic classes: a study of the FDA Adverse Event Reporting System (FAERS).

INTRODUCTION

Acute kidney injury (AKI) is defined as the rapid decrease in the kidney’s excretory function, with the retention of nitrogenous (urea and creatinine) and non- nitrogenous waste products.48 AKI occurs in about 20% of hospitalized patients and more than half of critically ill patients.49, 50 Following AKI, patients face a greater risk of chronic kidney disease, dialysis dependence, and higher mortality.51, 52

27 Medications are one of the leading contributors to AKI among hospitalized patients.53 Among those medications, antibiotics are known to be associated with AKI. Common antibiotic classes associated with AKI include colistin, aminoglycosides, beta- lactams, and vancomycin.54 Drug-drug interactions leading to AKI are a great concern with antibiotic therapy. A meta-analysis demonstrated that vancomycin plus piperacillin-tazobactam combination therapy had higher odds of AKI than vancomycin monotherapy, piperacillin-tazobactam monotherapy, and vancomycin plus cefepime or carbapenem combination therapy.55

A previous study using FDA Adverse Event Reporting System (FAERS) demonstrated that vancomycin, trimethoprim-sulfamethoxazole, piperacillin-tazobactam, and ciprofloxacin were significantly associated with AKI.56 However, the study did not mention less commonly used colistin nor aminoglycosides as being associated with AKI. Due to the rise in infections caused by multidrug-resistance (MDR) bacteria, prescribing of these nephrotoxic antibiotics have increased.57 A retrospective study conducted in Spain reported on the increased use of colistin in the treatment of MDR gram-negative bacteria infections, specifically, Acinetobacter baumannii.58 Currently, colistin is being used to treat pediatric and adult cystic fibrosis caused by resistant Pseudomonas aeruginosa, along with MDR gram-negative bacteria contributing to ventilator-associated pneumonia and bacteremia.59 This poses a concern for clinicians when deciding the most appropriate and safe medication regimen for a patient who has failed other antibiotic therapy and at risk for developing AKI. To our knowledge, no studies using FAERS have specifically looked at antibiotics and compared their association with AKI. The objective of this study was to evaluate the association between AKI and many available antibiotics using FAERS.

28 METHODS

Data Source

FAERS is a publicly available database, which is composed of adverse event reports submitted to United States Food and Drug Administration (FDA) by healthcare professionals, consumers, and manufacturers.1 FAERS data contain drug information (drug name, active ingredient, route of administration, the drug’s reported role in the event) and reaction information. Each report has a primary suspected drug with one or more adverse drug reactions (ADR) and may include other drugs taken by the patient. The Institutional Review Board (IRB) of The University of Texas Health Science Center at San Antonio determined that this study does not require IRB approval.

Study Design

FAERS data from January 1, 2015 to December 31, 2017 were included in the study. Some reports were submitted to the FDA multiple times with updated information. Therefore, duplicate reports were removed by case number, with only the most recently submitted version included in the study. Reports containing drugs which were administered in oral, intramuscular, subcutaneous, intravenous, and parenteral routes were included in the study, while other routes of administration were excluded.

Drug Exposure Definition

Each antibiotic was identified in FAERS by generic and brand names listed in the Drugs@FDA Database.20, 21 Drugs with a reported role coded as “PS” (Primary Suspect

29 Drug) or “SS” (Secondary Suspect Drug) were evaluated for inclusion.22 Antibiotics with less than three ADR reports were excluded from data analysis.23

Adverse Drug Reaction Definition

FAERS defines ADRs using Preferred Terms (PT) from the Medical Dictionary for Regulatory Activities (MedDRA).24 Preferred Term “Acute kidney injury” was used to identify AKI cases.

Statistical Analysis

A disproportionality analysis was conducted by computing Reporting Odds Ratios (ROR) and corresponding 95% confidence intervals (95%CI) for the association between AKI and each antibiotic class or individual antibiotic.2 ROR was calculated as the ratio of the odds of reporting AKI versus all other ADRs for a given drug, compared with this reporting odds for all other drugs present in FAERS.2 An association was considered to be statistically significant if the lower limit of 95%CI was above 1.0.2 A higher ROR suggested a stronger association between the antibiotic and AKI. Data analysis was performed using Microsoft Access 2016, Microsoft Excel 2016 (Microsoft Corporation,

Redmond, WA), and JMP Pro 13.2.1 (SAS Institute, Cary, NC).

RESULTS

A total of 2,042,801 reports (including 20,138 AKI reports) were considered, after inclusion criteria were applied. Colistin had the greatest proportion of AKI reports, representing 25% of all colistin reports. AKI RORs (95%CI) for antibiotics were (in 30 descending order): colistin 33.10 (21.24-51.56), aminoglycosides 17.41 (14.49-20.90), vancomycin 15.28 (13.82-16.90), trimethoprim-sulfamethoxazole 13.72 (11.94-15.76), penicillin combinations 7.95 (7.09-8.91), clindamycin 6.46 (5.18-8.04), cephalosporins 6.07 (5.23-7.05), daptomycin 6.07 (4.61-7.99), macrolides 3.60 (3.04-4.26), linezolid 3.48 (2.54-4.77), carbapenems 3.31 (2.58-4.25), metronidazole 2.55 (1.94-3.36), tetracyclines 1.73 (1.26-2.36), and fluoroquinolones 1.71 (1.49-1.97) (Figure 5.1). AKI

RORs (95%CI) for vancomycin combinations were (in descending order): vancomycin plus cefepime 24.31 (15.26-38.73), vancomycin plus piperacillin-tazobactam 22.20

(17.25-28.56), vancomycin plus cefepime or carbapenem 7.40 (5.00-10.94), and vancomycin plus carbapenem 2.04 (0.91-4.59).

31 Figure 5.1: Reporting Odds Ratios (RORs) for AKI with antibiotics.

† CI = confidence interval; AKI = acute kidney injury

DISCUSSION

This study identified 14 antibiotic classes significantly associated with AKI in FAERS (Figure 5.1). Of the macrolides and fluoroquinolones, only azithromycin and moxifloxacin were not significantly associated with AKI. Levofloxacin was found to 32 have a ROR and its 95%CI less than 1, indicating that the odds of reporting levofloxacin with AKI in FAERS was significantly less than that of other drugs. The mechanisms to explain how certain antibiotics can cause AKI are not fully established. Colistin-induced kidney injury is thought to be primarily caused by damage to proximal tubule cells which ultimately lead to apoptosis in the nephron.60 High doses of vancomycin that result in an increased plasma trough concentration and extended duration are risk factors for causing nephrotoxicity. Experimental studies describe the mechanism of vancomycin induced kidney injury as increased production of free oxygen radicals which leads to mitochondrial dysfunction and eventually, cellular apoptosis.61 Aminoglycosides are preferentially taken up by proximal tubular epithelial cells and alter phospholipid metabolism which leads to necrosis. Aminoglycosides also cause renal vasoconstriction which can contribute to AKI.62 This study indicated that colistin had the highest ROR among antibiotics with

AKI followed by aminoglycosides, vancomycin, and trimethoprim-sulfamethoxazole. Consistent with a recently published article that analyzed AKI reports and associated medications in FAERS, vancomycin, piperacillin-tazobactam, aminoglycosides, and trimethoprim-sulfamethoxazole had some of the highest AKI RORs among the antibiotics evaluated in this study.56 However, there are differences this study found that are significant to address. In contrast to Welch et al. article, this study compared the ROR for each drug’s association with AKI, while the previous article compared total number of

AKI reports.56 According to the previous article’s methodology, rarely utilized antibiotics would not be considered, such as colistin. Colistin was excluded from the previously published study, which may project a harmful message that it is not commonly associated with AKI.56

33 A retrospective analysis of 50 patients and a case report demonstrated clindamycin-induced AKI.63, 64 A case control study found an increased risk of AKI associated with the use of fluoroquinolones.65 Our study was able to confirm these findings as well as rank the antibiotics by their ROR for AKI reports in FAERS. This study also looked at the significance of drug-drug interactions associated with increased risk of AKI. Consistent with previous literature, this study found that higher odds of AKI were reported when vancomycin plus piperacillin-tazobactam combination therapy was given compared with vancomycin monotherapy, piperacillin- tazobactam monotherapy, and vancomycin plus cefepime or carbapenem combination therapy.55 Luther et al. assessed vancomycin plus cefepime or carbapenem combination therapy as a whole, but did not assess vancomycin plus cefepime combination therapy and vancomycin plus carbapenem combination therapy individually.55 In our study, vancomycin plus cefepime combination therapy and vancomycin plus carbapenem combination therapy were assessed individually. AKI ROR for vancomycin plus cefepime combination was slightly higher than that for vancomycin plus piperacillin- tazobactam combination. AKI ROR for vancomycin plus carbapenem combination was the lowest. This finding is important because vancomycin plus piperacillin-tazobactam and vancomycin plus cefepime are common empiric antibiotic combinations used in hospitals to help treat severe infections. This study was able to compare classes of antibiotics, rank individual antibiotics by their ROR, and confirm that combination antibiotic therapy increased the odds of a patient experiencing AKI. Ideally, this study will help guide the prescribing of certain antibiotics in patients who are more susceptible to developing AKI.

