Towards the Creation of Polymer Composites Which Can Be

TOWARDS THE CREATION OF POLYMER COMPOSITES WHICH CAN BE

REFILLED WITH ANTIBIOTICS AFTER IMPLANTATION FOR INFECTION

TREATMENT

ERIKA LEAH CYPHERT

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Dissertation Advisor: Horst von Recum, Ph.D.

Department of Biomedical Engineering

CASE WESTERN RESERVE UNIVERSITY

January 2021

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Erika Leah Cyphert

Candidate for the Doctor of Philosophy degree*.

(signed) Steven Eppell, Ph.D. (chair of committee)

Horst von Recum, Ph.D.

Eben Alsberg, Ph.D.

Agata Exner, Ph.D.

Jonathan Pokorski, Ph.D.

(date) September 25, 2020

*We also certify that written approval has been obtained for any proprietary material contained therein.

To my grandparents with love –

Phil and Ann Cyphert

Ted and Dorothy Lippold

Florence Miller

TABLE OF CONTENTS

TABLE OF CONTENTS………………………………………………………………..4

LIST OF TABLES…………………………………………………………………..….10

LIST OF FIGURES…………………………………………………………………….13

LIST OF ABBREVIATIONS……………………………………………………….…21

ACKNOWLEDGEMENTS……………………………………………………………24

ABSTRACT…………………………………………………………………………..…27

CHAPTER 1: DIAGNOSIS AND BIOMATERIAL-BASED TREATMENTS FOR

PERIPROSTHETIC JOINT INFECTIONS………………………………………….29

1.1. LIMITATIONS OF CLINICAL TREATMENT OF PERIPROSTHETIC

INFECTION…………………………………………………………………30

1.1.1. INTRODUCTION…………………………………………….…30

1.1.2. ISOLATION OF MICROBIAL ORGANISMS………………....33

1.1.3. POSSIBLE UNDERLYING PATIENT COMORBIDITIES……33

1.1.4. BIOFILM FORMATION AND BACTERIAL RESISTANCE…34

1.1.5. REVISION PROCEDURES/INITIAL TREATMENT

FAILURES...... 36

1.1.6. SUMMARY……………………………………………………...37

1.2. NOVEL TREATMENT MODALITIES FOR PJIS………………………...38

1.2.1. LIMITATIONS WITH TRADITIONAL ANTIBIOTIC-LADEN

PMMA BONE CEMENT……………………………………..…38

1.2.2. COMMERCIALLY AVAILABLE ALTERNATIVE

BIOMATERIALS FOR ANTIBIOTIC-LADEN PMMA BONE

CEMENT………………………………………………………...40

1.2.3. TITANIUM NANOTUBE ARRAYS……………………………41

1.2.4. IMPLANT COATINGS WITH ANTIBACTERIAL EFFECT….42

1.2.5. POLYMERS/HYDROGELS…………………………………….43

1.2.6. CYCLODEXTRIN (CD)-BASED DRUG DELIVERY………...44

1.2.7. CONCLUSIONS………………………………………………...48

1.3. ACKNOWLEDGEMENTS…………………………………………....……48

CHAPTER 2: ANTIBIOTIC REFILLING OF PMMA BONE CEMENT

THROUGH INCORPORATION OF CYCLODEXTRIN FOR TREATMENT OF

PERIPROSTHETIC JOINT INFECTION…………………………………………...49

2.1. ABSTRACT…………………………………………………………………50

2.2. INTRODUCTION…………………………………………………………..51

2.3. MATERIALS AND METHODS……………………………………………55

2.4. RESULTS…………………………………………………………………...62

2.5. DISCUSSION……………………………………………………………….75

2.6. CONCLUSIONS……………………………………………………………79

2.7. ACKNOWLEDGEMENTS…………………………………………………80

CHAPTER 3: TREATMENT OF BROAD-SPECTRUM PERIPROSTHETIC

JOINT INFECTIONS USING COMBINATORIAL ANTIBIOTIC PMMA

COMPOSITE…………………………………………………………………...………81

3.1. ABSTRACT…………………………………………………………………82

3.2. INTRODUCTION…………………………………………………………..83

3.3. MATERIALS AND METHODS……………………………………………87

3.4. RESULTS AND DISCUSSION………………………………………….…91

3.5. CONCLUSIONS…………………………………………………………..115

3.6. ACKNOWLEDGEMENTS………………………………………………..116

CHAPTER 4: POLYMERIZED CYCLODEXTRIN CAN BE REFILLED WITH

ANTIBIOTICS IN THE PRESENCE OF BACTERIAL BIOFILMS…………….117

4.1. ABSTRACT……………………………………………………………..…118

4.2. INTRODUCTION…………………………………………………………119

4.3. MATERIALS AND METHODS…………………………………………..122

4.4. RESULTS AND DISCUSSION…………………………………………...126

4.5. CONCLUSIONS…………………………………………………………..135

4.6. ACKNOWLEDGEMENTS………………………………………………..136

CHAPTER 5: ANTIBIOTIC REFILLING OF PMMA BONE CEMENT

COMPOSITE THROUGH BONE AND MUSCLE TISSUE TO TREAT

PERIPROSTHETIC JOINT INFECTIONS………………………………………...137

5.1. ABSTRACT………………………………………………………………..137

5.2. INTRODUCTION…………………………………………………………138

5.3. MATERIALS AND METHODS………………………………………..…141

5.4. RESULTS AND DISCUSSION…………………………………………...148

5.5. CONCLUSIONS…………………………………………………………..167

5.6. ACKNOWLEDGEMENTS………………………………………………..168

CHAPTER 6: PROCESSING TECHNIQUE OF PMMA BONE CEMENT

COMPOSITE DICTATES REFILLING CAPACITY FOR TREATMENT OF

PERIPROSTHETIC JOINT INFECTIONS………………………………………...169

6.1. ABSTRACT………………………………………………………………..170

6.2. INTRODUCTION…………………………………………………………171

6.3. MATERIALS AND METHODS…………………………………………..174

6.4. RESULTS………………………………………………………………….179

6.5. DISCUSSION……………………………………………………………...199

6.6. CONCLUSIONS…………………………………………………………..205

6.7. ACKNOWLEDGEMENTS………………………………………………..206

CHAPTER 7: CONCLUSIONS AND FUTURE DIRECTIONS…………………..207

7.1. CONCLUSIONS…………………………………………………………..207

7.2. FUTURE DIRECTIONS…………………………………………………..219

7.3. ACKNOWLEDGEMENTS………………………………………………..227

APPENDIX…………………………………………………………………………….228 8

PERMISSIONS………………………………………………………………...228

REFERENCES………………………………………………………………………...243

LIST OF TABLES

TABLE 2-1 Mechanical testing of cylindrical PMMA-CD composite samples

ultimate compressive strength, work to peak load, and strain to

peak load…...... 69

TABLE 2-2 Quantification of RMP refilling from agarose refilled polymer

samples: pure PMMA versus β-CD disks (top) and pure PMMA

versus PMMA-CD composites with 5 and 10 wt% CD

microparticles (bottom)…………………………………………..72

TABLE 3-1 Quantification of various parameters of cylindrical PMMA-CD

composite samples containing either only gentamicin or

tobramycin (control), 10 wt% empty (nondrug-filled) CD

microparticles with gentamicin or tobramycin, and 10 wt% RMP-

filled CD microparticles with gentamicin or tobramycin that were

micro-CT scanned……………………………………………....100

TABLE 3-2 Quantification of mechanical properties of PMMA-CD cylindrical

composite samples containing either (a) tobramycin or gentamicin

alone (controls), (b) 10 wt% empty (nondrug-filled) CD

microparticles with either free gentamicin or tobramycin, or (c) 10

wt% RMP-filled CD microparticles with either free gentamicin or

tobramycin……………………………………………..…….…105 10

TABLE 3-3 Quantification of amount of RMP initially filled (top) and amount

of RMP refilled through agarose phantom model into PMMA-CD

composite bead samples containing either 5 or 10 wt% empty

(nondrug-filled) CD microparticles with either free gentamicin or

tobramycin (bottom)……………………………………………109

TABLE 4-1 Mass of antibiotics filled into CD and dextran polymer disks in

agarose phantom model………………………………………...128

TABLE 4-2 Bacterial quantification of CD and dextran polymer disks with

mature biofilms…………………………………………………131

TABLE 6-1 Micro-CT quantification of porosity for hand- and vacuum-mixed

PMMA-CD composite cylinders containing either no

microparticles or 10 wt% or 15 wt% empty (nondrug-filled) CD

microparticles…………………………………………………...186

TABLE 6-2 Quantification of mechanical properties for hand- and vacuum-

mixed PMMA-CD composites containing either no CD

microparticles or 10 or 15 wt% empty (nondrug-filled) or RMP-

filled CD microparticles………………………………………...190

TABLE 7-1 Summary of relationship of amount of drug injected into refilling

model and amount filled into PMMA-CD composite to the

duration of antimicrobial activity possible…………………..…210

TABLE 7-2 Analysis of impact of parameters controlling mechanical strength

and refilling capacity of PMMA-CD composite……………..…214

LIST OF FIGURES

FIGURE 1-1 Schematic depicting challenges and downstream effects of

inappropriate periprosthetic joint infection diagnosis……………31

FIGURE 1-2 Proliferation of a single resistant bacterium when repeatedly

treated with antibiotics……………………………………..…….36

FIGURE 1-3 Free-radical polymerization of PMMA bone cement upon

combination of liquid (methyl methacrylate monomer) and

powdered component (benzoyl peroxide initiator and pre-

polymerized PMMA beads)…………………………………..….38

FIGURE 1-4 Interior of antibiotic-laden PMMA before and after releasing some

antibiotic from the sample……………………………………….39

FIGURE 1-5 Structure of soluble CD monomer, soluble pre-polymerized CD,

and cross-linked insoluble CD polymer microparticles….………44

FIGURE 1-6 Antibiotic refilling of PMMA versus cross-linked CD polymer

when implanted in soft tissue…………………………………….45

FIGURE 1-7 Daily release profile from cyclodextrin (CD) affinity-based

polymers as compared to drug release through diffusion (i.e.

PMMA). Polymers that use affinity-based release may be refilled

after implantation (b)…………………………………………….46 13

FIGURE 2-1 Schematic outlining synthesis of PMMA-CD composite and

subsequent evaluations…………………………………………...55

FIGURE 2-2 Zone of inhibition study of antimicrobial activity of three

antibiotics (RMP, tobramycin, and gentamicin) freely added into

PMMA (plain) beads against S. aureus………………………….63

FIGURE 2-3 Zone of inhibition study of antimicrobial activity of antibiotic-

filled (RMP, tobramycin, and gentamicin) CD microparticles as

PMMA-CD composite beads against S. aureus………………….64

FIGURE 2-4 Drug release profiles and antimicrobial susceptibility zone of

inhibition testing with a) tobramycin and b) gentamicin freely

added into PMMA beads and from c-g) antibiotic-CD

microparticles in PMMA-CD composite beads………………….66

FIGURE 2-5 Representative compressive load-displacement curves of PMMA-

CD composite cylinders. Load-displacement curves of all groups

superimposed (left) and zoomed in curve of free RMP cement

(right)………………………………………………………….…68

FIGURE 2-6 Image of PMMA beads (no CD microparticles) and CD disks prior

to RMP refilling in agarose model (top) and image of polymer

samples 48 hours after being refilled with RMP in model

(bottom)…………………………………………………………..71

FIGURE 2-7 Image of RMP refilling in PMMA-CD composite beads removed

from agarose model 48 hours after RMP injection………………73

FIGURE 2-8 Zone of inhibition study of antimicrobial activity of RMP refilled

PMMA-CD composite beads against S. aureus……………….…75

FIGURE 3-1 Schematic depicting drug combination PMMA-CD composite

formulations……………………………………………………...86

FIGURE 3-2 Zone of inhibition studies of drug combinations of PMMA-CD

composites and monotherapy drug control samples against S.

aureus (top), S. epidermidis (middle), and E. coli (bottom)…..…92

FIGURE 3-3 Gentamicin and tobramycin daily drug release and subsequent

antimicrobial activity of each release aliquot against S. aureus, S.

epidermidis, and E. coli from PMMA-CD composite beads…….97

FIGURE 3-4 Representative three-dimensional renderings of solid and pore

volume fractions of PMMA-CD composite cylinders……….....101

FIGURE 3-5 Representative stress versus strain curves from the compression

testing of combinatorial PMMA-CD composite samples…....…106

FIGURE 3-6 Images of refilled PMMA-CD combinatorial composite beads after

48 hours of being refilled with RMP…………………………...109

FIGURE 3-7 Zone of inhibition studies of PMMA-CD composite samples that

were refilled with RMP for 48 hours against S. aureus (top), S.

epidermidis (middle), and E. coli (bottom)……………………..112

FIGURE 4-1 Validation of refilling model with CD and dextran (control)

polymer disks (a). Refilling model with CD polymer without

biofilm (biofilm control) and CD polymer disks with either an

immature or mature biofilm (b)………………………………...122

FIGURE 4-2 Agarose-based tissue-mimicking refilling model with antibiotic

filled polymer disks after 52 hours (RMP) and 45 hours (MC) of

refilling (top). Removed polymer disks after 52 hours (RMP) and

45 hours (MC) of refilling (bottom)…………………………….127

FIGURE 4-3 Zone of inhibition study of antimicrobial activity of refilled RMP

and MC CD polymer disks with immature or mature biofilms

against S. aureus………………………………………………..134

FIGURE 5-1 Schematic depicting set-up for hard tissue (femur) refilling

model……………………………………………………………149

FIGURE 5-2 Images of set-up of femur refilling model containing PMMA-CD

composite and sliced bone after 48 hours of RMP refilling……150

FIGURE 5-3 Stereomicroscope images of PMMA-CD composite with and

without CD microparticles after 48 hours of RMP refilling in

femur model………………………………………………….…151

FIGURE 5-4 Quantification of the depth of RMP refilling of PMMA-CD

composite in femur after intraosseous infusion……………...…153

FIGURE 5-5 Zone of inhibition study of antimicrobial activity of PMMA-CD

composite removed from femur model after 48 hours of RMP

refilling against S. aureus………………………………………155

FIGURE 5-6 Schematic depicting set-up for soft tissue PMMA-CD composite

refilling and infection models…………………………………..158

FIGURE 5-7 Images of set-up of soft tissue PMMA-CD composite refilling

model before, immediately after, and 48 hours following injection

of RMP…………………………………………………….……159

FIGURE 5-8 Stereomicroscope images of PMMA-CD composite beads before

and after 1-3 cycles of refilling in soft tissue model……………160

FIGURE 5-9 Zone of inhibition of antimicrobial activity of PMMA-CD

composite beads after 1-3 cycles of refilling with RMP against S.

aureus……………………………………………………...……161

FIGURE 5-10 Remaining CFU counts of PMMA-CD composite beads with a 48

hour biofilm and tissue surrounding implanted bead 48 hours

following RMP injection (relative to controls without RMP)….163

FIGURE 5-11 Zone of inhibition study of antimicrobial activity of PMMA-CD

composite beads with 48 hours biofilm after RMP refilling against

S. aureus……………………………………………………...…165

FIGURE 6-1 Stereomicroscope images of sliced PMMA-CD composite

cylinders stained with methylene blue depicting both the

outside/exterior (a) and inside cut plane (b) of hand- and vacuum-

mixed cylinders. Detection of methylene blue stained pixels in

stereomicroscope images of exterior (c) and interior (d) PMMA-

CD using Matlab……………………………………………..…181

FIGURE 6-2 SEM micrographs of inside cut plane of hand-mixed PMMA-CD

composite cylinders without CD microparticles (top row), hand-

mixed containing 15 wt% CD microparticles (second row),

vacuum-mixed without CD microparticles (third row), and

vacuum-mixed containing 15 wt% CD microparticles (fourth

row)………………………………………………………..……183

FIGURE 6-3 Representative three-dimensional renderings of solid (top) and

pore fraction (bottom) of hand- and vacuum-mixed PMMA-CD

composite cylinders…………………………………………….185

FIGURE 6-4 Representative stress versus strain curves of hand- and vacuum-

mixed PMMA-CD composites either without CD microparticles,

10 or 15 wt% empty CD microparticles, or RMP-filled CD

microparticles………………………………………………...…189

FIGURE 6-5 Stereomicroscope images depicting the depth of diffusion of RMP

into hand- and vacuum-mixed PMMA-CD composite cylinders

after different periods of refilling with RMP through agarose…193

FIGURE 6-6 Quantification of the depth of refilling of RMP into hand- (left)

and vacuum-mixed PMMA-CD composites after 4-48 hours of

refilling through agarose………………………………………..195

FIGURE 6-7 Antimicrobial activity inside cut face (a) and exterior face of

vacuum- (b) and hand-mixed (c) PMMA-CD cylinders refilled

with RMP against S. aureus…………………………………….198

FIGURE 7-1 Schematic outlining the relationship between customizable

parameters of PMMA-CD composite delivery platform and

downstream effects of parameters on functionality and properties

of PMMA-CD composite……………………………………….208

FIGURE 7-2 Zone of inhibition study of antimicrobial activity of RMP refilled

PMMA-CD composites with different PMMA compositions

against S. aureus……………………………………………..…212

FIGURE 7-3 Tissue harvested from mice 14 days following inoculation with S.

aureus and treatment with PMMA-CD composite bead filled with

RMP (circled)…………………………………………………...220

FIGURE 7-4 Zone of inhibition of antimicrobial activity of tetracycline from

PMMA-CD composites………………………………………...223

FIGURE 7-5 NMR spectra of free tetracycline powder non-heated (top) and

heated at 100oC for 1 hour (bottom)……………………………224

FIGURE 7-6 Mass spectrometry spectra of free tetracycline powder non-heated

(left) and heated at 100oC for 1 hour (right)……………………225

LIST OF ABBREVIATIONS

*In alphabetical order

ANOVA Analysis of variance

ASTM American Society for Testing and Materials

CD Cyclodextrin

CT Computed Tomography

DICOM Digital Imaging and Communications in Medicine

DMF Dimethylformamide

DMSO Dimethyl sulfoxide

E. coli Escherichia coli

GFP Green Fluorescent Protein

HDI Hexamethylene diisocyanate

HV High Viscosity

IL Interleukin

LB Luria-Bertani

MC Minocycline

MIC Minimum Inhibitory Concentration

MMA Methyl Methacrylate

MRSA Methicillin-resistant Staphylococcus aureus

MSIS Musculoskeletal Infection Society

P. aeruginosa Pseudomonas aeruginosa

PBS Phosphate Buffered Saline pDEX Polymerized Dextran

PJI Periprosthetic Joint Infection

PLGA Poly(lactic-co-glycolic acid)

PMMA Poly(methyl methacrylate)

PNM Polymorphonuclear Neutrophil

PTFE Poly(tetrafluoroethylene)

QSAR Quantitative Structure-Activity Relationship

RMP Rifampicin

S. aureus Staphylococcus aureus

SEM Scanning Electron Microscopy

S. epidermidis Staphylococcus epidermidis

TNF-α Tumor Necrosis Factor Alpha

ZOI Zone of Inhibition

ACKNOWLEDGEMENTS

I would like to express my endless gratitude to Dr. Horst von Recum for all of his support, motivation, and guidance throughout this process and over the past 8.5 years. I never imagined when I joined your lab as a freshman that it would have such an impact on my life and that I would be hanging around Case nearly a decade later. It’s a testament to you as a mentor and advisor that I have always had an incredibly positive experience in the lab over all of these years. It is difficult to articulate how much you have really helped me and I am eternally grateful for your willingness to mentor and guide me through my own independent projects once Andrew graduated. I couldn’t have asked for a more supportive advisor allowing me to pursue my own research interests, international research endeavors, and travel and running excursions. You probably don’t remember me as the young freshman who passed out in the cell culture room, but I hope this makes you laugh as you reflect on how far I have progressed over the years.

I would also like to express my sincere thanks to all of my committee members for their kind words of support and insightful advice. To Dr. Eppell for all of your insight on orthopedics and PMMA. To Dr. Exner for all of your invaluable career and experimental guidance. To Drs. Pokorski and Alsberg, I appreciate your willingness to remain on my committee despite the long distance and your patience during the Skype calls. My project has truly benefitted from all of your collective wisdom and expertise.

To my mentors abroad in Japan and Poland: Dr. Masamichi Nakayama, Dr. Wojciech

Święszkowski, Dr. Monika Bil, and Dr. Ewa Kijeńska-Gawrońska, I appreciate your willingness to support me and continual patience for training me despite the language barrier. I won’t forget your kindness, acceptance, and curiosity about American culture.

I would like to thank all of the members of the von Recum lab. It is safe to say that the lab has been my “home” for many years and I really feel like we are a family. A special thanks to Andrew Fu for seeing my potential as a naïve and unexperienced freshman and willingness to take me on as a student. Your willingness to hire me has completely changed my life. To Sean Zuckerman for teaching me everything I know about bacteria and how to write a paper, to Edgardo Rivera for teaching me polymer synthesis, and to

Nate Rohner for teaching me animal handling. To Greg Learn for teaching me mechanical testing, CT imaging, and animal handling, machining countless samples, and all of our productive brainstorming conversations that have dramatically shaped my projects. None of this work would have been possible without all of your assistance! To my fellow grad students: Greg Learn and Kathleen Young, thanks for all of the laughs and memories. I won’t ever forget all of our Bibibop Thursdays, Polish store excursions,

Team EKG triathlon races, adventures in Poland, drives through Alabama (lol), and trying new fun(?) foods. Thanks for being my emotional support over the past couple of years. To all of my undergraduate and high school students: Sunny “Chao-yi” Lu, Dylan

Marques, Nora “Ningjing” Zhang, Fang Zhang, Saar Dolev, Sara Hurley, Grace Cousens, and Ben Grandstaff, this work is a collection of all of your hard work and would not have 25

happened without your endless enthusiasm and dedication. I am extremely grateful to all of you and I wish you all the best!

Finally, a special thanks to my family especially my parents, brother, and grandparents for all of your endless support over the years. While I realize that you don’t fully understand what I do, you have always expressed your curiosity and interest and have supported all of my crazy plans. Last but not least, I would like to thank Bella and Tatra, the craziest goldendoodles around. Tatra, thanks for putting a smile on my face every day and thanks for being my COVID-19 quarantine writing/running buddy.

Towards the Creation of Polymer Composites Which Can Be Refilled With

Antibiotics After Implantation For Infection Treatment

ERIKA LEAH CYPHERT

Periprosthetic joint infection (PJI) is one of the two leading causes of failure in arthroplasties and can develop inside of the bone as well as surrounding soft tissue. PJIs can be challenging to treat as initial diagnosis can be delayed and treatment typically involves administration of systemic antibiotics followed by removal of infected device.

Traditionally, antibiotics have been directly mixed into poly(methyl methacrylate)

(PMMA) bone cement to locally treat infections. However, this strategy often results in insufficient elution of drug for treatment of chronic infections and a limited range of antibiotics are compatible due to heat generated during polymerization. To overcome these shortcomings with antibiotic-PMMA, this work focused on development of PMMA composite material containing differing amounts of polymerized cyclodextrin (CD) microparticles. It was hypothesized that addition of CD microparticles to PMMA would enable post-implantation antibiotic refilling to occur and broaden the range of antibiotics compatible with PMMA to more effectively treat chronic PJIs without the need to remove the implant or expose patients to systemic antibiotics. This work first explores the emerging polymer technologies used to treat PJIs and then investigates a range of properties of PMMA containing CD microparticles and either a single or combination of

antibiotics to treat broad-spectrum infections. Specifically, the porosity, compressive strength, antibiotic filling/refilling, antibiotic release, and antimicrobial properties were evaluated. Impact on functionality of CD and PMMA-CD composites in the presence of bacterial biofilm was explored. Furthermore, ex vivo models were developed to simulate how antibiotic refilling would reasonably occur in either soft or hard tissue. PMMA-CD composites were able to demonstrate refilling capacity with several antibiotics, resulting in lasting antimicrobial activity (> 60 days) to treat chronic, broad-spectrum PJIs.

Composites were able to retain their mechanical properties upon addition of CD microparticles and strength was improved through use of vacuum-mixing, suggesting potential for load-bearing applications. Composites maintained their refilling properties even in presence of bacterial biofilms, demonstrating that they may not necessarily need to be removed if a biofilm forms. PMMA-CD composites have potential to serve as versatile delivery platform that is refillable and patient customizable to more effectively treat chronic PJIs.

CHAPTER 1. DIAGNOSIS AND BIOMATERIAL-BASED TREATMENTS FOR

PERIPROSTHETIC JOINT INFECTIONS

*Portions of this chapter are reprinted by permission from Springer Nature Customer

Service Centre GmbH: Springer Nature, Current Rheumatology Reports, 2018, 20:33.

Authors: Erika L. Cyphert*, Ashley E. Levack*, Mathias P. Bostrom, Christopher J.

Hernandez, Horst A. von Recum, Alberto Carli (*Equal contributions)

Author contributions: *ELC and AEL contributed equally to this work. ELC prepared the figures, completed literature search, and contributed with the writing. AEL contributed with the writing. AC, CJH, MPB, and HAV consulted with the conceptual development of the narrative and critically revised the text.

-Section 1.1. is not part of the above publication and was solely written by Erika L.

Cyphert

-Section 1.2. was adapted from the above publication where in the publication Erika L.

Cyphert developed the figures, completed literature search, and equally contributed with the writing with Ashley E. Levack

1.1. LIMITATIONS OF CLINICAL TREATMENT OF PERIPROSTHETIC

INFECTION

1.1.1. INTRODUCTION

Periprosthetic joint infections (PJIs) can result following primary total knee or hip arthroplasties and can be challenging to treat depending on their severity at diagnosis, the surgeon’s ability to effectively isolate the microbial organisms1–3, the patient’s underlying comorbidities1,4–9, if a bacterial biofilm is established2–5,10–12, and the methods used during the initial or revision procedure1,2,4,5,7,8,13. All of these factors are critical in enabling surgeons to prescribe the proper antibiotic treatment protocol and influence the patient’s likelihood of a successful outcome. If the infections are not quickly diagnosed, they can often lead to multiple revision surgeries, substantial osteolysis (loss of bone integrity), and in extreme cases amputation or death.

Figure 1-1: Schematic depicting the challenges and downstream effects of inappropriate periprosthetic joint infection diagnosis.

One of the primary challenges associated with PJI treatment is that they are difficult to correctly and quickly diagnose, which is predominantly due to the lack of a universal set of criteria that surgeons can use to consistently assess and diagnose patients with

PJIs2,14,15. Organizations such as the Musculoskeletal Infection Society (MSIS) and the

Knee Society have developed classification systems such as the MSIS system in an attempt to develop a universal set of criteria for the diagnosis of PJIs2,14,15, however not all surgeons adhere to this system and it is not always consistent in its ability to accurately diagnose PJIs. The MSIS system consists of two major and six minor criteria 31

and to diagnose a PJI 50% of the major and 66% of the minor criteria must be fulfilled2.

Major criteria include communication between the sinus tract (resulting from infection) and joint or two separate positive cultures from the periprosthetic region having phenotypically identical microorganisms16. Minor criteria include elevated serum C- reactive protein and erythrocyte sedimentation rate, elevated polymorphonuclear neutrophil percentage (PNM%) in synovial fluid, elevated white blood cell count in synovial fluid, single positive periprosthetic culture, periprosthetic tissue with positive histological analysis, or positive change in leukocyte esterase test strip16. A study conducted by Zmistowski et al. has demonstrated that effective diagnosis of PJI typically involves a comparison of both serum and joint white blood counts where white blood cells other than lymphocytes are used in the comparison15.

In a multicenter study conducted by Koh et al., 303 patients were diagnosed with PJIs using MSIS criteria following total knee arthroplasty2. Only 65% of patients met MSIS criteria for PJI diagnosis2. In those patients that did not meet MSIS criteria, 86% were diagnosed with PJIs based upon abnormal clinical findings (i.e. redness, local heat, joint pain) and abnormal radiographic findings2. Based upon this study, Koh et al. concluded that the criteria of the MSIS system should have less rigid constraints regarding which combination of symptoms results in PJI diagnosis2.

As an alternative to the MSIS and Knee Society criteria, surgeons have used a wide variety of tests to diagnose PJIs. In these cases, PJI diagnosis often involves some

combination of the following tests: radiographic and nuclear imaging1,2,4,8,13, synovial fluid aspiration1, hematologic testing (erythrocyte sedimentation rate, C-reactive protein)1,8,15, clinical discretion2,8,13, physical findings (redness, local heat, etc.)2,14, positive pre- or intra-operative culture4,6,8,12–15, purulence around prosthesis8,14,15, and histopathologic exam14.

1.1.2. ISOLATION OF MICROBIAL ORGANISMS

Once a PJI has been diagnosed, one of the most important components of the infection treatment plan is to isolate the microbial organism(s) causing the infection in order to select the most appropriate antibiotic therapy and to discern the severity of the infection.

However, the isolation process has been shown to be plagued with false negative results

(7-23% cases intra-operative cultures)1–4,13. Factors that can contribute to negative intra- operative cultures include: inappropriate incubation time and culture media1,2, prior antibiotic use2,3, and formation of bacterial biofilms2. Therefore, proper care must be taken to avoid antibiotic prophylaxis prior to acquiring the sample and to determine the appropriate incubation conditions.

1.1.3. POSSIBLE UNDERLYING PATIENT COMORBIDITIES

The effects of PJI can be exacerbated if the patient has an underlying comorbidity such as malnutrition6, abnormal coagulation (coagulopathy)8, diabetes mellitus4,7, immunosuppressive diseases4, or cardiac arrhythmia9, which have all been shown to be positively correlated to a patient’s risk of developing a PJI and other associated 33

complications. A study conducted by Gunningberg et al. has shown a positive correlation between a patient’s pre-operative serum albumin levels, as a measure of malnutrition, and their risk of developing a PJI6. If a patient presents with decreased serum albumin levels preoperatively, indicative of malnutrition, they have an increased risk of developing a

PJI, poor wound healing, increased hospital stays, and other complications6. Another comorbidity that can complicate the treatment of PJIs is abnormal blood coagulation8. A study by Saxena et al. has shown that >50% of patients with PJIs have abnormal blood coagulation, even without being administered anticoagulation drugs8. It is hypothesized that PJI patients can have abnormal coagulation due to increased cytokine levels (tumor necrosis factor-α, TNF-α; interleukin, IL-1 and IL-6) produced by macrophages and bacterial toxins at the infection site that can activate coagulation and initiate an abnormal coagulation cascade (down-regulation of anti-coagulants) and excess blood loss during surgery8. If a patient has a cardiac arrhythmia such as atrial fibrillation, their risk of developing a PJI and wound complication is also increased due to their use of anticoagulant medication9.

1.1.4. BIOFILM FORMATION AND BACTERIAL RESISTANCE

Treatment of PJIs can be further complicated if the infection has developed a bacterial biofilm on the surface of the implanted device. The two primary microorganisms found in biofilms stemming from PJIs are Staphylococcus aureus (S. aureus) and Staphylococcus epidermidis (S. epidermidis)11. A study by Molina-Manso et al. has shown that out of

nine antibiotics (vancomycin, rifampicin, tigecycline, ciprofloxacin, clindamycin, cloxacillin, daptomycin, trimethoprim/sulfamethoxazole, fosfomycin) tested against 32 clinical isolates of S. aureus and S. epidermidis collected from PJIs, none of the antibiotics were capable of completely eradicating the bacterial biofilm11. Bacteria within the biofilm are often difficult to treat since they are less metabolically active than their planktonic form11.

Eradication of biofilms is more challenging if the bacteria within them have developed antibiotic resistance. Figure 1-2 depicts a simplistic cartoon schematic regarding how antibiotic resistance can spread if a single bacterium has a mutation in the binding site of the antibiotic and is repeatedly treated with antibiotics. Bacterial resistance is of particular interest when aminoglycoside antibiotics such as gentamicin or tobramycin are used to treat PJIs10. These antibiotics are two of the most common drugs that are impregnated into poly(methyl methacrylate) (PMMA) bone cement to treat PJIs and are primarily used because of their high-heat stability that can withstand the heat generated during the polymerization of methyl methacrylate10. Despite the widespread use of aminoglycosides, they are very susceptible towards developing a resistance towards

Methicillin-resistant Staphylococcus aureus (MRSA) isolates10. The prevalence of drug- resistant bacteria is exacerbated with the use of tobramycin and gentamicin in PMMA bone cement due to the nature of the drug release where after long periods sub-inhibitory concentrations are released10. Therefore, there is a great need to find alternative drugs to

incorporate into PMMA bone cement that are less susceptible to developing resistance towards PJI staphylococci10.

Figure 1-2: Proliferation of a single resistant bacterium when repeatedly treated with antibiotic. One mechanism in which bacterial resistance can spread is if a slight mutation develops in the antibiotic binding site on the bacterium and antibiotics are administered.

1.1.5. REVISION PROCEDURES/INITIAL TREATMENT FAILURES

If standard antibiotic treatment (systemic and oral administration) fails to adequately treat the PJI, a surgical revision procedure is required in a one- or two-stage process. In a one- stage process, the initial implant is removed and immediately replaced with a new implant. The two-stage process is a slower procedure where the initial implant is removed, the infection is locally treated with an antibiotic filled cement spacer for 12-26 weeks, and the implant is replaced after this time4,13. Typically, two-stage revision procedures are preferred over one-stage, reflected in their average survival rates (one- stage: 82%, two-stage: 91%)13. However, from a patient’s perspective the two-stage

procedure can have serious limitations on their mobility and comfort and the delayed re- implantation procedure can be challenging due to scar formation and disuse of the tissue surrounding the implant13. One of the predominant limitations of the two-stage revision procedure with antibiotic-filled spacer is that there often is a burst release of drug initially from the spacer and at later times very little drug is released at levels below the minimum inhibitory concentration (MIC) of the drug which can drive the development of drug- resistant bacteria17,18.

1.1.6. SUMMARY

While PJIs occur in only 0.6-2.2% of total joint arthroplasties3, they can be complex and invasive to treat due to a variety of factors. Success in the treatment of PJIs primarily depends upon the overall health of the patient, lack of presence of bacterial biofilm or drug-resistant bacteria, ease of isolating the pathogenic organism, and rapid diagnosis following the initial onset of the infection. The lack of a uniform and universally implemented system for diagnosing PJIs has substantially contributed to the relatively slow and sometimes incorrect diagnosis of PJIs. Overall, the surgeon’s ability to effectively treat PJIs is limited to their knowledge of all of the above factors.

1.2. NOVEL TREATMENT MODALITIES FOR PJIS

1.2.1. LIMITATIONS WITH TRADITIONAL ANTIBIOTIC-LADEN PMMA BONE

CEMENT

While assisting in implant fixation, PMMA bone cement has been historically been simultaneously utilized as a vehicle for enabling localized delivery of antimicrobial agents to the periprosthetic joint environment. Over the past 50 years, antibiotics have been directly added to the PMMA in their powdered form during the preparation of the bone cement19. As shown in Figure 1-3, PMMA is generally surgically prepared by combining a liquid component comprised of methyl methacrylate monomer with a powdered component comprised of benzoyl peroxide (initiator) and pre-polymerized

PMMA beads, resulting in a free-radical polymerization.

Figure 1-3: Free-radical polymerization of PMMA bone cement upon combination of liquid (methyl methacrylate monomer) and powdered component (benzoyl peroxide initiator and pre-polymerized PMMA beads). 38

Despite the widespread use of antibiotic-laden PMMA to prevent and treat PJIs following arthroplasties, it has several limitations associated with its drug release kinetics.

Specifically, due to the relatively low porosity of PMMA (~1%) it is difficult for drugs to effectively diffuse out unless they reside in the outer periphery of the PMMA20.

Specifically, one study from Martinez-Moreno demonstrated at after 1344 hours nearly

95% of antibiotic initially added to the PMMA remained entrapped, deep within the center of the PMMA21. It has been demonstrated that antibiotics are only released from the outer 50-100 μm of PMMA22. Figure 1-4 depicts the interior of PMMA containing antibiotic (orange diamonds) before and after releasing drug from the sample. After prolonged periods of release, drug has only diffused out of the outer periphery of the

PMMA (~100 μm).

Figure 1-4: Interior of antibiotic-laden PMMA (orange diamonds) before and after releasing some antibiotic from the sample. Drug is released primarily from the outer ~100 μm, with the rest remaining trapped deep in the interior. 39

With drug remaining entrapped in the PMMA, it can result in prolonged sub-therapeutic release of the antibiotic, further driving antibiotic-resistance. Typically, antibiotic released from PMMA exhibits a “burst” release early on followed by an extended period of sub-therapeutic release. Ideally, to effectively eradicate infections a consistent therapeutic dose of the antibiotic is required for a period lasting a minimum of several weeks, however, this is highly dependent upon the species of bacteria and antibiotic used.

Additionally, PMMA cannot be filled with additional antibiotics after it is implanted in the patient. Therefore, the window of therapeutic activity of the implanted sample is limited by the initial amount of drug added to the system and if an infection develops much later on (i.e. months – years) more drug cannot be added to the system without removing the implant. Given these limitations of PMMA bone cement, a variety of alternative biomaterials have been investigated for treatment of PJIs.

1.2.2. COMMERCIALLY AVAILABLE ALTERNATIVE BIOMATERIALS FOR

ANTIBIOTIC-LADEN PMMA BONE CEMENT

A variety of commercially available biodegradable bone graft antibiotic-eluting substitutes are attractive alternatives to PMMA for filling bone voids and treating PJIs.

Calcium sulfate has been molded into radiopaque beads which resorb in 30-60 days, eliminating the necessity of a secondary removal surgery. The innate advantage of calcium sulfate over PMMA is that preparation does not cause high polymerization temperatures, theoretically enabling a wider range of antibiotics to be incorporated into it.

Furthermore, in vitro antibiotic release studies from calcium sulfate have exhibited 40

equivalent or better elution characteristics compared to PMMA23,24. Unfortunately, clinical studies using calcium sulfate beads have identified several cases of persistent serious wound drainage and heterotopic ossification secondary to possible hypersensitivity reactions25,26. Clinical data supporting regular use of calcium sulfate for antibiotic delivery in joint replacement surgery is limited. Commercially prepared calcium sulfate with tobramycin (OSTEOSET, Wright Medical) has been approved for use in other countries; however, antibiotic-laden calcium sulfate is not yet approved for use in the United States.

Additional biodegradable bone substitute materials have been investigated for the treatment of orthopedic infections, including bioactive glass, and various combined biocomposites27. However, the FDA has not approved most agents for antibiotic delivery, since evidence for their clinical efficacy is lacking25. Moving forward, a lingering limitation of existing resorbable bone substitutes is that the rate of resorption is typically faster than the ingrowth of surrounding bone. Hence, resorbable bone substitutes only provide mechanical support for a relatively short period of time, before bone can effectively remodel and therefore are not appealing solutions for revision joint replacement constructs which are expected to last for the long-term.

1.2.3. TITANIUM NANOTUBE ARRAYS

Local delivery of antibiotics using nanotube arrays processed on the surface of titanium represents another innovative method of preventing PJIs. In vitro experiments have

demonstrated sustained release of gentamicin for a duration of 11 days from the surface of titanium alloys coated with nanotube arrays28. Similarly, a non-antibiotic antimicrobial peptide with broad-spectrum activity has demonstrated sustained release from titanium nanotubes for 7 days.

1.2.4. IMPLANT COATINGS WITH ANTIBACTERIAL EFFECT

As an alternative to drug-eluting antibacterial coatings and nanotubes, several groups have developed orthopedic coatings that intrinsically repel or kill bacteria. Silver nanoparticles are one of the most common non-antibiotic antibacterial coatings29–40. In these settings, silver ions leach from the coating, diffuse into bacterial cells, and damage enzymes in the cell37. Specifically, several nanosilver formulations have been developed in which the silver is slowly solubilized over an extended period of time, preventing excessive local concentrations37. Only 7% by weight of the loaded silver was released from their coatings after 28 days, demonstrating the long-term antibacterial activity potential of their coating33. Other similar coatings have incorporated silver nanoparticles into hydroxyapatite and chitosan to create an antibacterial coating that promotes osseointegration31,32. These coatings have shown over 90% reduction in both E. coli and

S. aureus surrounding the implant32. Nevertheless, many of these novel materials have yet to be clinically evaluated in an orthopedic setting41.

1.2.5. POLYMERS/HYDROGELS

Synthetic polymers such as poly(lactic acid), poly(caprolactone), and poly(lactic-co- glycolic acid) (PLGA) have been studied in in vitro and in vivo models to provide sustained release of antibiotics via diffusion and hydrolytic bulk degradation25. The most studied synthetic polymer, PLGA, has shown greater efficacy for antibiotic delivery compared to systemic administration and local delivery using PMMA25. Alternatively, poly(caprolactone)-poly(quaternary ammonium salt) micelles are copolymers that have an intrinsic antibacterial effect. When combined with the antibacterial drug triclosan, investigators demonstrated that these micelles had the ability to function as carriers for drug delivery and had a synergistic antimicrobial effect against E. coli42. Finally, polymeric drug-eluting “sleeves” or covers for orthopedic devices have been patented by

Boston Scientific Inc. and Control Delivery Systems Inc.; however, they have not been widely implemented in clinical settings43,44.

Recent innovations have allowed researchers to coat an implant with a quickly resorbable hydrogel that contains one or more antibiotics, thereby delivering the therapy without impeding osseointegration45. Drug delivery using alginate-based hydrogel matrices has been compared to that using composite gelatin and hydroxyapatite matrices. In vitro sustained release of ciprofloxacin increased from 5 hours in the composite matrix to 10 days in the alginate matrix46.

1.2.6. CYCLODEXTRIN (CD)-BASED DRUG DELIVERY

Cyclodextrins are another class of drug delivery polymers with the potential for use in orthopedic applications. Cyclodextrin (CD) is a cyclic oligosaccharide comprised of 6-8 glucose monomers with a hydrophobic interior and a relatively hydrophilic exterior47.

When CD is cross-linked as an insoluble polymer, pharmaceuticals can form an inclusion complex where the CD “pockets” facilitate a controlled and prolonged release of the drug47. Figure 1-5 depicts a cartoon rendering of the structure and simplified synthesis of cross-linked CD polymer microparticles.

Figure 1-5: Structure of soluble CD monomer, soluble pre-polymerized CD, and cross-linked insoluble CD polymer microparticles.

Drug release from CD polymers is based on chemical binding affinity between the drug and the CD pockets. The affinity-based release from CD pockets does not follow the burst release pattern observed with the more common diffusion-based release delivery systems, thereby allowing for slow drug release even from thin coatings48–50. In its insoluble form, CD has been incorporated into a number of non-orthopedic implants

including hernia meshes51, vascular grafts52, and stents53 to deliver antibiotics over an extended period of time54–61. Furthermore, recent in vitro efforts have shown that CD biomaterials may be refilled with pharmaceuticals even when the material is covered by a biofilm, allowing for additional windows of therapeutic treatment long after implantation62. Figure 1-6 depicts how CD has the ability to be refilled with antibiotics following implantation into soft tissue, whereas traditionally-used materials for PJI treatments do not have this ability (i.e. PMMA bone cement) and Figure 1-7 depicts the difference in drug release profiles for CD (affinity-based) versus PMMA (diffusion only- based) polymer delivery systems.

Figure 1-6: Antibiotic refilling of PMMA versus cross-linked CD polymer when implanted in soft tissue.

Injected drug is able to selectively bind to embedded CD polymer relative to surrounding tissue due to the thermodynamic driving force of binding affinity of drug molecules for CD “pockets”. With PMMA, there is no thermodynamic driving force for the drug to favorably bind to the polymer relative to the surrounding tissue and it thus does not exhibit refilling capabilities.

This refilling functionality enables CD to be implanted without drug and filled with a selected drug or antibiotic “on demand” via a simple external bolus injection into nearby 45

tissue. Refilling enables clinicians to administer antibiotics selectively to indicated patients (i.e. only if the patients present with infection), rather than exposing all patients to systemic antibiotics. Preliminary in vivo animal studies have demonstrated the effectiveness of refilling for CD-coated hernia meshes61.

Figure 1-7: Daily release profile from CD polymers as compared to drug release through diffusion-only

(e.g. PMMA). Diffusion-based release results in a large bolus of drug released in the first few days

(potentially cytotoxic) and sub-inhibitory release at later times. In contrast, release from CD polymers is affinity-based, extending release within the desired treatment window for longer periods of time and avoiding an initial bolus effect (a). Polymers that use affinity-based release (i.e. CD) may be refilled after implantation, further extending the presence of the pharmaceutical within the treatment window.

For orthopedic applications, CD has been functionalized to hydroxyapatite56,58,61, incorporated into a chitosan nanoparticle coating54, and coated on bone screws and

Kirschner wires55. Specifically, Thi et al. and Lepretre et al. grafted hydroxypropyl-β-CD 46

onto microporous hydroxyapatite discs and loaded the discs with ciprofloxacin and vancomycin56,58. In this study, antibiotics were released into a simulated in vivo environment to evaluate the drug distribution at the infection site in an effort to develop more effective treatments for osteomyelitis58. Implants functionalized with CD showed a sustained antibiotic release over 150-300 hours. Lepretre et al. demonstrated that the CD- functionalized implants had increased bacteriostatic activity against S. aureus and improved osteoblast cytocompatibility56. Similarly, Taha et al. grafted CD onto hydroxyapatite-coated titanium hip implants and loaded both tobramycin and rifampicin in combination61. The goal of the dual drug delivery system was to have broad-spectrum coverage against both S. aureus and Enterobacter cloacae. The system was capable of sustained release of both antibiotics. Furthermore, Taha et al. have also demonstrated antibacterial activity of gentamicin-loaded plasma-sprayed hydroxyapatite-coated titanium functionalized with CD61. In another orthopedic application, Mattioli-Belmonte et al. developed a chitosan/CD nanoparticle coating for titanium54. The chitosan nanoparticles were functionalized with sulfobutyl ether β-CD and were capable of a 20- fold reduction in two S. aureus strains in vitro and a sustained release over 7 days.

Additionally, the nanoparticle coating demonstrated no significant cytotoxicity towards osteoblasts54. Furthermore, Thatiparti et al. directly coated both stainless steel bone screws and Kirschner fracture fixation wires with CD cross-linked with hexamethylene diisocyanate (HDI) and 2-isocyanatoethyl 2,6-diisocyanatohexanoate55. The devices were loaded with rifampicin (RMP), novobiocin, and vancomycin and demonstrated a 47

sustained release over 200 hours. Coated devices were capable of inhibiting the growth of

S. aureus over 28 days. While the coatings appear to be a promising technology, they have yet to be evaluated in translational in vivo and clinical settings.

1.2.7. CONCLUSIONS

PJI is a rare, but devastating complication of total joint arthroplasty that can negatively change patients’ lives and accrue significant cost to the healthcare system. Current clinical treatments for PJI involve morbid surgery, ineffective antibiotic delivery vehicles that have been shown to fail up to 22% of the time. Novel treatment modalities for PJI mostly target optimizing the implant surface to prevent bacterial biofilm formation or providing prolonged intra-articular antibiotic dosing to eradicate planktonic bacteria. CD polymers in particular have facilitated great advancements in other disorders and show early promise as an eventual option for clinical use for the treatment of PJIs.

1.3. ACKNOWLEDGEMENTS

This work was supported by the National Institute of Arthritis and Musculoskeletal and

Skin Diseases of the National Institutes of Health (NIH) under award number T32

AR007281. This content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

CHAPTER 2. ANTIBIOTIC REFILLING OF PMMA BONE CEMENT

THROUGH INCORPORATION OF CYCLODEXTRIN FOR TREATMENT OF

PERIPROSTHETIC JOINT INFECTION

*This chapter was reprinted (adapted) with permission from: Advanced Healthcare

Authors: Erika L. Cyphert, Greg D. Learn, Sara K. Hurley, Chao-yi Lu, Horst A. von

Recum

Author contributions: ELC – experimental design, data analysis and collection (all experiments), preparation of figures, writing/revision manuscript. GDL – experimental design (compression testing and micro-CT), data analysis and collection (compression testing and micro-CT), fabrication of cylindrical mold, writing manuscript (compression testing and micro-CT), revision manuscript. SKH and C-yL - data collection. HAvR – experimental design, data analysis, revision manuscript.

2.1. ABSTRACT

Poly(methyl methacrylate) (PMMA) bone cement has been used in several biomedical applications including as antibiotic-filled beads, temporary skeletal spacers, and cement for orthopedic implant fixation in arthroplasties. To mitigate development of PJIs following surgery, antibiotics are mixed into PMMA to achieve local delivery. However, since implanted PMMA is often structural, incorporated antibiotics must not compromise mechanical properties; limiting the selection of compatible antibiotics. Furthermore, antibiotics cannot be added to resolve future PJIs once PMMA is implanted. Finally, delivery from PMMA has been shown to be suboptimal as incorporated antibiotics exhibit early burst release with most of the drug remaining permanently trapped. This prolonged sub-therapeutic dosage can drive antibiotic resistance in pathogens. To overcome limitations of antibiotic-laden PMMA, insoluble cyclodextrin (CD) microparticles were incorporated into PMMA to provide more sustained delivery of a broader range of drugs, without impacting mechanics. PMMA-CD composites were synthesized and filled with one of three antibiotics and evaluated using zone of inhibition, drug release, and compression studies. Additionally, ability of PMMA-CD composites to serve as a refillable antibiotic delivery depot was explored. Findings suggested that addition of CD microparticles to PMMA promoted post-implantation antibiotic refilling and enabled incorporation of previously incompatible antibiotics while preserving favorable mechanical properties.

2.2. INTRODUCTION

Poly(methyl methacrylate) (PMMA) bone cement has been clinically implemented for a variety of applications including in orthopedic surgery as a temporary spacer following surgical debridement in two-part revision procedures63 and at the interface between the surface of a metallic implant and the patient’s skeleton to promote implant fixation and to assist in load transfer to bone in arthroplasties64. Beyond these hard-tissue applications,

PMMA cement has also been utilized in the treatment of soft tissue PJIs in the form of small antibiotic impregnated beads strung together in long chains and embedded in tissue65–68.

There is a prevalent need for an effective antibiotic-laden PMMA for the treatment of

PJIs. Specifically, over 1 million arthroplasties, including total hip and knee replacements, are performed annually in the United States69. With the aging U.S. population, this annual number is projected to increase to 3.48 million by 203070. While the incidence of metallic implant-related infections such as PJIs and osteomyelitis is below 1%, these complications are a considerable source of patient morbidity and mortality71. Eradication of PJIs is challenging as it requires long-term systemic pharmaceutical intervention and probable implant removal with one-stage or two-stage replacement72. However, since systemic antibiotic administration can lead to systemic toxicity and drive bacterial antibiotic resistance, local antibiotic delivery methods are

preferred and are primarily implemented through the use of antibiotic-filled PMMA bone cement73.

Despite the lack of a standardized protocol for its usage, antibiotic-laden PMMA is still widely used in arthroplasties to mitigate PJIs74. Specifically, antibiotic-filled PMMA beads are used by surgeons in more than 50% of two-stage arthroplasty revision procedures, and antibiotics are incorporated into 32% and 69% of temporary PMMA spacers in revision procedures of hips and knees, respectively74. Commercial antibiotic- filled PMMA typically contains either aminoglycosides (tobramycin and gentamicin) or glycopeptides (vancomycin). The concentration of antibiotic selected depends upon the intended use of the PMMA. Low doses of incorporated antibiotics may lead to the development of antibiotic-resistant bacteria over extended periods of time75, whereas high doses have been shown to have a detrimental effect on the compressive76 and shear77 mechanical properties of the PMMA.

PMMA cement is surgically prepared by mixing a pre-polymerized powder with a liquid monomer, forming a dough which quickly solidifies, attaining temperatures of 60-80oC in the process78. The high curing temperature results from the exothermic free-radical polymerization of PMMA, and necessitates that antibiotics used in PMMA be thermally stable to remain functional. Aminoglycosides (e.g., tobramycin and gentamicin) are one of the few classes of antibiotics compatible with PMMA that meet this requirement79.

While aminoglycoside-filled PMMA formulations are regularly used in clinical

applications, they nevertheless have inherent limitations. Specifically, Moojen et al. found sub-therapeutic release of gentamicin and tobramycin from PMMA spacers after 1 week (below MIC)80, demonstrating limited utility for treatment of advanced infections.

Beyond such transient release periods for incorporated therapeutics, PMMA has not demonstrated antibiotic refilling capabilities, meaning that there is a limited and suboptimal window of time during which the drug-filled PMMA can treat infections once implanted. The lack of antibiotic refillability precludes PMMA from being utilized in the treatment of chronic or latent infections and necessitates the use of revision surgeries to treat delayed or complicated infections.

Various strategies have been attempted to improve upon the antibiotic release kinetics

(i.e., amount of drug released and duration of release) from antibiotic-filled PMMA. To improve the permeability and subsequent antibiotic elution from PMMA, soluble porogens (e.g., xylitol) have been added as fillers to PMMA81,82. Additionally, an antimicrobial peptide, Dhvar-5, has been incorporated into PMMA and was shown to enable a more consistent and prolonged release of gentamicin and increased the total amount of gentamicin released by increasing the micro-porosity of the PMMA83. Yet, when manipulating the porosity and permeability of the PMMA, it is important to consider the delicate balance between the drug release characteristics of the PMMA and its mechanical properties, as voids in the PMMA can act as stress risers.

Cyclodextrins (CDs) are cyclic oligosaccharides composed typically of 6 to 8 glucose units arranged in a ring, forming a toroid structure containing a nonpolar interior pocket and a relatively polar exterior47. Insoluble polymerized CD has demonstrated the ability to form inclusion complexes (i.e., be “filled”) with a variety of antibiotics including

RMP, vancomycin, and erythromycin, resulting in an extended, affinity-based release when compared to release systems governed by diffusion alone (e.g., PMMA)48–50,84.

Therefore, the incorporation of polymerized CD into PMMA may extend and enable more control over the drug release profile of the filled antibiotic. CD polymers have also been shown to be capable of being refilled with drug once implanted85, making them an attractive option to incorporate into implanted medical devices, such as in the treatment of PJIs.

The hypothesis of the present study was that incorporation of CD microparticles into

PMMA would increase and prolong the effectiveness of incorporated antibiotics and enable drug refilling while reducing the impact of antibiotics on PMMA mechanics. In this work, polymerized CD microparticles were added into PMMA in an effort to 1) improve the drug release properties of PMMA, 2) enable a wider variety of antibiotics to be incorporated into PMMA without compromising mechanical properties, and 3) to permit future drug refilling of PMMA following implantation. Figure 2-1 provides a graphical overview of the analyses carried out on PMMA samples.

Figure 2-1: Schematic outlining the scope of the work. PMMA-CD composites were synthesized and evaluated in a variety of studies and compared to properties of PMMA without CD microparticles.

2.3. MATERIALS AND METHODS

2.3.1. MATERIALS

Lightly epichlorohydrin cross-linked β- and γ-CD pre-polymers (116 kDa molecular weight) were purchased from CycloLab (Budapest, Hungary). Ethylene glycol diglycidyl ether was purchased from Polysciences Inc. (Warrington, PA). Rifampicin (RMP) was purchased from Research Products International (Mt. Prospect, IL). Gentamicin and tobramycin were purchased from Fisher Scientific (Pittsburgh, PA). Simplex HV (high

viscosity) radiopaque bone cement from Stryker Orthopaedics (Mahwah, NJ) was kindly provided by Dr. Mathias Bostrom (Hospital for Special Surgery, NY). Green fluorescent protein (GFP)-labeled S. aureus stock culture was kindly provided by Dr. Edward

Greenfield (Case Western Reserve University, OH). PTFE stock for cylindrical specimen molds was purchased from McMaster-Carr (Aurora, OH). All other reagents and solvents were purchased from Fisher Scientific in the highest grade available.

2.3.2. SYNTHESIS OF INSOLUBLE POLYMERIZED CYCLODEXTRIN (CD)

MICROPARTICLES

1 gram of lightly cross-linked epichlorohydrin CD (β- or γ-CD) was dissolved in 4 mL of

0.2 M potassium hydroxide. 1.6 mL of ethylene glycol diglycidyl ether cross-linker was added to the dissolved polymer and the solution was vortexed for 2 minutes. The polymer solution was poured into a heated mixture (60oC) of 50 mL of light mineral oil with a mixture of Tween 85 (24%) and Span 85 (76%) (500 μL) and stirred at 1000 rpm. The emulsion was heated at 60oC for 4 hours. Microparticles were serially washed and centrifuged (600 rpm, 5 minutes) with solvents of gradually increasing polarity (2x each with light mineral oil, hexanes, acetone, MilliQ water). Washed microparticles were frozen with 35 mL MilliQ water and lyophilized for 3 days. Dried microparticles were stored in a desiccator prior to use in further experiments. Average microparticle diameter was previously established as ~250 μm)86.

2.3.3. SYNTHESIS OF INSOLUBLE POLYMERIZED β-CD DISKS

Polymerized β-CD disks were synthesized according to a previously established protocol49. Briefly 1 gram of β-CD pre-polymer was dissolved in 4 mL dimethylformamide (DMF) and cross-linked with hexamethylene diisocyanate (HDI) in a molar ratio of 1:0.16 (glucose residue:HDI). The solution was cured at 70oC for 45 minutes and punched into 6 mm disks. Disks were sequentially washed over several days using 100% DMF, 50:50 DMF:MilliQ water, and 100% MilliQ water to remove unreacted DMF and pre-polymer.

2.3.4. MICROPARTICLE DRUG FILLING

40 mg samples of β- or γ-CD microparticles were placed in 1.5 mL Eppendorf tubes and filled with 1 mL of 5 mg/mL of RMP in DMF or 1 mL of 15 mg/mL of tobramycin and gentamicin in phosphate buffered saline (PBS). Samples were placed on a rotisserie shaker for 3 days at room temperature, centrifuged for 8 minutes, and washed with MiliQ water (RMP: 3x; tobramycin and gentamicin: 1x). Washed microparticles were frozen and lyophilized for 3 days. Dried drug-filled microparticles were stored in a desiccator prior to use in further experiments.

2.3.5. SYNTHESIS OF PMMA-CD SPHERICAL BEADS AND CYLINDERS

4 gram samples of Simplex HV surgical grade bone cement powder were reacted at room temperature with 2 mL of methyl methacrylate monomer for each batch of PMMA, according to manufacturer instructions. Upon adding the liquid monomer, antibiotics

(1.25% wt/wt gentamicin, tobramycin, or RMP/PMMA) or insoluble polymerized empty and drug-filled CD microparticles (β- or γ-CD, 5 or 10 wt%) were hand-mixed into the

PMMA mixture until a soft dough formed. PMMA dough containing CD microparticles was pressed into a thin sheet and formed into 6 mm spherical beads, and dried at room temperature.

Cylindrical samples were fabricated and tested in accordance with ASTM F451-1687.

PMMA was formed into cylindrical samples 6 mm in diameter and 12 mm in height with the use of a custom-machined two-part PTFE mold. Uncured PMMA dough was hand- mixed and firmly finger-pressed into the cylindrical wells of the PTFE molds. Dough was allowed to cure for a minimum of 1 hour prior to removal of the solidified specimens from the mold. Samples were visually inspected and only samples that contained no visible defects larger than 0.5 mm in major diameter were considered for testing. Both ends of each sample were sanded flat and smooth using wet 240 grit (P280) silicon carbide sandpaper maintained perpendicular to the cylinder axis. Individual samples were weighed and their dimensions were measured (length and diameter) to the nearest 0.01 mm using digital calipers prior to testing for determining stress and strain, with diameter taken at the cylinder mid-length.

2.3.6. PERSISTENCE ZONE OF INHIBITION STUDY

PMMA-CD composite spherical beads (6 mm diameter) with antibiotic freely added in

(1.25% wt/wt drug/PMMA powder: gentamicin, tobramycin, RMP) and PMMA spherical

beads with either 5 or 10 wt% β- or γ-CD microparticles filled with each of the three antibiotics were placed on the center of a Trypticase soy agar plate with 70 μL of S. aureus culture49. Each condition was repeated in triplicate. Agar plates were incubated at

37oC overnight and zones of bacterial clearance surrounding each drug-filled PMMA-CD spherical bead were measured (from the edge of the bead to the edge of the bacterial clearance) using calipers and averaged. PMMA-CD spherical beads were transferred to a freshly seeded agar plate and the measurement process was repeated until the zones of clearance were no longer visible.

2.3.7. QUANTIFICATION AND ACTIVITY OF DRUG RELEASE ALIQUOTS

Drug-filled PMMA spherical beads with each of the three antibiotics (with and without

CD microparticles) were placed in 1 mL of release solution (PBS: gentamicin and tobramycin; or PBS with 10% DMF: RMP) and agitated at 37oC. Due to the hydrophobic nature of RMP, a hydrophobic sink (DMF) was added to the release media. After set time points (2 hours, 4 hours, 1 day, 2 days, etc.) the entire release solution (1 mL) was removed and replaced with 1 mL fresh solution to maintain infinite sink conditions. Drug concentration in the release solution was quantified using 200 μL samples of release media with UV absorbance spectroscopy (400 nm: gentamicin and tobramycin; 473 nm:

RMP) using a Biotek 96-well plate reader (H1: Winooski, VT). In order to quantify gentamicin and tobramycin concentration with absorbance spectroscopy, a ninhydrin

reaction was performed prior to scanning the samples88,89. Drug concentrations were determined using standard curves.

The subsequent antimicrobial activity at each time point for the drug release samples

(each aliquot) was also evaluated against S. aureus. A 6 mm filter paper disk was soaked in each drug release aliquot solution and a zone of inhibition study was carried out for a period of 1 day against S. aureus90. This experiment enabled the antimicrobial activity of each release aliquot to be compared to the amount of drug released at a given time point to ensure that the amount of drug being released exceeded the MIC.

2.3.8. COMPRESSION TESTING

Cylindrical PMMA-CD samples were loaded in compression at 20 mm/min on a mechanical testing frame (Material Testing System MTS 810, MTS Systems

Corporation). Data was sampled at 200 Hz and analyzed for ultimate load, ultimate compressive strength, stiffness, compressive modulus, strain to ultimate load, and work to ultimate load. For free RMP samples, due to the lack of a local maxima beyond the elastic limit in their load-displacement curves, yield load was defined a 2.0% offset load.

For all other groups, ultimate load was defined at the local maxima beyond the elastic limit in load-displacement curves.

2.3.9. PMMA ANTIBIOTIC REFILLING IN AGAROSE TISSUE PHANTOM STUDY

Tissue-mimicking agarose models were created by dissolving 0.075 wt/vol% agarose in

PBS and bringing it to a boil. 5 mL hot agarose was added to each well of a Costar 6-well cell culture plate (Corning Life Sciences, Corning, NY). Once the agarose solidified, a single PMMA spherical bead (with or without CD microparticles) and a polymerized β-

CD disk were placed on top of the agarose. An additional 5 mL of hot agarose was placed on top and allowed to set at room temperature. A 6 mm well was punched into the center of the agarose above the implanted sample(s). 100 μL of drug solution (6 mg/mL in methanol: RMP) was injected into the preformed well in the agarose and the samples were covered and agitated at 37oC for 48 hours. Drug-filled PMMA bead implants were removed from the agarose after 48 hours and placed on a freshly seeded S. aureus plate to evaluate the duration of their antimicrobial activity using a persistence zone of inhibition study (see Section 2.3.6). Each condition was repeated in triplicate.

2.3.10. QUANTIFICATION OF ANTIBIOTIC REFILLING EFFICIENCY

PMMA spherical beads filled with drug (with and without CD microparticles) from the agarose refilling study (see Section 2.3.9) were placed in 4.2 mL of DMF and agitated at

37oC overnight in an effort to leach out all of the filled drug. Antibiotic solution from the dissolved PMMA spherical beads was diluted and the concentration of drug was quantified using absorbance spectroscopy and standard curves.

2.3.11. STATISTICAL ANALYSIS

All data displayed is presented as the mean of each condition tested in triplicate (zone of inhibition, drug release, agarose refilling) or n ≥7 (compression testing) with the standard deviation as the error bars. Statistical analysis tests were carried out in Microsoft Excel

2016. A two-tailed Student’s T-test with unequal variances was used to evaluate the zone of inhibition (n = 3) and RMP agarose refilling (n = 3) data. A one-tailed Student’s T-test with unequal variances was used to evaluate the compressive strength (n = 7). T-test p values < 0.05 were considered to be statistically significant.

2.4. RESULTS

2.4.1. PERSISTENCE ZONE OF INHIBITION STUDY (PRE-FILLED PMMA-CD

BEADS)

The antimicrobial activity of each antibiotic (RMP, gentamicin, tobramycin) added either freely or filled in CD microparticles (β- or γ-CD, 5 or 10 wt%) into PMMA spherical beads during polymerization was evaluated against S. aureus in triplicate. Figure 2-2 demonstrates that clinically used formulations (i.e. gentamicin or tobramycin freely added into PMMA) resulted in a very short duration of antibacterial activity (only ~12 days of bacteria clearance). Meanwhile, RMP, an antibiotic currently incompatible with

PMMA in its free state, was able to clear bacteria for a much more substantial duration of time (i.e., > 100 days). Furthermore, the RMP zones of inhibition were significantly larger than those of gentamicin and tobramycin for all time points after 4 days (p < 0.05; 62

except for days 5 and 12). Thus, confirming the long-lasting and high potency of RMP against S. aureus91.

Figure 2-2: Persistence zone of inhibition study of antimicrobial activity of three antibiotics (RMP, tobramycin, and gentamicin) freely added into PMMA beads against S. aureus.

Similar zone of inhibition studies were also carried out for PMMA-CD composite spherical beads filled with the three antibiotics. Figure 2-3 depicts the duration of antimicrobial activity of each of the three antibiotics filled into both types of CD (β- or γ-

CD) in 5 or 10 wt%. The addition of 10 wt% RMP-filled CD microparticles (β- or γ-) to

PMMA beads yielded antimicrobial activity for more than 60 days, making this formulation a potential candidate for the treatment of chronic osteomyelitis and PJIs92. In the case of gentamicin and tobramycin filled PMMA-CD beads, incorporation of drug-

filled CD microparticles resulted in comparable antimicrobial activity to PMMA samples with the same drugs freely added in.

Figure 2-3: Persistence zone of inhibition study of duration of antimicrobial activity of PMMA-CD composites filled with RMP, tobramycin, or gentamicin against S. aureus.

2.4.2. QUANTIFICATION AND ACTIVITY OF DRUG RELEASE ALIQUOTS

Antibiotics released from PMMA spherical beads containing either free antibiotics or antibiotic-filled CD microparticles were collected at discrete time points over several weeks for gentamicin and tobramycin formulations. In an effort to validate the therapeutic efficacy of each of the release sample aliquots, antimicrobial susceptibility testing was completed at each time point against S. aureus bacteria90. Figure 2-4 depicts the combined release and activity (antimicrobial susceptibility testing) plots for gentamicin and tobramycin freely added into PMMA and complexed with different amounts of CD microparticles. The black line represents the normalized drug release profile (mass of drug released/mass of entire PMMA bead) and the red line indicates the antimicrobial zone of inhibition at each time point for each aliquot. Most of the plots demonstrated a relatively decreasing linear trend where the mass of drug released and antimicrobial activity decreased as the release time points increased, confirming the relationship between drug concentration and antimicrobial susceptibility. Generally, release aliquots were considered to have antimicrobial activity if their zone of inhibition

(measured from the edge of the bead to the edge of bacteria clearance) was > 1 mm.

When the antibiotics were freely added into PMMA beads (without CD) drug release and antimicrobial therapeutic activity was completed within 2-3 weeks with all samples.

Figure 2-4: Daily drug release profiles and antimicrobial susceptibility zone of inhibition testing with a) tobramycin and b) gentamicin freely added into PMMA beads and from c-g) antibiotic-filled PMMA-CD beads (β- and γ-CD, 5 and 10 wt%).

The release of RMP was not reported due to its high hydrophobicity and affinity for

CD93. As a result, RMP exhibited little to no elution from PMMA in PBS with 10 wt%

DMF and other slightly more hydrophobic sinks, such as Tween.

2.4.3. COMPRESSION TESTING

To evaluate the mechanical strength of different PMMA formulations with and without

CD microparticles, seven different PMMA groups (pure PMMA, PMMA with free tobramycin, PMMA with free RMP, PMMA-CD with empty microparticles, and PMMA-

CD with RMP-filled microparticles) were fabricated into cylinders (n≥7). Gentamicin and tobramycin are currently added into PMMA in clinically used formulations, so mechanical properties of PMMA-CD filled with these drugs were not evaluated in this study, as it would not provide additional information.

Figure 2-5 depicts the load-displacement curves for all 7 groups (left) and the zoomed-in curve for free RMP (right). Table 2-1 includes the quantification of the ultimate compressive strength, modulus, work to peak load, and strain to peak load.

Figure 2-5: Representative compressive load-displacement curves of PMMA-CD cylinders superimposed

(left) and zoomed in curve of free RMP in PMMA (right).

Comparable compressive strength was observed for pure PMMA, PMMA with free tobramycin, and PMMA with either 5 or 10 wt% RMP-filled CD microparticles. A decreased compressive strength resulted upon addition of either 5 or 10 wt% empty CD microparticles and the compressive strength dramatically decreased upon addition of free

RMP. While most PMMA samples demonstrated plastic deformation characteristics,

PMMA with free RMP demonstrated solely elastic deformation characteristics.

Table 2-1: Mechanical testing of cylindrical PMMA-CD samples. Statistically significant difference of samples relative to free tobramycin (§), 5 wt% empty β-CD (†), and 10 wt% empty β-CD (⸸).

Compression Test Ultimate Modulus (MPa) Work to Peak Strain to Peak Specimen Compressive Load (J) Load Strength (MPa) Plain PMMA 79.58 ± 1.00§ 2409.74 ± 99.61§ 1.009 ± 0.075§ 0.055 ± 0.003§ Free Tobramycin 77.03 ± 1.17 2318.07 ± 68.84 0.803 ± 0.130 0.058 ± 0.001 Free RMP 1.88 ± 0.15§ 48.68 ± 12.14§ 0.015 ± 0.008§ 0.042 ± 0.010§

5wt% empty β-CD 70.19 ± 1.44§ 2120.04 ± 50.46§ 0.879 ± 0.108 0.057 ± 0.002 10wt% empty β- 67.37 ± 1.05§† 1993.50 ± 0.853 ± 0.036 0.057 ± 0.002 CD 45.95§† 5wt% RMP-β-CD 76.22 ± 2.27† 2285.81 ± 0.971 ± 0.036§† 0.056 ± 0.002§ 101.67† 10wt% RMP-β-CD 75.80 ± 1.59⸸ 2302.55 ± 36.28⸸ 0.976 ± 0.065§⸸ 0.056 ± 0.003

Since there was a statistically significant difference between the compressive strength of pure PMMA and PMMA with free tobramycin (p = 0.0005) and both formulations are clinically used, PMMA with free tobramycin was used as the clinically relevant comparison for other antibiotic-incorporated experimental groups. Furthermore, PMMA with free tobramycin was selected as the control group since all experimental groups had antibiotics or additives (i.e., CD microparticles) incorporated into them. Addition of either 5 or 10 wt% empty CD microparticles resulted in a significant decrease in the compressive strength (5 wt%: 70.19 ± 1.44 MPa; 10 wt%: 67.37 ± 1.05 MPa) relative to free tobramycin (5 wt%: p = 3.2 x 10-7; 10 wt%: p = 9.1 x 10-10). Interestingly, addition of either 5 or 10 wt% RMP-filled CD microparticles resulted in a comparable compressive strength (5 wt%: 76.22 ± 2.27 MPa; 10 wt%: 75.80 ± 1.59 MPa) to free tobramycin (5 wt%: p = 0.21; 10 wt%: p = 0.064). Conversely, addition of free RMP

resulted in a dramatic decrease in compressive strength (1.88 ± 0.15 MPa) relative to free tobramycin (77.03 ± 1.17 MPa). The addition of RMP-filled CD microparticles helped to significantly improve the mechanical strength of PMMA relative to free RMP and yields a bone cement with strength comparable to clinically used formulations (i.e., free tobramycin). It should be noted that with time, mechanical properties of PMMA formulations continued to increase beyond the reported values. Preliminary data indicated that PMMA cylinders with 5 wt% RMP-filled CD microparticles attained a compressive strength of 93 MPa at ~23 days (data not shown), which was 17 MPa above their strength at 2 days (76 MPa).

2.4.4. AGAROSE TISSUE-MIMICKING REFILLING STUDY (PURE PMMA VS. CD

DISKS) AND QUANTIFICATION

The ability of polymerized β-CD disks to be refilled with antibiotics in an agarose tissue- mimicking model was compared against the ability of PMMA spherical beads (without

CD) to be refilled in the same manner. Figure 2-6 depicts the samples prior to being refilling with RMP (top) and after 48 hours of refilling (bottom). Presence of RMP was indicated by the red/orange color. Qualitatively, it was concluded that a greater amount of

RMP was refilled into the CD disks compared to pure PMMA beads (without CD). This observation was quantitatively confirmed through a drug leaching study where RMP refilled samples were placed in DMF. Table 2-2 (top) includes the quantification of RMP refilling in CD disks compared to pure PMMA beads. CD disks were capable of being

refilled with 16.55 ± 3.00 μg RMP, nearly 15x more drug than what was refilled into pure

PMMA (statistically significant; p = 0.012). When the mass of drug refilling was normalized against the mass of each polymer sample [mass of drug refilling (mg)/mass of

PMMA bead (mg)], the difference in refilling between CD disks and pure PMMA was even more striking with CD disks refilling nearly 80x more drug than what was possible in pure PMMA per mg of polymer sample (statistically significant; p = 0.0061).