34 LIMITATIONS

A causal relationship between a drug and an ADR cannot be established by FAERS. Significant bias may occur because of the spontaneous and voluntary reporting of ADRs.35, 36 Media attention and recent publication of an ADR in the literature might affect reporting behaviors. The association between a drug and an ADR is confounded by comorbid conditions and concomitant drugs. In spite of the limitations, FAERS has a large sample size and is suitable for discovering new and rare drug-ADR associations and drug-drug interactions.

CONCLUSION

This study found 14 classes of antibiotics that were significantly associated with AKI. Colistin had the highest AKI ROR among the antibiotics evaluated in this study. While this study confirmed previous literature of certain antibiotics associated with AKI, it also compared antibiotic classes and analyzed RORs for drug-drug interactions. Results obtained from FAERS should be interpreted with caution in the context of data limitations. Antibiotic stewardship is needed to prevent AKI and to improve health outcomes.

35 Chapter Six: Hypoglycemia Associations with Antibiotics

The study described in this chapter has been submitted as a manuscript to a journal and is currently under review:

Kennedy KE, Teng C, Patek TM, Frei CR. Hypoglycemia associated with antibiotics alone and in combination with sulfonylureas and meglitinides: an epidemiologic surveillance study of the FDA Adverse Event Reporting System (FAERS).

Authors’ contributions:

Study concept and design: Teng, Frei. Statistical analysis: Teng. Interpretation of data: All authors. Drafting of the manuscript: Kennedy, Frei, Teng. Critical revision of the manuscript for important intellectual content: All authors. Study supervision: Frei.

This study will be presented as a poster:

Kennedy KE, Teng C, Patek TM, Frei CR. Hypoglycemia risk with antibiotics: an epidemiologic surveillance study of the FDA Adverse Event Reporting System

(FAERS).

INTRODUCTION

In July of 2018, the FDA published a drug safety warning for the potential risk of developing hypoglycemia from fluoroquinolones.66 Even though hypoglycemia is a common adverse drug reaction (ADR) of insulin and sulfonylureas, it can become serious and lead to coma, seizure, life-threatening arrhythmias, myocardial infarction, and death.67 It is important to know if an antibiotic could increase the risk of developing

36 hypoglycemia, especially in patients at risk of developing hypoglycemia, such as patients with diabetes, and especially those taking glucose-lowering medications. Other antibiotics have been reported to be associated with hypoglycemia in the literature. A study of FAERS demonstrated a relationship between linezolid and hypoglycemia.68 Another FAERS study demonstrated that tigecycline was associated with hypoglycemia. As a result, the FDA added a warning with this information to the tigecycline package insert.69 In a cohort study of insurance claims, diabetic patients on three oral fluoroquinolones, levofloxacin, ciprofloxacin, or moxifloxacin, had higher odds of experiencing hypoglycemia than those on two macrolides (clarithromycin or azithromycin). This study also demonstrated that patients taking moxifloxacin had a higher risk of hypoglycemia than those taking ciprofloxacin.70 In a retrospective chart review study, dysglycemia occurred more frequently in patients receiving levofloxacin or ciprofloxacin than those receiving ceftriaxone.71 A few case reports have also been published that report hypoglycemia from doxycycline and trimethoprim- sulfamethoxazole use.72-77 Lastly, two case reports demonstrated that cefditoren is associated with hypoglycemia.78, 79 However, very few of these studies accounted for the concomitant use of glucose-lowering medications when assessing risk of hypoglycemia with these antibiotics.

Drug-drug interactions involving antibiotics can put patients at an even greater risk of hypoglycemia. Several antibiotics interact with sulfonylureas to increase the risk of hypoglycemia. A cohort study of insurance claims demonstrated that when glipizide or glyburide were prescribed, clarithromycin, levofloxacin, trimethoprim-sulfamethoxazole, metronidazole, and ciprofloxacin were all associated with higher risks of hypoglycemia than non-interacting antibiotics.80 A case report also demonstrated severe hypoglycemia associated with a clarithromycin-repaglinide drug interaction.81 The clarithromycin drug 37 label warns that the concomitant use of oral hypoglycemic agents and/or insulin with clarithromycin may cause hypoglycemia.82 There is no current literature that has systematically compared antibiotics and the risk of developing hypoglycemia. The objective of this study was to evaluate both the association between antibiotics and hypoglycemia, and the influence of concomitant glucose-lowering medications, using the FDA Adverse Event Reporting System

(FAERS).1

METHODS

Data Source

FAERS is a publicly available database composed of adverse event reports that were submitted to United States Food and Drug Administration (FDA).1 FAERS data contain drug information, including drug name, active ingredient, route of administration, the drug’s reported role in the event, and reaction information. Each report has a primary suspected drug with one or more ADRs and may include other drugs taken by the patient.

Study Design

FAERS data from January 1, 2004 to December 31, 2017 were included in the study. If a report was submitted to the FDA multiple times with updated information, only the most recently submitted version was included in this study. Duplicate reports were also removed by matching age, sex, event date, and reporter country.

38 Drug Exposure Definition

Each antibiotic was identified in FAERS by generic and brand names listed in the Drugs@FDA Database.20, 21 Drugs with a reported role coded as “PS” (Primary Suspect Drug) or “SS” (Secondary Suspect Drug) were evaluated for inclusion.22 Antibiotics with less than three ADR reports were excluded from this data analysis.23

Adverse Drug Reaction Definition

FAERS defines ADRs using Preferred Terms (PT) from the Medical Dictionary for Regulatory Activities (MedDRA).24 The following Preferred Terms: “Blood glucose decreased”, “Hypoglycaemia”, “Hypoglycaemic coma”, “Hypoglycaemic encephalopathy”, “Hypoglycaemic seizure”, “Hypoglycaemic unconsciousness”, and “Shock hypoglycaemic” were used to identify hypoglycemia cases for this study.

Statistical Analysis

A disproportionality analysis was conducted by computing Reporting Odds Ratios (ROR) and corresponding 95% confidence intervals (95%CI) for the association between hypoglycemia and each antibiotic class or individual antibiotic.2 ROR was calculated as the ratio of the odds of reporting hypoglycemia versus all other ADRs for a given drug, compared with this reporting odds for all other drugs present in FAERS.2 An association was considered to be statistically significant if the lower limit of 95% CI was above 1.0.2 A higher ROR suggested a stronger association between the antibiotic and hypoglycemia. An adjusted ROR was calculated after removing reports of potentially confounding sulfonylureas (chlorpropamide, gliclazide, glimepiride, glipizide, glyburide, tolazamide, and tolbutamide) and meglitinides (repaglinide and nateglinide) from the database. Data 39 analysis was performed using Microsoft Access 2016, Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA), SAS 9.4, and JMP Pro 13.2.1 (SAS Institute, Cary, NC).

RESULTS

A total of 2,334,959 reports (including 18,466 hypoglycemia reports) were considered, after inclusion criteria were applied. Cefditoren had the greatest proportion of hypoglycemia reports, representing 10% of all cefditoren reports. Statistically significant hypoglycemia RORs (95% CI) for antibiotics were: cefditoren 14.03 (8.93-22.03), tigecycline 3.32 (1.95-5.65), clarithromycin 2.41 (1.89-3.08), ertapenem 2.07 (1.14-3.75), moxifloxacin 2.06 (1.59-2.65), levofloxacin 1.66 (1.37-2.01), and linezolid 1.54 (1.07- 2.20) (Figure 6.1).

40 Figure 6.1: Reporting Odds Ratios (RORs) for hypoglycemia with antibiotics.

† CI = confidence interval

An adjusted ROR was calculated after removing reports of concomitant sulfonylureas or meglitinides. The adjusted RORs for agents with a significant risk associated with hypoglycemia were: cefditoren 14.25 (9.08-22.39), tigecycline 3.34 (1.96-5.68), ertapenem 1.93 (1.03-3.60), and clarithromycin 1.56 (1.15-2.11) (Figure 6.2).

41 Figure 6.2: Adjusted Reporting Odds Ratios (aRORs) for hypoglycemia with antibiotics in patients not on a sulfonylurea or a meglitinide.

† CI = confidence interval

Another adjusted ROR was calculated to quantify the impact of a drug-drug interaction between clarithromycin and repaglinide. The hypoglycemia ROR (95% CI) for clarithromycin taken with repaglinide was 20.91 (2.52-173.69), while the

42 hypoglycemia ROR (95% CI) for clarithromycin taken without repaglinide was 2.38 (1.86-3.04).

DISCUSSION

Seven antibiotics were found to be significantly associated with hypoglycemia. These antibiotics included: cefditoren, tigecycline, clarithromycin, ertapenem, moxifloxacin, levofloxacin, and linezolid. Four antibiotics were found to be significantly associated with hypoglycemia in patients not concomitantly taking a sulfonylurea or a meglitinide. These antibiotics included: cefditoren, tigecycline, ertapenem, and clarithromycin. While several of the antibiotics have prior studies or case reports indicating an association with hypoglycemia, ertapenem has no previous literature regarding hypoglycemia. Trimethoprim-sulfamethoxazole was not significantly associated with hypoglycemia, even though case reports have reported hypoglycemia as an ADR of trimethoprim-sulfamethoxazole.72-74 Finally, doxycycline was not associated with developing hypoglycemia, despite some prior case reports to the contrary.75-77 Potential mechanisms and implications for clinical practice are discussed below. Concerning the fluoroquinolones, a retrospective cohort study that evaluated the risk of developing hypoglycemia with gatifloxacin, levofloxacin, ciprofloxacin, and azithromycin found only gatifloxacin and levofloxacin had a significantly greater risk of developing hypoglycemia.83 Before that study was published, gatifloxacin was removed from the market due to an increased risk of dysglycemia.84 The July 2018 FDA drug safety warning required all drugs in the fluoroquinolone class to have additional labeling that warned patients and providers of the risk of developing hypoglycemia.66 However, the FDA drug safety warning was based on only 56 reports in FAERS. Levofloxacin accounted for 79% of the total reports. Forty-seven of these patients were found to be

43 diabetic patients. Thirty-five of the 47 diabetic patients were also taking a sulfonylurea medication. However, in our study, only levofloxacin and moxifloxacin were significantly associated with hypoglycemia, and ciprofloxacin was not. Furthermore, when patients on sulfonylureas or meglitinides were removed from the sample, there was no significant association for any of the fluoroquinolones and hypoglycemia. The mechanisms of how antibiotics could cause hypoglycemia are not fully understood. However, there are some proposed mechanisms. Fluoroquinolones may activate the L-type voltage-dependent Ca2+ channel.85 Fluoroquinolones may also increase insulin secretion by inhibiting the K+ATP channels in pancreatic beta cells.86 Sulfonylureas stimulate insulin secretion by inhibiting the K+ATP channels in pancreatic beta cells.87 This compounded mechanism could explain why hypoglycemia was commonly seen in patients taking fluoroquinolones with a concomitant sulfonylurea. Cefditoren had the highest ROR of 14.03, with a confidence interval of 8.93-

22.03. Cefditoren is a cephalosporin that contains pivalic acid, and antibiotics containing a pivalic acid moiety have been shown to decrease serum carnitine concentration. Severe carnitine deficiency can lead to the inability to produce glucose, which could lead to hypoglycemia.78 Finally, clarithromycin interacts with several medications, including repaglinide, an oral glucose-lowering medication. Clarithromycin is a CYP3A4 inhibitor, and repaglinide is metabolized by CYP3A4. Even low doses of clarithromycin can increase the plasma concentration of repaglinide. This increased plasma concentration can lead to severe hypoglycemia.88 This finding prompted us to calculate the ROR of hypoglycemia in patients taking clarithromycin and repaglinide concomitantly. The hypoglycemia ROR for clarithromycin taken with repaglinide was 20.91, with a confidence interval of 2.52- 173.69, while hypoglycemia ROR for clarithromycin taken without repaglinide was 2.38, 44 with a confidence interval of 1.86-3.04. Clarithromycin also interacts with sulfonylureas and may increase the sulfonylurea concentration by inhibiting P-glycoprotein in the intestinal wall.89 Clinicians should consider avoiding antibiotics with high hypoglycemia ROR including cefditoren, tigecycline, clarithromycin, ertapenem, moxifloxacin, levofloxacin, and linezolid, in patients at a high risk for developing hypoglycemia, including those on glucose-lowering medications, such as sulfonylureas and meglitinides. There are important study limitations to acknowledge, many of which are well known with regard to the FAERs database. A causal relationship between a drug and an ADR cannot be determined by FAERS. Significant bias may occur, because of the spontaneous and voluntary reporting of ADRs.35, 36 Also, the severity of the ADR is not reported through FAERS. Media attention and recent publication of an ADR in the literature might affect the reporting behaviors. Despite the limitations, FAERS has a large sample size and is suitable for discovering new and rare drug-ADR associations.

CONCLUSIONS

Many patients on antibiotics, including fluoroquinolones, are at increased risk of hypoglycemia when they are also taking sulfonylureas or meglitinides. Patients on cefditoren, tigecycline, ertapenem, and clarithromycin are at increased risk for hypoglycemia even if they are not taking sulfonylureas or meglitinides. The association between ertapenem and hypoglycemia has not been previously reported. Finally, patients taking concomitant clarithromycin and repaglinide are at much greater risk of hypoglycemia than those taking clarithromycin without repaglinide.

45 Chapter Seven: Rhabdomyolysis Associations with Antibiotics

The study described in this chapter has been submitted as a manuscript to

International Journal of Medical Sciences and has been accepted by the journal:

Teng C, Baus C, Wilson JP, Frei CR. Rhabdomyolysis associations with antibiotics: a pharmacovigilance study of the FDA Adverse Event Reporting System

(FAERS). Int J Med Sci.

Authors’ contributions:

Study concept and design: Teng, Frei. Statistical analysis: Teng. Interpretation of data: All authors. Drafting of the manuscript: Teng. Critical revision of the manuscript for important intellectual content: All authors. Study supervision: Frei.

INTRODUCTION

Rhabdomyolysis is a clinical syndrome of skeletal muscle injury associated with myoglobinuria, electrolyte abnormalities, and acute kidney injury.90 The severity of rhabdomyolysis can range from an asymptomatic elevation in creatinine kinase (CK) to severe life-threatening symptoms associated with high CK levels and renal failure. The most common presenting symptoms include myalgia, weakness, and tea-colored urine.

Some antibiotics have been shown to play a role in developing rhabdomyolysis. Multiple case reports have shown an association between daptomycin and rhabdomyolysis.91-93 The use of macrolides alone or in combination with statins was reported to be related to rhabdomyolysis.94, 95 A review of 53 cases showed rhabdomyolysis associated with concomitant use of statins and azithromycin.96 46 Additionally, case reports indicated that trimethoprim-sulfamethoxazole induced rhabdomyolysis occurred in both immunocompetent and immunocompromised patients.97, 98 There have been case reports linking the use of linezolid to serious cases of rhabdomyolysis.99, 100 Fluoroquinolones have also been shown to be associated with rhabdomyolysis in case reports when fluoroquinolones were used alone or in combination with statins.101, 102 Rhabdomyolysis is listed as an adverse drug reaction (ADR) of cefdinir in its United States Food and Drug Administration (FDA) label.103 Rhabdomyolysis was also associated with certain medications, such as statins.104

Antibiotics are frequently prescribed home medications, so it is imperative to know which classes and specific agents within classes are more likely to cause rhabdomyolysis. Although some antibiotics are known to be associated with rhabdomyolysis, no study has systematically compared rhabdomyolysis associations for common antibiotics. The objective of this study was to evaluate the association between rhabdomyolysis and common antibiotics using the FDA Adverse Event Report System (FAERS).