Figure 2-6: Image of PMMA bead (no CD microparticles) and CD disk prior to RMP refilling in agarose model (top) and image of polymer samples 48 hours after being refilled with RMP in model (bottom). RMP was indicated by the red/orange color. Scale bars = 2 mm.

Table 2-2: Quantification of RMP refilling from agarose refilled polymer samples: pure PMMA versus CD disks (top) and pure PMMA versus PMMA-CD composites (bottom).

Average mass refilled drug Average normalized mass (μg) refilled drug/mass of PMMA sample (mg/mg) β-CD disks vs. pure PMMA β-CD disks 16.55 ± 3.00 0.0978 ± 0.0131 Pure PMMA 1.13 ± 0.28 0.0012 ± 0.0003 PMMA w/particles vs. pure PMMA Pure PMMA 0.85 ± 0.57 0.0009 ± 0.0005 5% β-CD particles 1.56 ± 0.12 0.0019 ± 0.0004 5% γ-CD particles 1.66 ± 0.28 0.0017 ± 0.0001 10% β-CD particles 2.01 ± 0.40 0.0027 ± 0.0005 10% γ-CD particles 2.77 ± 0.86 0.0033 ± 0.0011

2.4.5. AGAROSE REFILILNG STUDY (PMMA-CD COMPOSITES) AND

QUANTIFICATION

Similar agarose-based refilling studies were completed on PMMA-CD composite spherical beads in an effort to demonstrate that the addition of CD microparticles to

PMMA induced a refilling property which was not inherently present in pure PMMA (see

Figure 2-6). “Refilling,” in the agarose model, was defined as being filled with antibiotic after implantation when both the polymerization of PMMA was complete and the sample did not contain any drug when it was initially implanted. Figure 2-7 depicts the PMMA beads following 48 hours of agarose refilling with RMP with increasing amounts of CD microparticles. It was observed that as the percentage of CD microparticles added into the

PMMA beads increased, the amount of refilling noticeably increased, as qualitatively indicated by the presence of the red/orange color. DMF was used to leach out the drug

from the refilled PMMA-CD composite beads in an effort to quantify and confirm these observations.

Figure 2-7: Images of RMP refilling in PMMA-CD beads removed from agarose model (pure PMMA, or

PMMA-CD composites) 48 hours after RMP injection. Scale bars = 2 mm.

Table 2-2 (bottom) displays the quantification of RMP refilling in PMMA-CD composites compared to pure PMMA samples (far left images on Figure 2-7). PMMA beads with 5 wt% of either β- or γ-CD were able to refill comparable amounts of RMP

(1.56 μg: β-CD and 2.77 μg γ-CD; p = 0.27). RMP refilling in PMMA beads with 5 wt%

CD microparticles (β- or γ-) demonstrated a slightly higher mass of refilled drug

compared to pure PMMA, but not statistically significant (1.56 μg: β-CD; 1.66 μg: γ-CD;

0.85 μg: pure PMMA) p = 0.086 (normalized β-CD vs pure PMMA) and p = 0.12

(normalized γ-CD vs pure PMMA). When 10 wt% CD microparticles (β- or γ-) were added to PMMA beads, RMP refilling was significantly higher than with 5 wt% CD microparticles (2.01 μg: β-CD; 2.77 μg: γ-CD; 0.85 μg: pure PMMA) p = 0.016

(normalized β-CD vs pure PMMA) and p = 0.053 (normalized γ-CD vs pure PMMA).

The small amount of drug refilled into pure PMMA was attributed to passive drug diffusion into the surface of the PMMA bead, rather than CD-induced refilling processes.

Therefore, the addition of either 5 or 10 wt% CD microparticles to PMMA beads increased and induced RMP refilling into PMMA beads.

2.4.6. PERSISTENCE ZONE OF INHIBITION STUDY (REFILLED PMMA-CD BEADS)

The duration of antimicrobial activity PMMA-CD beads refilled with RMP was evaluated against S. aureus in a persistence zone of inhibition study. Figure 2-8 depicts the zone of inhibition studies for PMMA-CD composites containing either β- or γ-CD microparticles (5 or 10 wt%). Addition of 10 wt% CD microparticles (either β- or γ-CD) resulted in 30-40 days of antimicrobial activity against S. aureus, which had the potential to serve as therapy beyond the window provided by the initial drug filled into PMMA-CD beads when they are implanted.

Figure 2-8: Persistence zone of inhibition study of duration of antimicrobial activity of RMP refilled

PMMA-CD composite beads against S. aureus.

2.5. DISCUSSION

CD microparticles were incorporated into PMMA bone cement in an effort to improve upon antibiotic elution kinetics and induce antibiotic refilling without compromising the mechanical properties of PMMA. Analysis of the elution of aminoglycosides (i.e., gentamicin and tobramycin) from PMMA (with and without CD microparticles) indicated that their antimicrobial activity and release was completed within 2-3 weeks independent of the presence of CD microparticles (see Figure 2-4). This finding demonstrated that beyond 2-3 weeks, PMMA formulations containing aminoglycosides were no longer capable of releasing a therapeutically relevant amount of antibiotic (capable of clearing bacteria), but rather tend to release antibiotic at sub-therapeutic levels for a prolonged period of time while promoting the development of drug-resistant bacteria94. The poor

release kinetics and subsequent antimicrobial susceptibility to gentamicin and tobramycin further confirmed the need to develop a PMMA antibiotic delivery system that would be compatible with a wider-range of antibiotics (beyond aminoglycosides) and would facilitate a more controlled and prolonged release of antibiotic. While our drug release studies demonstrated a slight “burst” release of antibiotic early on, the high local concentration of antibiotic released initially was not anticipated to present a cytotoxicity concern95,96. Previous studies have shown that gentamicin, tobramycin, and RMP do not have any direct detectable cytotoxicity toward primary human osteoblasts in high concentrations96. CD microparticles additionally enabled refilling of PMMA following implantation. Substantial refilling was observed in PMMA-CD composites with 10 wt%

CD microparticles relative to very little refilling obtained with pure PMMA, demonstrating the capacity of CD-laden PMMA to be refilled following implantation.

The resulting antimicrobial activity of the refilled PMMA has the potential to serve as an additional window of therapy beyond that of what was obtained with PMMA-CD composite that was initially filled with RMP upon curing and implantation. When taken in combination, the duration of antimicrobial activity of PMMA-CD filled with RMP was substantially longer relative to the duration of activity that was obtained with clinically used formulations (i.e., gentamicin or tobramycin freely added in PMMA), suggesting the potential to treat advanced chronic PJIs96–98.

Furthermore, the refillability property of PMMA-CD composite makes it conducive

towards the treatment of potentially lifelong PJIs. Since many PJIs may not present until several months following the end of the initial surgery, it is often difficult for surgeons to locally deliver additional antibiotics from the implanted PMMA. As a result, the standard of care is to systemically administer antibiotics99. If the infection still persists, it is often necessary to remove the implanted device and perform a surgical debridement. Our

PMMA-CD composite system is expected to be long-lasting in vivo, unaffected by any degradative processes upon implantation, and, thus, durable for long-term antibiotic refilling applications. PMMA cement is typically used as a non-degradable implant material17 and CD microparticles have been shown to have long-term in vivo stability

(out to at least 28 days)86. Additionally, our drug release study samples did not display any indication of degradation after nearly 60 days in solution.

Incorporation of free RMP in PMMA cylinders resulted in a cement that more closely resembled an elastomeric material rather than the hard cement formed after incorporation of RMP in CD microparticles (both 5 and 10 wt%). The poor mechanics of PMMA with free RMP were attributed to free RMP acting as a plasticizer. It was hypothesized that the carbonyl groups of RMP acted to increase the free volume between the PMMA chains100.

As a result, RMP was able to block interactions between adjacent PMMA chains, promoting their relative mobility and decreasing the strength of PMMA100. The increased strength of the PMMA cylinders upon addition of RMP-filled CD microparticles was hypothesized to result from the inclusion complex formed between CD and RMP. While

RMP was complexed with CD, the CD was able to block the plasticizing groups of RMP 77

from interacting with PMMA polymer chains, thus inhibiting the plasticization effect. As a result, the cement retained its inherent mechanical strength.

Considering the increased duration of antimicrobial activity relative to clinically used formulations (60+ days), refilling capabilities (30-40 days additional activity), and compressive strength (~76 MPa) of PMMA-CD composites with 10 wt% CD microparticles, we envisioned that our PMMA-CD composite could be applied in several different clinical scenarios. The inherent mechanical strength of our PMMA-CD composite relative to clinical PMMA-antibiotic formulations, suggested that our PMMA could be used for weight or load-bearing applications such as in total knee or hip arthroplasties for the treatment of PJIs.

One limitation of the study was the relatively low concentration of antibiotic freely added into PMMA. In an effort to account for the poor mechanics of PMMA with free RMP, we incorporated a relatively low amount of drug (1.25% wt/wt drug/PMMA) and subsequently incorporated the same low amount of each of the other antibiotics into the samples. Since upwards of 5% wt/wt drug/PMMA may be used in clinical cases, it is possible that gentamicin and tobramycin may demonstrate an increased antimicrobial activity in PMMA than what were observed with only 1.25% wt/wt drug/PMMA.

Nevertheless, the duration of antimicrobial activity that we observed from PMMA-CD composites containing 10 wt% RMP-filled CD microparticles well surpassed the duration of activity observed in PMMA with concentrations of gentamicin or tobramycin even

beyond those used clinically80.

2.6. CONCLUSIONS

Through this study it was demonstrated that the addition of up to 10 wt% CD microparticles to PMMA bone cement allowed for a more consistent and prolonged duration of antimicrobial activity with antibiotics, such as RMP (compared to when antibiotic was freely incorporated into the PMMA) while preserving the mechanical properties of the cement. Simultaneously, the addition of CD microparticles enabled the

PMMA to be refilled with antibiotics following implantation for additional windows of therapeutic activity. The combined duration of antimicrobial activity resulting from

PMMA containing 10 wt% CD microparticles that were initially filled and later refilled with RMP was nearly 100 days, making the PMMA-CD composite a potential candidate for the treatment of chronic PJIs. RMP cannot currently be added freely into PMMA due to the severe detriments the drug has on the compressive strength. However, our PMMA-

CD composite with RMP-filled CD microparticles retained PMMA’s compressive strength. Therefore, through the addition of CD microparticles to PMMA, a wider range of antibiotics (beyond traditional aminoglycosides and glycopeptides) may potentially be used with PMMA than what are currently compatible. Furthermore, since the addition of up to 10 wt% RMP-filled CD microparticles resulted in a comparable compressive strength to that observed in clinically-used PMMA containing free tobramycin, our

PMMA formulations with CD microparticles demonstrated the potential to be used in

weight-bearing arthroplasty applications for treatment and prevention of PJIs.

2.7. ACKNOWLEDGMENTS

The authors gratefully acknowledge support through National Science Foundation (NSF)

Graduate Research Fellowship Program Grant No. CON501692 (E.L.C.), NIH NIAMS

Ruth L. Kirschstein NRSA T32 AR007505 Training Program in Musculoskeletal

Research (G.D.L.), CWRU’s Macromolecular Engineering Department’s Research

Experience for Undergraduates funded through NSF DMR-1559708 (S.K.H.), and NIH

R01GM12477 (H.A.vR.). They would also like to thank Grace Cousens (Laurel School,

Cleveland, OH) and Benjamin Grandstaff (Gilmour Academy, Cleveland, OH) for assistance in data collection, Chris Tuma for technical assistance, and the core facility resources provided by the Advanced Manufacturing and Mechanical Reliability Center and the Case Center for Imaging Research at CWRU.

CHAPTER 3. TREATMENT OF BROAD-SPECTRUM PERIPROSTHETIC

JOINT INFECTIONS USING COMBINATORIAL ANTIBIOTIC PMMA

COMPOSITE

*This chapter reprinted (adapted) with permission from Biomacromolecules. 21: 854-

Authors: Erika L. Cyphert, Chao-yi Lu, Dylan W. Marques, Greg D. Learn, Horst A. von Recum

Author contributions: ELC – experimental design, data analysis and collection (all experiments), preparation of figures, writing/revision manuscript. C-yL – data analysis and collection, manuscript preparation. DWM – data collection. GDL – experimental design (compression testing), data analysis and collection (compression testing), writing manuscript (compression testing). HAvR – experimental design, data analysis, revision manuscript.

3.1. ABSTRACT

Antibiotics are commonly added to PMMA bone cement by surgeons to locally treat PJIs inside and out of the bone. However, this strategy is of limited value in high-risk patients where PJIs can be chronic and otherwise hard to treat. When only one drug is incorporated and applied toward polymicrobial PJIs that contain multiple bacterial species, there is a high risk that bacteria can develop antibiotic resistance. To combat these limitations, we developed a combination antibiotic PMMA-CD composite system composed of RMP-filled CD microparticles added into PMMA containing a second antibiotic. Different antibiotic combination formulations were evaluated through zone of inhibition, drug activity, antibiotic release/refilling, and mechanical studies. Our combination antibiotic PMMA-CD composite system achieved up to an 8-fold increase in the duration of antimicrobial activity in comparison to clinically used antibiotic-filled

PMMA. Inclusion of CD microparticles allowed for refilling of additional antibiotics after simulated implantation in agarose, resulting in additional windows of therapeutic activity. Mechanical testing showed that our tested formulations did have a small, but significant decrease in mechanical properties when compared to unmodified controls.

While further studies are needed to determine whether the tested PMMA-CD composite formulations would be still suitable for load-bearing applications (e.g., treatment of PJIs in total arthroplasties), our composites nevertheless demonstrated that they would be amenable for a variety of non-load-bearing applications (e.g., as antimicrobial beads for the treatment of soft tissue PJIs and as temporary spacer in treatment of two-stage PJI 82

arthroscopic revision surgeries).

3.2. INTRODUCTION

PJIs following procedures such as total hip and knee arthroplasties have a relatively low occurrence, ranging from 0.3 to 1.28%101–103. Nevertheless, patients that acquire PJIs often must undergo intensive systemic antibiotic regiments and unpalatable revision surgeries that place a substantial burden on them, both physically and financially101,104,105.

Revision procedures have been shown to increase the patient’s risk for further infection, morbidity, and mortality101,106. The aggressiveness of PJIs can be exacerbated if the infection is composed of both Gram-positive and Gram-negative pathogens. While Gram- positive Staphylococci species, specifically Staphylococcus aureus (S. aureus) and

Staphylococcus epidermidis (S. epidermidis), account for the majority of PJIs, up to 15% of PJIs are caused by Gram-negative species107. Escherichia coli (E. coli) accounts for upwards of 20-30% of Gram-negative PJIs107.

In these complex cases, monotherapy antibiotic treatment may not provide broad- spectrum antimicrobial coverage or bacteria may become resistant toward the antibiotic over an extended period of time. In an effort to mitigate the risk of developing drug resistance, antibiotics are often administered in combination108,109. The decreased risk of drug resistance with combination therapies has been attributed to the different modes of action of the antibiotics against bacteria and the increased difficulty for bacterial populations to acquire simultaneous resistance to each unique compound108. For example,

one clinical study showed that 42% of patients treated with monotherapy for infection developed antibiotic-resistant bacteria, while only 17% of patients treatment with a combination of two antibiotics developed resistant bacteria110. In addition to decreasing the risk for drug resistance, combination therapies have also been found to be more suitable for treating polymicrobial infections due to their broader antimicrobial coverage110,111. Furthermore, combination antibiotic therapy can often result in a synergistic antibacterial effects as a result of complex interactions between the drugs and the bacteria. For example, one drug may increase the uptake of the other drug by enhancing the permeability of the bacterial cell wall. This may explain the synergism observed when aminoglycoside and β-lactam antibiotics are used in combination112.

Another potential mechanism for synergism to occur is to use drugs that target specific metabolic pathways leading to secondary effects that can further enhance the potency of the drugs112. This can be observed when trimethoprim and sulfonamide are used against

E. coli mutants112.

To locally deliver antibiotics in orthopedic procedures (e.g., total arthroplasties) for treatment and prevention of PJIs, antibiotics are often directly added to PMMA bone cement during the mixing process113. Typically, only one antibiotic, such as gentamicin or tobramycin, is used in an effort to provide antimicrobial coverage114. The primary disadvantages to this strategy are (a) these antibiotics may not provide adequate broad- spectrum coverage, (b) using only one drug increases the risk of developing drug resistance, and (c) that the window of antimicrobial activity is limited by the initial 84

amount of antibiotic present within the PMMA, as it is not feasible to replenish the drug reservoir once depleted following implantation115–117. Furthermore, there is often a burst release of the antibiotic from the PMMA, which can result in a suboptimal delivery profile for the treatment of prolonged PJIs113. To overcome the poor release kinetics of antibiotics from PMMA, others have developed a variety of strategies such as using liposomes118, adding porogens119, or using calcium phosphate spheres120. Additionally, non-antibiotic-based solutions such as nanodiamonds121 and zinc-containing composites122 have been explored for preventing and treating PJIs. However, there are no existing delivery systems capable of providing broad-spectrum, long-term, and refillable therapy.

To create a formulation of PMMA cement that can effectively treat a range of PJIs and that has the ability to be refilled with antibiotics after implantation, we evaluated a

PMMA-CD composite system comprised of insoluble CD microparticles and combinations of antibiotics. The microparticles added multiple features to the PMMA including prolonged antibiotic release and refilling, which we have previously shown, as well as the capability of tuning the release rate of one drug separately from the other115.

Figure 3-1 depicts a schematic of the composition of the different PMMA-CD composites containing both different antibiotic combinations and differing amounts of CD microparticles.

Figure 3-1: Schematic depicting drug combination composite formulations of PMMA-CD bone cement.

Pre-polymerized CD was cross-linked into insoluble microparticles that were filled with RMP. Different amounts of RMP-filled CD microparticles were incorporated into PMMA during polymerization along with either gentamicin or tobramycin (non-encapsulated in CD).

In this study, we have developed several PMMA-CD bone cement composite formulations containing combination antibiotics (i.e., either gentamicin or tobramycin with RMP-filled CD microparticles) with different amounts of CD microparticles to evaluate their efficacy in treating broad-spectrum bacterial species, as well as their suitability for treating long-term or latent infections. We hypothesized that our PMMA-

CD composite system containing combination antibiotics and CD microparticles would allow for an extended antimicrobial effect that was effective against a broader range of bacterial species in comparison to clinically used monotherapy drug-PMMA formulations. Specifically, we have evaluated the duration of antimicrobial activity of

different combination antibiotic PMMA-CD composite formulations against S. aureus, S. epidermidis, and E. coli through persistence zone of inhibition studies. Antibiotic release studies were also completed. To evaluate the capacity of the PMMA-CD composites to be refilled at later time points (after initial implantation), antibiotic refilling studies were completed in agarose to simulate the diffusion of drug through the tissue. Finally, the porosity and mechanical properties of different formulations were evaluated to determine if they met the standards to be used for PJI treatment in load-bearing arthroplasty applications.

3.3. MATERIALS AND METHODS

3.3.1. MATERIALS

Lightly epichlorohydrin cross-linked β-CD pre-polymer was purchased from CycloLab

(Budapest, Hungary). Ethylene glycol diglycidyl ether was purchased from Polysciences

Inc. (Warrington, PA). RMP was purchased from Research Products International (Mt.

Prospect, IL). Gentamicin and tobramycin were purchased from Fisher Scientific

(Pittsburgh, PA). Stryker Simplex HV (high viscosity) radiopaque bone cement produced by Stryker Orthopaedics (Mahwah, NJ) was purchased from eSutures. Green fluorescent protein (GFP)-labeled S. aureus stock culture was kindly provided by Dr. Edward

Greenfield (CWRU, OH). E. coli (Migula) Castellani and Chalmers (ATCC 25922) and

S. epidermidis (Winslow and Winslow) Evans (ATCC 12228) stock cultures were purchased from American Type Culture Collection (ATCC, Manassas, VA).

Poly(tetrafluoroethylene) (PTFE) stock for cylindrical specimen molds was purchased from McMaster-Carr (Aurora, OH). All other reagent and solvents were purchased from

Fisher Scientific in the highest grade available.

3.3.2. SYNTHESIS OF INSOLUBLE POLYMERIZED CD MICROPARTICLES

See Chapter 2, section 2.3.2.

3.3.3. CD MICROPARTICLE ANTIBIOTIC FILLING

See Chapter 2, section 2.3.4.

3.3.4. SYNTHESIS OF PMMA-CD COMPOSITE SPHERICAL BEADS AND

CYLINDERS

Small spherical PMMA-CD composite beads were synthesized by combining 2 gram samples of Stryker Simplex HV radiopaque surgical-grade bone cement powder and 100 mg of either gentamicin or tobramycin powder (5 wt/wt% drug/PMMA) (comparable to concentration used clinically) and either 100 mg (5 wt/wt% CD/PMMA) or 200 mg (10 wt/wt% CD/PMMA) of CD microparticles and mixed until homogeneous. The powder mixture was reacted with 1 mL of methyl methacrylate monomer for each batch of

PMMA (in accordance with the manufacturer’s instructions) to form a soft dough.

PMMA dough with free drug and CD microparticles (either filled with RMP or non-drug- filled (empty)) was then formed into 6 mm diameter beads and dried at room temperature115. Cylindrical PMMA-CD composite samples were synthesized according to

a previously published protocol115.

3.3.5. PERSISTENCE ZONE OF INHIBITION STUDY (PREFILLED BEADS)

Persistence zone of inhibition testing was carried out according to a previously established protocol (see Chapter 2, section 2.3.6), where PMMA-CD composite spherical beads containing 2 antibiotics were either placed on trypticase soy agar (S. aureus and S. epidermidis) or Luria-Bertani (LB) broth agar (E. coli) Petri dishes with 70

μL of the respective bacterial culture.

3.3.6. QUANTIFICATION AND ACTIVITY ZONE OF INHIBITION OF DRUG

RELEASE ALIQUOTS

Release and antimicrobial activity of gentamicin and tobramycin from combinatorial

PMMA-CD composite and monotherapy controls against S. aureus, S. epidermidis, and

E. coli was carried out according to a previously described methodology (see Chapter 2, section 2.3.7.).

3.3.7. MICROCOMPUTED TOMOGRAPHY (MICRO-CT) SCANS AND POROSITY

QUANTIFICATION

Micro-CT scans were collected to evaluate the porosities of PMMA-CD composite samples and were collected in accordance with a previously published protocol115.

Briefly, cylindrical PMMA-CD composite specimens were placed individually in polypropylene tubes prior to scanning. Scans were conducted using a Siemens Inveon

PET-CT scanner (Siemens Medical Solutions, Malvern, PA) controlled by a Siemens

Inveon Acquisition Workplace software on a PC. Each sample was scanned using identical scanning and reconstruction settings and was exported in DICOM format from

Siemens Inveon Research Workplace software. PMMA pore segmentation/thresholding, thresholding of solid fraction, and model export were completed in 3D Slicer software

(BWH and 3D Slicer contributors, version 4.8)123. An upper threshold of -200 was used for segmentation of pores, and a lower threshold of 0 was used for thresholding the solid fraction. Netfabb Standard 2018 (Autodesk, Inc.) was used to calculate volumetric measurement of models built in 3D Slicer. Pore volume fraction was calculated by dividing the total volume segmented by the volume of pores.

3.3.8. COMPRESSION TESTING

See Chapter 2, section 2.3.8.

3.3.9. PMMA-CD COMPOSITE ANTIBIOTIC REFILLING IN AGAROSE PHANTOM

MODEL

See Chapter 2, section 2.3.9.

3.3.10. QUANTIFICATION OF ANTIBIOTIC REFILLING EFFICIENCY

See Chapter 2, section 2.3.10.

3.3.11. STATISTICAL ANALYSIS

Data from all studies was displayed as the mean of each condition evaluated in triplicate 90

(persistence zone of inhibition, drug release, activity zone of inhibition, micro-CT, agarose phantom refilling) or n ≥7 (compression testing) with error bars depicting the standard deviation. All statistical tests were completed in Microsoft Excel 2016. Two- tailed Student’s t-tests with unequal variances were conducted for persistence zone of inhibition study (n = 3), drug release study (n = 3), and activity zone of inhibition study

(n = 3). Single-factor ANOVA with Tukey’s post hoc tests (α = 0.05) was used to analyze the compression data (n = 7 for gentamicin-only control, n = 7 for tobramycin-only control, n = 8 for 10 wt% empty CD with tobramycin, n = 9 for 10 wt% RMP-filled CD with tobramycin, n = 7 for 10 wt% empty CD with gentamicin, n = 10 for 10 wt% RMP- filled CD with gentamicin), the porosity data obtained from micro-CT (n = 3), and the quantification of RMP refilling (n = 3). P-values of less than 0.05 from the analyses were considered to be statistically significant.

3.4. RESULTS AND DISCUSSION

3.4.1. PERSISTENCE ZONE OF INHIBITION STUDY

The duration of antimicrobial activity of PMMA-CD composite samples with drug combinations (RMP-filled CD microparticles with either gentamicin or tobramycin) and

PMMA samples containing only a single drug (RMP-filled CD microparticles only or either free gentamicin or tobramycin only) against S. aureus, S. epidermidis, and E. coli was evaluated. Figure 3-2 depicts the results of the persistence zone of inhibition study.

Figure 3-2: Persistence zone of inhibition studies of PMMA-CD composites with various drug combinations (5 or 10 wt% RMP-filled CD microparticle combined with either gentamicin or tobramycin) and monotherapy drug control samples (either RMP alone, or gentamicin or tobramycin alone) against S. aureus (top), S. epidermidis (middle), and E. coli (bottom).

Generally, regardless of the species of bacteria, PMMA-CD composites containing only 92

RMP-filled CD microparticles (5 or 10 wt%) exhibited the shortest duration of antimicrobial activity followed by PMMA containing only gentamicin or tobramycin. In all cases, the longest antimicrobial effect was observed in PMMA-CD composite samples containing combination antibiotics and these samples typically exhibited significantly larger zones of inhibition during early time points than samples containing only RMP- filled CD microparticles.

Against S. aureus, the longest duration of antimicrobial activity was observed with

PMMA-CD composites containing both 10 wt% RMP-filled CD microparticles and gentamicin lasting nearly 70 days. PMMA-CD composite samples containing both 5 wt%

RMP-filled CD microparticles and gentamicin had only a slightly shorter duration of about 60 days. A comparable duration of activity was observed for PMMA-CD composite beads containing either 5 or 10 wt% RMP-filled CD microparticles with tobramycin (i.e., ~60 days). The shortest duration of activity was observed for PMMA-

CD composite beads containing only RMP-filled CD microparticles (10 wt% CD: 24 days; 5 wt% CD: 15 days). Whereas, PMMA beads containing only gentamicin or tobramycin had intermediate durations of clearance of approximately 50 and 40 days, respectively.

Against S. epidermidis, the longest duration of antimicrobial activity was observed in the

PMMA-CD composites containing both 10 wt% RMP-filled CD microparticles and gentamicin, lasting approximately 50 days. PMMA-CD composite samples containing

gentamicin and 5 wt% RMP-filled CD microparticles had a slightly shorter duration of activity. The activity was similar for PMMA-CD composite samples containing both tobramycin and either 5 or 10 wt% RMP-filled CD microparticles lasting ~50 days.

Conversely, PMMA-CD composite samples containing only either 5 or 10 wt% RMP- filled CD microparticles demonstrated a dramatic decrease in the duration of activity compared to drug combination PMMA-CD composite samples, specifically only ~10 and

18 days, respectively. As with S. aureus, samples containing only gentamicin or tobramycin had intermediate durations of activity around 40 days.

For E. coli, the longest duration of antimicrobial activity was determined to be PMMA-

CD composite antibiotic combination samples with 10 wt% RMP-filled CD microparticles and tobramycin, lasting nearly 55 days. A slightly shorter duration was observed in PMMA-CD composite samples with 10 wt% RMP-filled CD microparticles and gentamicin (i.e., 50 days). Both PMMA-CD composite samples with 5 wt% RMP- filled CD microparticles and either free gentamicin or tobramycin demonstrated similar durations of clearance lasting nearly 50 days. In stark contrast, monotherapy PMMA-CD samples containing only 5 or 10 wt% RMP-filled CD only demonstrated clearance for about 7 days. Samples with only gentamicin or tobramycin had an antimicrobial activity duration of approximately 30 and 31 days, respectively.

Overall, an increase in the duration of antimicrobial activity was observed in the persistence zone of inhibition studies against all three species of bacteria (i.e., S. aureus,

S. epidermidis, and E. coli) when drug combination PMMA-CD composites were utilized. Specifically, for S. aureus, there was up to a 4-fold increase observed in the duration of antimicrobial activity between the monotherapy version of PMMA and drug combination PMMA-CD composites, for S. epidermidis, a 5-fold increase, and for E. coli, an 8-fold increase. We hypothesized that it was possible that this improvement could be attributed to the additional model of action against bacteria from the second antibiotic and/or the decrease in the probability for the bacterial culture to include strains that were resistant to the therapy108. There may also have been a synergistic effect between the two drugs that could have allowed for an enhanced antimicrobial activity depending upon each drug’s interaction with the bacteria112. Specifically, the combination of RMP and gentamicin has been shown to have an enhanced antimicrobial effect against specific staphylococcal species124.

Based upon our persistence zone of inhibition studies, the incorporation of both gentamicin and 10 wt% RMP-filled CD microparticles into PMMA-CD composites had the longest antimicrobial effect against Gram-positive species of bacteria (i.e., S. aureus and S. epidermidis) compared to that of other composite formulations evaluated. Against

E. coli, PMMA-CD composites containing both tobramycin and 10 wt% RMP-filled CD microparticles had the longest duration of activity. These findings suggested that for treating or preventing polymicrobial PJIs from arthroplasties, PMMA-CD composite containing both gentamicin and 10 wt% RMP-filled CD microparticles would be amenable for use for Gram-positive dominant infections, and PMMA-CD composite 95

containing both tobramycin and 10 wt% RMP-filled CD microparticles would be amenable for use for Gram-negative dominant infections.

3.4.2. QUANTIFICATION AND ACTIVITY OF DRUG RELEASE ALIQUOTS

After the persistence zone of inhibition studies were performed, the amount and resultant activity of the released drug were determined to see if it correlated with zone of inhibition persistence. The amount of released gentamicin or tobramycin was assessed by quantifying release at discrete time points from PMMA-CD composite beads using a ninhydrin reaction. Simultaneously, the antimicrobial activity of each respective release aliquot (for each time point) was also evaluated using an activity zone of inhibition study against S. aureus, S. epidermidis, and E. coli to ensure that the amount of drug released at each time point was capable of clearing all three species of bacteria. Daily tobramycin or gentamicin release plots and their respective antimicrobial activity are depicted in Figure

3-3.

Figure 3-3: Gentamicin and tobramycin daily drug release and subsequent antimicrobial activity (based on activity zone of inhibition) of each release aliquot against S. aureus (red), S. epidermidis (blue), and E. coli

(green) from PMMA-CD composite beads.

For each condition, there is a stacked graph where the top is the activity zone of inhibition study and the bottom is the corresponding quantification of the daily amount of 97

either gentamicin or tobramycin released from each sample. The quantification of drug released was normalized by dividing the mass of the drug released by the mass of the

PMMA-CD composite sample and was plotted on a semi-logarithmic scale. Generally, both the activity and release plots for each condition demonstrated a similar downward trend. Specifically, the activity zone of inhibition study matched the daily amount of drug released in the sense that when a higher concentration of drug was detected in the release solution, a larger-sized zone of inhibition resulted.

In general, the addition of RMP-filled CD microparticle to PMMA-CD composites resulted in an increase in the duration of the release of gentamicin from PMMA.

Specifically, samples containing only gentamicin (without any CD microparticles) released the drug for only 7 days. Whereas, PMMA-CD composites containing either 5 or

10 wt% RMP-filled CD microparticles in addition to gentamicin resulted in gentamicin release lasting 28 and 35 days, respectively. The amount of gentamicin released from

PMMA-CD composite samples containing 10 wt% RMP-filled CD microparticles was significantly greater than the amount released from the PMMA control samples containing only gentamicin (without any CD microparticles) after 1, 2, and 4 days (p <

0.05). Whereas, there were no significant differences found in the amount of gentamicin released from PMMA-CD composite samples containing both 5 wt% RMP-filled CD microparticles and gentamicin to those containing only gentamicin (p > 0.05).

Correspondingly, the activity zone of inhibition study of aliquots from PMMA-CD

composites containing both gentamicin and CD microparticles demonstrated that the aliquots had a longer duration of activity when compared to that of PMMA samples with only gentamicin. More specifically, PMMA samples with only gentamicin had a clearance time of up to 28 days for all three bacterial species. Whereas, PMMA-CD composite samples containing both RMP-filled CD microparticles and gentamicin had clearance times lasting 42 days (5 wt%) and 49 days (10 wt%).

In contrast to PMMA samples with only gentamicin, the addition of RMP-filled CD microparticles into PMMA did not result in an increase in the duration of the release of tobramycin from the PMMA. When freely added into PMMA without CD microparticles, tobramycin was released over 91 days. Conversely, tobramycin was only released for 63 or 70 days when 10 or 5 wt% RMP-filled CD microparticles were added to the PMMA.

The corresponding activity zone of inhibition study aligned with these findings.

Specifically, there was a longer clearance time for PMMA samples containing only tobramycin (91 days) compared to PMMA-CD composites containing both tobramycin and RMP-filled CD microparticles (70 days: 5 wt% CD; 63 days: 10 wt% CD).

3.4.3. MICRO-CT SCANS AND POROSITY QUANTIFICATION

Micro-CT scans were collected in triplicate for a total of six groups of PMMA-CD composite samples in an attempt to elucidate how the addition of CD microparticles and free drug affected the sample porosity. Scans were analyzed to quantify the pore volume fraction, average pore volume, number of pores, and solid fraction voxel mean and

standard deviation. Results are compiled in Table 3-1, and representative three- dimensional renderings of the solid and pore volume fractions are depicted in Figure 3-4.

It was important to note that the CD microparticles were co-registered with air (pores) during segmentation (radiodensity pure CD polymer = -400 HU, upper pore segmentation threshold = -200 HU; data not shown); therefore, the results that were observed regarding the porosity were partially due to the CD microparticles themselves.

Table 3-1: Quantification of various parameters of cylindrical PMMA-CD composite samples that were micro-CT scanned. Statistically significant difference of samples containing both tobramycin (and CD microparticles) relative to free tobramycin (†), and statistically significant difference of samples containing both gentamicin (and CD microparticles) relative to free gentamicin (‡).

Micro-CT Pore Volume Average Pore Average # Solid Fraction Solid Fraction Group Fraction (%) Volume (nL) Discrete Pores Voxel Mean Voxel St Dev (n = 3) (HU) (HU) Free tobra 0.655 ± 0.309 0.531 ± 0.161 3362 ± 712 1418±57 488±20 control Free genta 0.853 ± 0.313 0.609 ± 0.171 4101 ± 569 1408±24 486±13 control 10wt% empty β- 1.797 ± 0.089† 0.940 ± 0.120 5589 ± 575† 1297±21† 503±11 CD w/free tobra 10wt% RMP- 1.006 ± 0.686 0.540 ± 0.244 4787 ± 1195 1352±22 489±3 filled β-CD w/free tobra 10wt% empty β- 2.381 ± 0.367‡ 1.020 ± 0.224 6825 ± 503‡ 1211±52‡ 477±26 CD w/free genta 10wt% RMP- 1.302 ± 0.187 0.621 ± 0.039 5970 ± 450 1266±38‡ 487±13 filled β-CD w/free genta

100

Figure 3-4: Representative three-dimensional renderings of the solid and pore volume fractions of PMMA-

CD composite cylinders containing (a) free gentamicin only (control), (b) 10 wt% empty (non-drug filled)

CD microparticles with free gentamicin, and (c) 10 wt% RMP-filled CD microparticles with free gentamicin.

Generally, the addition of CD microparticles increased the pore volume fraction of

PMMA-CD composite samples. For PMMA containing only gentamicin, the pore volume fraction was 0.9%. However, upon addition of 10 wt% empty or RMP-filled CD microparticles the pore volume fraction increased to 2.4% and 1.3%, respectively. A similar trend was observed in PMMA samples containing tobramycin (pore volume fraction: tobramycin only = 0.7%, +10 wt% empty CD microparticles = 1.8%, +10 wt%

RMP-filled CD microparticles = 1.0%). 101

The addition of CD microparticles generally increased the number of pores in each sample. For PMMA samples containing only tobramycin, an average of 3362 discrete pores was found in each sample. Through the addition of 10 wt% empty CD microparticles into samples containing tobramycin, the average number of discrete pores increased significantly to 5589 (p < 0.05). For PMMA samples containing only gentamicin, an average of 4101 discrete pores was found in each sample. Through the addition of empty 10 wt% CD microparticles, the average number of pores increased significantly to 6825 pores (p < 0.05).