METHODS

Data Source

FAERS is a publicly available database, which is composed of adverse event reports that were submitted to the FDA.1 FAERS data contain drug information (drug name, active ingredient, route of administration, the drug’s reported role in the event) and reaction information. Each report has a primary suspected drug with one or more ADRs and may include other drugs taken by the patient.

47 Study Design

FAERS data from January 1, 2004 to December 31, 2017 were included in the study. Some reports were submitted to FDA multiple times with updated information. Therefore, duplicate reports were removed by case number, with only the most recently submitted version included in the study. Another step of removing duplicate reports was performed by matching age, sex, event date, and reporter country.

Drug Exposure Definition

Each antibiotic was identified in FAERS by generic and brand names listed in the Drugs@FDA Database.20, 21 Drugs with a reported role coded as “PS” (Primary Suspect Drug) were evaluated for inclusion. Antibiotics with less than three ADR reports were excluded from data analysis.23

Adverse Drug Reaction Definition

FAERS defines ADRs using Preferred Terms (PT) from the Medical Dictionary for Regulatory Activities (MedDRA).24 Preferred Term “Rhabdomyolysis” was used to identify rhabdomyolysis cases. Standardised MedDRA Queries (SMQs) are groupings of

MedDRA terms, usually at the PT level, which relate to an adverse drug reaction. Hepatic disorders (SMQ) was used to identify cases with hepatic disorders. Acute renal failure

(SMQ) and the term “Renal failure acute” were used to identify acute kidney injury cases.

48 Statistical Analysis

A disproportionality analysis was conducted by computing Reporting Odds Ratios (ROR) and corresponding 95% confidence intervals (95%CI) for the association between rhabdomyolysis and each antibiotic class or individual antibiotic or each statin.2 ROR was calculated as the ratio of the odds of reporting rhabdomyolysis versus all other ADRs for a given drug, compared with this reporting odds for all other drugs present in

FAERS.2 An association was considered to be statistically significant if the lower limit of 95%CI was above 1.0.2 A higher ROR suggested a stronger association between the antibiotic and rhabdomyolysis. An adjusted ROR was calculated after removing reports of potentially confounding statins from the data analysis. These drugs include atorvastatin, fluvastatin, lovastatin, pitavastatin, pravastatin, rosuvastatin, and simvastatin. Rhabdomyolysis RORs for meropenem reports with hepatic disorders and without hepatic disorders were calculated. Rhabdomyolysis RORs for piperacillin- tazobactam reports with acute kidney injury and without acute kidney injury were computed. Data analysis was performed using Microsoft Access 2016, Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA), SAS 9.4, and JMP Pro 13.2.1 (SAS Institute, Cary, NC).

RESULTS

A total of 2,334,959 reports (including 7,685 rhabdomyolysis reports) were considered, after inclusion criteria were applied. Daptomycin had the greatest proportion of rhabdomyolysis reports, representing 5.5% of all daptomycin reports. Statistically significant rhabdomyolysis RORs (95% CI) for antibiotics were (in descending order): daptomycin 17.94 (14.08-22.85), cefditoren 8.61 (3.54-20.94), cefaclor 7.16 (2.28-22.49),

49 erythromycin 5.93 (3.17-11.10), norfloxacin 4.50 (1.44-14.07), clarithromycin 3.95 (2.77-5.64), meropenem 3.19 (1.51-6.72), azithromycin 2.94 (1.96-4.39), cefdinir 2.84 (1.06-7.62), piperacillin-tazobactam 2.61 (1.48-4.61), trimethoprim-sulfamethoxazole 2.53 (1.52-4.21), linezolid 2.49 (1.47-4.21), ciprofloxacin 2.10 (1.51-2.92) (Figure 7.1). Meropenem had seven rhabdomyolysis reports, including four reports with hepatic disorders. Rhabdomyolysis RORs (95% CI) for meropenem reports with hepatic disorders and without hepatic disorders were 9.77 (3.61-26.46) and 1.68 (0.54-5.23), respectively. Cefditoren had five rhabdomyolysis reports, including one report of 5-year- old girl with decreased carnitine level. Cefaclor had three rhabdomyolysis reports, including two reports on concomitant Nonsteroidal Anti-Inflammatory Drugs (NSAIDS), which were indomethacin and flurbiprofen. Piperacillin-tazobactam had twelve rhabdomyolysis reports, including three reports with acute kidney injury. Rhabdomyolysis RORs (95% CI) for piperacillin-tazobactam reports with acute kidney injury and without acute kidney injury were 4.91 (1.57-15.37) and 2.26 (1.17-4.35), respectively.

50 Figure 7.1: Reporting Odds Ratios (RORs) for rhabdomyolysis with antibiotics.

† CI = confidence interval

An adjusted ROR was calculated after removing reports of concomitant statins. The adjusted RORs for agents with a significant association with rhabdomyolysis were: daptomycin 14.38 (10.85-19.05), cefditoren 8.81 (3.62-21.44), cefaclor 7.27 (2.31-22.85), meropenem 3.42 (1.62-7.21), erythromycin 3.20 (1.32-7.72), cefdinir 2.93 (1.10-7.86), piperacillin-tazobactam 2.56 (1.41-4.63), clarithromycin 2.51 (1.58-3.99), azithromycin

2.43 (1.55-3.82), trimethoprim-sulfamethoxazole 2.32 (1.34-4.01), linezolid 2.11 (1.16- 3.81), ciprofloxacin 1.80 (1.25-2.59) (Figure 7.2). The ROR association rank had little change when adjusted to exclude statins.

51 Figure 7.2: Adjusted Reporting Odds Ratios (aRORs) for rhabdomyolysis with antibiotics.

† CI = confidence interval

All statins were significantly associated with rhabdomyolysis. RORs (95% CI) for statins were (in descending order): simvastatin 76.29 (70.92-82.06), lovastatin 50.95 (34.59-75.05), pitavastatin 38.92 (27.86-54.35), fluvastatin 24.51 (16.89-35.58), pravastatin 23.66 (17.83-31.40), rosuvastatin 22.94 (20.75-25.37), atorvastatin 15.09 (13.77-16.53).

52 DISCUSSION

Our results demonstrated significant rhabdomyolysis associations (from strongest to weakest) with daptomycin, cefditoren, cefaclor, erythromycin, norfloxacin, clarithromycin, meropenem, azithromycin, cefdinir, piperacillin-tazobactam, trimethoprim-sulfamethoxazole, linezolid, and ciprofloxacin. Our results confirmed previously known rhabdomyolysis associations with daptomycin, macrolides, trimethoprim-sulfamethoxazole, linezolid, fluoroquinolones, and cefdinir.91-103 Meropenem, cefditoren, cefaclor, and piperacillin-tazobactam were found to be associated with rhabdomyolysis, which were not reported in the literature. In a case report, meropenem induced prolonged severe hypokalemia and hypomagnesaemia, leading to chronic muscle weakness.105 Meropenem-induced hypokalemia and hypomagnesaemia might play a role in meropenem-induced muscle injury and rhabdomyolysis. In our study, four out of seven meropenem-associated rhabdomyolysis reports had hepatic disorders. Rhabdomyolysis ROR for meropenem reports with hepatic disorders was higher than those without hepatic disorders. Since meropenem is hepatically metabolized,106 hepatic disorders might be a risk factor for meropenem- induced rhabdomyolysis. A case report demonstrated that cefditoren pivoxil induced hypocarnitinemia and hypoglycemia in a 6-year-old girl with Fukuyama-type congenital muscular dystrophy.78 In our study, a 5-year-old girl on cefditoren had both rhabdomyolysis and decreased carnitine level. Maybe decreased carnitine level was associated with cefditoren-induced rhabdomyolysis. There were three cefaclor-associated rhabdomyolysis reports in our study. Two of these reports had concomitant NSAIDS. NSAIDS are known to be associated with acute kidney injury.107 Since rhabdomyolysis is often associated with acute kidney injury90 and 53 cefaclor is renally excreted, NSAIDS might play a role in cefaclor-induced rhabdomyolysis. In our study, twelve piperacillin-tazobactam reports listed rhabdomyolysis as an ADR. Three of these reports had acute kidney injury as well. Rhabdomyolysis ROR for piperacillin-tazobactam reports with acute kidney injury was higher than those without acute kidney injury. Piperacillin-tazobactam was known to be associated with acute kidney injury.56 Since rhabdomyolysis is commonly associated with acute kidney injury90 and piperacillin-tazobactam is renally excreted, acute kidney injury might be associated with piperacillin-tazobactam-induced rhabdomyolysis.