Solid fraction voxel mean provides a description of the average radiodensity of the solid material (i.e., excluding pores) in a given PMMA-CD composite formulation. Previous work has shown that the addition of RMP resulted in an increase in the solid fraction voxel mean. In general, in this study, our results suggested that the incorporation of microparticles (either empty or drug-filled) into PMMA already containing drug reduced the average radiodensity of the solid fraction (p < 0.05), the only exception being the

PMMA-CD composite containing both 10 wt% RMP-filled CD microparticles and tobramycin (p > 0.05). Filling of CD microparticles with RMP tended to increase the average radiodensity of the solid fraction (p > 0.05).

Solid fraction voxel standard deviation provides a description of the inconsistency in the volumetric distribution of the individual solid components (especially the radiopacifier component barium sulfate) within the PMMA. Most samples demonstrated similar

102

standard deviations around 480, suggesting that all of the PMMA-CD composite formulations investigated possessed a similar spatial heterogeneity (p > 0.05).

Our observation in the porosity of the PMMA-CD composites may be linked to the results from the drug release studies. The slight but insignificant increase in pore volume fractions of PMMA-CD composite containing both gentamicin and RMP-filled CD microparticles observed from the micro-CT scans may be related to the 4- to 5-fold increase in the duration of the release of gentamicin from PMMA-CD composite samples compared to that of monotherapy (clinically used) PMMA formulations (i.e., only gentamicin). Studies previously completed by other groups have indicated that porosity may be a contributing factor to the total amount of drug released and the duration of sustained drug release from PMMA systems22,125,126. This was also observed in our studies where the increase in porosity resulted in an increased duration of the release of the antibiotic. PMMA-CD composite samples containing both gentamicin and 10 wt%

RMP-filled CD microparticles had a greater pore volume fraction and contained a greater number of pores than samples containing only gentamicin. The drug release studies also demonstrated that PMMA-CD composites containing both gentamicin and CD microparticles had a longer duration of release compared to PMMA containing only gentamicin. The addition of 10 wt% RMP-filled CD microparticles into samples containing tobramycin did not significantly change the pore volume fraction or the number of pores (p > 0.05). These results were in compliance with release studies, which demonstrated that the addition of 10 wt% RMP-filled CD microparticles did not increase 103

the duration of release for PMMA containing tobramycin. Interestingly, linear fits with

R-squared values of 0.999 were observed when the pore volume fraction was plotted against the duration of drug release and when the number of pores was plotted against the duration of drug release for PMMA-CD composites containing both 10 wt% RMP-filled

CD microparticles and tobramycin or gentamicin and samples containing only tobramycin. These high R-squared values demonstrated strong negative correlations between the pore volume fraction and duration of drug release and between the number of pores in the sample and the duration of drug release. Nevertheless, it was important to note that the porosity results were partially due to the CD microparticles themselves as the pores and CD microparticles were co-registered.

3.4.4. COMPRESSION TESTING

Cylindrical PMMA-CD composite specimens of the same formulations that were CT- scanned were tested in unconfined compression to determine the effect of the presence of

CD microparticles and free drug on the mechanical properties of PMMA. Findings are summarized in Table 3-2, and representative (i.e., closest to the average values for ultimate compressive strength) stress-strain curves are shown in Figure 3-5.

104

Table 3-2: Quantification of the mechanical properties of PMMA-CD composite cylindrical samples.

Statistically significant difference of samples containing both tobramycin (and CD microparticles) relative to free tobramycin (†) and samples containing both gentamicin (and CD microparticles) relative to free gentamicin (‡).

Compression Ultimate Modulus (MPa) Normalized Strain to Failure Test Specimen Compressive Work to Failure (%) Strength (MPa) (J/cm3)

Free tobra control 78.17 ± 3.99 2399.51 ± 82.39 2.600 ± 0.178 5.25 ± 0.17 Free genta 71.45 ± 1.31 2303.89 ± 63.98 2.256 ± 0.117 5.09 ± 0.15 control 10wt% empty β- 48.09 ± 2.45† 1431.86 ± 139.42† 1.711 ± 0.098† 5.51 ± 0.19 CD w/free tobra 10wt% RMP- 61.07 ± 2.28† 1970.84 ± 87.04† 2.032 ± 0.096† 5.25 ± 0.13 filled β-CD w/free tobra 10wt% empty β- 49.80 ± 1.32‡ 1571.45 ± 58.79‡ 1.760 ± 0.063 5.61 ± 0.14‡ CD w/free genta 10wt% RMP- 49.79 ± 1.60‡ 1521.21 ± 59.68‡ 1.762 ± 0.069‡ 5.64 ± 0.19‡ filled β-CD w/free genta

105

Figure 3-5: Representative stress versus strain curves from the compression testing of PMMA-CD composite samples.

Free gentamicin and tobramycin formulations both exceeded a 70 MPa ultimate compressive strength, while all other groups were below that (~61 MPa for 10 wt%

RMP-filled CD microparticles with tobramycin, ~ 49 MPa for all others). For reference, plain PMMA, fabricated and evaluated in the same manner but without antibiotics or additives, has an ultimate compressive strength of nearly 80 MPa (data not shown).

PMMA containing free tobramycin had a higher ultimate compressive strength and normalized work to failure (p < 0.05) than PMMA containing free gentamicin. The addition of CD microparticles, whether RMP-filled or empty significantly decreased the compressive strength, modulus, and normalized work to failure (p < 0.05) of PMMA containing free tobramycin. Likewise, the addition of CD microparticles, whether RMP-

106

filled or empty, significantly lowered the compressive strength and modulus (p < 0.05) of

PMMA containing free gentamicin. Incorporation of RMP into CD microparticles significantly increased the compressive strength, modulus, and normalized work to failure (p < 0.05) of CD microparticle-laden PMMA containing free tobramycin, whereas the incorporation of RMP into CD microparticles demonstrated no significant effect on compressive strength, modulus, or strain to failure (p > 0.05) of CD microparticle-laden

PMMA containing free gentamicin.

PMMA samples that contained only gentamicin were found to be weaker than samples that contained only tobramycin. It was hypothesized that this may be due to the difference in wettability of the two drugs by the liquid monomer through hydrogen bonding. After the addition of empty CD microparticles, the compressive strength difference between the two was not as striking. This may be due to an affinity-based interaction between the liquid monomer and the empty CD microparticles. One possible explanation for the low strength of the combinatorial antibiotic PMMA-CD composite formulations could be that, in comparison to the monotherapy formulations, there was a greater volume of dry solids for the same volume of methyl methacrylate liquid monomer, making the liquid volume less able to wet all of the dry solids during mixing.

This ultimately could result in polymerization of shorter (lower-molecular-weight)

PMMA chains. If so, it was hypothesized that this weakness could be alleviated by increasing the volume of liquid monomer used for mixing these composite formulations or alternatively by incorporating fewer CD microparticles. It was important to note, 107

however, that preliminary studies have indicated that PMMA continued to strengthen with additional time (data not shown). Additionally, it was evident that the hand-mixing procedure in this study always resulted in the incorporation of some pores (at least 0.5 vol%) within the PMMA; therefore, vacuum-mixing (as is done in clinical settings) may further elevate the strength of our composite PMMA-CD materials (see Chapter 6).

Mechanical properties of PMMA-CD composites, such as compressive strength, have been shown to be affected by the porosity of the composite. Higher porosity has been associated with lower compressive strength126. This general trend has also been observed when the number of pores or pore volume fraction was plotted against the compressive strength of the sample. Interestingly, a strong negative linear correlation was found when the number of pores was plotted against the compressive strength (R2 = 0.872).

3.4.5. REFILILNG PMMA-CD COMPOSITES AND QUANTIFICATION IN AGAROSE

PHANTOM MODEL

PMMA-CD composite beads containing both free tobramycin or gentamicin and empty

CD microparticles were refilled with RMP using the tissue-mimicking agarose phantom model62,115. Figure 3-6 depicts the images of the refilled PMMA-CD composite samples after 48 hours of RMP refilling.

108

Figure 3-6: Images of combinatorial antibiotic PMMA-CD composite beads refilled with RMP for 48 hours in the agarose phantom model.

Table 3-3 displays the quantification of the amount of RMP refilled into the composite samples containing CD microparticles. Quantification was carried out on composite samples that were either initially empty and refilled in the agarose phantom model or that contained CD microparticles that were initially filled with RMP prior to being added to the PMMA.

Table 3-3: Quantification of the amount of RMP initially filled (top) and the amount of RMP refilled through the agarose phantom model into combinatorial antibiotic PMMA-CD composite beads (bottom).

Average mass RMP (μg) Average normalized mass RMP/mass of PMMA sample (μg/mg) PMMA w/pre-filled particles 5% β-CD particles w/free genta 4.49 ± 0.31 0.0715 ± 0.0031 5% β-CD particles w/free tobra 2.83 ± 0.65 0.0454 ± 0.0029 10% β-CD particles w/free genta 19.48 ± 2.61 0.3589 ± 0.0303 10% β-CD particles w/free tobra 5.81 ± 1.07 0.0952 ± 0.0224 PMMA w/re-filled particles

5% β-CD particles w/free genta 5.30 ± 0.67 0.0803 ± 0.0083 5% β-CD particles w/free tobra 3.64 ± 0.47 0.0581 ± 0.0016 10% β-CD particles w/free genta 5.93 ± 1.12 0.1080 ± 0.0102 10% β-CD particles w/free tobra 5.21 ± 0.16 0.0624 ± 0.0075 109

The quantification demonstrated the potential for refilling RMP into composite samples containing empty CD microparticles following implantation in agarose in the presence of other drugs already incorporated into PMMA (i.e., gentamicin or tobramycin).

Specifically, no statistically significant difference was found between the normalized amount of RMP filled in the pre-filled and re-filled conditions for PMMA-CD composites containing both gentamicin and 5 wt% CD microparticles (p > 0.05).

Similarly, there was no significant difference in the amount of normalized RMP filled in

PMMA-CD composite samples containing both tobramycin and either 5 or 10 wt% CD microparticle for refilled conditions (p > 0.05).

Based upon the results from the quantification of RMP filled into pre-filled CD microparticles, samples with 10 wt% RMP-filled CD microparticles had between 2 and 5 times more RMP in the composite system relative to samples containing only 5 wt%

RMP-filled CD microparticles. However, the persistence zone of inhibition studies revealed a relatively minor difference between the durations of antimicrobial activity between PMMA-CD composite samples containing either 5 or 10 wt% RMP-filled CD microparticles. The small difference in the duration of the activity of composite samples was attributed to a nonhomogeneous distribution of RMP-filled CD microparticles in the samples rather than the loss of RMP’s antimicrobial activity or the duration of drug release form the CD microparticles. The antimicrobial activity of the PMMA-CD composite samples was likely limited by the amount of antibiotic located near the surface of the composite sample that was in direct contact with the agar plate. The proximity of 110

the antibiotic to the surface of the sample can be directly related to its ability to diffuse out. Meaning that the RMP-filled CD microparticle needed to be well distributed in the composite sample such that the amount of drug-filled CD microparticles located at the surface of the sample in direct contact with the agar plate was distinguishable between samples containing either 5 or 10 wt% CD microparticles.

3.4.6. PERSISTENCE ZONE OF INHIBITION STUDY (REFILLED PMMA-CD

COMPOSITE BEADS)

Persistence zone of inhibition studies against S. aureus, S. epidermidis, and E. coli were also conducted using PMMA-CD composite samples that were removed from the agarose-refilling model after 48 hours of RMP refilling (i.e., RMP-refilled samples with either 5 or 10 wt% CD microparticles and either free gentamicin or tobramycin), and the results are displayed in Figure 3-7.

111

Figure 3-7: Persistence zone of inhibition studies of combinatorial antibiotic PMMA-CD composite samples refilled with RMP for 48 hours in agarose model. Duration of antimicrobial activity was evaluated against S. aureus (top), S. epidermidis (middle), and E. coli (bottom).

In general, refilled PMMA-CD composite samples containing either 5 or 10 wt% empty

112

CD microparticles were able to allow for an additional window of antimicrobial therapy beyond what could be obtained from samples containing either gentamicin or tobramycin with 5 or 10 wt% RMP pre-filled CD microparticles. More specifically, against S. aureus, refilled composite samples with gentamicin or tobramycin and CD microparticles had a clearance duration between 40 and 51 days. These samples had durations of clearance between 24 and 42 days against S. epidermidis. Against E. coli, refilled composite samples containing either gentamicin or tobramycin with CD microparticles demonstrated antimicrobial activity duration ranging from 28 to 50 days.

One aspect of the superior antimicrobial activity obtained with the composite drug combination PMMA-CD formulation was its inherent ability to be refilled with antibiotics following implantation through the use of CD microparticles. Previous studies have reported that there was a negligible amount of antibiotic refilled into PMMA samples containing no CD microparticles following implantation relative to what was capable of being refilled in samples containing CD microparticles115. The agarose phantom refilling studies conducted demonstrated that the composite drug combination formulations were able to be refilled with a comparable normalized amount of antibiotic to what was initially found in pre-filled samples. The subsequent persistence zone of inhibition study suggested that these refilled RMP PMMA-CD composite samples also retained their antimicrobial effect and that additional windows of antimicrobial activity were achieved after implantation. Furthermore, the refilling studies indicated the potential for the combinatorial antibiotic PMMA-CD composite system to be implanted in the 113

patient with empty CD microparticles where the patient can be provided with a local bolus of antibiotic into the tissue, near the implant as necessary. This would allow for a patient-specific therapy and would help to minimize the patient’s unnecessary exposure to antibiotics.

Without refilling, we were able to achieve between 30 and 60 days of antibiotic release and subsequent antimicrobial activity with the majority of drug combination PMMA-CD composite formulations that had pre-filled RMP CD microparticles. Through refilling composite PMMA-CD systems with empty CD microparticles, we were able to obtain up to 50 more days of antimicrobial activity. Approximately 76% of total hip arthroplasty and 68% of total knee arthroplasty PJIs occur within 60 days after surgery. After 90 days, the risk for PJIs decreases approximately 2-4% each month127. To have an efficient drug delivery system that is capable of treating bacterial infection from joint replacement surgeries, the duration of antimicrobial activity of the system should reach a minimum of

60 days, which is achievable through our refillable combinatorial antibiotic PMMA-CD composite system.

Refilling of CD microparticles in PMMA is not limited to RMP, meaning that other drugs can be added into the system post-implantation, allowing for a patient-specific therapy where drug combinations can be customized or optimized to treat PJIs depending upon the species of bacteria present locally. The PMMA-CD composite formulation would be able to target antibiotic-resistant strains of bacteria by refilling the CD microparticles

114

with antibiotics, which the bacteria have not been exposed to and potentially decreasing the need for revision surgeries due to PJIs.

3.5. CONCLUSIONS

In this study, a combinatorial antibiotic PMMA-CD composite system was developed to achieve a longer and more effective antimicrobial activity against a broader range of bacteria compared to clinically used monotherapy PMMA formulations for treatment of broad-spectrum PJIs. A 4- and 8-fold increase in the duration of antimicrobial activity against a variety of bacteria was observed (i.e., S. aureus, S. epidermidis, and E. coli) when 10 wt% RMP-filled CD microparticles were added into antibiotic-laden PMMA. In addition to the extended duration of antimicrobial activity, the PMMA-CD composite formulation also allowed for the refilling of new antibiotics into the system’s empty CD microparticles, meaning that antimicrobial activity has the potential to be prolonged and that patient-specific therapy can be achieved as new suitable antibiotics are added into the system. Finally, the compressive strength and porosity of the combinatorial antibiotic

PMMA-CD composite formulations were examined and a very small, but significant decrease in the mechanical properties and increase in porosity upon addition of CD microparticles was found. The resultant mechanical strength of the tested formulations of the combinatorial antibiotic PMMA-CD composite system suggested that the system may be most suitable for non-structural applications such as a temporary spacer in two-part arthroplasty revision procedures for treatment of PJIs, as well as antibiotic-filled beads,

115

and that further studies may be required to determine the composite’s use in load-bearing arthroplasty PJI applications.

3.6. ACKNOWLEDGEMENTS

The authors acknowledge financial support through the National Science Foundation

(NSF) Graduate Research Fellowship Program Grant no. CON501692 (E.L.C.), the

Center for Stem Cell and Regenerative Medicine Undergraduate Student Summer

Program (ENGAGE) (C-y.L.), the Support of Undergraduate Research & Creative

Endeavors (SOURCE) (D.W.M.), NIH NIAMS Ruth L. Kirschstein NRSA T32

AR007505 Training Program in Musculoskeletal Research (G.D.L.), and NIH

R01GM121477 (H.A.vR.).

116

CHAPTER 4. POLYMERIZED CYCLODEXTRIN CAN BE REFILLED WITH

ANTIBIOTICS IN THE PRESENCE OF BACTERIAL BIOFILMS

*This chapter was adapted with permission from Acta Biomaterialia. 2017. 57: 95-102.

Authors: Erika L. Cyphert, Sean T. Zuckerman, Julius N. Korley, Horst A. von Recum

Author contributions: ELC – experimental design, data analysis and collection (all experiments), preparation of figures, writing/revision manuscript. STZ – experimental design (biofilm models), data analysis (biofilm models), writing/revision manuscript.

JNK and HAvR – experimental design, data analysis, revision manuscript.

117

4.1. ABSTRACT

Current post-operative standard of care for surgical procedures, including device implantations (i.e. arthroplasties), dictates prophylactic antimicrobial therapy for prevention of PJIs, but a small percentage of patients still develop PJIs. Systemic antimicrobial therapy needed to treat PJIs can lead to downstream tissue toxicities and generate drug-resistant bacteria. To overcome issues associated with systemic drug administration, a polymer, polymerized CD, incorporating specific drug affinity has been developed with the potential to be filled or refilled with antimicrobials, post-implantation, even in the presence of a bacterial biofilm. This polymer could be used as an implant coating or stand-alone drug delivery device, and can be translated to a variety of applications, such as implanted or indwelling medical devices (i.e. arthroplasties), and/or surgical site infections (i.e. PJIs). The antibiotic filling of empty CD polymer was analyzed in an in vitro filling/refilling agarose model mimicking post-implantation tissue conditions. Filling in the absence of bacteria was compared to filling in the presence of bacterial biofilms of varying maturity to demonstrate proof-of-concept necessary prior to in vivo experiments. Antibiotic filling into biofilm-coated CD polymers was comparable to drug filling seen in same CD polymers without biofilm demonstrating that CD polymers retain their ability to be filled with antibiotics even in the presence of a biofilm.

Additionally, post-implantation filled antibiotics exhibited sustained bactericidal activity in a zone of inhibition assay demonstrating post-implantation capacity to deliver filled antibiotics in a timeframe necessary to eradicate bacteria in biofilms. This work has 118

shown that CD polymers can be filled with high levels of antibiotics post-implantation independent of biofilm presence potentially enabling device rescue, rather than removal of the implant, in the case of treatment of PJIs.

4.2. INTRODUCTION

PJIs are a major complication that can result following total knee and hip joint arthroplasties, often resulting in implant removal and potential serious side-effects128. S. aureus infections are one of the most frequent types of infections that can develop following these procedures and in severe cases can result in osteomyelitis129. S. epidermidis is less pathogenic than S. aureus, but both can form biofilms on the surface of implanted materials130–135. Biofilms are complex clusters of bacteria composed of live and dead bacteria, proteins, polysaccharides, and an extracellular matrix formed by the bacteria131,132. A biofilm begins developing when bacteria adhere to proteins that have adsorbed to the surface of the material. Over time, several layers of bacteria can cluster together on the surface of the material. A dense polysaccharide layer surrounds a mature biofilm forming a diffusional barrier that combined with a metabolic reduction in bacteria means traditional antibiotic treatments are often unsuccessful and can lead to resistant bacterial populations136–138. Oftentimes the only treatment for the biofilm-coated implant and surrounding necrotic tissue is surgical removal138. This current work has demonstrated the ability of a drug delivery polymer to be filled with antibiotic after implantation to potentially treat infected arthroplasty implants even when coated with a

119

biofilm, thereby rescuing the implant and possibility avoiding additional revision surgical procedures and implant removal.

Prophylactic antimicrobial therapy before and after surgery has led to rates of infection that are low compared to the total number of procedures: 1-4% of 1 million knee and hip arthroplasties completed in 2010 in the United States139,140; however, these infections still added up to nearly 2 million total cases of hospital-acquired infections costing the U.S. >

$11 billion annually128. Systemic prophylactic antimicrobial therapy can also potentially drive drug resistance in the bacteria141.

For those patients who develop a chronic PJI including osteomyelitis, treatment is very difficult often leading to osteolysis and significantly increased treatment costs and tissue morbidity25. To address these concerns, engineers have developed a variety of localized delivery approaches including antibiotic-filled PMMA bone cement25; however, all patients receiving the antimicrobial implant still receive antimicrobials independent of whether they develop a PJI. Ideally only patients who develop PJIs would be exposed to post-operative antimicrobials. Therefore, in this work a customizable CD polymer delivery system was developed that allows the physician to fill with antimicrobial only when an infection presents. Briefly, the physician would perform the surgical procedure as normal and implant a small amount of CD polymer without any drug at the end of the procedure prior to closing. If the patient is to develop an infection, the physician can then inject a small dosage of antimicrobial into the tissue near the location of the implanted

120

CD polymer. Some of the injected drug would immediately act on the planktonic bacteria while the remaining drug would be able to fill into the affinity “pockets” of the polymerized CD system. For patients at high risk of PJIs, the CD polymer could be refilled with drug prior to implantation and then refilled with either additional drug or a different drug if an infection presents.

In this work, the ability to refill antibiotics into CD-based polymers across the diffusional barrier of mature biofilms is reported for the first time, building upon the previous work of refilling CD-based polymers with the chemotherapeutic drug, doxorubicin85. This work demonstrated the versatility of the CD-based delivery system by refilling RMP and minocycline (MC). Specifically, in this paper, empty CD polymer disks were filled with antibiotics in a post-implantation tissue-mimicking agarose phantom. For the “non- affinity” control, a linear glucose polymer (polymerized dextran; pDEX) was selected since it represented a chemistry most similar to CD, but lacked the affinity interaction of the drug with the CD pocket (see Figure 4-1). Post-implantation drug filling into both non-affinity (pDEX) and affinity (CD) polymers was quantified, and the bioactivity of the filled drug was shown via a zone of inhibition assay. Since many patients do not present with PJIs immediately following arthroplasties, the effect of bacteria and biofilm development on post-implantation filling of CD polymer disks was also investigated.

Specifically, CD polymer disks were exposed to bacteria prior to filling using a modified protocol for developing mature biofilms142,143.

121

Figure 4-1: Set-up of agarose-based tissue-mimicking refilling models with (b) and without (a) biofilm.

4.3. MATERIALS AND METHODS

4.3.1. MATERIALS

See Chapter 3, section 3.3.1.

4.3.2. SYNTHESIS OF CD AND PDEX POLYMER DISKS

CD polymer disks and polymerized dextran (pDEX) were synthesized according to a previously published protocol and cross-linked with HDI in a molar ratio of 1:0.16

(glucose residue: HDI)47,49,55. Briefly, 1 gram of each β-CD pre-polymer and dextran were dissolved in 4 mL dimethylformamide (DMF; CD) or dimethylsulfoxide (DMSO;

122

dextran). CD polymers were cured at 70oC for 45 minutes and pDEX for 120 minutes and punched into 5 mm disks. Disks were thoroughly washed over a period of several days in

100% solvent, 50:50 solvent:water, and 100% water prior to use in order to remove unreacted residual polymer and solvent from CD and pDEX.

4.3.3. PREPARATION OF AGAROSE TISUE PHANTOM

An agarose gel was created by dissolving 0.075% weight agarose in PBS and bringing the solution to a boil. 10 mL of the hot solution was placed in each well of a Costar 6- well tissue culture plate and allowed to gel at room temperature. Two 5 mm disks (CD or pDEX) were placed on top of the agarose and an additional 5 mL hot agarose solution was placed on top. Polymer disks were aligned in the agarose to the center far left and right. The entire agarose solution was allowed to solidify at room temperature and a 3 mm biopsy punch was used to form a well in the center of each agarose gel to inject the drug solution (Figure 4-1A).

4.3.4. BIOFILM FOMATION ON POLYMER DISKS

A 5 mm CD or pDEX polymer disk was placed in a 5 mL solution of 2x Trypticase soy broth (BBL) and 50 μL of freshly cultured GFP-labeled S. aureus was added. Each solution was incubated at 37oC for 24 or 72 hours to form an immature or mature biofilm, respectively. The biofilm-coated polymer disks were then implanted into agarose as above.

123

4.3.5. AGAROSE-BASED BIOFILM REFILILNG MODEL

Biofilm coated polymer disks were carefully removed from the Trypticase soy broth and blotted on a Kimwipe to remove culture media and non-adherent bacteria. After the initial layer was placed on one side of the agarose and the biofilm coated CD polymer disk was placed opposite. The final hot agarose layer (5 mL) was added and allowed to set at room temperature. A 3 mm well was punched in the center of each agarose gel for the drug filling solution (Figure 4-1B).

4.3.6. ANTIBIOTIC REFILLING WITH AGAROSE

Drug solutions of RMP dissolved in methanol (1 mg/mL; 50 μL) and MC dissolved in

PBS (11 mg/mL; 100 μL) were injected into each agarose gel. Methanol was used to solubilize water-insoluble RMP and despite its inherent toxicity in large quantities, the amount used in the filling model (~0.03% of the total agarose volume) does not impact bacterial viability for in vitro work.

4.3.7. QUANTIFICATION OF ANTIBIOTIC REFILLING

The mass of RMP filled into the implanted polymers was quantified by leaching antibiotics into a DMF solvent sink. Every couple of hours, the entire volume of release media was removed from the sample and replaced with fresh DMF to maintain infinite sink conditions. The concentration of drug in the solvent was quantified by scanning small (200 μL) samples of release media with UV absorbance spectroscopy (485 nm,

124

RMP) and a calibration curve. Table 4-1A displays the averaged refilling data for each

RMP refilling condition.

Due to the challenge of MC oxidizing once filled into the disks, the mass of MC filled into the implanted polymers was quantified using UV absorbance spectroscopy using a

9x9 data point area scan of both empty polymer disks and MC filled disks (390 nm). An initial calibration curve was created of the absorbance of known quantities of MC filled into disks with the background disk absorbance subtracted. This calibration curve was used to correlate the absorbance of the MC refilled disks to the known mass of MC.

Table 4-1B displays the averaged refilling data for each MC refilling condition.

4.3.8. PERSISTENCE ZONE OF INHIBITION STUDY (REFILLED DISKS)

See Chapter 2, section 2.3.6.

4.3.9. BACTERIAL QUANTIFICATION OF BIOFILMS

The bacterial load present on biofilm coated CD and pDEX polymer disks was quantified at the start of refilling through agarose. Briefly, biofilm coated disks were implanted in agarose and removed once the agarose solidified. Each disk was placed in a tube with 4 mL of Trypticase soy broth and homogenized for 30 seconds with an Omni TH homogenizer. 70 μL aliquots of the homogenized solution were then spread on Trypticase soy agar plates in duplicate and incubated overnight. Bacterial colonies were counted and averaged. Bacteria in biofilms was also quantified after 1 and 2 days during refilling of

125

RMP carried out as detailed above. After 1 or 2 days, the disks were removed from the agarose, homogenized and bacteria plated and counted 24 hours later. Further experiments were conducted with mature biofilm coated CD polymer disks incubated for

1, 2, 3, 7, 14, and 21 days in agarose refilling conditions without any drug in order to detect changes in bacterial colony viability after different periods of incubation.

4.3.10. STATISTICAL ANALYSIS

All data is displayed as the mean of each condition tested in triplicate with the standard deviation as the error bars. One-way ANOVA statistics were performed on the antibiotic refilling data (Table 4-1) with Microsoft Excel 2013. A two-tailed Student’s t-test assuming unequal variances was performed on the antibiotic filling data (n = 3) and bacterial quantification data (n = 6) (Table 4-2). A t-test value < 0.01 was determined to be statistically significant.

4.4. RESULTS AND DISCUSSION

4.4.1. CD ANTIBIOTIC REFILLING AND QUANTIFICATION

The efficacy of the post-implantation antibiotic filling was evaluated with affinity (CD) and non-affinity (pDEX) polymer disks. Figure 4-2 (top) shows the disks embedded in the agarose after 52 hours (RMP) and 45 hours (MC) of antibiotic filling as well as the removed implants (Figure 4-2 bottom).

126

Figure 4-2: Agarose-based refilling model with antibiotic filled disks after 52 hours (RMP) and 45 hours

(MC) of filling (top images). Removed polymer disks after 52 hours (RMP) and 45 hours (MC) of filling

(bottom images). Dark orange (RMP) and yellow (MC) colors indicated the presence of the drug. Scale bars of wells = 10 mm (red) and of disks = 2 mm (black).

For both antibiotics tested, CD polymer disks were refilled with more drug compared to non-affinity (pDEX) polymer disks as qualitatively indicated by the dark orange color of

RMP and yellow/green color of MC and quantitatively indicated by the results of Table

4-1. Table 4-1A demonstrated that over three time more RMP was refilled into the CD

(affinity) polymer compared to the non-affinity (pDEX) polymer (p = 0.0035). Similar results were observed for MC filling into CD versus pDEX polymers where nearly three times more MC was filled into the CD affinity polymer compared to non-affinity control

(pDEX; p = 0.056) (Table 4-1B). The increased antibiotic refilling into CD polymers was attributed to the affinity “pockets” present in the polymer that were not present in the 127

conventional, non-affinity polymer (pDEX). The hypothesis was that the CD “pockets” in the polymer created a sink for the drug to stick into thereby maintaining a steady concentration gradient from the bulk agarose leading drug to concentrate into the CD polymer relative to that in the agarose system. Further, the increased filling of antibiotics into affinity-based CD polymer was attributed to the relatively high binding energies of

RMP (-4.39 kcal/mol)93 and MC for CD.

Table 4-1: Mass of antibiotics refilled into affinity (CD) and non-affinity (pDEX) polymer disks through agarose model with and without the presence of a bacterial biofilm. a. Rifampicin Average loaded drug mass (n=3) (μg) No biofilm CD 2.8 ± 0.6 pDEX 1.0 ± 0.4 24 hour biofilm CD 3.1 ± 0.6 pDEX 3.3 ± 0.4 72 hour biofilm CD 2.9 ± 0.2 pDEX 1.8 ± 0.2 b. Minocycline Average loaded drug mass (n=3) (μg) No biofilm CD 38.1 ± 15.6 pDEX 13.6 ± 0.9 24 hour biofilm CD 50.5 ± 1.2 pDEX 40.4 ± 2.0 72 hour biofilm CD 39.6 ± 9.0 pDEX 50.0 ± 7.3

Previous RMP release data demonstrated that under physiological release conditions (pH

7.4) RMP was gradually released from CD over 1200 hours144 and over 384 hours for

MC (manuscript in preparation). Regarding drug release through tissue (i.e., agarose refilling model), it was hypothesized that the CD delivery system would be capable of demonstrating the same, slow and consistent release profile for RMP and MC over > 1 128

month to treat infections, such as PJIs, over an extended period of time. Conversely, release from the pDEX polymer has typically exhibited a “burst” release profile with the majority of the release occurring over only several hours or days145. Therefore, with the

CD delivery system a more consistent and prolonged release profile of antibiotics can be obtained.

Further, the affinity-based CD polymer system has been shown to exhibit low degradation and be mechanically robust under in vivo physiological conditions144. Low degradation would be desirable for the treatment of long-term severe PJIs and ensures that the polymer has the mechanical integrity to withstand repeated filling and drug release cycles without significant changes to release kinetics. The mechanical integrity and viscoelastic properties of CD polymer systems have been extensively analyzed using thermo-gravimetric analysis, differential scanning calorimetry, and rheology with and without sterilization to confirm that the moduli of CD polymers are within the range of native tissue and that neither heat nor ethylene oxide sterilization significantly alter the mechanical integrity of the polymer49,144.

4.4.2. POST-IMPLANTATION REFILLING IN THE PRESENCE OF BIOFILM

After demonstrating that affinity interactions drive higher filling of both RMP and MC into CD polymer disks, the effect of bacterial biofilm’s diffusional resistance on filling polymers with drug was investigated. CD polymer disks were exposed to S. aureus for 24 hours to form an immature biofilm or 72 hours to form a mature biofilm143,146 prior to

129

embedding in the agarose system. There was no difference in the RMP refilling into CD polymer disks with the immature biofilm (3.3 ± 0.4 μg) compared to control CD polymer disks without any bacterial exposure (3.1 ± 0.6 μg; p = 0.6).

For CD polymer disks with mature biofilms, more RMP was present in the control disk without bacterial exposure (2.9 ± 0.2 μg) compared to the disk with the biofilm (1.8 ± 0.2

μg), but the difference was not statistically significant (p = 0.15). Two-way ANOVA showed that the CD polymer disks with the mature biofilm filled less RMP than the disks with a less mature biofilm (p < 0.005). No statistically significant differences were observed for MC filling into CD polymer disks with mature, immature, or no biofilms.

Overall, the presence of bacteria did not significantly affect the post-implantation filling of RMP or MC into the CD affinity polymer disks.

4.4.3. BACTERIAL QUANTIFICATION OF BIOFILMS

Bacteria present in the mature biofilm coated CD polymer disks was quantified in an effort to correlate drug refilling to the number of bacteria present in the biofilm. Each condition was tested with n = 6 replicates and averaged. Table 4-2 shows quantification of bacteria from CD polymer disks with mature biofilms after 1 and 2 days of RMP refilling time as expressed as a percentage of bacteria adherent on control CD polymer disks incubated in the agarose model for the respective time but without any antibiotic filling. The bacterial count for RMP filled CD polymer disks decreased as the refilling time increased and after two days the majority of the bacteria in the biofilm had been

130

eradicated. Bacteria in mature biofilms adhered on the surface of CD polymers was quantified after 1, 2, and 3 days of incubation in the refilling model without antibiotics and showed that there was no change in bacterial counts when antibiotics were not injected to day 3.

Table 4-2: Bacterial quantification of CD and pDEX polymer disks with mature biofilms (n = 6) after t = 1 and 2 days of RMP refilling as a percentage of control disks incubated for 1 or 2 days without antibiotic refilling. All data shown is reported as the average percentage of bacterial colonies remaining relative to control with the standard deviation.

% Remaining control colonies after drug % Remaining control colonies after drug exposure (n= 6) t = 24 hours exposure (n= 6) t = 24 hours CD 63.2 ± 50.0 8.9 ± 4.4 pDEX 39.8 ± 31.2 95.1 ± 82.3

Further, the effect of RMP filling into non-affinity pDEX control polymers with mature biofilms was evaluated. pDEX polymer disks had significantly more bacteria in the mature biofilm (p = 0.008) at the time of implantation into the agarose. Both CD and pDEX polymer disks with mature biofilms showed a decrease in bacteria at day 1 of

RMP filling. At day 2 the pDEX polymer disks with the mature biofilm showed similar bacteria counts as day 1 compared to the CD polymer disks, which showed a decrease.

Due to the large degree of variation in bacteria present in less mature biofilms (consistent with previous observations), the primary focus was on quantifying the bacteria present in mature biofilms143,146. While the mature biofilm did decrease the amount of drug present in the CD polymer disk slightly (p = 0.15; Table 4-1A), quantitative culturing revealed a 131

lower bacterial load on these disks. The hypothesis was that some of the initial antibiotic in the bolus injection was immediately used to eradicate the planktonic S. aureus around the biofilm and the rest of the antibiotic was filled into the CD polymers to be gradually released to eradicate the less metabolically active bacteria present in the biofilm, rather than the biofilm resisting drug diffusion into the CD disk. This hypothesis was further validated by the observation that the agarose background absorbance signal of the immature (t = 24 hour) was comparable to the signal of the mature (t = 72 hour) biofilm coated CD disk. Since the signal was similar and the more mature biofilm coated CD disk contained less drug, some of the initial injected drug must have been used to eradicate some of the bacteria in the mature biofilm. The presence of bacteria still remaining on the

CD polymer surface after the post-implantation filling underscores the importance of sustained antibiotic delivery where the drug released back into the biofilm over the period of days or weeks is needed to continue eradicating the bacteria present on the polymer or implant surface. Additional studies quantifying bacteria from mature biofilms on CD polymers after 1-3 days of incubation verified that RMP and not unfavorable survival conditions in the agarose model was responsible for eradicating the majority of bacteria in the biofilm. Further, the post-implantation refilling model used in this work supported bacterial biofilm viability for at least 7 days (13.3% of initial biofilm colonies) with minimal viability after 14 days.

In the bacterial quantification studies of the mature biofilm on pDEX control polymers, it was hypothesized that the initial decrease in bacteria count for CD and pDEX polymers 132

from t = 0 to day 1 was driven by the antibiotic bolus whereas the decrease from day 1 to

2 seen only for CD polymers with mature biofilms was driven by the affinity filling

(Tables 4-1 and 4-2). The lack of decrease in bacteria counts (over 2 days) for pDEX polymers with mature biofilms supported this hypothesis.

4.4.4. PERSISTENCE ZONE OF INHIBITION STUDY

To demonstrate that post-implantation filled antibiotic was active and capable of continued bacterial eradication, post-implantation filled RMP and MC CD polymer disks were removed from the agarose refilling model, and evaluated with a persistence zone of inhibition assay against S. aureus. Four separate conditions were tested in triplicate.

Figure 4-3 shows that post-implantation refilled RMP from CD polymer disks without a biofilm was able to eradicate S. aureus for 19 days. CD polymer disks with immature biofilms were also able to clear S. aureus out 19 days, but CD polymer disks with mature biofilms only cleared bacteria for 11 days, consistent with lower drug refilling (Table 4-

1A). The difference in bacterial eradication between RMP from immature and mature biofilms demonstrated that increased drug refilling into CD polymer (Table 4-1A) resulted in an extended antimicrobial timeframe. Figure 4-3 also showed that refilled MC was active for 19 days independent of the presence or maturity of biofilm. Overall, the results from this study confirmed that post-implantation refilled CD polymers were capable of releasing a therapeutic dose of active antibiotic sufficient to inhibit the growth of bacteria for nearly 3 weeks.

133

Figure 4-3: Persistence zone of inhibition study of post-implantation refilled RMP and MC filled CD polymer disks with and without immature (left) or mature biofilms (right) against S. aureus (n= 3).

Current clinical treatment for device infections typically involves a bolus antibiotic injection147, resulting in a spike in local antibiotic concentration on the scale of hours.

Any bacteria that remain on the surface of the device post-antibiotic bolus can continue growing to form a biofilm, and the one-time antibiotic dose can also contribute to the development of drug-resistant bacteria. In this work, the decreased bacterial load observed on polymer disks following bolus injection demonstrated how this one-time antimicrobial injection was insufficient to eradicate the infection (Table 4-2). The result

134

typically involves revision surgery and removal of the infected device, significantly increasing overall cost and putting the patient at risk for a second surgical procedure. The work presented here could improve the outcome by rescuing the infected device (e.g. arthroplasty implant) by providing a localized drug sink capable of delivering a steady concentration of antibiotic that can eradicate bacteria even with the decreased metabolic rates observed when a biofilm is present. Previously, biofilms had been observed to pose a diffusional barrier preventing lethal concentrations of drug to reach the bacteria. This work showed that drug does freely diffuse through a biofilm; however, the capacity to trap that drug and re-administer it at therapeutic doses had previously been missing. For the bacteria in the biofilm that are less metabolically active than planktonic, a re- administration of the drug on the scale of days to weeks is necessary to eradicate the bacteria and subsequent infection.

4.5. CONCLUSIONS

Prophylactic antimicrobial treatment is not clinically sufficient since patients may still present with PJI up to a year after arthroplasty surgeries. In these cases, the PJI will likely be well established and potentially have formed a biofilm. The in vitro evaluation of post- implantation drug filling presented in this work indicated neither bacteria nor biofilm impaired antibiotic filling or the CD polymer’s ability to eradicate bacteria. The delivery system in this work utilized two components to treat established infections. A bolus antibiotic injection served to eradicate the planktonic bacteria surrounding the initial

135

infection, while the implanted CD polymer served as an antibiotic depot slowly re- administering drug to eradicate less metabolically active bacteria that are present in a biofilm. The work presented here demonstrates for the first time that antimicrobials can be filled into CD polymers independent of the presence of a biofilm and opens the door to potentially rescuing infected arthroplasty implant devices rather than removing them if a

PJI or biofilm develops.

4.6. ACKNOWLEDGEMENTS

The authors would like to thank Christopher Hernandez (Case Western Reserve

University) for assistance in the development of the reproducible in situ agarose filling model. This work was funded by NIH NIGMS SBIR 1R43GM100525 and NSF

CAREER Award CBET-0954489.

136

CHAPTER 5. ANTIBIOTIC REFILLING OF PMMA BONE CEMENT

COMPOSITE THROUGH BONE AND MUSCLE TISSUE TO TREAT

PERIPROSTHETIC JOINT INFECTIONS

5.1. ABSTRACT

While PJIs result in a small percentage of patients following arthroplasties, they pose a considerable treatment challenge if they develop and spread into bone and surrounding soft tissue. In advanced cases, patients may face prolonged immobility, high treatment costs, and exposure to systemic antibiotics for extended periods of time. Proactive treatment for PJIs involves locally administering antibiotics prophylactically using antibiotic-laden PMMA bone cement. Drug elution from antibiotic-laden PMMA generally is insufficient to treat prolonged infections. Previous work has demonstrated efficacy of incorporating CD microparticles into PMMA to improve antibiotic release, enable extended antimicrobial activity, and allow for drug refilling following implantation in an agarose-based tissue mimicking model. To better simulate how antibiotic refilling may be possible in more physiologically-relevant tissue models, this work investigated development of bone and muscle tissue refilling models to treat two aspects of PJIs. A hard tissue refilling model was developed by embedding PMMA-CD composite in rabbit femur and administering antibiotic via an intraosseous infusion technique. A soft tissue refilling model was developed by implanting PMMA-CD composite beads in bovine muscle tissue and administering antibiotic via injection

137

directly into the tissue with and without presence of biofilm. Duration of antimicrobial activity of refilled PMMA-CD was evaluated. PMMA-CD composite in both hard and soft tissue models was capable of being refilled with antibiotics and resulted in prolonged

(14-40+ days) of antimicrobial activity, demonstrating potential efficacy of using

PMMA-CD composite for treatment of prolonged PJIs in bone or soft tissue that can be refilled after implantation on a patient-specific basis.

5.2. INTRODUCTION

PJIs can be a devastating complication resulting from some arthroplasties and can involve infection of both the soft tissue surrounding the prosthetic implant as well as inside of the bone in severe cases148. To mitigate the development of and to treat PJIs, antibiotic-laden

PMMA bone cement is often utilized where it is either directly implanted in the bone (i.e. arthroplasty fixation) or in the form of beads that are embedded in the surrounding infected soft tissue to locally deliver antibiotics63,64,78,115,149–151. When antibiotic-laden

PMMA is directly implanted into the bone, there is a unique challenge to administer additional antibiotic doses into the implanted PMMA if the infection is prolonged. As more drug cannot be added to the PMMA after it is implanted in bone, the duration of antimicrobial activity is limited based upon the initial amount of antibiotic added to the

PMMA upon fabrication152,153. When antibiotic-laden PMMA beads are implanted and packed into soft tissue near the infected surgical site in arthroplasty revision procedures151,154,155, the duration of antimicrobial activity of antibiotic-PMMA beads can

138

be limited if bacterial biofilms or resistant bacteria adhere to the surface of the bead154,155.

Specifically, if antibiotic is released from the PMMA beads at sub-therapeutic levels, bacteria can adhere and form a thick biofilm matrix that is impenetrable to many antibiotics, rendering the treatment unsuccessful155.

Traditionally, to prepare antibiotic-laden PMMA, antibiotics such as aminoglycosides and glycopeptides are often incorporated into the cement during polymerization114.

However, when antibiotics are added to PMMA in this manner, it often results in an insufficient release of drug. Typically, a “burst” release is observed initially where a substantial amount of drug is released within a very short period of time and is followed by a prolonged period of antibiotic release at sub-therapeutic levels113. Oftentimes much of the drug added initially can remain trapped in the PMMA permanently with one study showing upwards of 85% of drug remaining trapped over an extended period of time156.

In conjunction with the antibiotic-filled PMMA, additional antibiotics may be administered either intravenously or intraosseously to treat PJIs and orthopedic infections in an effort to supplement the poor antibiotic release kinetics from PMMA157–159.

Intraosseous (IO) infusion offers a unique advantage over traditional intravenous administration as it can provide a higher local tissue antibiotic concentration while reducing off-target systemic effects157. IO infusion is generally used in clinical scenarios in which venous access is compromised and involves placing a specialized needle (16 or

18 gauge; 2.5-3 cm long) directly into the medullary cavity in the bone marrow (distal femur or proximal tibia) and remains in place as a port (up to 72 hours) to provide direct 139

venous access for therapeutics160,161.

In an effort to improve upon the antibiotic release kinetics and to enable post- implantation antibiotic refilling to occur, a PMMA composite containing polymerized

CD microparticles was previously developed (see Chapters 2-3)115,150. While much success has been observed using the agarose-based refilling model, it is not the most representative of bone or muscle tissue. To better recapitulate what PMMA-CD composite refilling may look like to treat two aspects of PJIs where 1) PMMA is embedded in bone for treatment of osteomyelitis and 2) PMMA beads are embedded in soft muscle tissue for treating infections surrounding the bone, both hard and soft tissue antibiotic refilling models were developed in this work.

To simulate how antibiotic refilling of PMMA-CD composite embedded in femur could reasonably occur, a refilling model was developed in which PMMA-CD composite was embedded in the bone and IO infusion was used to provide direct access to the bone marrow to administer antibiotics and allow for refilling of embedded PMMA-CD. The antibiotic refilling ability of PMMA with and without CD microparticles embedded in femur was evaluated through post-refilling analysis of the depth of antibiotic diffusion into PMMA-CD composite and the resultant duration of antimicrobial activity. To translate the refilling model for the application of PMMA beads to treat soft muscle tissue

PJIs, PMMA beads with and without CD microparticles were implanted in bovine muscle tissue and the tissue was directly injected with a bolus of antibiotic. Extent of refilling

140

capacity of PMMA-CD composite beads was evaluated through administration of iterative antibiotic doses and subsequently investigating the resultant duration of antimicrobial activity of refilled PMMA-CD beads. To explore the possibility of refilling

PMMA-CD composite beads through muscle tissue if they are initially covered in bacterial biofilm, a modified soft-tissue refilling model was developed where biofilms were formed on composite beads prior to implantation in the tissue. Duration of antimicrobial activity of refilled PMMA-CD composite beads with biofilm was evaluated and remaining colony forming unit counts on the bead and in the surrounding tissue were quantified after antibiotic treatment. Implementation of the hard and soft tissue refilling models developed in this work could help to provide insight into novel treatment options to improve outcomes for patients with chronic or prolonged PJIs.

5.3. MATERIALS AND METHODS

5.3.1. MATERIALS

Lightly epichlorohydrin cross-linked β-CD pre-polymer (116 kDa molecular weight) was purchased from CycloLabs (Budapest, Hungary). Ethylene glycol diglycidyl ether was purchased from Polysciences Inc. (Warrington, PA). RMP was purchased from Research

Products International (Mt. Prospect, IL). Simplex® HV (high viscosity) radiopaque bone cement (Stryker Orthopaedics, Mahwah, NJ) was purchased from eSutures (Mokena, IL).

New Zealand white rabbit knee joint tissue was graciously donated by the Animal

Resource Center (ARC) at Case Western Reserve University (CWRU, Cleveland, OH).

141

Bovine abdominal muscle tissue was purchased from a local butcher (Cleveland, OH).

Cook® Medical Intraosseous (IO) infusion needles (16 gauge, 4 cm) (Bloomington, IN) were purchased from eSutures. Green fluorescent protein (GFP)-labeled Staphylococcus aureus bacteria was kindly provided by Dr. Ed Greenfield (CWRU, Cleveland, OH).

5.3.2. CD MICROPARTICLE SYNTHESIS

See Chapter 2, section 2.3.2.

5.3.3. PMMA-CD COMPOSITE PREPARATION

For both bone and muscle tissue refilling models, PMMA was prepared using a similar methodology, according to manufacturer instructions. Specifically, 2 gram samples of

Simplex® HV surgical grade bone cement powder was mixed with 1 mL methyl methacrylate monomer until a soft dough formed. For PMMA samples containing CD microparticles, 15 wt% CD microparticles (wt/wt) were added to each batch of PMMA prior to addition of liquid monomer and mixed until homogeneous. For bone refilling model PMMA samples, the soft dough was directly finger-pressed and cured in the bone

(displacing bone marrow) and had dimensions of approximately 6 mm tall and 4 mm diameter. For muscle tissue refilling model PMMA samples, the soft dough was pressed into a thin layer (~2 mm thick) and small beads were punched out (6 mm diameter) and allowed to cure at room temperature.

142

5.3.4. FEMUR PMMA-CD COMPOSITE REFILLING MODEL

To simulate how we envisioned PMMA-CD composite antibiotic refilling in the scenario of treating arthroplasty PJIs, we set out to develop a PMMA-CD refilling model of the femur. Femurs were harvested from New Zealand white female rabbits and the femoral head epiphysis and metaphysis and patellar epiphysis and metaphysis were removed. The remaining diaphysis shaft was measured and cut in half yielding two femoral samples from 1 bone. Bone marrow remained intact throughout. The innermost cut end of each bone was filled (6 mm deep) with PMMA-CD composite (either with or without 15 wt%

CD microparticles) and allowed to harden. The opposite end of the bone was capped with a thin layer of PMMA (without CD microparticles). An IO infusion needle (16 gauge, 4 cm) was used to drill a hole into the femur 11 mm away from the end of the bone filled with PMMA-CD composite and remained in place to serve as a port to directly inject antibiotic into the medullary cavity. The IO infusion needle was secured in place using mounting putty and cyanoacrylate glue was used to lightly cover both ends of the bone to prevent leakage of the drug. The femur section containing the IO infusion needle was placed on top of 5 mL of solidified agarose (0.075% wt/vol) in a 6-well plate and covered with an additional 8 mL agarose to fully cover the bone sample62. Once the agarose solidified, 100 μL of RMP (12 mg/mL in methanol) was injected directly into the IO needle into the bone marrow of the femur. The plate was covered and placed in a 37oC incubator for 48 hours. After incubation, bone segments were removed from the agarose,

IO infusion needles were removed, and the bone was sliced in half vertically in order to 143

visualize the depth of refilling of RMP into the embedded PMMA-CD composite. Each condition was carried out in quadruplicate, consisting of both halves of the right and left femur of a single rabbit.

5.3.5. ANALYSIS OF BONE REFILLING PMMA-CD COMPOSITE – DEPTH OF RMP

REFILLING AND DURATION OF ANTIMICROBIAL ACTIVITY

In an effort to evaluate the extent of antibiotic refilling possible of PMMA-CD composite when it was embedded in the femur, refilled PMMA-CD composite (from section 5.3.4) was explanted from the bone, sliced in half vertically and analyzed using two metrics.

First, stereomicroscope images were collected of the interior cut face of the PMMA-CD composite while it was still embedded in the bone and of the front and back of the explanted PMMA-CD composite (with and without 15 wt% CD microparticles) to provide direct visualization of refilling. Depth of RMP refilling was directly quantified from the stereomicroscope images using ImageJ where up to 8 measurements were collected on each half of the sample measured from the edge of the sample until the end of the visible “orange” region (based on visual inspection) and averaged. Then, to determine the duration of antimicrobial activity possible from refilled PMMA-CD composite, explanted PMMA-CD samples were evaluated in a persistence zone of inhibition study against S. aureus (see section 2.3.6.).

144

5.3.6. SOFT TISSUE PMMA-CD COMPOSITE REFILLING MODEL

To recapitulate the scenario in which antibiotic-filled PMMA-CD composite beads are used to treat soft tissue infections, a refilling model was developed to mimic RMP refilling through bovine abdominal muscle tissue. Small sections of abdominal tissue (2 cm x 2 cm x 1 cm) were prepared, a small horizontal incision was made in the side (1 cm thick) of the tissue and a PMMA-CD composite bead (with or without 15 wt% CD microparticles) was implanted. A small amount of cyanoacrylate glue was used to seal off the incision. The tissue containing the PMMA-CD composite bead was then embedded in agarose (13 mL total) in a 6-well plate, as described previously (see section 2.3.9.)62,115.

Once the agarose set, two 100 μL injections of RMP (12 mg/mL dissolved in methanol;

200 μL total) were injected directly into the tissue at different locations approximately 6 mm away from the implanted PMMA-CD composite bead. The plate was then covered and incubated at 37oC for 48 hours. Following 48 hours of incubation, tissue was either removed from the agarose and the PMMA-CD beads were explanted and analyzed (for 1 cycle of refilling) or an additional 200 μL bolus of RMP (12 mg/mL in methanol) was injected into the tissue (two 100 μL injections directly into tissue) and incubated for an additional 48 hours (for 2 cycles of refilling, or repeated twice for 3 cycles of refilling).

Each condition was completed in triplicate.

145

5.3.7. ANALYSIS OF REFILLING PMMA-CD COMPOSITE IN SOFT TISSUE –

STEREOMICROSCOPE IMAGES, DURATION OF ANTIMICROBIAL ACTIVITY

Following 1-3 cycles of refilling, PMMA-CD composite beads were explanted from the soft tissue and analyzed using two methods to determine the extent of antibiotic refilling possible in soft tissue. First, images of both the front and back of each PMMA-CD composite bead (with and without 15 wt% CD microparticles) were collected using a stereomicroscope to provide a direct visualization of the antibiotic refilling. Then, the duration of antimicrobial activity possible from each refilled sample was determined through a persistence zone of inhibition study against S. aureus bacteria according to a previously described methodology (see section 2.3.6.).

5.3.8. SOFT TISSUE INFECTION PMMA-CD COMPOSITE REFILLING MODEL

In an effort to evaluate the possibility of allowing for antibiotic refiling to occur in soft tissue in the scenario in which a bacterial biofilm has formed on the surface of the implanted PMMA-CD composite bead, a modified soft tissue infection refilling model was developed. Prior to implantation in the soft tissue sample (2 cm x 2 cm x 1 cm),

PMMA-CD composite beads (with and without 15 wt% CD microparticles) were submerged in 1.5 mL of S. aureus culture and incubated for 48 hours at 37oC. Beads were lightly dried off and embedded into the soft tissue sample and the tissue was held together with cyanoacrylate glue and embedded in agarose (5 mL bottom, 8 mL top) in a

6-well plate and allowed to solidify. Tissue was incubated for 24 hours at 37oC to allow

146

for the infection to further develop. After this time, two 100 μL injections of RMP (12 mg/mL in methanol; 200 μL total) were injected directly into the tissue at different locations ~6 mm away from the implanted PMMA-CD composite bead and the plate was covered and incubated at 37oC for 48 hours. As a control, a subset of tissues containing

PMMA-CD composite (with and without 15 wt% CD microparticles) was not injected with any RMP to provide growth control counts of the S. aureus CFU without the presence of drug (72 hours incubation total prior to explanting). After 48 hours of incubation with drug (or blank methanol control), tissue was removed from the agarose and the PMMA-CD composite bead was explanted and analyzed.

5.3.9. ANALYSIS OF REFILLING PMMA-CD COMPOSITE WITH BIOFILM – EXTENT

OF INFECTION REMAINING, DURATION OF ANTIMICROBIAL ACTIVITY

To evaluate the extent of antibiotic refilling possible in the presence of a bacterial biofilm on the surface of the PMMA-CD composite bead through soft tissue, the explanted refilled beads were analyzed using two metrics.

First, the extent of infection remaining in the tissue and on the surface of the PMMA-CD composite bead after treatment with RMP or methanol was analyzed through quantification of CFU counts. All of the tissue surrounding the PMMA-CD composite beads was homogenized in 15 mL of 2x Trypticase soy broth. PMMA-CD composite beads were sonicated for 30 minutes in 3 mL of 2x Trypticase soy broth. For both solutions (from tissue and PMMA-CD composite beads), 70 μL aliquots of solution were

147

plated out on Trypticase soy agar plates and incubated overnight. CFU counts were completed using ImageJ (2018). Extent of infection remaining after treatment with antibiotic was compared relative to the CFU count in control samples without any antibiotic.

Separately, to analyze the duration of antimicrobial activity of PMMA-CD composite beads refilled with RMP through biofilm, refilled beads were explanted from the tissue and evaluated in a persistence zone of inhibition study against S. aureus (see section

2.3.6.).

5.3.10. STATISTICAL ANALYSIS

All data was reported as the mean and standard deviation of a minimum of 3-4 samples.

Statistical analyses were carried out in Microsoft Excel 2016. Two-tailed Student’s t-tests with unequal variances were conducted for depth of antibiotic diffusion (n = 4), persistence zone of inhibition (n = 3), and CFU count (n = 3). P-values of less than 0.05 were considered to be statistically significant (p < 0.05 = *, p < 0.01 = **).

5.4. RESULTS AND DISCUSSION

5.4.1. FEMUR PMMA-CD COMPOSITE REFILLING MODEL

The aim of this work was to evaluate the possibility of allowing for PMMA-CD composite (containing 15 wt% CD microparticles) to be refilled with antibiotic, RMP, following implantation in bone for the treatment of arthroplasty PJIs. Previous work has

148

demonstrated the ability of PMMA-CD composite to be refilled in an agarose-based tissue phantom, but this model did not sufficiently replicate the properties of diffusion through hard tissue, such as bone. To allow for appreciable refilling to occur in the

PMMA-CD composite, it is necessary that the PMMA-CD be in contact with a high concentration of antibiotic for a period of time (i.e. that the drug is not immediately cleared away). Since antibiotic cannot freely diffuse through bone, it is necessary to penetrate the bone and inject drug directly into the patient’s bone marrow using IO infusion to get a high enough concentration of antibiotic in direct contact with the implanted PMMA-CD. Figure 5-1 depicts a schematic of the general set-up of the femur

PMMA-CD composite refilling model developed.

Figure 5-1: Schematic of the set-up of femur PMMA-CD composite refilling model where PMMA containing CD microparticles was embedded in the bone and an IO infusion needle was placed near the

149

implanted PMMA-CD to serve as a port to inject antibiotics. Extent of antibiotic refilling of PMMA-CD using the model was evaluated through cutting bone in half, explanting PMMA-CD composite, imaging, and evaluating duration of antimicrobial activity against S. aureus.

Once the femur model was prepared, a bolus of RMP was injected into the IO port and the bone segment was incubated for 48 hours. Figure 5-2 depicts an image of the prepared femur segment with PMMA-CD composite embedded, insertion of the IO needle, and the interior of the femur segment after 48 hours of refilling with RMP.

Figure 5-2: Images of prepared femur segment with embedded PMMA-CD composite (left), femur segment embedded in agarose with IO needle inserted (middle), and interior of the femur segment 48 hours following refilling with RMP (right). Red/orange color along the periphery of PMMA-CD was indicative of

RMP refilling.

150

Refilling of PMMA-CD composite embedded in the femur model was compared to refilling of PMMA without CD microparticles, as a control. Figure 5-3 depicts representative images of the interiors of the PMMA-CD composite (with and without 15 wt% CD microparticles) that was embedded in the femur model and refilled with RMP for 48 hours.

Figure 5-3: Representative stereomicroscope images of the interior of PMMA-CD composite with and without (plain) 15 wt% CD microparticles after 48 hours of refilling with RMP in femur model.

Figure 5-3 qualitatively confirmed that refilling of PMMA-CD composites was possible using the IO infusion technique as indicated by the presence of the red/orange color along the periphery of the interior of the sliced PMMA-CD. Specifically, RMP has a red/orange 151

color and the presence of this color along the periphery of the PMMA was indicative of refilling. It was important to note the lack of the “orange” periphery on the plain PMMA samples (without CD microparticles), highlighting the necessity of the inclusion of CD microparticles into the PMMA to enable refilling and demonstrating that the “orange” periphery was not due to staining from blood or bone marrow in the femur model, but rather drug. To quantify the amount of refilling possible in samples with and without CD microparticles using the IO infusion technique, stereomicroscope images were collected of the interior of all embedded PMMA-CD samples and ImageJ was used to measure the depth of RMP diffusion into PMMA-CD. This measurement provided insight regarding how much PMMA-CD material is feasibly used in drug refilling and delivery functions.

Figure 5-4 provides quantification of the depth of RMP refilling in PMMA-CD composite with and without CD microparticles and depicts how representative measurements were completed. Up to 8 measurements were completed on each sample from the edge to the interior where the “orange” color ended, based on visual inspection.

Measurements were completed and averaged on both halves from 4 independent samples of each condition.

152

Figure 5-4: Quantification of the depth of RMP diffusion into PMMA-CD composite with and without CD microparticles in the femur model and images depicting how representative measurements were collected.

Red lines on photos indicate where diffusion was measured from. **Indicates p < 0.01. Ruler marks on images = 1 mm.

Composite samples containing 15 wt% CD microparticles refilled RMP an average depth of 295 ± 50 μm, whereas samples without CD microparticles refilled RMP a comparably negligible amount (60 ± 8 μm) (p = 1.5 x 10-6). Thus, demonstrating that in order to achieve antibiotic refilling in PMMA implanted in bone, incorporation of CD microparticles was critical. While the depth of RMP refilling on the periphery of PMMA- 153

CD composite was mostly uniform on individual samples, refilling depth did vary on some samples. Lack of uniformity of RMP refilling depth was generally attributed to inhomogeneity in the distribution of CD microparticles in the PMMA-CD composite. For example, if greater amounts of CD microparticles were clustered at the periphery, a greater depth of RMP diffusion would theoretically result.

To determine the duration of time in which RMP refilled PMMA-CD composite (with and without CD microparticles) was able to provide therapeutically-relevant antimicrobial activity, refilled samples were evaluated in a persistence zone of inhibition study against S. aureus. PMMA-CD samples were sliced in half and placed interior-side down on agar plates seeded with S. aureus. Figure 5-5 displays how the size of the zone of inhibition changed when PMMA-CD refilled samples were challenged to a fresh lawn of bacteria every day.

154

Figure 5-5: Persistence zone of inhibition study of PMMA-CD composite (with and without CD microparticles) refilled with RMP in femur model against S. aureus. *Indicates p < 0.05.

Results from the persistence zone of inhibition study suggested that use of the IO infusion technique resulted in ~14 days of additional antimicrobial activity after PMMA-CD composite was implanted in the bone. Thus, demonstrating the potential of the refilling model to provide the patient with an alternative treatment to PJIs where it may not be necessary to physically remove the infected arthroplasty implant, but simply locally administer more antibiotic as needed. Currently, if a PJI develops, the first line of treatment is systemic antibiotics followed by one- or two-stage revision procedures that result in significant trauma, mobility restrictions, and cost to the patient162,163. The ability to refill PMMA-CD composite after it has been implanted in bone to treat prolonged PJIs,

155

could help to dramatically improve PJI treatment and patient outcomes without the need for revision surgeries. Furthermore, the PMMA-CD composite system has the potential to be patient-customizable where the antibiotic administered can be tailored to target a specific causative pathogen.

Specifically, in the persistence zone of inhibition study, PMMA containing 15 wt% CD microparticles was able to clear S. aureus for > 14 days, whereas PMMA without CD microparticles (plain) was only able to clear S. aureus for < 7 days. Over the first 6 days of the study, PMMA with 15 wt% CD microparticles had a significantly larger zone of inhibition than the control, plain PMMA (p < 0.05; except after 1 day where p > 0.05).

Antimicrobial activity resulting from the plain PMMA was attributed to the small amount of RMP that passively diffused into the periphery of the PMMA. Whereas, the prolonged duration of antimicrobial activity resulting from PMMA-CD composite was attributed to the greater amount of drug that was able to be refilled (related to the increased depth of diffusion – see Figure 5-4) as well as the more complex affinity-based drug release mechanism of CD. Rather than just passively diffusing out of the plain PMMA, incorporation of CD into the PMMA enabled drug to bounce in and out of CD’s slightly hydrophobic inner “pockets” due to its affinity for CD prior to diffusing out of the

PMMA. Thus, enabling a more prolonged and consistent release of drug from the PMMA and subsequent antimicrobial activity.

It is important to note this femur refilling model was designed under the assumption that

156

the PJI will predominantly be developing internally (i.e. osteomyelitis) rather than externally to the bone164,165.

5.4.2. SOFT TISSUE PMMA-CD COMPOSITE REFILLING MODEL

Figure 5-6 depicts a schematic of the set-up of the soft tissue PMMA-CD composite refilling and infection refilling models where 1) the ability of PMMA-CD beads containing 15 wt% CD microparticles to be refilled with RMP through soft tissue and 2) the ability of the same beads to be refilled in the presence of a bacterial biofilm was analyzed. The soft tissue refilling model (without biofilm) was carried out over three consecutive iterations of refilling to probe the maximum amount of RMP that could be refilled into the implanted PMMA-CD beads and to evaluate if refilling could be repeatedly performed in the model. This experiment functioned to simulate the scenario in which the patient may require multiple doses of antibiotic for persistent soft tissue

PJIs. Figure 5-7 provides a visual representation of the set-up of the implantation of the

PMMA-CD bead in the tissue in agarose, injection of RMP, and incubation in tissue.

157

Figure 5-6: Preparation of tissue sample for implantation of PMMA-CD bead for refilling and infection models. Success of refilling PMMA-CD bead through soft tissue (with and without infection) was analyzed through persistence zone of inhibition studies and CFU counts.

158

Figure 5-7: Images of implantation of PMMA-CD bead in tissue in agarose, initial injection of RMP into tissue, and tissue following 48 hours of incubation and refilling with RMP.

PMMA-CD beads with and without 15 wt% CD microparticles were implanted in the tissue and refilled with either 1, 2, or 3 consecutive bolus injections of RMP. Following incubation with RMP, PMMA-CD beads were explanted from the tissue and imaged using stereomicroscopy. Figure 5-8 depicts the images of explanted PMMA-CD beads after 1-3 cycles of RMP refilling.

159

Figure 5-8: Stereomicroscopy images of PMMA-CD beads after 1-3 cycles of RMP refilling in soft tissue.

Figure 5-8 revealed that as additional injections of RMP were subsequently administered in the soft tissue model, the amount of RMP refilled in the PMMA-CD beads with 15 wt% CD microparticles increased, as indicated by the increased red/orange color.

PMMA-CD initially had a white/opaque color, therefore, when exposed to the red/orange

RMP, the color change of the PMMA-CD indicated RMP refilling. Without CD microparticles (i.e. plain PMMA), the PMMA demonstrated very little to no RMP refilling across three cycles. To evaluate the duration of antimicrobial activity possible from PMMA-CD samples refilled over 3 cycles, a persistence zone of inhibition study was carried out against S. aureus. Figure 5-9 displays the results of the persistence zone of inhibition study for PMMA-CD beads refilled over 1-3 cycles with RMP.

160

Figure 5-9: Persistence zone of inhibition study of PMMA-CD beads after 1-3 cycles of RMP refilling in soft tissue against S. aureus.

Results from the persistence zone of inhibition study (Figure 5-9) demonstrated that through repetitive refilling in soft tissue, PMMA-CD composite beads had the ability to provide additional windows of antimicrobial therapy (~25-40+ days). As PJIs can be prolonged and present between < 1 month to more than 2 years following initial surgery, development of a refillable, patient-customized PMMA antibiotic delivery system that can be refilled through soft tissue provides a welcome alternative to many existing, more traumatic therapies166.

Specifically, plain PMMA beads (without CD microparticles) were only able to clear 161

bacteria for ~8 days independent of the number of RMP refilling cycles. This suggested that there was a threshold for the amount of antibiotic that could passively diffuse into plain PMMA that was reached after a single cycle of refilling with RMP. In contrast,

PMMA-CD beads were able to clear bacteria for ~25 days (1 or 2 cycles of RMP refilling) or > 40 days (3 cycles of RMP refilling). The increased duration of antimicrobial activity resulting from additional cycles of refilling, demonstrated that with

CD microparticles, the threshold for RMP refilling of PMMA was substantially higher than that of plain PMMA and that the threshold was not reached after 1 or 2 cycles of refilling. Generally, the size of the zone of inhibition for plain PMMA beads was significantly smaller than that of PMMA-CD composite refilled for the same number of cycles (p < 0.05).

5.4.3. SOFT TISSUE INFECTION PMMA-CD COMPOSITE REFILLING MODEL

When biomaterials remain implanted in a patient over a prolonged period of time, they pose the risk of forming bacterial biofilms on their surface, particularly if they are being used in antimicrobial applications167–169. As a result, we were interested in exploring how soft tissue refilling into PMMA-CD composite beads may or may not be impacted if a bacterial biofilm has formed on the surface of the PMMA-CD. S. aureus biofilms were statically formed on the surface of PMMA-CD beads over 48 hours prior to implantation in tissue. Once implanted, the tissue was incubated for 1 day to allow for the infection to develop and either a bolus of RMP (dissolved in methanol) or blank methanol was

162

injected directly into the tissue and incubated for an additional 48 hours. PMMA-CD beads were removed from the tissue after this time and CFU counts were performed on both the PMMA-CD bead (via sonication) and the surrounding tissue (via homogenization). Figure 5-10 displays the CFU counts remaining on either the PMMA-

CD bead (with and without CD microparticles) and the tissue surrounding each bead after treatment with RMP as a percentage of colonies remaining relative to no treatment (i.e. blank methanol).

Figure 5-10: Percent remaining CFU counts of S. aureus on PMMA-CD beads (with and without 15 wt%

CD microparticles) and respective tissue after treatment with RMP relative to no treatment. *Indicates p <

0.05.

Results from Figure 5-10 demonstrated that regardless of the presence of CD microparticles in the PMMA, RMP was able to eradicate a substantial amount of bacteria

163

present both in the tissue and in the biofilm on the surface of the PMMA bead.

Specifically, upon treatment with RMP only 0.003% and 0.0009% of bacteria remained on the surface of the PMMA-CD bead and in the tissue, respectively. Whereas, 0.012% and 0.0026% of bacteria remained on the plain PMMA bead and in the tissue. While there was a small decrease in the bacterial load of PMMA-CD composite samples compared to plain PMMA, it is important to consider that both types of PMMA beads nevertheless resulted in significant decreases in bacterial load relative to controls without

RMP treatment (> 99.975% initial bacteria eradicated after RMP treatment). Upon exploring the remaining bacterial load after RMP treatment, the refilled PMMA-CD beads were also evaluated for their duration of antimicrobial activity in a persistence zone of inhibition study (Figure 5-11), as a reflection of their capacity to be refilled in the presence of a bacterial biofilm.

164

Figure 5-11: Persistence zone of inhibition of PMMA-CD beads explanted from soft tissue infection model after being refilled with RMP for 48 hours. *Indicates p < 0.05.

Figure 5-11 helped to demonstrate that while the bacterial load present on PMMA-CD beads and in tissue after treatment with RMP was similar in PMMA with and without CD microparticles, composition of PMMA did have a dramatic impact on the amount of

RMP filled into PMMA beads in the presence of infection. Specifically, PMMA-CD beads refilled in the presence of infection were able to clear bacteria for 40+ days, whereas plain PMMA beads (without CD microparticles) were only able to clear bacteria for < 7 days. Demonstrating that with PMMA-CD beads, not only was it possible to clear most of the biofilm on the bead and in the tissue, but it was possible to refill antibiotic in

165

the implanted bead and obtain a prolonged duration of antimicrobial activity. The slightly decreased bacterial load in Figure 5-10 using PMMA-CD composite supported evidence from Figure 5-11 that a greater amount of drug was filled in the PMMA-CD bead, enabling a longer release to more effectively clear the infection, relative to plain PMMA.

These results provided further support regarding the potential therapeutic efficacy of

PMMA-CD composite beads to treat soft-tissue PJIs, enabling prolonged antimicrobial activity on demand, even in the presence of biofilm. To elaborate, PMMA-CD beads have the ability to be refilled with drug on a patient-specific basis where the type, amount, and frequency of antibiotic bolus can be tailored based on the causative pathogen to obtain the desired duration of antimicrobial activity.

The large error bars present in Figure 5-11 were a reflection of the inherent variability of biofilm formation on the surface of PMMA-CD beads between samples. Distribution of

CD microparticles and surface roughness were likely some of the factors that contributed to this variability of biofilm formation across samples as roughness dictates the density of bacterial biomass that can adhere to the surface of the PMMA-CD170–172. The structure and density of the bacterial load in each biofilm can impact downstream properties of the

PMMA such as the refilling capacity and rate of drug release, thereby introducing variance into the duration of antimicrobial activity (i.e. Figure 5-11).