LIMITATIONS

A causal relationship between a drug and an ADR cannot be determined by FAERS. Significant bias may occur because of the spontaneous and voluntary reporting of ADRs.35, 36 Media attention and recent publication of an ADR in the literature might affect the reporting behaviors. The association between a drug and an ADR is confounded by comorbidities and concomitant drugs. Despite the limitations, FAERS has a large sample size and is suitable for discovering new and rare drug-ADR associations.

CONCLUSIONS

This study confirms prior evidence for rhabdomyolysis associations with daptomycin, macrolides, trimethoprim-sulfamethoxazole, linezolid, fluoroquinolones, and cefdinir. This study also identifies previously unknown rhabdomyolysis association with meropenem, cefditoren, cefaclor, and piperacillin-tazobactam. Results obtained from

54 FAERS should be interpreted with caution in the context of data limitations. Antibiotic stewardship is needed to prevent rhabdomyolysis and to improve health outcomes.

55 Chapter Eight: Conclusions and Future Directions

CONCLUSIONS

This dissertation evaluated the association between antibiotics and certain ADRs, which were CDI, TdP-QTP, AKI, hypoglycemia, and rhabdomyolysis. Several new antibiotic-ADR associations were found, which were associations between amikacin and

TdP-QTP, between ertapenem and hypoglycemia, and between meropenem and rhabdomyolysis. Patients on vancomycin and piperacillin-tazobactam had higher association with AKI than those on vancomycin alone and those on piperacillin- tazobactam alone. The fluoroquinolone class was associated with hypoglycemia, when patients were also taking sulfonylureas or meglitinides. Moxifloxacin and levofloxacin were associated with hypoglycemia, when patients were also taking sulfonylureas or meglitinides. Ciprofloxacin was not associated with hypoglycemia. The association between penicillin combinations and CDI in patients 65 years or older was higher than that in patients less than 65 years old.

STRENGTHS AND LIMITATIONS

The strengths of FAERS study are the large sample size and the real-world setting. There are several limitations of FAERS. A causal relationship between a drug and an ADR cannot be established by using FAERS. Underreporting and lack of overall drug use data may lead to significant bias because ADRs are spontaneously and voluntarily reported. The association between a drug and an ADR is confounded by concomitant drugs and comorbid diseases. Media attention to a particular ADR and recent drug approval may affect reporting behaviors.

56 TRANSLATIONAL APPLICATION

Translation research is focused on applying research findings from laboratory and clinical studies to clinical practice. This work identified new antibiotic-ADR associations, which will help clinicians to select the most appropriate antibiotics for their patients. For example, this work discovered a new association between ertapenem and hypoglycemia. Clinicians should avoid using ertapenem in patients with a high risk of developing hypoglycemia. This work confirmed higher risk of AKI with vancomycin and piperacillin-tazobactam combinations. Clinicians should take caution when choosing this combination. This work indicated that ciprofloxacin was not associated with hypoglycemia, while moxifloxacin and levofloxacin were associated with hypoglycemia when taken with sulfonylureas or meglitinides. Clinicians should choose ciprofloxacin if a fluoroquinolone is needed and the patient is at a high risk of hypoglycemia. This work demonstrated that the association between penicillin combinations and CDI in patients 65 years or older was higher than that in patients less than 65 years old. Clinicians should monitor CDI signs and symptoms if the patient is over 65 years old and is taking a penicillin combination.

FUTURE DIRECTIONS

Future research will be focused on evaluation of new ADR associations discovered by FAERS studies using electronic health records. The biological mechanisms of ADR will be studied using human cellular models and small animals. ADR and drug- drug interactions involving other drugs besides antibiotics will also be studied using FAERS and electronic health records.

57 Appendix

APPENDIX 1. STUDY MEDICATIONS

Generic name Brand name(s) Amikacin Amikin Gentamicin U-gencin, Bristagen, Apogen, Gentafair Tobramycin Nebcin Doripenem Doribax Ertapenem Invanz Imipenem-cilastatin Primaxin Meropenem Merrem Meropenem- Vabomere vaborbactam Cefadroxil Duricef, Ultracef Cefazolin Ancef, Kefzol Cephalexin Keflex, Keflet, Keftab, Panixine Disperdose Cefaclor Ceclor, Raniclor Cefotetan Cefotan Cefoxitin Mefoxin Cefprozil Cefzil Cefuroxime Zinacef, Ceftin, Kefurox Cefdinir Omnicef Cefditoren Spectracef Cefixime Suprax Cefotaxime Claforan Cefpodoxime Vantin, Banan Ceftazidime Fortaz, Ceptaz, Tazidime, Tazicef, Pentacef Ceftazidime- Avycaz avibactam Ceftibuten Cedax Ceftriaxone Rocephin Cefepime Maxipime Ceftaroline Teflaro Ciprofloxacin Cipro, Proquin XR Delafloxacin Baxdela Gemifloxacin Factive Levofloxacin Levaquin Moxifloxacin Avelox Norfloxacin Noroxin 58 Ofloxacin Floxin Dalbavancin Dalvance Oritavancin Orbactiv Telavancin Vibativ Vancomycin Vancocin, Firvanq, Vancoled, Vancor Clindamycin Cleocin Lincomycin Lincocin Daptomycin Cubicin Azithromycin Zithromax, Zmax Clarithromycin Biaxin Erythromycin Robimycin, R-P Mycin, Ilotycin, Ery-tab, Eryc, E-base, PCE, E- mycin, Ilosone, E.E.S., Wyamycin E, Pediamycin, Eryped, Erythrocin, Ethril, Pfizer-E, Bristamycin, Erypar Fidaxomicin Dificid Aztreonam Azactam Nitrofurantoin Furadantin, Ivadantin, Macrodantin, Macrobid, Furalan Fosfomycin Monurol Metronidazole Flagyl, Metro, Protostat, Satric, Metromidol Linezolid Zyvox Tedizolid Sivextro Amoxicillin Amoxil, Larotid, Moxatag, Polymox, Trimox, Utimox, Wymox, Dispermox Ampicillin Amcill, Omnipen-N, Penbritin, Pfizerpen-A, Polycillin, Principen, Probampacin, Totacillin Amoxicillin- Augmentin clavulanate Cloxacillin Tegopen, Cloxapen Dicloxacillin Pathocil, Dynapen, Dycill Nafcillin Unipen, Nallpen Oxacillin Prostaphlin, Bactocill Penicillin G Bicillin, Duracillin A.S., Pentids, Permapen, Pfizerpen Penicillin V Beepen-VK, Betapen-VK, Ledercillin VK, Penapar-VK, Pen- Vee K, Pfizerpen VK, Uticillin VK, V-cillin, Veetids Piperacillin Pipracil Ticarcillin-clavulanate Timentin Ampicillin-sulbactam Unasyn Piperacillin- Zosyn tazobactam Bacitracin Baciim Colistin Coly-mycin S Polymyxin B Aerosporin, Polymycin B Sulfamethoxazole- Septra, Bactrim, Sulfatrim, Sulfamethoprim, Cotrim, 59 trimethoprim Sulmeprim, Uroplus, Trimeth/sulfa Demeclocycline Declomycin Doxycycline Vibramycin, Vibra-tabs, Doryx, Monodox, Periostat, Atridox, Oracea, Doxychel, Doxy, Acticlate, Zenavod, Xyrosa Minocycline Minocin, Arestin, Solodyn, Dynacin, Ximino, Minolira Tetracycline Tetrex, Achromycin, Tetracyn, Tetrachel, Panmycin, Sumycin, Retet, Tetramed, Bristacycline, Cyclopar, Robitet Tigecycline Tygacil

60 Glossary

Acronym Definition ADR Adverse drug reaction AKI Acute kidney injury CA-CDI Community-associated Clostridium difficile infection CDI Clostridium difficile infection CI Confidence interval CK Creatinine kinase CMC Cephalosporins, monobactams, and carbapenems FAERS FDA Adverse Event Reporting System FDA Food and Drug Administration HA-CDI Hospital-acquired Clostridium difficile infection HLGT High Level Group Term HLT High Level Term IC Information Component IQR Interquartile range IRB Institutional Review Board LLT Lowest Level Term MDR Multidrug-resistant MedDRA Medical Dictionary for Regulatory Activities NSAIDS Nonsteroidal Anti-Inflammatory Drugs PPE Palmarplantar erythrodysesthesia PRR Proportional Reporting Ratios PT Preferred Term QTP QT Prolongation ROR Reporting Odds Ratio RRR Relative Reporting Ratios SMQ Standardised MedDRA Queries SOC System Organ Classes TdP Torsades de pointes

61 References

1. Food and Drug Administration. FDA Adverse Event Reporting System (FAERS). Available from https://www.fda.gov/drugs/surveillance/fda-adverse-event- reporting-system-faers. Accessed August 12, 2019. 2. Bate A, Evans SJ. Quantitative signal detection using spontaneous ADR

reporting. Pharmacoepidemiol Drug Saf. 2009;18(6):427-36. 3. Harpaz R, DuMouchel W, Shah NH, Madigan D, Ryan P, Friedman C. Novel data-mining methodologies for adverse drug event discovery and analysis. Clin Pharmacol Ther. 2012;91(6):1010-21. 4. Böhm R. Primer on Disproportionality Analysis. Available from http://openvigil.sourceforge.net/doc/DPA.pdf. Accessed August 12, 2019. 5. Fukuda A, Tahara K, Hane Y, Matsui T, Sasaoka S, Hatahira H, et al.

Comparison of the adverse event profiles of conventional and liposomal formulations of doxorubicin using the FDA adverse event reporting system. PLoS One. 2017;12(9):e0185654. 6. Wittayanukorn S, Qian J, Johnson BS, Hansen RA. Cardiotoxicity in targeted therapy for breast cancer: A study of the FDA adverse event reporting system (FAERS). J Oncol Pharm Pract. 2017;23(2):93-102. 7. Yue Z, Shi J, Li H, Li H. Association between Concomitant Use of Acyclovir or

Valacyclovir with NSAIDs and an Increased Risk of Acute Kidney Injury: Data Mining of FDA Adverse Event Reporting System. Biol Pharm Bull. 2018;41(2):158-62.

62 8. Gupta AK, Carviel J, Gupta MA, Shear NH. Assessing dutasteride-associated sexual dysfunction using the U.S. Food and Drug Administration Adverse Event Reporting System. J Eur Acad Dermatol Venereol. 2018;32(8):1373-6. 9. Serebruany V, Cherepanov V, Fortmann S, Kim MH. Mortality and oral anticoagulants in the Food and Drug Administration Adverse Event Reporting System. Open Heart. 2017;4(2):e000629.

10. Hazell L, Shakir SA. Under-reporting of adverse drug reactions : a systematic review. Drug Saf. 2006;29(5):385-96.

11. Sakaeda T, Tamon A, Kadoyama K, Okuno Y. Data mining of the public version of the FDA Adverse Event Reporting System. Int J Med Sci. 2013;10(7):796-803. 12. Reveles KR, Lee GC, Boyd NK, Frei CR. The rise in Clostridium difficile infection incidence among hospitalized adults in the United States: 2001-2010. Am J Infect Control. 2014;42(10):1028-32.

13. Centers for Disease Control and Prevention. Data Summary of HAIs in the US: Assessing Progress 2006-2016. Available from: https://www.cdc.gov/hai/data/archive/data-summary-assessing-progress.html. Accessed August 12, 2019. 14. Reveles KR, Pugh MJV, Lawson KA, Mortensen EM, Koeller JM, Argamany JR,

et al. Shift to community-onset Clostridium difficile infection in the national Veterans Health Administration, 2003-2014. Am J Infect Control.

2018;46(4):431-5. 15. Rupnik M, Wilcox MH, Gerding DN. Clostridium difficile infection: new developments in epidemiology and pathogenesis. Nat Rev Microbiol. 2009;7(7):526-36.

63 16. Abt MC, McKenney PT, Pamer EG. Clostridium difficile colitis: pathogenesis and host defence. Nat Rev Microbiol. 2016;14(10):609-20. 17. Brown KA, Khanafer N, Daneman N, Fisman DN. Meta-analysis of antibiotics and the risk of community-associated Clostridium difficile infection. Antimicrob Agents Chemother. 2013;57(5):2326-32. 18. Deshpande A, Pasupuleti V, Thota P, Pant C, Rolston DD, Sferra TJ, et al.

Community-associated Clostridium difficile infection and antibiotics: a meta- analysis. J Antimicrob Chemother. 2013;68(9):1951-61.

19. Slimings C, Riley TV. Antibiotics and hospital-acquired Clostridium difficile infection: update of systematic review and meta-analysis. J Antimicrob Chemother. 2014;69(4):881-91. 20. Food and Drug Administration. Drugs@FDA: FDA Approved Drug Products. Available from https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm.

Accessed August 12, 2019. 21. Evoy KE, Teng C, Encarnacion VG, Frescas B, Hakim J, Saklad S, et al. Comparison of Quetiapine Abuse and Misuse Reports to the FDA Adverse Event Reporting System With Other Second-Generation Antipsychotics. Subst Abuse. 2019;13:1178221819844205.

22. McConeghy KW, Soriano MM, Danziger LH. A Quantitative Analysis of FDA Adverse Event Reports with Oral Bisphosphonates and Clostridium difficile.

Pharmacotherapy. 2016;36(10):1095-101. 23. Evans SJ, Waller PC, Davis S. Use of proportional reporting ratios (PRRs) for signal generation from spontaneous adverse drug reaction reports. Pharmacoepidemiol Drug Saf. 2001;10(6):483-6.

64 24. MedDRA MSSO. Introductory Guide for Standardised MedDRA Queries (SMQs) Version 21.0. Available from http://www.meddra.org/sites/default/files/guidance/file/smq_intguide_21_0_engli sh.pdf. Accessed August 12, 2019. 25. Brook I, Wexler HM, Goldstein EJ. Antianaerobic antimicrobials: spectrum and susceptibility testing. Clin Microbiol Rev. 2013;26(3):526-46.

26. Modi SR, Collins JJ, Relman DA. Antibiotics and the gut microbiota. J Clin Invest. 2014;124(10):4212-8.

27. Jump RL, Polinkovsky A, Hurless K, Sitzlar B, Eckart K, Tomas M, et al. Metabolomics analysis identifies intestinal microbiota-derived biomarkers of colonization resistance in clindamycin-treated mice. PLoS One. 2014;9(7):e101267. 28. McDonald LC, Killgore GE, Thompson A, Owens RC, Jr., Kazakova SV, Sambol

SP, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med. 2005;353(23):2433-41. 29. Deneve C, Bouttier S, Dupuy B, Barbut F, Collignon A, Janoir C. Effects of subinhibitory concentrations of antibiotics on colonization factor expression by moxifloxacin-susceptible and moxifloxacin-resistant Clostridium difficile strains.

Antimicrob Agents Chemother. 2009;53(12):5155-62. 30. Vardakas KZ, Trigkidis KK, Boukouvala E, Falagas ME. Clostridium difficile

infection following systemic antibiotic administration in randomised controlled trials: a systematic review and meta-analysis. Int J Antimicrob Agents. 2016;48(1):1-10.

65 31. He M, Miyajima F, Roberts P, Ellison L, Pickard DJ, Martin MJ, et al. Emergence and global spread of epidemic healthcare-associated Clostridium difficile. Nat Genet. 2013;45(1):109-13. 32. Guh AY, Mu Y, Baggs J, Winston LG, Bamberg W, Lyons C, et al. Trends in incidence of long-term-care facility onset Clostridium difficile infections in 10 US geographic locations during 2011-2015. Am J Infect Control. 2018;46(7):840-2.

33. Lessa FC, Mu Y, Bamberg WM, Beldavs ZG, Dumyati GK, Dunn JR, et al. Burden of Clostridium difficile infection in the United States. N Engl J Med.

2015;372(9):825-34. 34. Stevens DL, Bisno AL, Chambers HF, Dellinger EP, Goldstein EJ, Gorbach SL, et al. Practice guidelines for the diagnosis and management of skin and soft tissue infections: 2014 update by the Infectious Diseases Society of America. Clin Infect Dis. 2014;59(2):e10-52.