166

5.5. CONCLUSIONS

Conventional treatments for PJIs that have developed in the bone and soft tissue surrounding the prosthetic implant involve radical measures such as months of systemic antibiotics and removal and replacement of the prosthetic, rendering the patient immobile for an extended period of time. The presented work demonstrated the ability of a PMMA-

CD composite material to be refilled with antibiotics after being implanted in bone or muscle tissue. Refilled PMMA-CD had an extended duration of antimicrobial activity from 14-40+ days and PMMA-CD composites retained the ability to be refilled with antibiotics even in the presence of a biofilm and tissue infection. Of critical importance, this work demonstrated that PMMA-CD had the ability to provide sustained and on- demand antimicrobial therapy without the necessity to remove the implant if an infection develops. Furthermore, through this work, IO infusion appeared to be a viable clinically- used technique to enable refilling of PMMA-CD after implantation in bone, reporting for the first time the ability to refill PMMA in bone. In summary, PMMA-CD composite is a promising material that is versatile and can be applied to treat two aspects of PJIs – internal osteomyelitis and surrounding soft tissue infections that can be refilled with antibiotics in a patient-customizable manner regardless of if it is implanted in bone or soft tissue.

167

5.6. ACKNOWLEDGEMENTS

The authors acknowledge financial support through the National Science Foundation

(NSF) Graduate Research Fellowship Program Grant no. CON501692 (E.L.C.), NIH

NIAMS Ruth L. Kirschstein NRSA T32 AR007505 Training Program in Musculoskeletal

Research (G.D.L.), and NIH R01GM121477 (H.A.vR.). Ningjing (Nora) Zhang, Dylan

W. Marques, Fang Zhang for assistance with data collection. Greg D. Learn for experimental design of refilling models and dissection of rabbit femurs.

168

CHAPTER 6. PROCESSING TECHNIQUE OF PMMA BONE CEMENT

COMPOSITE DICTATES REFILILNG CAPACITY FOR TREATMENT OF

PERIPROSTHETIC JOINT INFECTIONS

*This chapter was reprinted (adapted) with permission from: Cyphert EL et al. ACS

Chemical Society.

Authors: Erika L. Cyphert, Greg D. Learn, Dylan W. Marques, Chao-yi Lu, Horst A. von Recum

(compression testing), slicing PMMA samples, writing manuscript (compression testing).

DWM – data analysis and collection. C-yL – data collection. HAvR – experimental design, data analysis, revision manuscript.

169

6.1. ABSTRACT

Antibiotic-laden PMMA bone cement is used in a variety of applications including load- bearing arthroplasties and non-load bearing orthopedic revision procedure temporary spacers and antibiotic beads to treat infections. Depending upon the surgical preparation technique, properties of PMMA can widely vary. Primary objective of this work was to perform an in-depth structure-function analysis regarding how processing of PMMA impacted material and structural properties (i.e. porosity) and downstream functional properties (i.e. drug refilling, strength). PMMA with CD microparticles was generated via hand- or vacuum-mixing and characterized for material and structural properties including porosity and internal morphology, and functional properties of drug refilling, compressive strength, and antimicrobial activity. CD microparticles were incorporated into PMMA to enable functional refilling property and to determine new information on drug distribution and distance or depth of PMMA in which refilled drug was able to penetrate. Vacuum-mixing of PMMA resulted in improved mechanical strength, and allowed for incorporation of greater amounts of CD microparticles, but less homogeneity relative to hand-mixing. Refilling studies indicated shallow penetration of drug into

PMMA samples without CD. However, PMMA with CD microparticles showed increased depth of drug penetration, indicating drug deeper within device could be delivered, resulting in more drug available for delivery and more opportunity for later antibiotic refilling on a patient-specific basis. Knowledge of structure-function relationships can assist and provide valuable information in design and optimization of 170

PMMA-CD for specific load-bearing or non-load-bearing applications.

6.2. INTRODUCTION

Arthroplasties are one of the most common orthopedic procedures performed annually in the U.S. with more than 700,000 knee and 330,000 hip replacements in 2010156. To ensure mechanical fixation of the metallic implant to the patient’s native bone, PMMA bone cement is frequently used and is prepared on-site at the time of the procedure. While arthroplasties are nearly 95% successful, the most common reasons for revision surgery include PJIs and mechanical failure (aseptic loosening and instability)173. Consequently, both the antimicrobial activity and mechanical strength of the PMMA play a key role in the success of the implant when the cement is used in load-bearing applications, such as arthroplasties. Beyond arthroplasty load-bearing applications, PMMA has often been used in non-load bearing applications such as strings of antibiotic-laden PMMA beads that are embedded in the soft tissue surrounding the PJI and as temporary spacers in PJI revision procedures26,115,149.

To prepare antibiotic-laden PMMA, antibiotics such as aminoglycosides (gentamicin and tobramycin) are frequently incorporated into PMMA during preparation21. There are several methods in which the PMMA can be prepared including: hand-mixing, vacuum- mixing, centrifugation174, and low frequency ultrasound156 which can result in a variety of porosities and antibiotic elution rates. To treat prolonged infections and prevent the development of drug-resistant bacteria, it is imperative that a therapeutically-relevant

171

amount of antibiotic be released from the PMMA at or above its MIC for an extended period of time156. However, many PMMA systems exhibit poor elution with an initial

“burst” release and nearly 85% of antibiotic remaining trapped after an extended duration of time156.

Previous studies have demonstrated the relationship between the porosity of PMMA and the resultant mechanical properties, where a large porosity yields a weak PMMA175. In an effort to improve the mechanical strength of the PMMA via a reduction in its porosity, while reducing surgeons’ exposure to irritating methyl methacrylate (MMA) vapor, vacuum-mixing has been regularly utilized as an alternative to open-air hand-mixing techniques176. Studies have demonstrated a greater than three-fold decrease in the porosity of the PMMA following vacuum-mixing relative to hand-mixing177.

Specifically, hand-mixed PMMA has been shown to contain a greater number of pores that can serve as points of stress concentration, ultimately weakening the PMMA177.

Nevertheless, while dramatic decreases in porosity have been observed with vacuum- mixing, this technique has been shown to result in a non-uniform distribution of microporosity where large pores are formed near the interface of the implant stem and the

PMMA175. A non-uniform distribution of pores can ultimately hinder drug distribution and bone in-growth in the PMMA and therefore the use of vacuum-mixing is somewhat controversial among surgeons for arthroplasty load-bearing applications175.

To improve upon the duration and consistency of antibiotic release from PMMA, we

172

have previously explored the incorporation of different amounts of insoluble polymerized

CD microparticles in the hand-mixed PMMA (see Chapters 2, 3, and 5)115,150. In an effort to improve upon the mechanical strength of CD microparticle laden hand-mixed PMMA while allowing for greater amounts of CD microparticles to be incorporated, we were interested in exploring the preparation technique of vacuum-mixing on the structural and functional properties of PMMA-CD composite. The goal of this present work was to provide an in-depth structure-function analysis of the material and structural properties of the PMMA-CD composite prepared via either hand- or vacuum-mixing and to examine how these parameters impacted the downstream functional properties of mechanical strength, antibiotic refilling, and antimicrobial activity of the PMMA-CD composite.

Furthermore, this analysis strived to investigate the depth of PMMA material that was actively involved in drug delivery. Specifically, the depth or distance/thickness in which the antibiotic was able to diffuse into the PMMA-CD material during refilling.

In this work, PMMA containing either 5, 10, or 15 wt% CD microparticles was synthesized via either hand- or vacuum-mixing and the distribution of the CD microparticles in the different PMMA-CD samples and the material and structural properties and morphology of PMMA-CD was subsequently analyzed via methylene blue staining, scanning electron microscopy (SEM), and micro-CT. To analyze how these material and structural properties impacted the functional properties of antibiotic refilling and antimicrobial activity, a series of antibiotic refilling studies into hand- and vacuum- mixed samples were then carried out using RMP over several durations of time and the 173

antimicrobial activity and depth that the drug was able to diffuse into each sample type over a given amount of time was evaluated. Average ultimate compressive strength of hand- and vacuum-mixed PMMA-CD composites was analyzed to elucidate how preparation technique and structure of PMMA-CD composite impacted its ability to be used in certain load-bearing applications. It was hypothesized that the use of vacuum- mixing would enhance the compressive strength of the CD microparticle-laden PMMA through a reduction in porosity, relative to hand-mixed PMMA containing CD.

6.3. MATERIALS AND METHODS

6.3.1. MATERIALS

See Chapter 2, section 2.3.1.

6.3.2. SYNTHESIS OF INSOLUBLE β-CD MICROPARTICLES

See Chapter 2, section 2.3.2.

6.3.3. RMP FILLING CD MICROPARTICLES

See Chapter 2, section 2.3.4.

6.3.4. SYNTHESIS OF HAND- AND VACUUM-MIXED PMMA-CD CYLINDERS

2 gram samples of Simplex® HV surgical grade bone cement powder were combined at

22oC with 1 mL MMA monomer for each batch of PMMA, in accordance with manufacturer instructions. Prior to adding the liquid monomer, empty or RMP-filled β-

174

CD microparticles (5, 10, or 15 wt%) were mixed into the powder under homogeneous.

For hand-mixed samples, the mixture of PMMA powder and CD microparticles was hand-mixed (open-air) with liquid monomer until a soft dough formed. For vacuum- mixed samples, the mixture of PMMA powder and CD microparticles was added into a

MixEvac® 3 Bone Cement Mixer with the liquid monomer. Vacuum pressure was turned on and PMMA was mixed for 30 seconds, the dough was scraped off of the sides, and the pressure was turned back on and mixed for an additional 45 seconds. For all samples once dough formed it was firmly finger-pressed into custom machined two-part PTFE or aluminum cylindrical molds (6 mm diameter, 12 mm height).

6.3.5. ANALYSIS OF DISTRIBUTION OF CD MICROPARTICLES IN HAND- AND

VACUUM-MIXED PMMA-CD COMPOSITE WITH STEREOMICROSCOPE

Methylene blue staining was performed in order to directly visualize the distribution of

CD microparticles in PMMA-CD composite as a vehicle to determine new information on potential drug distribution in PMMA-CD. PMMA-CD cylindrical samples containing either 5, 10, or 15 wt% empty CD microparticles (each condition in triplicate) were sliced in half along their vertical axis using a low-speed saw (Buehler Isomet Low Speed Saw with a diamond wafer blade). Samples were then submerged in and stained with methylene blue for 30 seconds (2.31 mg/mL dissolved in 23% ethanol) and rinsed with water. Methylene blue selectively stained CD microparticles while the surrounding

PMMA remained unstained178. Both the inside and outside of each sliced sample were

175

viewed and imaged under a stereomicroscope (AmScope Stereomicroscope, equipped with AmScope MU1000 camera). Images were processed in Matlab 2018b (Mathworks,

Natick, MA) via an in-house written program. Images were segmented into regions by referencing pixel values with that of a reference image of stained CD microparticles in

PMMA-CD. Pixels were classified using the “Nearest Neighbor rule,” which calculated the Euclidean distance between the pixel and each color and marked and classified the pixel179. Pixels that were classified as the methylene blue stained CD microparticles were segmented out and displayed independently of the other pixels.

6.3.6. SCANNING ELECTRON MICROSCOPY (SEM)

SEM was used to visualize the morphology and structure of the inside of the PMMA-CD cylinders with and without CD microparticles to better understand the distribution of CD microparticles, undissolved PMMA beads (contained in manufactured PMMA powder), and radiopacifier based upon preparation technique. Sliced PMMA-CD cylinders containing either 15 wt% CD microparticles or no microparticles (both hand- and vacuum-mixed) were placed on an SEM stub using carbon tape and the inside cut planes of samples were sputter-coated with palladium under vacuum (~25 nm). Images were collected using a JSM-6510 series JEOL scanning electron microscope at 23-370x magnification with an excitation voltage of 5 kV.

176

6.3.7. MICRO-CT SCANS AND POROSITY ANALYSIS PMMA-CD HAND- AND

VACCUM-MIXED CYLINDERS

See Chapter 3, section 3.3.7.

6.3.8. COMPRESSION TESTING PMMA-CD HAND- AND VACCUM-MIXED

CYLINDERS

See Chapter 2, section 2.3.8.

6.3.9. AGAROSE-BASED REFILLING STUDIES PMMA-CD HAND- AND VACCUM-

MIXED CYLINDERS

Tissue-mimicking agarose models were used to evaluate the functional property of antibiotic refilling of PMMA-CD cylinders with and without CD microparticles to probe depth of drug diffusion/penetration and to develop an understanding regarding the amount of refilled drug which would be realistically clinically available based upon preparation technique. Agarose models were created according to a previously established protocol (see Chapter 2, section 2.3.9)62,115,150. 100 μL of drug solution (6 mg/mL RMP dissolved in methanol) was injected into the pre-formed well in the agarose and plates were covered and agitated at 37oC for either 4, 18, 24, 30, 42, or 48 hours.

Drug filled PMMA-CD cylinder samples were removed from the agarose after their respective time points. Each condition was repeated in triplicate.

177

6.3.10. ANALYSIS OF DEPTH OF DIFFUSION AND RMP REFILLING INTO PMMA-

CD HAND-AND VACUUM-MIXED CYLINDERS

In an effort to better understand how much material actively participated and would theoretically be useable for drug refilling as well as understand the time-course of antibiotic refilling in PMMA-CD, a series of refilling studies were carried out and the

PMMA-CD was sliced in half and imaged with stereomicroscopy. Specifically, cylindrical PMMA-CD samples that had been refilled for either 4, 18, 24, 30, 42, or 48 hours were sliced in half using a low-speed saw (each in triplicate). Stereomicroscopy images of the inside of both halves of sliced samples were collected and analyzed using

ImageJ (2018). The depth or distance of diffusion of RMP into PMMA-CD cylinders was measured from the edge of the sample to the inside of the PMMA-CD cylinder where the refilling ended (as identified upon visual inspection, RMP had red/orange color – measured where “orange” drug outline ended). Two measurements were collected on each side of the sample and averaged along all four sides of the sample (8 measurements per sample, each condition in triplicate).

6.3.11. ANALYSIS OF ANTIMICROBIAL ACTIVITY OF RMP REFILLING HAND- AND

VACUUM-MIXED PMMA-CD COMPOSITE CYLINDERS

See Chapter 2, section 2.3.6.

178

6.3.12. STATISTICAL ANALYSIS

All calculations were reported as the average of a minimum of three samples with the error bars as the standard deviation. Microsoft Excel 2016 was used to complete all of the calculations. For analysis of methylene blue staining (n = 3), depth of antibiotic diffusion

(n = 3, 8 measurements each), micro-CT (n = 3), and compression testing (No CD hand- mixed: n = 8, 10 wt% empty CD hand-mixed: n = 10, 10 wt% RMP-filled CD hand- mixed: n = 4, No CD vacuum-mixed: n = 8, 10 wt% empty CD vacuum-mixed: n = 7, 10 wt% RMP-filled CD vacuum-mixed: n = 4, 15 wt% empty CD vacuum-mixed: n = 8,

15wt% RMP-filled CD vacuum-mixed: n = 10), one tailed Student’s T-tests with unequal variances were performed. A Spearman’s rho correlation was performed on depth or distance of antibiotic diffusion measurements to evaluate correlation between depth of diffusion and time.

6.4. RESULTS

6.4.1. VISUALIZATION/ANALYSIS OF DISTRIBUTION OF CD MICROPARTICLES

WITHIN PMMA (HAND- AND VACCUM-MIXED)

PMMA-CD cylinders were sliced in half, stained with methylene blue, and imaged with stereomicroscopy in an effort to allow for the direct macro-scale visualization of the distribution and structure of the CD microparticles within the PMMA-CD and to provide some insight regarding how the preparation technique impacted the homogeneity of additives (such as CD microparticles or drugs) in the PMMA. Figure 6-1 depicts 179

stereomicroscope images of PMMA-CD cylinders (a – inside cut plate, b – exterior of cylinder) containing either 5, 10, or 15 wt% CD microparticles prepared via hand- or vacuum-mixing stained with methylene blue. Presence of CD microparticles was indicated by the section of PMMA-CD that was stained blue. As the weight percentage of

CD microparticles in the PMMA-CD increased, there was a corresponding gradual increase in the amount of methylene blue staining upon visual inspection as expected. CD microparticles appeared to be more uniformly distributed in PMMA cylinders that were hand-mixed and CD microparticles demonstrated clumping in vacuum-mixed PMMA cylinders.

Upon stereomicroscopy imaging a Matlab program was used to segment out and quantify the relative percentage of methylene blue staining of the exterior/outside and inside cut plane of each sample. Figure 6-1 (c-d) provides a visual representation of how the Matlab program segmented out the blue pixels. Specifically, the original stereomicroscope image is shown (left) along with the segmented out blue pixels that the program detected

(middle) and the two images were then superimposed (right).

180

Figure 6-1: Stereomicroscope images of sliced PMMA-CD cylinders stained with methylene blue depicting both the outside/exterior (a) and inside cut plane (b) of both hand- and vacuum-mixed PMMA-

CD cylinders containing 5, 10, or 15 wt% CD microparticles. Methylene blue staining indicated the presence of CD microparticles. Images representative of n = 3 replicates per experimental group. Scale bars

= 2 mm. Schematic depicting detection of methylene blue stained pixels in stereomicroscope image using

Matlab for quantification of outside/exterior (c) and inside cut plane (d) of staining. The exterior image is of vacuum-mixed PMMA-CD with 15 wt% CD microparticles and the interior image is of hand-mixed

PMMA-CD with 15 wt% CD microparticles. These images are representative of n = 3 replicates per experimental groups. The original methylene blue stained image is shown (left) with the segmented out stained CD microparticle pixels (middle) and the superimposed image (right). Scale bars = 2 mm.

SEM images were collected to supplement methylene blue staining to visualize the internal morphology and structure of PMMA-CD and to provide more information on the homogeneity (or lack thereof) of CD microparticles, undissolved PMMA beads, and radiopacifier in PMMA based upon preparation technique. Figure 6-2 displays SEM

181

images of PMMA-CD cylinders containing either no or 15 wt% CD microparticles that were hand- or vacuum-mixed. Both macro- (left column; 23-50x) and micro-scale (right column; 140-370x) images were collected for each sample. Generally, for samples not containing CD microparticles the surface was more uniform and smooth, whereas upon the addition of CD microparticles, the surface contained micro-cracks and appeared

“crumbly.” Independent of the preparation technique and the presence of CD microparticles, all PMMA types contained substantial amounts of undissolved PMMA beads (indicated by the arrows) that typically had a diameter < 50 μm180. CD microparticles were distinguished from undissolved PMMA beads if their diameter was

≥ ~100 μm86.

182

Figure 6-2: SEM micrographs of inside cut plane of hand-mixed PMMA cylinders containing no CD microparticles (top row), hand-mixed containing 15 wt% CD microparticles (second row), vacuum-mixed containing no CD microparticles (third row), and vacuum-mixed containing 15 wt% CD microparticles

(fourth row). All images on left column provide a macro-perspective (23-50x magnification) and the micro-

183

scale perspective in the right column (140-370x magnification).

6.4.2. MICRO-CT SCANS AND POROSITY QUANTIFICATION PMMA-CD HAND-

AND VACUUM-MIXED CYLINDERS

Micro-CT scans were analyzed for PMMA-CD composites to determine and visualize the relative porosity, structure, and homogeneity of the distribution of pores in hand- versus vacuum-mixed PMMA-CD. Figure 6-3 depicts a 3-dimensional rendering of representative hand- and vacuum-mixed PMMA-CD cylinders with and without CD microparticles displaying both the solid and pore volume fraction of the PMMA-CD.

Pores tended to be concentrated towards the end of the cylinder that was at the base during polymerization as a result of trapped air and the slightly elevated temperature at the top of the mold (relative to chilled 4oC mold). Given that PMMA-CD tends to cure at the location of the highest temperature first and volumetrically shrinks during polymerization it was not possible to completely eliminate pores, regardless of preparation technique175.

184

Figure 6-3: Representative 3D solid (top) and pore fraction (bottom) renderings of hand- and vacuum- mixed PMMA-CD cylinders with and without empty CD microparticles.

Table 6-1 provides a quantification of the pore volume fraction, average pore volume, average number of discrete pores, solid fraction voxel mean, and solid fraction standard deviation. As shown in Figure 6-3, without the presence of CD microparticles, hand- mixed PMMA-CD cylinders generally contained large voids of trapped air in the center of the sample, whereas there were fewer pores in the vacuum-mixed samples and they were aggregated towards the periphery of the PMMA-CD. These findings were in agreement with previous reports in which voids were smaller in vacuum-mixed PMMA-

CD samples177 and pores tended to form near the PMMA-air interface due to polymerization shrinkage175. Interestingly, upon the addition of CD microparticles, the pores were more uniformly distributed throughout the hand-mixed sample. With the 185

vacuum-mixed samples, the pores were larger, more aggregated, and were less evenly distributed in the PMMA. Since CD microparticle were co-registered with air (pores) during segmentation (radiodensity pure CD polymer = -400 HU, upper pore segmentation threshold = -200 HU; data not shown), this data suggested that the microparticles were clumping under the force of the vacuum.

Table 6-1: Micro-CT quantification of porosity for hand- and vacuum-mixed PMMA-CD cylinders. †

Indicates value statistically significant relative to no CD hand-mix control and ‡ indicates value statistically significant relative to no CD vacuum-mix control.

Micro-CT Average Pore Average Pore Average # Solid Fraction Solid Fraction Group (n = 3) Volume Volume (nL) Discrete Voxel Mean Standard Fraction (%) Pores (HU) Deviation (HU) Hand-mixed No β-CD 0.467 ± 0.103 0.485 ± 0.154 3001 ± 658 1511.86 ± 501.32 ± 9.99 22.43 10wt% empty 1.645 ± 0.233† 0.714 ± 0.196† 6657 ± 949† 1351.49 ± 550.61 ± 9.30† β-CD 23.51† 15wt% empty 1.906 ± 0.440† 0.770 ± 0.177 7101 ± 362† 1270.60 ± 504.54 ± 14.76 β-CD 27.10† Vacuum-mixed No β-CD 1.115 ± 0.269 1.359 ± 0.272 2413 ± 140 1553.10 ± 506.45 ± 10.58 29.96 10wt% empty 0.736 ± 0.097 2.087 ± 0.102‡ 1008 ± 176‡ 1481.92 ± 467.57 ± β-CD 14.66‡ 10.42‡ 15wt% empty 1.485 ± 0.737 2.913 ± 2.356 1714 ± 550 1422.17 ± 486.67 ± 5.97‡ β-CD 21.62‡

The 3D models reflected the porosity quantification provided in Table 6-1. Specifically, with hand- and vacuum-mixed PMMA-CD cylinders containing no CD microparticles, hand-mixed samples had a significantly smaller pore volume fraction (p = 0.02) and average pore volume (p = 0.007) than vacuum-mixed samples but contained a

186

comparable number of pores (p = 0.13) and comparable solid fraction voxel mean (p =

0.29) and standard deviation (p = 0.07). Generally, as CD microparticles were added to the PMMA cylinders, there was a dramatic increase in the pore volume fraction (1.3- to

4-fold increase) and the average pore volume of the samples (1.6- to 2-fold increase) relative to samples prepared in the same manner without CD microparticles (except for

1.5-fold decrease in vacuum-mixed samples containing 10 wt% CD microparticles).

Furthermore, there was a statistically significant increase in the number of pores in hand- mixed samples containing CD microparticles compared to those without (p < 0.05). There were no dramatic changes observed in either solid fraction voxel mean or standard deviation in samples containing CD microparticles.

Unlike PMMA samples without CD microparticles, vacuum-mixed samples containing

CD microparticles had lower pore volume fractions than their hand-mixed counterparts

(hand-mixed: 10 wt% CD = 1.645%, 15 wt% CD = 1.906%; vacuum-mixed: 10 wt% CD

= 0.736%, 15 wt% CD = 1.485%). Interestingly, while vacuum-mixed cylinders containing CD microparticles had lower pore volume fractions, they had significantly larger pore volumes than their hand-mixed counterparts (p < 0.05) (hand-mixed: 10 wt%

CD = 0.714 nL, 15 wt% CD = 0.770 nL; vacuum-mixed: 10 wt% CD = 2.087 nL, 15 wt%

CD = 2.913 nL) and significantly less pores (p < 0.05). Specifically, there was a 4- to 6- fold decrease in the number of pores in vacuum-mixed PMMA-CD relative to hand- mixed samples (hand-mixed: 10 wt% CD = 6657 pores, 15 wt% CD = 7101 pores; vacuum-mixed: 10 wt% CD = 1008 pores, 15wt% CD = 1714 pores). These results 187

suggested that vacuum-mixed samples with CD microparticles contained fewer but larger pores than their hand-mixed counterparts.

6.4.3. COMPRESSION TESTING PMMA-CD HAND- AND VACUUM-MIXED

CYLINDERS

The average compressive strength of PMMA-CD samples prepared by either hand- or vacuum-mixing containing either 10 or 15 wt% empty or RMP-filled CD microparticles was evaluated and representative stress-strain curves (representative based on average ultimate compressive strength) from each condition are displayed in Figure 6-4. Table 6-

2 provides a quantification of various mechanical properties of the PMMA-CD including ultimate compressive strength, modulus, and work to peak load.

188

Figure 6-4: Representative stress versus strain curves of hand- and vacuum-mixed PMMA-CD containing either no CD microparticles or 10 or 15 wt% empty (non-drug-filled) or RMP-filled CD microparticles under compression. A minimum of 4 samples per condition were evaluated.

189

Table 6-2: Quantification of the mechanical properties for hand- and vacuum-mixed PMMA-CD. †

Indicates value statistically significant relative to no CD hand-mix control, ‡ indicates value statistically significant relative to no CD vacuum-mix control, and ⸙ indicates value statistically significant relative to corresponding 10 or 15 wt% empty (non-drug filled) CD hand- or vacuum-mixed samples.

Compression Test Ultimate Compressive Modulus (MPa) Normalized Work to Specimen Strength (MPa) Peak Load (J/cm3) Hand-mixed No β-CD 70.61 ± 1.52 2003.28 ± 97.35 2.494 ± 0.297 10wt% empty β-CD 59.33 ± 2.40† 1824.63 ± 47.62† 2.070 ± 0.131† 10wt% RMP β-CD 64.36 ± 1.07†⸙ 2018.19 ± 44.34⸙ 2.234 ± 0.026†⸙ Vacuum-mixed No β-CD 80.80 ± 3.39 2187.21 ± 216.29 3.361 ± 0.489 10wt% empty β-CD 61.61 ± 1.98‡ 1853.59 ± 47.41‡ 2.177 ± 0.102‡ 10wt% RMP β-CD 65.52 ± 2.04‡⸙ 2057.28 ± 10.31⸙ 2.288 ± 0.103‡ 15wt% empty β-CD 55.47 ± 1.90‡ 1687.56 ± 70.18‡ 1.959 ± 0.080‡ 15wt% RMP β-CD 65.46 ± 2.87‡⸙ 2040.72 ± 52.05‡⸙ 2.254 ± 0.130‡⸙

Without the presence of CD microparticles, vacuum-mixed PMMA cylinders had a significantly greater average compressive strength than that of hand-mixed (vacuum- mixed = 80.80 MPa, hand-mixed = 70.61 MPa, p < 0.001). As expected, upon introduction of empty CD microparticles the average compressive strength in both hand- and vacuum-mixed samples significantly decreased (10 wt% = ~1.2-1.3-fold decrease, 15 wt% = ~1.5-fold decrease). In agreement with results from previous studies115,150,

PMMA-CD containing RMP-filled (drug-filled) CD microparticles was significantly stronger in all cases than PMMA-CD containing the same amount of empty (non-drug filled) CD microparticles (p < 0.05) (10 wt% RMP-filled CD hand-mixed = +8.48%, 10 wt% RMP-filled CD vacuum-mixed = +6.35%, 15 wt% RMP-filled CD vacuum-mixed =

+18.01%). Interestingly, the average compressive strength of PMMA-CD cylinders with

190

10 wt% RMP-filled CD microparticles hand-mixed and 15 wt% RMP-filled CD microparticles vacuum-mixed were equivalent (p > 0.05).

In terms of the modulus, upon addition of empty CD microparticles there was a significant decrease in the modulus for both hand- and vacuum-mixed PMMA-CD cylinders (p < 0.05) (10 wt% empty CD hand-mixed = -8.92%, 10 wt% empty CD vacuum-mixed = -15.25%, 15 wt% empty CD vacuum-mixed = -22.84%), but not as dramatic of a change when the CD microparticles were first filled with RMP (10 wt%

RMP-filled CD hand-mixed = +0.74%, 10 wt% RMP-filled CD vacuum-mixed = -5.94%,

15 wt% RMP-filled CD vacuum-mixed = -6.70%). Similar trends were observed in the work to peak load in that the addition of empty CD microparticles significantly decreased the work relative to PMMA-CD cylinders without CD microparticles (p < 0.05) (10 wt % empty CD hand-mixed = -17.00%, 10 wt% empty CD vacuum-mixed = -35.23%, 15 wt% empty CD vacuum-mixed = -41.71%), whereas RMP-filled CD microparticles did not impact the work to peak load as substantially (10 wt% RMP-filled CD hand-mixed = -

10.42%, 10 wt% RMP-filled CD vacuum-mixed = -31.92%, 15 wt% RMP-filled CD vacuum-mixed = -32.94%).

6.4.4. ANALYSIS OF THE DEPTH OF DIFFUSION OF RMP DURING REFILLING

HAND- AND VACUUM-MIXED PMMA-CD COMPOSITES

In an effort to better understand how the time-course of antibiotic refilling (exposure time to antibiotic) impacted the overall depth or distance of diffusion of antibiotic refilling into

191

PMMA cylinders containing CD microparticles and to determine the amount of PMMA-

CD material realistically useable for drug delivery, a series of antibiotic refilling studies were carried out in agarose and stereomicroscopy images were collected of the inside of the refilled PMMA cylinders. Stereomicroscope images of PMMA-CD cylinders refilled with RMP in the agarose model after select time points are shown in Figure 6-5. Arrows indicate examples of location where measurements for the depth of RMP diffusion were collected. Slicing of cylindrical samples allowed for direct visualization of relative depth or thickness of PMMA material that RMP was able to be “refilled” and penetrate into the

PMMA-CD composite cylinders after curing.

192

Figure 6-5: Stereomicroscope images depicting the depth of diffusion of RMP into hand- and vacuum- mixed PMMA cylinders containing 15 wt% CD microparticles. Representative refilling images

(representative of n = 3 per experimental group) are shown after either 24 or 48 hours. Arrows indicate the relative depth of RMP refilling upon visual inspection of “orange” periphery of RMP. Scale bars = 2 mm.

Figure 6-6 displays the quantification of the depth or distance of RMP diffused into

PMMA-CD during refilling for each condition after different durations of exposure time to RMP. As expected, RMP was only refilled on the periphery of the PMMA-CD cylinders and did not penetrate very deep into the PMMA-CD, even in samples containing 15 wt% CD microparticles that had an increased porosity. The maximum depth or distance of diffusion was observed after 48 hours of refilling with hand-mixed samples containing 15 wt% CD microparticles (184.27 μm). Generally, as the duration of 193

refilling time increased, the depth of diffusion gradually increased (Spearman’s rho correlation: p = 0.65-1.0). Similarly, as the amount of CD microparticles in the PMMA-

CD increased, the depth of diffusion increased. Surprisingly, even after only 4 hours of exposure to RMP, refilling was visible in PMMA-CD cylinders containing 10 wt%

(hand-mixed: 51.85 μm) or 15 wt% CD microparticles (hand-mixed: 62.98 μm; vacuum- mixed: 24.19 μm). On the other hand, refilling was not visible in either hand- or vacuum- mixed samples containing 5 wt% CD microparticles until 48 hours of RMP exposure

(hand-mixed: 24.17 μm; vacuum-mixed: 44.98 μm). Depth of RMP refilling was generally significantly greater in hand-mixed PMMA-CD samples relative to vacuum- mixed samples containing the same amount of CD microparticles (p < 0.05), which was attributed to differences in porosity between hand- and vacuum-mixed cylinders.

Specifically, after 48 hours of refilling, the depth of RMP diffusion was nearly 1.6x greater in PMMA-CD hand-mixed cylinders than in vacuum-mixed samples (hand- mixed: 10 wt% CD = 137.29 μm, 15 wt% CD = 184.27 μm; vacuum-mixed: 10 wt% CD

= 82.54 μm, 15 wt% CD = 115.15 μm).

194

Figure 6-6: Quantification of the depth or distance of diffusion of RMP refilled into hand- (left) and vacuum-mixed (right) PMMA-CD cylinders containing either 5, 10, or 15 wt% CD microparticles after 4,

18, 24, 30, 42, or 48 hours of exposure to RMP during refilling.

6.4.5. ANALYSIS OF ANTIMICROBIAL ACTIVITY OF RMP REFILLED HAND- AND

VACUUM-MIXED PMMA-CD COMPOSITE CYLINDERS

Once the depth of RMP diffusion into refilled PMMA-CD cylinders was determined, we next examined the resultant antimicrobial activity of both the inside cut face (Figure 6-

7a) as well as the exterior face (Figure 6-7b) of the PMMA-CD (containing either 5 or 15 wt% CD) that was refilled with RMP over either 4, 18, 30, or 48 hours against S. aureus.

Generally, results from Figure 6-7 demonstrated that the duration of RMP refilling time and content/amount of CD microparticles directly impacted the antimicrobial activity of refilled PMMA-CD against S. aureus. Specifically, as the duration of RMP refilling time 195

and content of CD microparticles increased, the size of the zone of inhibition and the duration of activity against S. aureus tended to increase. PMMA-CD cylinders refilled with RMP for 48 hours containing 15 wt% CD microparticles demonstrated the greatest and longest-term antimicrobial effect. Hand-mixed PMMA-CD tended to have a greater antimicrobial effect than vacuum-mixed PMMA-CD of the same CD composition.

Antimicrobial activity from the inside cut face (Figure 6-7a) was measured after 1 day of exposure to S. aureus and activity from the exterior face (Figure 6-7b-c) was measured over 1-5 days of exposure to S. aureus. Results from Figure 6-7a indicated that the interior of hand- and vacuum-mixed PMMA-CD containing 5 wt% CD microparticles was able to clear bacteria if it was refilled with RMP for 48 hours. Additionally, the interior of vacuum-mixed PMMA-CD containing 15 wt% CD microparticles was able to demonstrate antimicrobial activity after it had been refilled with RMP for only 4 hours.

The interior of hand- and vacuum-mixed PMMA-CD containing 15 wt% CD microparticles generally demonstrated larger sized zones of inhibition against S. aureus as the duration of refilling time with RMP increased.

Figure 6-7b-c indicated that the exterior face of hand- and vacuum-mixed PMMA-CD containing 15 wt% CD microparticles was able to consistently clear S. aureus over 5 days when it was refilled with RMP for 48 hours. For shorter durations of refilling time (i.e. 4,

18, 30 hours), the duration of activity of the exterior face of PMMA-CD against S. aureus was generally less than what was observed when the same PMMA-CD samples were

196

refilled with RMP for 48 hours (< 5 days). Generally, hand-mixed PMMA-CD cylinders demonstrated a longer-term and stronger antimicrobial effect (i.e. larger sized zone of inhibition) than vacuum-mixed cylinders of the same composition, refilled under the same conditions. Finally, the exterior face of PMMA-CD cylinders (5 wt% CD, refilled with RMP for 48 hours) was able to clear S. aureus over 2-3 days. Release of refilled

RMP from PMMA-CD was not evaluated due to RMP’s strong hydrophobicity and affinity for CD.

197

Figure 6-7: Antimicrobial activity of inside cut face (a) of RMP refilled PMMA-CD cylinders against S. aureus. PMMA-CD cut cylinders were refilled for either 4, 18, 30, or 48 hours and contained either 5 or 15 wt% CD microparticles. Antimicrobial activity was reflected in the size of the zone of inhibition after 24 hours of exposure to S. aureus. Antimicrobial activity of the exterior face of vacuum-mixed (b) and hand- mixed (c) RMP refilled PMMA-CD cylinders against S. aureus. Antimicrobial activity was reflected in the size of the zone of inhibition after exposure to S. aureus over a period of 5 days.

198

6.5. DISCUSSION

Traditionally, antibiotic-laden PMMA has been used in several load-bearing

(arthroplasties) and non-load-bearing applications (beads for soft tissue infections and temporary arthroplasty revision spacers)26,63,115,149. This study has provided an in-depth analysis regarding a variety of properties of antibiotic-laden PMMA containing an additive (i.e. CD microparticles) and has related how surgical processing of the PMMA directly impacted its intended end-use applications. Specifically, the innovation of this study lies in the structure-function analysis regarding how surgical processing altered the material and structural properties of PMMA-CD composite and how these properties impacted the functional ability of PMMA-CD to be refilled with drug, clear the infection, and be used in a certain application based upon its strength. This is the first study, to our understanding, regarding how the preparation and structure of PMMA with an additive dictated the ability of the material to be refilled with drug and its mechanical strength.

SEM studies indicated heterogeneity (presence of large clusters) of undissolved PMMA powder beads near the surface of the PMMA-CD regardless of the preparation technique

(both hand- and vacuum-mixed PMMA-CD) and incorporation of CD microparticles.