35. Moore N, Thiessard F, Begaud B. The history of disproportionality measures (reporting odds ratio, proportional reporting rates) in spontaneous reporting of adverse drug reactions. Pharmacoepidemiol Drug Saf. 2005;14(4):285-6. 36. Montastruc JL, Sommet A, Bagheri H, Lapeyre-Mestre M. Benefits and strengths of the disproportionality analysis for identification of adverse drug reactions in a

pharmacovigilance database. Br J Clin Pharmacol. 2011;72(6):905-8. 37. Roden DM. Clinical practice. Long-QT syndrome. N Engl J Med.

2008;358(2):169-76. 38. Poluzzi E, Raschi E, Motola D, Moretti U, De Ponti F. Antimicrobials and the risk of torsades de pointes: the contribution from data mining of the US FDA Adverse Event Reporting System. Drug Saf. 2010;33(4):303-14.

66 39. Abo-Salem E, Fowler JC, Attari M, Cox CD, Perez-Verdia A, Panikkath R, et al. Antibiotic-induced cardiac arrhythmias. Cardiovasc Ther. 2014;32(1):19-25. 40. Antzelevitch C, Sun ZQ, Zhang ZQ, Yan GX. Cellular and ionic mechanisms underlying erythromycin-induced long QT intervals and torsade de pointes. J Am Coll Cardiol. 1996;28(7):1836-48. 41. Anderson ME, Mazur A, Yang T, Roden DM. Potassium current antagonist

properties and proarrhythmic consequences of quinolone antibiotics. J Pharmacol Exp Ther. 2001;296(3):806-10.

42. Volberg WA, Koci BJ, Su W, Lin J, Zhou J. Blockade of human cardiac potassium channel human ether-a-go-go-related gene (HERG) by macrolide antibiotics. J Pharmacol Exp Ther. 2002;302(1):320-7. 43. Bischoff U, Schmidt C, Netzer R, Pongs O. Effects of fluoroquinolones on HERG currents. Eur J Pharmacol. 2000;406(3):341-3.

44. Damle B, Labadie RR, Cuozzo C, Alvey C, Choo HW, Riley S, et al. Lack of an effect of standard and supratherapeutic doses of linezolid on QTc interval prolongation. Antimicrob Agents Chemother. 2011;55(9):4302-7. 45. Keller GA, Alvarez PA, Ponte ML, Belloso WH, Bagnes C, Sparanochia C, et al. Drug-Induced QTc Interval Prolongation: A Multicenter Study to Detect Drugs

and Clinical Factors Involved in Every Day Practice. Curr Drug Saf. 2016;11(1):86-98.

46. Lorberbaum T, Sampson KJ, Woosley RL, Kass RS, Tatonetti NP. An Integrative Data Science Pipeline to Identify Novel Drug Interactions that Prolong the QT Interval. Drug Saf. 2016;39(5):433-41. 47. Kumar V, Khosla S, Stancu M. Torsade de Pointes Induced by Hypokalemia from Imipenem and Piperacillin. Case Rep Cardiol. 2017;2017:4565182. 67 48. Bellomo R, Kellum JA, Ronco C. Acute kidney injury. Lancet. 2012;380(9843):756-66. 49. Zeng X, McMahon GM, Brunelli SM, Bates DW, Waikar SS. Incidence, outcomes, and comparisons across definitions of AKI in hospitalized individuals. Clin J Am Soc Nephrol. 2014;9(1):12-20. 50. Hoste EA, Bagshaw SM, Bellomo R, Cely CM, Colman R, Cruz DN, et al.

Epidemiology of acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med. 2015;41(8):1411-23.

51. Chawla LS, Bellomo R, Bihorac A, Goldstein SL, Siew ED, Bagshaw SM, et al. Acute kidney disease and renal recovery: consensus report of the Acute Disease Quality Initiative (ADQI) 16 Workgroup. Nat Rev Nephrol. 2017;13(4):241-57. 52. Linder A, Fjell C, Levin A, Walley KR, Russell JA, Boyd JH. Small acute increases in serum creatinine are associated with decreased long-term survival in

the critically ill. Am J Respir Crit Care Med. 2014;189(9):1075-81. 53. Bentley ML, Corwin HL, Dasta J. Drug-induced acute kidney injury in the critically ill adult: recognition and prevention strategies. Crit Care Med. 2010;38(6 Suppl):S169-74. 54. Khalili H, Bairami S, Kargar M. Antibiotics induced acute kidney injury:

incidence, risk factors, onset time and outcome. Acta Med Iran. 2013;51(12):871- 8.

55. Luther MK, Timbrook TT, Caffrey AR, Dosa D, Lodise TP, LaPlante KL. Vancomycin Plus Piperacillin-Tazobactam and Acute Kidney Injury in Adults: A Systematic Review and Meta-Analysis. Crit Care Med. 2018;46(1):12-20.

68 56. Welch HK, Kellum JA, Kane-Gill SL. Drug-Associated Acute Kidney Injury Identified in the United States Food and Drug Administration Adverse Event Reporting System Database. Pharmacotherapy. 2018;38(8):785-93. 57. Waterer GW, Wunderink RG. Increasing threat of Gram-negative bacteria. Crit Care Med. 2001;29(4 Suppl):N75-81. 58. Berlana D, Llop JM, Fort E, Badia MB, Jodar R. Use of colistin in the treatment

of multiple-drug-resistant gram-negative infections. Am J Health Syst Pharm. 2005;62(1):39-47.

59. Nation RL, Li J. Colistin in the 21st century. Curr Opin Infect Dis. 2009;22(6):535-43. 60. Gai Z, Samodelov SL, Kullak-Ublick GA, Visentin M. Molecular Mechanisms of Colistin-Induced Nephrotoxicity. Molecules. 2019;24(3). 61. Bamgbola O. Review of vancomycin-induced renal toxicity: an update. Ther Adv

Endocrinol Metab. 2016;7(3):136-47. 62. Mingeot-Leclercq MP, Tulkens PM. Aminoglycosides: nephrotoxicity. Antimicrob Agents Chemother. 1999;43(5):1003-12. 63. Wan H, Hu Z, Wang J, Zhang S, Yang X, Peng T. Clindamycin-induced Kidney Diseases: A Retrospective Analysis of 50 Patients. Intern Med.

2016;55(11):1433-7. 64. Subedi P, Chowdhury A, Tanovic K, Dumic I. Clindamycin: An Unusual Cause

of Acute Kidney Injury. Am J Case Rep. 2019;20:248-51. 65. Bird ST, Etminan M, Brophy JM, Hartzema AG, Delaney JA. Risk of acute kidney injury associated with the use of fluoroquinolones. CMAJ. 2013;185(10):E475-82.

69 66. Food and Drug Administration. FDA updates warnings for fluoroquinolone antibiotics on risks of mental health and low blood sugar adverse reactions. Available from https://www.fda.gov/news-events/press-announcements/fda- updates-warnings-fluoroquinolone-antibiotics-risks-mental-health-and-low-blood- sugar-adverse. Accessed August 12, 2019. 67. Frier BM. Hypoglycaemia in diabetes mellitus: epidemiology and clinical

implications. Nat Rev Endocrinol. 2014;10(12):711-22. 68. Viswanathan P, Iarikov D, Wassel R, Davidson A, Nambiar S. Hypoglycemia in

patients treated with linezolid. Clin Infect Dis. 2014;59(8):e93-5. 69. Kadoyama K, Sakaeda T, Tamon A, Okuno Y. Adverse event profile of tigecycline: data mining of the public version of the U.S. Food and Drug Administration adverse event reporting system. Biol Pharm Bull. 2012;35(6):967- 70.

70. Chou HW, Wang JL, Chang CH, Lee JJ, Shau WY, Lai MS. Risk of severe dysglycemia among diabetic patients receiving levofloxacin, ciprofloxacin, or moxifloxacin in Taiwan. Clin Infect Dis. 2013;57(7):971-80. 71. Mohr JF, McKinnon PS, Peymann PJ, Kenton I, Septimus E, Okhuysen PC. A retrospective, comparative evaluation of dysglycemias in hospitalized patients

receiving gatifloxacin, levofloxacin, ciprofloxacin, or ceftriaxone. Pharmacotherapy. 2005;25(10):1303-9.

72. Eshraghian A, Omrani GR. Cotrimoxazole-induced hypoglycemia in outpatient setting. Nutrition. 2014;30(7-8):959. 73. Lee AJ, Maddix DS. Trimethoprim/sulfamethoxazole-induced hypoglycemia in a patient with acute renal failure. Ann Pharmacother. 1997;31(6):727-32.

70 74. Johnson JA, Kappel JE, Sharif MN. Hypoglycemia secondary to trimethoprim/sulfamethoxazole administration in a renal transplant patient. Ann Pharmacother. 1993;27(3):304-6. 75. Tan CH, Shelley C, Harman KE. Doxycycline-induced hypoglycaemia. Clin Exp Dermatol. 2016;41(1):43-4. 76. Odeh M, Oliven A. Doxycycline-induced hypoglycemia. J Clin Pharmacol.

2000;40(10):1173-4. 77. Basaria S, Braga M, Moore WT. Doxycycline-induced hypoglycemia in a

nondiabetic young man. South Med J. 2002;95(11):1353-4. 78. Ito M, Fukuda M, Suzuki Y, Wakamoto H, Ishii E. Carnitine-related hypoglycemia caused by 3 days of pivalate antibiotic therapy in a patient with severe muscular dystrophy: a case report. BMC Pediatr. 2017;17(1):73. 79. Makino Y, Sugiura T, Ito T, Sugiyama N, Koyama N. Carnitine-associated

encephalopathy caused by long-term treatment with an antibiotic containing pivalic acid. Pediatrics. 2007;120(3):e739-41. 80. Parekh TM, Raji M, Lin YL, Tan A, Kuo YF, Goodwin JS. Hypoglycemia after antimicrobial drug prescription for older patients using sulfonylureas. JAMA Intern Med. 2014;174(10):1605-12.

81. Khamaisi M, Leitersdorf E. Severe hypoglycemia from clarithromycin- repaglinide drug interaction. Pharmacotherapy. 2008;28(5):682-4.

82. Clarithromycin [package insert]. North Chicago, IL: Abbott Laboratories; 2000. Available from https://www.accessdata.fda.gov/drugsatfda_docs/label/2012/050662s044s050,506 98s026s030,050775s015s019lbl.pdf. Accessed August 12, 2019.

71 83. Aspinall SL, Good CB, Jiang R, McCarren M, Dong D, Cunningham FE. Severe dysglycemia with the fluoroquinolones: a class effect? Clin Infect Dis. 2009;49(3):402-8. 84. Food and Drug Administration (9/9/2008). Determination that TEQUIN (gatifloxacin) was withdrawn from sale for reasons of safety or effectiveness. Available from https://www.federalregister.gov/documents/2008/09/09/E8-

20938/determination-that-tequin-gatifloxacin-was-withdrawn-from-sale-for- reasons-of-safety-or. Accessed August 12, 2019.

85. Bito M, Tomita T, Komori M, Taogoshi T, Kimura Y, Kihira K. The mechanisms of insulin secretion and calcium signaling in pancreatic beta-cells exposed to fluoroquinolones. Biol Pharm Bull. 2013;36(1):31-5. 86. Saraya A, Yokokura M, Gonoi T, Seino S. Effects of fluoroquinolones on insulin secretion and beta-cell ATP-sensitive K+ channels. Eur J Pharmacol.

2004;497(1):111-7. 87. Ashcroft FM. Mechanisms of the glycaemic effects of sulfonylureas. Horm Metab Res. 1996;28(9):456-63. 88. Niemi M, Neuvonen PJ, Kivisto KT. The cytochrome P4503A4 inhibitor clarithromycin increases the plasma concentrations and effects of repaglinide.

Clin Pharmacol Ther. 2001;70(1):58-65. 89. Eberl S, Renner B, Neubert A, Reisig M, Bachmakov I, Konig J, et al. Role of p-

glycoprotein inhibition for drug interactions: evidence from in vitro and pharmacoepidemiological studies. Clin Pharmacokinet. 2007;46(12):1039-49. 90. Zimmerman JL, Shen MC. Rhabdomyolysis. Chest. 2013;144(3):1058-65. 91. Kazory A, Dibadj K, Weiner ID. Rhabdomyolysis and acute renal failure in a patient treated with daptomycin. J Antimicrob Chemother. 2006;57(3):578-9. 72 92. Papadopoulos S, Ball AM, Liewer SE, Martin CA, Winstead PS, Murphy BS. Rhabdomyolysis during therapy with daptomycin. Clin Infect Dis. 2006;42(12):e108-10. 93. Patel SJ, Samo TC, Suki WN. Early-onset rhabdomyolysis related to daptomycin use. Int J Antimicrob Agents. 2007;30(5):472-4. 94. Brener ZZ, Bilik I, Khorets B, Winchester JF, Bergman M. Rhabdomyolysis

following clarithromycin monotherapy. Am J Med Sci. 2009;338(1):78. 95. Patel AM, Shariff S, Bailey DG, Juurlink DN, Gandhi S, Mamdani M, et al. Statin

toxicity from macrolide antibiotic coprescription: a population-based cohort study. Ann Intern Med. 2013;158(12):869-76. 96. Strandell J, Bate A, Hagg S, Edwards IR. Rhabdomyolysis a result of azithromycin and statins: an unrecognized interaction. Br J Clin Pharmacol. 2009;68(3):427-34.

97. Ainapurapu B, Kanakadandi UB. Trimethoprim-sulfamethoxazole induced rhabdomyolysis. Am J Ther. 2014;21(3):e78-9. 98. Jen SP, Sharma R. Trimethoprim-sulphamethoxazole-associated rhabdomyolysis in an HIV-infected patient. Int J STD AIDS. 2011;22(7):411-2. 99. Carroll MW, Choi H, Min S, Hwang S, Park H, Song T, et al. Rhabdomyolysis in

a patient treated with linezolid for extensively drug-resistant tuberculosis. Clin Infect Dis. 2012;54(11):1624-7.

100. Lechner AM, Past E, Porsche U, Kern JM, Hoppe U, Pretsch I. Two cases of serious rhabdomyolysis during linezolid treatment. Infection. 2017;45(4):563-6. 101. John F, Oluronbi R, Pitchumoni CS. Levofloxacin-induced rhabdomyolysis: a case report. J Med Case Rep. 2016;10(1):235.

73 102. Goldie FC, Brogan A, Boyle JG. Ciprofloxacin and statin interaction: a cautionary tale of rhabdomyolysis. BMJ Case Rep. 2016;2016. 103. Cefdinir [package insert]. Morris Plains, NJ: Parke-Davis; 1999. Available from https://www.accessdata.fda.gov/drugsatfda_docs/label/1999/50739S2LBL.PDF. Accessed August 12, 2019. 104. Omar MA, Wilson JP. FDA adverse event reports on statin-associated

rhabdomyolysis. Ann Pharmacother. 2002;36(2):288-95. 105. Margolin L. Impaired rehabilitation secondary to muscle weakness induced by

meropenem. Clin Drug Investig. 2004;24(1):61-2. 106. Craig WA. The pharmacology of meropenem, a new carbapenem antibiotic. Clin Infect Dis. 1997;24 Suppl 2:S266-75. 107. Zhang X, Donnan PT, Bell S, Guthrie B. Non-steroidal anti-inflammatory drug induced acute kidney injury in the community dwelling general population and

people with chronic kidney disease: systematic review and meta-analysis. BMC Nephrol. 2017;18(1):256.

74 Vita

Chengwen Teng was born in Yangzhou, Jiangsu, China. After graduating from Yangzhou High School of Jiangsu Province, in 2006, he entered Wuhan University in Wuhan, Hubei, China. He received the degree of Bachelor of Science in biology from

Wuhan University in June 2010. In August 2010, he entered the University of Southern Mississippi in Hattiesburg, Mississippi. He received the degree of Master of Science in biological sciences from the University of Southern Mississippi in August 2012. He attended the University of Alabama at Birmingham in Birmingham, Alabama from August 2012 to August 2013 to complete pre-pharmacy coursework. In August 2013, he entered the University of Oklahoma College of Pharmacy in Oklahoma City, Oklahoma. In June 2017, he received the degree of Doctor of Pharmacy with the valedictorian honor from the University of Oklahoma College of Pharmacy. In July 2017, he enrolled in the Joint Translational Science PhD program at the University of Texas at Austin, the University of Texas Health Sciences Center at San Antonio, and the University of Texas at San Antonio in San Antonio, Texas.

Permanent address (or email): [email protected] This dissertation was typed by Chengwen Teng.

75