Undissolved PMMA beads have been shown to arise from the process of air-drying on the surface of the PMMA177. Ultimately, the presence of the undissolved PMMA beads coupled with the micro-cracks most likely contributed to the significant decrease in compressive strength in PMMA-CD. Micro-cracks generally result from points of stress

199

concentration and can propagate to form a macro fissure and complete fracture of

PMMA181. It is likely that micro-cracks formed in PMMA-CD composites due to the increased amount of dry component (i.e. CD microparticles) to the dry PMMA powder upon polymerization while using the same volume of liquid monomer that is used with dry powder without CD microparticles. If the ratio of the liquid monomer was not enough to effectively wet or coat the dry components (i.e. PMMA powder and CD microparticles), micro-cracks could result125.

The results of the porosity studies helped to quantitatively confirm what was observed in

Figure 6-1 with methylene blue staining, that vacuum-mixing tended to result in clumping of CD microparticles in PMMA-CD. Since the pores in micro-CT scans were co-registered with CD microparticles, the increase in both the average pore volume and number of discrete pores in vacuum-mixed samples further substantiated the clumping of the CD microparticles in PMMA-CD. This finding directly corresponded with the work of Hoey et al., which demonstrated that vacuum-mixing altered the distribution of the pores in PMMA, such that pores were evenly distributed in hand-mixed samples, but upon vacuum-mixing, large pores were preferentially distributed near the surface of the

PMMA175. While the results of the micro-CT study did not provide an exact measure of the porosity of samples containing CD microparticles (due to the co-registration), they did provide some insight regarding how the CD microparticles altered the distribution of the pores in PMMA-CD based upon the preparation technique. We believe that the interaction of the additive (CD microparticles) in PMMA-CD with the vacuum force 200

during preparation that resulted in clumping of the CD microparticles could be used as a model or vehicle to further probe into the distribution/homogeneity of how other additives to PMMA (i.e. drugs) may behave in PMMA when vacuum-mixed.

Unfortunately, the CD microparticles could not be isolated from the pores in PMMA using micro-CT in this study because it was impossible to predict how the radiodensity of the CD microparticles may be altered upon implantation in the PMMA.

We analyzed how the functional property of the PMMA-CD composites to be refilled with antibiotic after implantation depended upon the duration of time that the PMMA-CD composite was exposed to the drug, preparation technique, content of CD microparticles, and the resultant structural property (i.e. porosity). Specifically, we evaluated how the depth of diffusion of antibiotic (RMP as a model drug) into the PMMA-CD cylinders changed based upon these factors to elucidate how much (i.e. thickness) of PMMA-CD material was usable for drug delivery functional purposes. Uniform refilling of RMP was not observed in the PMMA-CD, but rather RMP only diffused into the periphery of the cylinders. While it may seem counterintuitive, it is ideal for the majority of the antibiotic to be located near the surface of the PMMA-CD rather than homogeneously distributed throughout. Due to the relatively low porosity of the PMMA (pores only ~0.5% volume of PMMA), it is generally difficult for drug to diffuse back out particularly if it is in the center of the implant. Specifically, previous studies have shown that even after prolonged periods of time, upwards of 85% of antibiotic can remain permanently trapped within the

PMMA matrix156,182,183. The closer the drug is in proximity to the surface of the PMMA, 201

the more readily it can diffuse back out, and the more available it will be for opposing

PJIs in the long term. Thus demonstrating that only the first ~100-200 μm of the PMMA-

CD was viable for drug delivery purposes. While longer antibiotic exposure durations

(i.e. 24-48 hours) resulted in an increased depth of RMP diffusion, some diffusion was apparent in samples after only 4 hours of exposure, suggesting that extraordinarily long antibiotic exposure time may not be necessary to obtain effective RMP refilling as these samples demonstrated antimicrobial activity (see Figure 6-7). A relatively low concentration of RMP was used during the refilling study in this work (i.e. 6 mg/mL), a more concentrated RMP refilling solution (i.e. 12 mg/mL) would likely result in an increased depth of diffusion into PMMA-CD and subsequent increased duration of antimicrobial activity. This study served as a proof-of-concept to analyze the depth of diffusion and refilling at the lower end of drug concentration that could be injected and demonstrated that refilling at a low drug concentration did result in some antimicrobial activity.

Nevertheless, through this structure-function analysis we identified the necessity of considering the compromise between the functional properties of the amount of refilling

(depth) possible and the resultant mechanical properties of the PMMA-CD composites as the mechanics dictate the end-use application of the material. To elaborate, for load- bearing applications, such as in arthroplasties, the mechanical strength is much more critical than in the use of PMMA-CD beads for the treatment of soft tissue infections. We identified an inverse relationship between the depth of antibiotic refilling possible over 202

time and the strength of the PMMA-CD composites. Specifically, vacuum-mixed (less porous) PMMA-CD composites tended to be stronger than hand-mixed, however, the depth of RMP refilling was greater in hand-mixed (more porous) PMMA-CD cylinders.

As the results of Figure 6-7 indicated, the antimicrobial activity of the refilled PMMA-

CD samples was related to the depth of RMP diffusion during refilling. Therefore, for an increased duration of antimicrobial activity, an increased depth of RMP diffusion during refilling would be necessary. Differences in strength and depth of RMP diffusion between PMMA-CD hand- and vacuum-mixed cylinders containing the same content of

CD microparticles were primarily attributed to differences in porosity that resulted from the two mixing techniques. Thus, highlighting that, in the preparation of PMMA-CD composite, it is critical for the surgeon to weigh the priority of the mechanical strength of the PMMA-CD composite to its potential refilling depth and long-term therapeutic efficacy for its end-use application.

Ultimately, the results from this study suggested that the use of vacuum-mixing enabled a greater amount of CD microparticles to be incorporated into PMMA-CD while retaining the mechanical strength that was possible with hand-mixing. However, in order to best retain the mechanical properties of the PMMA-CD, it was important that the microparticles be initially filled with drug (i.e. RMP). Our previous studies were in agreement with this finding that RMP-filled CD microparticles in PMMA were stronger than those that were empty in PMMA115,150. Quantitative structure-activity relationship

(QSAR) modeling was carried out in which the relative binding affinity of benzoyl 203

peroxide (used as an initiator in PMMA) and methyl methacrylate into CD was determined184. Both molecules demonstrated high binding affinity energies with CD

(benzoyl peroxide = -16.68 kcal/mol, methyl methacrylate = -8.70 kcal/mol), which suggested that perhaps when empty CD microparticles were added to PMMA during polymerization, both the benzoyl peroxide initiator and methyl methacrylate could have selectively formed inclusion complexes with CD pockets, thus limiting the efficiency of the polymerization reaction. As a result, the average ultimate compressive strength of the

PMMA-CD was weakened. However, when the CD microparticles were first filled with

RMP before they were added to the PMMA, the presence of the RMP effectively blocked the ability of either benzoyl peroxide or methyl methacrylate to readily complex with the

CD “pockets,” thereby allowing the reaction to proceed as normal, attaining a higher average ultimate compressive strength. Given that hand-mixed PMMA-CD containing 10 wt% RMP-filled CD microparticles had a comparable strength to vacuum-mixed PMMA-

CD containing 15 wt% RMP-filled CD microparticles, vacuum-mixing allowed for an increased amount of CD microparticles to be incorporated into the PMMA without compromising the mechanical properties. The ability to incorporate greater quantities of

CD microparticles into PMMA would theoretically enable a greater amount of drug to be filled/refilled and allow for a deeper penetration of refilling into the PMMA-CD, thereby enabling the treatment of more prolonged PJIs.

204

6.6. CONCLUSIONS

Through this work we have completed an in-depth structure-function analysis regarding the impact of processing technique and composition of PMMA-CD composite on the structural properties of porosity and functional properties of mechanical strength, refilling depth of RMP, and antimicrobial activity. These functional properties ultimately play a key role dictating the ideal end-use application of the PMMA-CD composite material.

Information from these structure-function relationships elucidated from this analysis can serve as a toolbox to help other researchers design and optimize the composition of

PMMA-CD composite to be utilized in a variety of load-bearing and non-load-bearing applications that meets the desired mechanical, refilling, and antimicrobial activity properties. Vacuum-mixing of PMMA-CD composite resulted in an increased average ultimate compressive strength relative to hand-mixing, enabling a greater amount of CD microparticles to be incorporated. However, vacuum-mixing resulted in a less homogeneous porosity, drug distribution, and depth of RMP refilling. Nevertheless, the heterogeneities in the structural properties of CD microparticle distribution and porosity were inconsequential relative to the ability to incorporate greater amounts of CD microparticles into PMMA-CD without impacting the functional property of mechanical strength. Specifically, a greater amount of CD microparticles in PMMA-CD could enhance the antibiotic refilling and delivery capacity and enable treatment of prolonged

PJIs. Thus, vacuum-mixing appeared to be a viable technique to enhance the mechanical strength of PMMA-CD composite such that it could more effectively be utilized in PJI 205

treatment in load-bearing applications, such as arthroplasties and hand-mixing appeared to be a more amenable technique to prepare PMMA-CD for non-load-bearing applications such as antibiotic-filled beads for treatment of soft tissue infections.

6.7. ACKNOWLEDGEMENTS

The authors would like to thank Susan Kozawa (CWRU) for assistance and use of the stereomicroscope and Kathleen Young (CWRU) for assistance with collection of SEM images. The authors gratefully acknowledge financial support through National Science

Foundation (NSF) Graduate Research Fellowship Program Grant No. CON501692

(E.L.C.), NIH NIAMS Ruth L. Kirschstein NRSA T32 AR007505 Training Program in

Musculoskeletal Research (G.D.L.), Support of Undergraduate Research & Creative

Endeavors (SOURCE) (D.W.M.), Center for Stem Cell and Regenerative Medicine

Undergraduate Student Summer Program (ENGAGE) (C-y.L.), and NIH

R01GM1214377 (H.A.vR.).

206

CHAPTER 7. CONCLUSIONS AND FUTURE DIRECTIONS

7.1. CONCLUSIONS

This work has explored and demonstrated the capability of PMMA-CD composite to allow for antibiotic refilling to occur, ultimately broadening the range of antibiotics that are compatible with PMMA bone cement. This PMMA-CD delivery platform has the potential to more effectively treat chronic PJIs without the need to remove the implant or unnecessarily expose the patient to systemic antibiotics. PMMA-CD composite has been shown to be very versatile and amenable for a variety of applications and retains its functionality even in the presence of bacterial biofilm. Key findings are expanded upon below:

Versatility of PMMA-CD composite delivery platform

One of the primary advantages of the PMMA-CD composite delivery platform developed throughout Chapters 2-6 is that it is a very versatile system that can be tailored for a variety of therapeutic applications. The PMMA-CD composite has demonstrated the ability to be refilled with antibiotics in a variety of tissue types (i.e. both soft muscle tissue and hard bone tissue), thus enabling its use in the treatment of infections inside and out of the bone. Figure 7-1 outlines the primary parameters in which the PMMA-CD delivery platform can be customized and the subsequent downstream effect of these customizable parameters.

207

Figure 7-1: Schematic outlining the relationship between customizable parameters of PMMA-CD composite drug delivery platform and the downstream effects of these parameters on the functionality and properties of the composite.

Figure 7-1 highlights how surgeons have the ability to adjust the frequency of the refilling dose, type and amount of CD and type of drug incorporated to target a specific pathogen or cell type, as well as the porosity of the PMMA-CD composite. All of these parameters have a direct impact on the amount of drug that can be refilled into the composite as an increased frequency of refilling, greater amount of CD, more porous, and strong affinity of the drug for CD will enable a greater amount of drug to be refilled than would otherwise be possible. When a greater amount of drug is able to be refilled into the composite system, it subsequently impacts and increases the duration of time in which the

208

drug is released and the duration of therapeutic activity that is possible from the released drug. Beyond the amount of drug refilled, the amount of CD used and the porosity have a direct impact on the resultant mechanical strength of the composite. Greater amounts of

CD microparticles and pores tended to decrease the strength of the composite (see

Chapters 2-3). Nevertheless, techniques such as vacuum-mixing can be utilized to improve the mechanical strength and reduce the porosity of the PMMA-CD composite

(see Chapter 6).

Knowledge of the relationship between the amount of drug injected into the tissue (or agarose-based tissue mimic) to the amount of drug that can be refilled into the PMMA-

CD composite, can help surgeons to optimize the frequency and concentration of refilling dose in order to obtain the desired duration of therapeutic activity. Table 7-1 summarizes the relationship of mass of drug injected into the refilling model (agarose-based) to the mass of drug refilled into the composite and the subsequent duration of antimicrobial activity of PMMA-CD.

209

Table 7-1: Summary of relationship between amount of drug injected into agarose refilling model and amount filled into PMMA-CD composite to the duration of antimicrobial activity possible with refilled composite. Parameters include efficiency of drug refilling into composite (amount filled/amount injected into model), normalized mass of drug filled into composite (mg drug/mg polymer), and duration of antimicrobial activity against S. aureus obtained from persistence zone of inhibition studies (see Chapter

2).

Composition of Efficiency of drug Normalized mass drug Duration of activity PMMA-CD composite refilling filled against S. aureus (%: amount drug (mg drug/mg (days) filled/amount injected) polymer) β-CD microparticles 5 wt% CD 0.26 0.0019 ~14 10 wt% CD 0.35 0.0027 ~30 γ-CD microparticles 5 wt% CD 0.28 0.0017 ~5 10 wt% CD 0.46 0.0033 ~40

Additionally, PMMA-CD composites have demonstrated the capability of accommodating multiple drugs simultaneously while retaining their refilling capacity

(see Chapter 3). This property makes the PMMA-CD composite amenable for treatment of broad-spectrum infections where multiple drugs may be required to eradicate both gram-positive and gram-negative pathogens and to reduce the likelihood of developing drug-resistance in the case of infection and cancer treatment185–188.

Finally, PMMA-CD composite has been shown to be refillable across a variety of types of PMMA. Specifically, Chapters 2-6 demonstrated the ability of Stryker Simplex HV bone cement to be refilled with antibiotics when combined with CD microparticles. As the type of cement utilized in the procedure is heavily dependent on the surgeon’s 210

discretion and a wide variety of bone cement formulations are clinically used (i.e. different viscosities, different compositions – amount of initiator), the refillability of

PMMA-CD composites comprised of three additional PMMA types was explored180,189–

191. Additional compositions of PMMA bone cement evaluated included: Stryker Simplex

P SpeedSet®, DePuy Synthes SmartSet MV (medium viscosity), and DePuy Synthes

SmartSet HV (high viscosity). Small beads (6 mm diameter) containing 10 wt% empty

CD microparticles were prepared for each type of bone cement listed above, refilled in an agarose-based model with RMP over 48 hours, and evaluated for their duration of antimicrobial activity against S. aureus in a persistence zone of inhibition study. Figure

7-2 displays the results of the persistence zone of inhibition study of RMP refilled

PMMA-CD composites with and without 10 wt% CD microparticles.

211

Figure 7-2: Persistence zone of inhibition study of RMP refilled PMMA-CD composite of different compositions with and without 10 wt% CD microparticles against S. aureus.

Results from Figure 7-2 indicated that regardless of the type/composition of clinical

PMMA bone cement used in the PMMA-CD composites, all types tested were able to demonstrate RMP refilling to obtain 20-40 days of antimicrobial activity, relative to the

~5 days of activity obtained with plain PMMA (no CD microparticles). Thereby, demonstrating the versatility of the PMMA-CD composite platform to conform to a range of clinically-used PMMA standards.

212

Importance of balancing mechanical properties and refilling and delivery properties of PMMA-CD composite

A key takeaway from this work was the inherent fine line that must be balanced when considering the capacity of the PMMA-CD composite for refilling/delivery and its mechanical properties. This consideration was heavily dependent upon the intended end- use or application for the PMMA-CD composite. To elaborate, for load-bearing applications, such as in arthroplasties, the mechanical strength was much more critical than in the use of PMMA beads for the treatment of soft tissue PJIs. In order to obtain a stronger PMMA-CD composite, techniques such as vacuum-mixing can be used to reduce the porosity. Additionally, fewer CD microparticles can be incorporated into the

PMMA as they act as stress risers192. Nevertheless, there was a clear tradeoff reducing the porosity and amount of CD microparticles. Specifically, by reducing the porosity, it was more difficult for drug to diffuse in to enable refilling and with fewer CD microparticles, the total amount that can be refilled into the composite was limited. A summary of these findings are outlined in Table 7-2.

213

Table 7-2: Analysis of impact of parameters controlling mechanical strength and refilling capacity of

PMMA-CD composite. Parameters analyzed included content of CD microparticles, preparation technique, pore volume fraction, ultimate compressive strength, and depth of RMP diffusion (see Chapter 6).

Composition/Preparation Pore volume fraction Ultimate compressive Depth of RMP (%) strength (MPa) diffusion (μm) Hand-mixed No CD (plain) 0.47 70.6 0 10 wt% CD 1.64 59.3 137 15 wt% CD 1.91 -- 184 Vacuum-mixed No CD (plain) 1.11 80.8 0 10 wt% CD 0.74 61.6 82 15 wt% CD 1.48 55.5 115

From the results of Table 7-2, it was evident that there was a strong inverse correlation between the ultimate compressive strength of the PMMA-CD composite and the depth of

RMP diffusion (correlation coefficient = -0.90). Therefore, highlighting that, in the preparation of PMMA-CD composite, it is critical for the surgeon to weigh the priority of the mechanical strength of the PMMA to its potential refilling capacity and long-term therapeutic efficacy on a case-by-case basis.

Implications for device rescue, rather than removal using PMMA-CD composite

Evidence from Chapters 4 and 6 regarding refilling of CD and PMMA-CD composite in the presence of a bacterial biofilm demonstrated that PMMA-CD composites were able to maintain their functionality of refilling and subsequently releasing drug despite the formation of a biofilm or infection. These results suggested that the PMMA-CD composites have the potential to provide a novel alternative strategy towards the

214

treatment of PJIs and general implant infections193,194. Specifically, the current standard of treatment of advanced PJIs is to administer antibiotics systemically, surgically debride the infection, and remove the infected device193,194. Whereas with PMMA-CD composite material it may be possible to rescue the device and treat the infection via refilling without the need to physically remove the implant. Typically, if biofilms form on the surface of prosthetic implants, they can develop resistance to antibiotics and can be incredibly challenging to treat as antibiotics have difficulty penetrating through the diffusional barrier of the biofilm194. With CD polymer and PMMA-CD composite, data has indicated that the refilling process helps to reduce the bacterial burden on the material and in the surrounding tissue and allows for drug to be filled into composite to provide sustained therapy to help to fully eradicate the PJI. The ability to keep the initial implant intact through the use of refilling PMMA-CD composite, despite infection, would help to dramatically limit the suffering and discomfort of patients diagnosed with PJIs.

Lasting impact of PMMA-CD composite delivery platform

In summary, based upon the findings from studies in Chapters 2-6, it is evident that for the treatment of soft tissue PJIs, that PMMA-CD beads comprised of up to 15 wt% CD microparticles, prepared via hand-mixing, appear offer the potential for the longest-term antimicrobial treatment. For soft tissue PJIs, PMMA-CD composite does not need to have any particular mechanical strength as it is not a load-bearing application. Therefore, it is

215

possible to increase the content of CD microparticles (up to 15 wt%) to enable for more drug to be refilled and use hand-mixing to further increase the porosity and drug uptake.

Conversely, for treatment of PJIs in load-bearing applications, such as arthroplasties, it is evident that an ideal composition of PMMA-CD composite would be a maximum of 10 wt% CD microparticles prepared via vacuum-mixing and refilled using IO infusion.

Beyond inclusion of 10 wt% CD microparticles, the strength dramatically decreases.

Furthermore, vacuum-mixing has been shown to increase the mechanical strength.

Therefore, this formulation maximizes the amount of CD microparticles that could be incorporated into the PMMA to allow for refilling to occur, without dramatically altering the mechanical properties.

The technology developed in this work stands as the first instance of reported refilling

PMMA material following implantation in either soft tissue or hard tissue. Knowledge from these studies can be used to develop other novel refillable delivery systems to tackle parallel aspects of orthopedic infections or more broadly infections or chronic maladies throughout the body.

Nevertheless, it is important to consider the global rise of antibiotic-resistant and multi- antibiotic-resistant bacteria when developing technologies for long-term antibiotic delivery. While the PMMA-CD composite system and its novel aspect of providing refillable and patient-customizable treatment offers an option available to surgeons using existing antimicrobial agents, the future of infection treatment may likely lie in

216

microbiome-based therapeutics that have yet to be discovered that promote homeostasis of the microbiota and reduce the risk of drug-resistance.

Limitations of PMMA-CD composite delivery platform

While the PMMA-CD composite delivery platform offers many advantages to existing

PMMA-antibiotic delivery systems, it is important to note several limitations of the system. Specifically, in order for noticeable drug refilling to occur, the composite material must be in contact with the drug for a relatively long period of time. Results from Chapter 6, suggested that there is a minimum of 4 hours of exposure to drug to enable visible antibiotic refilling to occur in the PMMA-CD composite. Depending on the application and its location in the body, it may not always be possible for the bolus of drug administered to remain in contact with the PMMA-CD for this length of time.

Furthermore, utilization of the IO infusion technique to refill PMMA-CD composite that has been implanted in bone may cause fairly substantial trauma to the patient as sedation is often required195. Beyond the trauma of inserting the IO port, there is also the risk that the patient can develop an infection at the port if it remains in place for a prolonged period of time as the port can be a source to introduce new infections195,196. Finally, as IO infusion can provide access to central circulation196, it is possible that this technique would result in rapid clearance of drug from bone marrow, rather than remaining in place to enable refilling of implanted PMMA-CD composite.

217

Finally, not all drugs explored in this work appeared to be compatible with the refilling technique. The success of refilling is heavily dependent upon the affinity of the drug used in refilling for the type of CD in the PMMA-CD composite (i.e. α, β, γ). If a drug does not strongly bind to CD, there is not an easy way to enable refilling PMMA-CD to occur with that particular drug.

218

7.2. FUTURE DIRECTIONS

Translation of PMMA-CD composite to in vivo subcutaneous and orthopedic infection models

In vitro studies of PMMA-CD composite (see Chapters 2-6) have demonstrated the ability of the composite to enable a controlled and prolonged release of antibiotic relative to existing PMMA systems where the drug is freely incorporated. Furthermore, these studies have highlighted the capacity of PMMA-CD to be refilled with antibiotics following implantation in bone or muscle tissue and maintained this capacity even in the presence of infection or biofilms. However, in order to eventually implement PMMA-CD composite in a clinical scenario, it will be necessary to translate and evaluate the composite in more physiologically relevant in vivo models. To evaluate the in vivo efficacy of the PMMA-CD for the treatment of two aspects of PJIs (internal osteomyelitis and surrounding soft tissue infection), it will be necessary to explore several small animal models.

To explore the efficacy of PMMA-CD beads to treat soft tissue infections, a subcutaneous implantation infection mouse model can be used. A pilot study using female Swiss Webster mice (n = 3) was completed where a PMMA-CD bead containing

10 wt% RMP-filled CD microparticles was subcutaneously implanted laterally into the dorsal region of each mouse (CWRU IACUC approval #: 2019-0063). Wounds were closed with nylon sutures and surgical glue and 100 μL of 1 x 104 CFU/mL S. aureus was

219

administered into the subcutaneous pocket around the implanted bead. Animals were monitored over 14 days and tissue and bead were harvested and analyzed for CFU counts. Figure 7-3 depicts images of harvested tissue after 14 days treatment of infection with RMP-filled PMMA-CD bead.

Figure 7-3: Representative images of PMMA-CD composite beads containing 10 wt% RMP-filled CD microparticles (circled) subcutaneously implanted in Swiss Webster mice 14 days after exposure to S. aureus.

Treatment with PMMA-CD containing 10 wt% RMP-filled CD microparticles, generally resulted in clearance of infection after 14 days. This was supported by evidence from

Figure 7-1 where there was a lack of cardinal signs of tissue infection. Specifically, there was no indication of erythema, edema, or exudate in the tissue surrounding the implanted

PMMA bead and the tissue was not friable197. Future work is required to confirm these findings relative to controls not receiving any antibiotics. Specifically, a more exhaustive

220

study will be required where a greater sized cohort of mice will be implanted with RMP- filled PMMA-CD composite bead and bacterial clearance will be compared to control animals not receiving antibiotics to ensure that the bacterial clearance is attributed to the action of the antibiotic rather than the animal’s immune system.

To evaluate the possibility of PMMA-CD composite to treat osteomyelitis or infections after it is implanted in bone, an orthopedic osteomyelitis model could be explored.

Specifically, a model could be adapted where PMMA-CD containing RMP and plain

PMMA (control) are implanted into a mouse femoral intramedullary canal surrounding a stainless steel Kirschner wire and the joint space inoculated with S. aureus198,199.

Furthermore, to evaluate the refilling capacity of PMMA-CD composite in vivo, additional conditions could be added to both the soft tissue subcutaneous implantation infection and osteomyelitis models where empty (non-drug filled) PMMA-CD composite is implanted and a local bolus of antibiotic (RMP) is either injected near the implanted bead (soft-tissue) or infused using an IO port in the bone (osteomyelitis). These experiments would help to confirm that the refilling results observed in both in vitro and ex vivo settings can be translated to a more complex in vivo environment. While the ex vivo soft tissue and bone refilling models demonstrated refilling efficacy, these set-ups are not the most realistic as they do not account for blood flow and other biological conditions. Therefore, in order for refilling to be reasonably implemented clinically,

221

extensive in vivo experiments evaluating refilling of PMMA-CD in soft tissue and bone would be required in relevant animal models (i.e. mice and rabbits).

PMMA-CD composite to “protect” activity of heat-sensitive antibiotics

In order to be incorporated into PMMA, antibiotics must remain stable and maintain their activity after exposure to high polymerization temperature of PMMA (~83oC)200. As a result, there are few antibiotics that are able to withstand these conditions, restricting the range of antibiotics in the surgeon’s arsenal to treat PJIs200,201. Incorporating antibiotics into CD “pockets” prior to polymerization in PMMA could help to stabilize and “protect” the antibiotics from thermal degradation during polymerization. Evidence has suggested that when molecules, such as antibiotics, form an inclusion complex with CD, the complex helps to shield the molecule and improve its chemical and physical stability202–

204. Tetracyclines are a class of broad-spectrum antibiotics that are currently incompatible with PMMA due to their loss of activity upon heating (> 16-fold loss of activity after autoclaving)205. Preliminary work has demonstrated that loading tetracycline into γ-CD microparticles prior to incorporation in PMMA, enabled tetracycline to retain its antimicrobial activity over an extended period of time (>60 days). Figure 7-4 displays a persistence zone of inhibition study against S. aureus where duration of antimicrobial activity of tetracycline was compared to when it was freely incorporated into PMMA (i.e. no CD microparticles) relative to when it was first incorporated into either 5 or 10 wt% γ-

CD microparticles that were then embedded in PMMA.

222

Figure 7-4: Persistence zone of inhibition of tetracycline released from plain PMMA or PMMA containing

5 or 10 wt% γ-CD microparticles against S. aureus.

Results from Figure 7-4 demonstrated that tetracycline is incorporated into PMMA without the presence of CD microparticles (i.e. plain PMMA), it has a very short duration of antimicrobial activity (~7 days). However, if it is first loaded into CD microparticles and then incorporated into PMMA, it is capable of maintaining prolonged antimicrobial activity (5 wt% CD = ~55 days, 10 wt% CD = ~70 days). Therefore, not only was the CD able to help protect the activity of tetracycline over 50+ days, it was also able to more effectively control and prolong its release relative to when it was freely incorporated into 223

PMMA. Nevertheless, these studies do not provide any insight into the underlying mechanism of how CD may be shielding and protecting the activity of tetracycline.

Therefore, further chemical analyses are necessary to understand how tetracycline thermally deactivates with and without presence of CD. Preliminary nuclear magnetic resonance (NMR) and mass spectrometry spectra of free tetracycline powder with and without 1 hour of heating at 100oC indicated no detectable key mass or structural changes in tetracycline upon heating (Figures 7-5 and 7-6).

Figure 7-5: NMR spectra of free tetracycline powder non-heated (top) and heated at 100oC for 1 hour

(bottom).

224

Figure 7-6: Mass spectrometry spectra of free tetracycline powder non-heated (left) and heated at 100oC for 1 hour (right).

As drug solutions tend to be more sensitive to thermal degradation than in solid powder form206, these preliminary experiments may not have been the most relevant to elucidate underlying mechanisms since no significant mass or structural changes were evident in tetracycline powder upon heating. Future work should pursue analysis of mass or structural changes of solubilized tetracycline in solution via NMR and mass spectrometry. Furthermore, extensive testing evaluating thermal stability of free tetracycline relative to tetracycline-filled CD microparticles including thermal gravimetric analysis (TGA), differential scanning calorimetry (DSC), and high performance liquid chromatography with evaporative light-scattering detection would be critical to probe into underlying interaction between tetracycline and CD and its impact on the thermal properties of tetracycline207–209. In addition to the thermal analysis studies, resultant antimicrobial activity of tetracycline heated with and without encapsulation in 225

CD should be analyzed using a minimum inhibitory concentration assay210. Successful demonstration CD’s ability to protect antimicrobial activity of a known heat-labile drug

(tetracycline as a model drug), would ultimately enable wider range of antibiotics to be incorporated into PMMA, enabling surgeons to better treat or target specific causative pathogens of PJIs.

PMMA-CD composite as platform for chemotherapeutic delivery

Beyond infection applications, the PMMA-CD composite delivery system developed in this work can serve as a general platform for delivery of other classes of drugs of interest.

Specifically, future work could explore incorporation of different chemotherapeutics (i.e. cisplatin, doxorubicin) to serve as a refillable and controlled delivery system to treat osteosarcoma locally211–214. While the first-line of treatment for osteosarcoma involves surgical resection of the primary tumor, tumors can often recur if the tumor margins are not completely removed215. As a result, different chemotherapeutic-eluting depots have been developed to provide localized therapy to eradicate remaining tumor cells following tumor resection212,213,216–218. Nevertheless, as recurrent osteosarcomas can result as late as

8 years following the initial treatment, there is a need to develop a chemotherapeutic delivery system that is capable of being refilled with drug to repeatedly treat the patient on-demand without the need to perform multiple implantations of the depot219. The

PMMA-CD composite antibiotic delivery system previously developed in this work

(Chapter 2-6) is versatile and could be readily filled with chemotherapeutic agents, rather

226

than antibiotics to provide a refillable drug depot to treat osteosarcoma. Future work could explore this new application of the PMMA-CD composite as well as investigate the potential for antibiotics to be co-delivered from the PMMA-CD composite with chemotherapeutics to both treat infection and kill cancer cells at the surgical site.

7.3. ACKNOWLEDGEMENTS

Thanks to Jaqueline Wallat for assistance with NMR and mass spectrometry analysis and data collection, and to Nathan Rohner, Chao-yi Lu, Dylan Marques, and Ningjing Zhang for assistance and data collection in pilot in vivo study.

227

APPENDIX

PERMISSIONS

PERMISSION FOR RE-PRINT OF CHAPTER 1

228

229

230

231

232

233

PERMISSION FOR RE-PRINT OF CHAPTER 2

234

235

236

237

238

239

PERMISSION FOR RE-PRINT OF CHAPTER 3

240

PERMISSION FOR RE-PRINT OF CHAPTER 4

241

PERMISSION FOR RE-PRINT OF CHAPTER 6

242

REFERENCES

(1) Shanmugasundaram, S.; Ricciardi, B. F.; Briggs, T. W. R.; Sussmann, P. S.;

Bostrom, M. P. Evaluation and Management of Periprosthetic Joint Infection-an

International, Multicenter Study. Hosp. Spec. Surg. 2014, 10, 36–44.

(2) Koh, I. J.; Cho, W.-S.; Choi, N. Y.; Parvizi, J.; Kim, T. K. How Accurate Are

Orthopedic Surgeons in Diagnosing Periprosthetic Joint Infection after Total Knee

Arthroplasty?: A Multicenter Study. Knee 2015, 22, 180–185.

(3) Ravn, C.; Furustrand Tafin, U.; Betrisey, B.; Overgaard, S.; Trampuz, A. Reduced

Ability to Detect Surface-Related Biofilm Bacteria after Antibiotic Exposure under

in Vitro Conditions. Acta Orthop. 2016, 87, 644–650.

(4) Azzam, K.; Parvizi, J.; Jungkind, D.; Hanssen, A.; Fehring, T.; Springer, B.;

Bozic, K.; Della Valle, C.; Pulido, L.; Barrack, R. Microbiological, Clinical, and

Surgical Features of Fungal Prosthetic Joint Infections: A Multi-Institutional

Experience. J Bone Jt. Surg Am 2009, 91 (Suppl 6), 142–149.

(5) Yano, K.; Minoda, Y.; Sakawa, A.; Kuwano, Y.; Kondo, K.; Fukushima, W.;

Tada, K. Positive Nasal Culture of Methicillin-Resistant Staphylococcus Aureus

(MRSA) Is a Risk Factor for Surgical Site Infection in Orthopedics. Acta Orthop.

2009, 80, 486–490.

(6) Gunningberg, L.; Persson, C.; Akerfeldt, T.; Stridsberg, M.; Leo Swenne, C. Pre-

243

and Postoperative Nutritional Status and Predictors for Surgical-Wound Infections

in Elective Orthopaedic and Thoracic Patients. Euro e-Journal Clin Nutr. Metab.

2008, 3, e93–e101.

(7) Namba, R. S.; Inacio, M. C. S.; Paxton, E. W. Risk Factors Associated with Deep

Surgical Site Infections After Primary Total Knee Arthroplasty. J Bone Jt. Surg

Am 2013, 95, 775–782.

(8) Saxena, A.; Baratz, M.; Austin, M. S.; Purtill, J. J.; Parvizi, J. Periprosthetic Joint

Infection Can Cause Abnormal Systemic Coagulation. J Arthroplast. 2011, 26, 50–

57.

(9) Bozic, K. J.; Ward, D. T.; Lau, E. C.; Chan, V.; Wetters, N. G.; Naziri, Q.; Odum,

S.; Fehring, T. K.; Mont, M. A.; Gioe, T. J.; et al. Risk Factors for Periprosthetic

Joint Infection Following Primary Total Hip Arthroplasty: A Case Control Study. J

Arthroplast. 2014, 29, 154–156.

(10) Anguita-Alonso, P.; Hanssen, A. D.; Osmon, D. R.; Trampuz, A.; Steckelberg, J.

M.; Patel, R. High Rate of Aminoglycoside Resistance among Staphylococci

Causing Prosthetic Joint Infection. Clin. Orthop. Relat. Res. 2005, 439, 43–47.

(11) Molina-Manso, D.; del Prado, G.; Ortiz-Perez, A.; Manrubia-Cobo, M.; Gomez-

Barrena, E.; Cordero-Ampuero, J.; Estaban, J. In Vitro Susceptibility to Antibiotics

of Staphylococci in Biofilms Isolated from Orthopaedic Infections. Int J

244

Antimicrob. Agents 2013, 41, 521–523.

(12) Renz, N.; Chevaux, F.; Borens, O.; Trampuz, A. Successful Treatment of

Periprosthetic Joint Infection Caused by Granulicatella Para-Adiacens with

Prosthesis Retention: A Case Report. BMC Musculoskelet. Disord. 2016, 17, 156–

160.

(13) Neumann, D. R. P.; Hofstaedter, T.; List, C.; Dorn, U. Two-Stage Cementless

Revision of Late Total Hip Arthroplasty Infection Using a Premanufactured

Spacer. J Arthroplast. 2012, 27, 1397–1401.

(14) Berbari, E. F.; Hanssen, A. D.; Duffy, M. C.; Steckelberg, J. M.; Ilstrup, D. M.;

Harmsen, W. S.; Osmon, D. R. Risk Factors for Prosthetic Joint Infection: Case-

Control Study. Clin Infect Dis 1998, 27, 1247–1254.

(15) Zmistowski, B.; Restrepo, C.; Huang, R.; Hozack, W. J.; Parvizi, J. Periprosthetic

Joint Infection Diagnosis - A Complete Understanding of White Blood Cell Count

and Differential. J Arthroplast. 2012, 27, 1589–1593.

(16) Parvizi, J.; Gehrke, T. Definition of Periprosthetic Joint Infection. J Arthroplast.

2014, 29, 1331.

(17) Nandi, S. K.; Mukherjee, P.; Roy, S.; Kundu, B.; De, D. K.; Basu, B. Local

Antibiotic Delivery Systems for the Treatment of Osteomyelitis. Mater Sci Eng C

2009, 29, 2478–2485.

245

(18) Levy, S. B. Antibiotic Resistance: Consequences of Inaction. Clin Infect Dis 2001,

3, S124–S129.

(19) Wang, J.-S.; Dunne, N. Bone Cement Fixation: Acrylic Cements. In Joint

Replacement Technology; Woodhead Publishing Series in Biomaterials, 2008; pp

212–251.

(20) Tai, C.-L.; Lai, P.-L.; Lin, W.-D.; Tsai, T.-T.; Lee, Y.-C.; Liu, M.-Y.; Chen, L.-H.

Modification of Mechanical Properties, Polymerization Temperature, and

Handling Time of Polymethylmethacrylate Cement for Enhancing Applicability in

Vertebroplasty. Biomed Res Int 2016, 2016, 7901562.

(21) Martinez-Moreno, J.; Mura, C.; Merino, V.; Nacher, A.; Climente, M.; Merino-

Sanjuan, M. Study of the Influence of Bone Cement Type and Mixing Method on

the Bioactivity and the Elution Kinetics of Ciprofloxacin. J Arthroplast. 2015, 30,

1243–1249.

(22) Neut, D.; Van De Belt, H.; Van Horn, J. R.; Van Der Mei, H. C.; Busscher, H. J.

The Effect of Mixing on Gentamicin Release from Polymethylmethacrylate Bone

Cements. Acta Orthop Scand 2003, 74 (6), 670–676.

(23) McConoughey, S. J.; Howlin, R. P.; Wiseman, J.; Stoodley, P.; Calhoun, J. H.

Comparing PMMA and Calcium Sulfate as Carrriers for the Local Delivery of

Antibiotics to Infected Surgical Sites. J Biomed Mater Res B Appl Biomater 2015,

246

103, 870–877.

(24) Aiken, S. S.; Cooper, J. J.; Florance, H.; Robinson, M. T.; Michell, S. Local

Release of Antibiotics for Surgical Site Infection Management Using High-Purity

Calcium Sulfate: An in Vitro Elution Study. Surg Infect 2015, 16, 54–61.

(25) Inzana, J.; Schwarz, E.; Kates, S.; Awad, H. Biomaterials Approaches to Treating

Implant-Associated Osteomyelitis. Biomater 2016, 81, 58–71.

(26) McPherson, E. J.; Dipane, M. V.; Sherif, S. M. Dissolvable Antibiotic Beads in

Treatment of Periprosthetic Joint Infection and Revision Arthroplasty: The Use of

Synthetic Pure Calcium Sulfate (Stimulan(R)) Impregnated with Vancomycin &

Tobramycin. Reconstr. Rev 2013, 3, 32–43.

(27) McLaren, A. C. Alternative Materials to Acrylic Bone Cement for Delivery of

Depot Antibiotics in Othopaedic Infections. Clin Orthop Relat Res 2004, 427,

101–106.

(28) Gulati, K.; Aw, M. S.; Losic, D. Drug-Eluting Ti Wires with Titania Nanotube

Arrays for Bone Fixation and Reduced Bone Infection. Nanoscale Res Lett 2011,

6, 571.

(29) Jia, Z.; Xiu, P.; Li, M.; Xu, X.; Shi, Y.; Cheng, Y.; Wei, S.; Zheng, Y.; Xi, T.; Cai,

H.; et al. Bioinspired Anchoring AgNPs onto Micro-Nanoporous TiO2 Orthopedic

Coatings: Trap-Killing of Bacteria, Surface-Regulated Osteoblast Functions and

247

Host Responses. Biomater 2016, 75, 203–222.

(30) Cochis, A.; Azzimonti, B.; Della Valle, C.; Chiesa, R.; Arciola, C. R.; Rimondini,

L. Biofilm Formation on Titanium Implants Counteracted by Grafting Gallium and

Silver Ions. J Biomed Mater Res A 2014, 103, 1176–1187.

(31) Yanovska, A. A.; Stanislavov, A. S.; Sukhodub, L. B.; Kuznetsov, V. N.;

Illiashenko, V. Y.; Danilchenko, S. N.; Sukhodub, L. F. Silver-Doped

Hydroxyapatite Coatings Formed on Ti-6Al-4V Substrates and Their

Characterization. Mater Sci Eng C 2014, 36, 215–220.

(32) Ciobanu, G.; Illisei, S.; Luca, C. Hydroxyapatite-Silver Nanoparticles Coatings on

Porous Polyurethane Scaffold. Mater Sci Eng C 2014, 35, 36–42.

(33) Pishbin, F.; Mourino, V.; Gilchrist, J. B.; McComb, D. W.; Kreppel, S.; Salih, V.;

Ryan, M. P.; Boccaccini, A. R. Single-Step Electrochemical Deposition of

Antimicrobial Orthopaedic Coatings Based on a Bioactive Glass/Chitosan/Nano-

Silver Composite System. Acta Biomater 2013, 9, 7469–7479.

(34) Yin, B.; Liu, T.; Yin, Y. Prolonging the Duration of Preventing Bacterial Adhesion

of Nanosilver-Containing Polymer Films through Hydrophobicity. Langmuir 2012,

28, 17019–17025.

(35) Zhang, X.; Wu, H.; Geng, Z.; Huang, X.; Hang, R.; Ma, Y.; Yao, X.; Tang, B.

Microstructure and Cytotoxicity Evaluation of Duplex-Treated Silver-Containing

248

Antibacterial TiO2 Coatings. Mater Sci Eng C 2014, 45, 402–410.

(36) Massa, M. A.; Covarrubias, C.; Bittner, M.; Fuentevilla, I. A.; Capetillo, P.; Von

Marttens, A.; Carvajal, J. C. Synthesis of New Antibacterial Composite Coating

for Titanium Based on Highly Ordered Nanoporous Silica and Silver

Nanoparticles. Mater Sci Eng C 2014, 45, 146–153.

(37) Sussman, E. M.; Casey, B. J.; Dutta, D.; Dair, B. J. Different Cytotoxicity

Responses to Antimicrobial Nanosilver Coatings When Comparing Extract-Based

and Direct-Contact Assays. J Appl Toxicol 2015, 35, 631–639.

(38) Song, C.; Chang, Y.; Cheng, L.; Xu, Y.; Chen, X.; Zhang, L.; Zhong, L.; Dai, L.

Preparation, Characterization, and Antibacterial Activity Studies of Silver-Loaded

Poly(Styrene-Co-Acrylic Acid) Nanocomposites. Mater Sci Eng C 2014, 36, 146–

151.

(39) Jaiswal, S.; Bhattacharya, K.; McHale, P.; Duffy, B. Dual Effects of B-

Cyclodextrin-Stabilised Silver Nanoparticles: Enhanced Biofilm Inhibition and

Reduced Cytotoxicity. J Mater Sci Mater Med 2015, 26, 52.

(40) Liu, X.; Xu, Y.; Wang, X.; Shao, M.; Xu, J.; Wang, J.; Li, L.; Zhang, R.; Guo, X.

Stable and Efficient Loading of Silver Nanoparticles in Spherical Polyelectrolyte

Brushes and the Antibacterial Effects. Colloids Surf B Biointerfaces 2015, 127,

148–154.

249

(41) Sullivan, M. P.; McHale, K. J.; Parvizi, J.; Mehta, S. Nanotechnology: Current

Concepts in Orthopaedic Surgery and Future Directions. Bone Jt. J 2014, 96-B,

569–573.

(42) Leng, M.; Hu, S.; Lu, A.; Cai, M.; Luo, X. The Anti-Bacterial Poly(Caprolactone)-

Poly(Quaternary Ammonium Salt) as Drug Delivery Carriers. Appl Microbiol

Biotechnol 2016, 100, 3049–3059.

(43) Watson, D. A.; Smith, T. J.; Chen, J.; Laporte, R.; Paul, A. Device and Method for

Treating Conditions of a Joint. Patent 6936270B2, 2005.

(44) Bucay-Couto, W.; Li, J. Long-Term Indwelling Medical Devices Containing

Slow-Releasing Antimicrobial Agents and Having a Surfactant Surface. U.S.

Patent 7749203B2, 2010.

(45) Getzlaf, M. A.; Lewallen, E. A.; Kremers, H. M.; Jones, D. L.; Bonin, C. A.;

Dudakovic, A.; Thaler, R.; Cohen, R. C.; Lewallen, D. G.; van Winjnen, A. J.

Multi-Disciplinary Antimicrobial Strategies for Improving Orthopaedic Implants

to Prevent Prosthetic Joint Infections in Hip and Knee. J Orthop Res 2015, 34,

177–186.

(46) Sangeetha, K.; Girija, E. K. Tailor Made Alginate Hydrogel for Local Infection

Prophylaxis in Orthopedic Applications. Mater Sci Eng C Mater Biol Appl 2017,

78, 1046–1053.

250

(47) Wang, N. X.; von Recum, H. A. Affinity-Based Drug Delivery. Macromol. Biosci.

2011, 11, 321–332. https://doi.org/10.1002/9781118747896.ch13.

(48) Bibby, D. C.; Davies, N. M.; Tucker, I. G. Mechanisms by Which Cyclodextrins

Modify Drug Release from Polymeric Drug Delivery Systems. Int J Pharm 2000,

197, 1–11.

(49) Thatiparti, T. R.; von Recum, H. A. Cyclodextrin Complexation for Affinity-Based

Antibiotic Delivery. Macromol. Biosci. 2010, 10 (1), 82–90.

(50) Cyphert, E. L.; Wallat, J. D.; Pokorski, J. K.; Von Recum, H. A. Erythromycin

Modification That Improves Its Acidic Stability While Optimizing It for Local

Drug Delivery. Antibiotics 2017, 6 (2). https://doi.org/10.3390/antibiotics6020011.

(51) Harth, K. C.; Rosen, M. J.; Thatiparti, T. R.; Jacobs, M. R.; Halaweish, I.;

Bajaksouzian, S.; Furlan, J.; von Recum, H. A. Antibiotic-Releasing Mesh Coating

to Reduce Prosthetic Sepsis: An In Vivo Study. J Surg Res 2010, 163, 337–343.

(52) Blanchemain, N.; Haulon, S.; Martel, B.; Traisnel, M.; Morcellet, M.; Hildebrand,

H. F. Vascular PET Prostheses Surface Modification with Cyclodextrin Coating:

Development of a New Drug Delivery System. Eur J Vasc Endovasc Surg 2005,

29, 628–632.

(53) Sobocinski, J.; Laure, W.; Taha, M.; Courcot, E.; Chai, F.; Simon, N.; Addad, A.;

Martel, B.; Haulon, S.; Woisel, P.; et al. Mussel Inspired Coating of a

251

Biocompatible Cyclodextrin Based Polymer onto CoCr Vascular Stents. ACS Appl

Mater Interfaces 2014, 6, 3575–3586.

(54) Mattioli-Belmonte, M.; Cometa, S.; Ferretti, C.; Iatta, R.; Trapani, A.; Ceci, E.;

Falconi, M.; De Giglio, E. Characterization and Cytocompatibility of an

Antibiotic/Chitosan/Cyclodextrins Nanocoating on Titanium Implants. Carbohydr.

Polym. 2014, 110, 173–182.

(55) Thatiparti, T. R.; Shoffstall, A. J.; von Recum, H. A. Cyclodextrin-Based Device

Coatings for Affinity-Based Release of Antibiotics. Biomater 2010, 31, 2335–

2347.

(56) Lepretre, S.; Chai, F.; Hornez, J.-C.; Vermet, G.; Neut, C.; Descamps, M.;

Hildebrand, H. F.; Martel, B. Prolonged Local Antibiotics Delivery from

Hydroxyapatite Functionalised with Cyclodextrin Polymers. Biomater 2009, 30,

6086–6093.

(57) Taha, M.; Chai, F.; Blanchemain, N.; Goube, M.; Martel, B.; Hildebrand, H. F.

Validating the Poly-Cyclodextrins Based Local Drug Delivery System on Plasma-

Sprayed Hydroxyapatite Coated Orthopedic Implant with Toluidine Blue O. Mater

Sci Eng C 2013, 33, 2639–2647.

(58) Hoang Thi, T. H.; Chai, F.; Lepretre, S.; Blanchemain, N.; Martel, B.; Siepmann,

F.; Hildebrand, H. F.; Siepmann, J.; Flament, M. P. Bone Implants Modified with

252

Cyclodextrin: Study of Drug Release in Bulk Fluid and into Agarose Gel. Int J

Pharm 2010, 400, 74–85.

(59) Temtem, M.; Pompeu, D.; Jaraquemada, G.; Cabrita, E. J.; Casimiro, T.; Aguiar-

Ricardo, A. Development of PMMA Membranes Functionalized with

Hydroxypropyl-b-Cyclodextrins for Controlled Drug Delivery Using a

Supercritical CO2-Assisted Technology. Int J Pharm 2009, 376, 110–115.

(60) Jacobsen, P. A. L.; Rafaelsen, J.; Nielsen, J. L.; Juhl, M. V.; Theilgaard, N.;

Larsen, K. L. Distribution of Grafted B-Cyclodextrin in Porous Particles for Bone

Tissue Engineering. Microporous Mesoporous Mater 2013, 168, 132–141.

(61) Taha, M.; Chai, F.; Blanchemain, N.; Neut, C.; Goube, M.; Maton, M.; Martel, B.;

Hildebrand, H. F. Evaluation of Sorption Capacity of Antibiotics and Antibacterial

Properties of a Cyclodextrin-Polymer Functionalized Hydroxyapatite-Coated

Titanium Hip Prosthesis. Int J Pharm 2014, 477, 380–389.

(62) Cyphert, E. L.; Zuckerman, S. T.; Korley, J. N.; von Recum, H. A. Affinity

Interactions Drive Post-Implantation Drug Filling, Even in the Presence of

Bacterial Biofilm. Acta Biomater. 2017, 57.

https://doi.org/10.1016/j.actbio.2017.04.015.

(63) McPherson, E. J.; Lewonowski, K.; Dorr, L. D. Techniques in Arthroplasty: Use of

an Articulated PMMA Spacer in the Infected Total Knee Arthroplasty. J

253

Arthroplast. 1995, 10, 87–89.

(64) Webb, J. C. J.; Spencer, R. F. The Role of Polymethylmethacrylate Bone Cement

in Modern Orthopaedic Surgery. BJJ 2007, 89-B, 851–857.

(65) Stone, P. A.; Mousa, A. Y.; Hass, S. M.; Dearing, D.; Campbell, J. R.; Parker, A.;

Thompson, S.; AbuRahma, A. F. Antibiotic-Loaded Polymethylmethacrylate

Beads for the Treatment of Extracavitary Vascular Surgical Site Infections. J Vasc.

Surg 2012, 55, 1706.

(66) Klemm, K. The Use of Antibiotic-Containing Bead Chains in the Treatment of

Chronic Bone Infections. Clin Microbio Infect 2001, 7, 28–31.

(67) Wentao, Z.; Lei, G.; Liu, Y.; Wang, W.; Song, T.; Fan, J. Approach to

Osteomyelitis Treatment with Antibiotic Loaded PMMA. Microb. Pathog. 2017,

102, 42–44.

(68) Mavroudis, C.; Katzmark, S. L.; Banzel, B. L.; Gray, L. A.; Polk, H. C. Successful

Treatment of Empyema Thoracis with Polymethylmethacrylate Antibiotic-

Impregnated Beads in the Guinea Pig. Ann. Thorac. Surg 1988, 46, 615–618.

(69) Number of all-listed procedures for discharges from short-stay hospitals, by

procedure, category and age: United States

http://www.cdc.gov/nchs/data/nhds/4procedures/2010pro4_numberprocedureage.p

df.

254

(70) Lavernia, C. J.; Alcerro, J. C.; Contreras, J. S. Knee Arthroplasty: Growing Trends

and Future Problems. Int J Clin Rheumatol 2010, 5, 565.

(71) Bohl, D. D.; Sershon, R. A.; Fillingham, Y. A.; Della Valle, C. J. Incidence, Risk

Factors, and Sources of Sepsis Following Total Joint Arthroplasty. J Arthroplast.

2016, 31, 2875–2879.

(72) Widmer, A. F. New Developments in Diagnosis and Treatment of Infection in

Orthopedic Implants. Clin Infect Dis 2001, 33, S94–S106.

(73) Diefenbeck, M.; Muckley, T.; Hofmann, G. O. Prophylaxis and Treatment of

Implant-Related Infections by Local Application of Antibiotics. Injury 2006, 37,

S95.

(74) Joseph, T. N.; Chen, A. L.; Di Cesare, P. E. Use of Antibiotic-Impregnated

Cement in Total Joint Arthroplasty. J Am Acad Orthop Surg 2003, 11, 38–47.

(75) van de Belt, H.; Neut, D.; Schenk, W.; van Horn, J. R.; van der Mei, H. C.;

Bussche, H. J. Staphylococcus Aureus Biofilm Formation on Different

Gentamicin-Loaded Polymethylmethacrylate Bone Cements. Biomater 2001, 22,

1607–1611.

(76) Pelletier, M. H.; Malisano, L.; Smitham, P. J.; Okamoto, K.; Walsh, W. R. The

Compressive Properties of Bone Cements Containing Large Doses of Antibiotics.

J Arthroplast. 2009, 24, 454.

255

(77) Laine, J. C.; Nguyen, T.-Q. D.; Buckley, J. M.; Kim, H. T. Effects of Mixing

Techniques on Vancomycin-Impregnated Polymethylmethacrylate. J Arthroplast.

2011, 26, 1562–1566.

(78) Bistolfi, A.; Massazza, G.; Verne, E.; Masse, A.; Deledda, D.; Ferraris, S.; Miola,

M.; Galetto, F.; Crova, M. Antibiotic-Loaded Cement in Orthopedic Surgery: A

Review. ISRN Orthop. 2011, 290851, 1.

(79) Scott, C. P.; Higham, P. A.; Dumbleton, J. H. Effectiveness of Bone Cement

Containing Tobramycin: An in Vitro Susceptibility Study of 99 Organisms Found

in Infected Joint Arthroplasty. BJJ 1999, 81-B, 440.

(80) Moojen, D. J.; Hentenaar, B.; Vogely, C. H.; Verbout, A. J.; Castelein, R. M.;

Dhert, W. J. In Vitro Release of Antibiotics from Commercial PMMA Beads and

Articulating Hip Spacers. J Arthroplast. 2008, 23, 1152.

(81) Slane, J. A.; Vivanco, J. F.; Rose, W. E.; Squire, M. W.; Ploeg, H.-L. The

Influence of Low Concentrations of a Water Soluble Poragen on the Material

Properties, Antibiotic Release, and Biofilm Inhibition of an Acrylic Bone Cement.

Mater Sci Eng C Mater Biol Appl 2014, 42, 168.

(82) McLaren, A. C.; McLaren, S. G.; Hickmon, M. K. Sucrose, Xylitol, and Erythritol

Increase PMMA Permeability for Depot Antibiotics. Clin Orthop Relat Res 2007,

461, 60.

256

(83) Faber, C.; Hoogendoorn, R. J. W.; Lyaruu, D. M.; Stallmann, H. P.; van Marle, J.;

van Nieuw Amerongen, A.; Smit, T. H.; Wuisman, P. I. J. M. The Effect of the

Antimicrobial Peptide, Dhvar-5, on Gentamicin Release from a

Polymethylmethacrylate Bone Cement. Biomater 2005, 26, 5717–5726.

(84) Cyphert, E. L.; von Recum, H. A. Emerging Technologies for Long-Term

Antimicrobial Device Coatings: Advantages and Limitations. Exp. Biol. Med.

2017, 242 (8). https://doi.org/10.1177/1535370216688572.

(85) Fu, A.; von Recum, H. Affinity-Based Delivery and Refilling of Doxorubicin for

Treatment of Glioblastoma Multiforme, 2013.

(86) Grafmiller, K. T.; Zuckerman, S. T.; Petro, C.; Liu, L.; von Recum, H. A.; Rosen,

M. J.; Korley, J. N. Antibiotic-Releasing Microspheres Prevent Mesh Infection in

Vivo. J Surg Res 2016, 206, 41–47.

(87) ASTM F451-16 Standard Specification for Acrylic Bone Cement. ASTM

International: West Conshohocken, PA 2016.

(88) Ismail, A.; Almonem Doolaanea, A.; Awang, M.; Mohamed, F. High Initial Burst

Release of Gentamicin Formulated as PLGA Microspheres Implant for Treating

Orthopaedic Infection. Int J Pharm Pharm Sci 2012, 4, 685.

(89) Che Nor Zarida, C.; Fauziah, O.; Arifah, A.; Azfar Rizal, A.; Nazri, M.; Ahmad

Hafiz, Z.; Rusnah, M.; Mohd Azam Khan, G.; Hasni Idayu, S. In Vitro Elution and

257

Dissolution of Tobramycin and Gentamicin from Calcium Phosphate. African J

Pharm. Pharmacol. 2011, 5, 2283–2291.

(90) Reller, L. B.; Weinstein, M.; Jorgensen, J. H.; Ferraro, M. J. Antimicrobial

Susceptibility Testing: A Review of General Principles and Contemporary

Practices. Clin Infect Dis 2009, 49, 1749–1755.

(91) Sanz-Ruiz, P.; Carbo-Laso, E.; Del Real-Romero, J. C.; Aran-Ais, F.; Ballesteros-

Iglesias, Y.; Paz-Jimenez, E.; Sanchez-Navarro, M.; Perez-Liminana, M. A.;

Vaquero-Martin, J. Microencapsulation of Rifampicin: A Technique to Preserve

the Mechanical Properties of Bone Cement. J Orthop Res 2017, 36 (1).

(92) Spellberg, B.; Lipsky, B. A. Systemic Antibiotic Therapy for Chronic

Osteomyelitis in Adults. Clin Infect Dis 2012, 54, 393–407.

(93) Chadha, R.; Saini, A.; Gupta, S.; Arora, P.; Thakur, D.; Jain, D. Encapsulation of

Rifampicin by Natural and Modified B-Cyclodextrins: Characterization and

Thermodynamic Parameters. J Incl Phenom Macrocylc Chem 2010, 67, 109–116.

(94) Hickok, N. J.; Shapiro, I. M. Immobilized Antibiotics to Prevent Orthopaedic

Implant Infections. Adv Drug Deliv Rev 2012, 64, 1165–1176.

(95) Duewelhenke, N.; Krut, O.; Eysel, P. Influence on Mitochondria and Cytotoxicity

of Different Antibiotics Administered in High Concentrations on Primary Human

Osteoblast Cell Lines. Antimicrob Agents Chemother 2007, 51, 54–63.

258

(96) Rod-Fleury, T.; Dunkel, N.; Assal, M.; Rohner, P.; Tahintzi, P.; Bernard, L.;

Hoffmeyer, P.; Lew, D.; Uckay, I. Duration of Post-Surgical Antibiotic Therapy

for Adult Chronic Osteomyelitis: A Single-Centre Experience. Int Orthop. 2011,

35, 1725–1731.

(97) Haidar, R.; Der Boghossian, A.; Atiyeh, B. Duration of Post-Surgical Antibiotics

in Chronic Osteomyelitis: Empiric or Evidence-Based? Int J Infect Dis 2010, 14,

e752–e758.

(98) Galanakis, N.; Giamarellou, H.; Moussas, T.; Dounis, E. Chronic Osteomyelitis

Caused by Multi-Resistant Gram-Negative Bacteria: Evaluation of Treatment with

Newer Quinolones after Prolonged Follow-Up. J Antimicrob Chemother 1997, 39,

241–246.

(99) Teterycz, D.; Ferry, T.; Lew, D.; Stern, R.; Assal, M.; Hoffmeyer, P.; Bernard, L.;

Uckay, I. Outcome of Orthopedic Implant Infections Due to Different

Staphylococci. Int J Infect Dis 2010, 14, e913.

(100) Narvaez Rincon, P. C.; Suarez Palacios, O. Y. Plasticizers. In: Polymers and

Polymeric Composites: A Reference Series; Springer: Berlin, Germany, 2015.

(101) Levack, A. E.; Cyphert, E. L.; Bostrom, M. P.; Hernandez, C. J.; von Recum, H.

A.; Carli, A. V. Current Options and Emerging Biomaterials for Periprosthetic

Joint Infection. Curr. Rheumatol. Rep. 2018, 20 (6).

259

https://doi.org/10.1007/s11926-018-0742-4.

(102) Springer, B. D.; Cahue, S.; Etkin, C. D.; Lewallen, D. G.; Mcgrory, B. J.

Arthroplasty Today Infection Burden in Total Hip and Knee Arthroplasties : An

International Registry-Based Perspective. Arthroplast. Today 2017, 3 (2), 137–

140. https://doi.org/10.1016/j.artd.2017.05.003.

(103) Pulido, L.; Ghanem, E.; Joshi, A.; Purtill, J.; Parvizi, J. Periprosthetic Joint

Infection. Clin Orthop Relat Res 2008, 466, 1710–1715.

https://doi.org/10.1007/s11999-008-0209-4.

(104) Kurtz, S. M.; Lau, E.; Watson, H.; Schmier, J. K.; Parvizi, J. Economic Burden of

Periprosthetic Joint Infection in the United States. J. Arthroplasty 2012, 27 (8), 61-

65.e1. https://doi.org/10.1016/j.arth.2012.02.022.

(105) Alp, E.; Cevahir, F.; Ersoy, S.; Guney, A. Incidence and Economic Burden of

Prosthetic Joint Infections in a University Hospital: A Report from a Middle-

Income Country. J Infect Public Heal. 2016, 9 (4), 494–498.

(106) Boddapati, V.; Fu, M. C.; Mayman, D. J.; Su, E. P.; Sculco, P. K.; McLawhorn, A.

S. Revision Total Knee Arthroplasty for Periprosthetic Joint Infection Is

Associated With Increased Postoperative Morbidity and Mortality Relative to

Noninfectious Revisions. J Arthroplast. 2018, 33 (2), 521–526.

(107) Crémet, L.; Corvec, S.; Bémer, P.; Bret, L.; Lebrun, C.; Lesimple, B.; Miegeville,

260

A. F.; Reynaud, A.; Lepelletier, D.; Caroff, N. Orthopaedic-Implant Infections by

Escherichia Coli: Molecular and Phenotypic Analysis of the Causative Strains. J.

Infect. 2012, 64 (2), 169–175. https://doi.org/10.1016/j.jinf.2011.11.010.

(108) Elek, S. D. Principles and Problems of Combined Antibiotic Therapy. Postgr.

Med. J. 1956, 32 (369), 324–327.

(109) Worthington, R. J.; Melander, C. Combination Approaches to Combat Multi-Drug

Resistant Bacteria. Trends Biotechnol. 2014, 31 (3), 177–184.

https://doi.org/10.1016/j.tibtech.2012.12.006.Combination.

(110) Gribble, M. J.; Chow, A. W.; Naiman, S. C.; Smith, J. A.; Bowie, W. R.; Sacks, S.

L.; Grossman, L.; Buskard, N.; Growe, G. H.; Plenderleith, L. H. Prospective

Randomized Trial of Piperacillin Monotherapy versus Carboxypenicillin

Aminoglycoside Combination Regimens in the Empirical Treatment of Serious

Bacterial Infections. Antimicrob. Agents Chemother. 1983, 24 (3), 388–393.

https://doi.org/10.1128/AAC.24.3.388.

(111) Rybak, M. J.; McGrath, B. J. Combination Antimicrobial Therapy for Bacterial

Infections. Drugs 1996, 52 (3), 390–405. https://doi.org/10.2165/00003495-

199652030-00005.

(112) Bollenbach, T. Antimicrobial Interactions: Mechanisms and Implications for Drug

Discovery and Resistance Evolution. Curr. Opin. Microbiol. 2015, 27, 1–9.

261

https://doi.org/10.1016/j.mib.2015.05.008.

(113) Hendriks, J. G. E.; Horn, J. R. Van; Mei, H. C. Van Der; Busscher, H. J.

Backgrounds of Antibiotic-Loaded Bone Cement and Prosthesis-Related Infection.

Biomaterials 2004, 25, 545–556. https://doi.org/10.1016/S0142-9612(03)00554-4.

(114) Jiranek, W. A.; Hanssen, A. D.; Greenwald, A. S. Antibiotic-Loaded Bone Cement

for Infection Prophylaxis in Total Joint Replacement. J. Bone Jt. Surg. 2006, 88-A

(11), 2487–2500.

(115) Cyphert, E. L.; Learn, G. D.; Hurley, S. K.; Lu, C.-Y.; von Recum, H. A. An

Additive to PMMA Bone Cement Enables Postimplantation Drug Refilling,

Broadens Range of Compatible Antibiotics, and Prolongs Antimicrobial Therapy.

Adv. Healthc. Mater. 2018. https://doi.org/10.1002/adhm.201800812.

(116) Kirkpatrick, D.; Trachtenberg, L.; Mangino, P.; Von Fraunhofer, J.; Seligson, D.

In Vitro Characteristics of Tobramycin-PMMA Beads: Compressive Strength and

Leaching. Orthopedics 1985, 8 (9), 1130–1133.

(117) Gomes, D.; Pereira, M.; Bettencourt, A. F. Osteomyelitis: An Overview of

Antimicrobial Therapy. Brazilian J. Pharm. Sci. 2013, 49 (1), 13–27.

https://doi.org/10.1590/s1984-82502013000100003.

(118) Ayre, W. N.; Birchall, J. C.; Evans, S. L.; Denyer, S. P. A Novel Liposomal Drug

Delivery System for PMMA Bone Cements. J. Biomed. Mater. Res. Part B 2016,

262

104B, 1510–1524. https://doi.org/10.1002/jbm.b.33488.

(119) Wu, K.; Chen, Y.; Hsu, Y.-M.; Chang, C.-H. Enhancing Drug Release From

Antibiotic-Loaded Bone Cement Using Porogens. J. Am. Acad. Orthop. Surg.

2016, 24, 188–195.

(120) Öhman, C.; Persson, C.; López, A.; Qin, T.; Engqvist, H.; Xia, W. Enhanced Drug

Delivery of Antibiotic-Loaded Acrylic Bone Cements Using Calcium Phosphate

Spheres. J. Appl. Biomater. Funct. Mater. 2015, 13 (3), 0–0.

https://doi.org/10.5301/jabfm.5000222.

(121) Shirani, A.; Hu, Q.; Su, Y.; Joy, T.; Zhu, D.; Berman, D. Combined Lubricant for

Artificial Joints. ACS Appl Mater Interfaces 2019, 43500.

(122) Su, Y.; Cockerill, I.; Wang, Y.; Qin, Y.-X.; Chang, L.; Zheng, Y.; Zhu, D. Zinc-

Based Biomaterials for Regeneration and Therapy. Trends Biotechnol 2019, 37,

428–441.

(123) Fedorov, A.; Beichel, R.; Kalpathy-Cramer, J.; Finet, J.; Fillion-Robin, J.-C.;

Pujol, S.; Bauer, C.; Jennings, D.; Fennessy, F.; Sonka, M.; et al. 3D Slicer as an

Image Computing Platform for the Quantitative Imaging Network. Magn Reson

Imaging 2012, 30 (9), 1323–1341. https://doi.org/10.1016/j.mri.2012.05.001.3D.

(124) Lowy, F. D.; Chang, D. S.; Lash, P. R. Synergy of Combinations of Vancomycin,

Gentamicin, and Rifampin against Methicillin-Resistant, Coagulase-Negative

263

Staphylococci. Antimicrob. Agents Chemother. 1983, 23 (6), 932–934.

(125) van de Belt, H.; Neut, D.; Uges, D. R. A.; Schenk, W.; van Horn, J. R.; van der

Mei, H. C.; Busscher, H. J. Surface Roughness, Porosity and Wettability of

Gentamicin-Loaded Bone Cements and Their Antibiotic Release. Biomaterials

2000, 21 (19), 1981–1987.

(126) Ozturk, A. U.; Baradan, B. A Comparison Study of Porosity and Compressive

Strength Mathematical Models with Image Analysis. Comput Mater Sci 2008, 43

(4), 974–979.

(127) Yokoe, D. S.; Avery, T. R.; Platt, R.; Huang, S. S. Reporting Surgical Site

Infections Following Total Hip and Knee Arthroplasty: Impact of Limiting

Surveillance to the Operative Hospital. Clin. Infect. Dis. 2013, 57 (9), 1282–1288.

https://doi.org/10.1093/cid/cit516.

(128) Schierholz, J.; Beuth, J. Implant Infections: A Haven for Opportunistic Bacteria. J

Hosp Infect 2001, 49, 87–93.

(129) Olson, M.; Horswill, A. Staphylococcus Aureus Osteomyelitis: Bad to the Bone.

Cell Host Microbe 2013, 13, 629–631.

(130) Naber, C. Staphylococcus Aureus Bacteremia: Epidemiology, Pathophysiology,

and Management Strategies. Clin Infect Dis 2009, 48, S231–S237.

(131) Arciola, C.; Campoccia, D.; Speziale, P.; Montanaro, L.; Costerton, J. Biofilm 264

Formation in Staphylococcus Implant Infections. A Review of Molecular

Mechanisms and Implications for Biofilm-Resistant Materials. Biomaterials 2012,

33, 5967–5982.

(132) Cha, J.; Yoo, J.; Chung, H.-S.; Park, S.-H.; Kim, H.; Lee, Y.; Chung, G.

Investigation of Biofilm Formation and Its Associated with the Molecular and

Clinical Characteristics of Methicillin-Resistant Staphylococcus Aureus. Osong

Pub Heal. Res Perspect 2013, 4, 225–232.

(133) Iordache, F.; Grumezescu, V.; Grumezescu, A.; Curutiu, C.; Ditu, L.; Socol, G.;

Ficai, A.; Trusca, R.; Holban, A. Gamma-Cyclodextrin/Usnic Acid Thin Film

Fabricated by MAPLE for Improving the Resistance of Medical Surfaces to

Staphylococcus Aureus Colonization. Appl Surf Sci 2015, 336, 407–412.

(134) Francolini, I.; Donelli, G.; Stoodley, P. Polymer Designs to Control Biofilm

Growth on Medical Devices. Rev Envirion Sci Biotechnol 2003, 2, 307–319.

(135) Desrousseux, C.; Sautou, V.; Descamps, S.; Traore, O. Modification of the

Surfaces of Medical Devices to Prevent Microbial Adhesion and Biofilm

Formation. J Hosp Infect 2013, 85, 87–93.

(136) Cohen, N.; Lobritz, M.; Collins, J. Microbial Persistence and the Road to Drug

Resistance. Cell Host Microbe 2013, 13, 632–642.

(137) Campoccia, D.; Montanaro, L.; Arciola, C. The Significance of Infection Related

265

to Orthopedic Devices and Issues of Antibiotic Resistance. Biomaterials 2006, 27,

2331–2339.

(138) Davies, D. Understanding Biofilm Resistance to Antibacterial Agents. Nat Rev

2003, 2, 114–122.

(139) Tande, A.; Patel, R. Prosthetic Joint Infection. Clin Microbiol Rev 2014, 27, 302–

345.

(140) Laudermilch, D.; Fedora, C.; Heyl, A.; Rao, N.; McGough, R. Outcomes of

Revision Total Knee Arthroplasty after Methicillin-Resistant Staphylococcus

Aureus Infection. Clin Orthop Relat Res 2010, 468, 2067–2073.

(141) Cavanaugh, D.; Berry, J.; Yarboro, S.; Dahners, L. Better Prophylaxis against

Surgical Site Infection with Local as Well as Systemic Antibiotics. J Bone Jt. surg

Am 2009, 91, 1907–1912.

(142) Zhao, G.; Ye, L.; Huang, Y. In Vitro Model of Bacterial Biofilm Formation on

Polyvinyl Chloride Biomaterial. Cell Biochem Biophys 2011, 61, 371–376.

(143) Orgaz, B.; Lobete, M.; Puga, C.; San Jose, C. Effectiveness of Chitosan against

Mature Biofilms Formed by Food Related Bacteria. Int J Mol Sci 2011, 12, 817–

828.

(144) Halpern, J. M.; Gormley, C. A.; Keech, M. A.; von Recum, H. A.

Thermomechanical Properties, Antibiotic Release, and Bioactivity of a Sterilized 266

Cyclodextin Drug Delivery System. J Mater Chem B 2014, 2 (18), 2764–2772.

(145) Thatiparti, T.; Averell, N.; Overstreet, D.; von Recum, H. Multiplexing

Interactions to Control Antibiotic Release from Cyclodextrin Hydrogels.

Macromol Biosci 2011, 11, 1544–1552.

(146) Leid, J.; Shirtliff, M.; Costerton, J.; Stoodley, P. Human Leukocytes Adhere to

Penetrate, and Respond to Staphylococcus Aureus Biofilms. Infect Immun 2002,

70, 6339–6345.

(147) Tzovaras, G.; Delikoukos, S.; Christodoulides, G.; Spyridakis, M.; Mantzos, F.;

Tepetes, K.; Athanassiou, E.; Hatzitheofilou, C. The Role of Antibiotic

Prophylaxis in Elective Tension-Free Mesh Inguinal Hernia Repair: Results of a

Single-Center Prospective Randomized Trial. Int J Clin Pr. 2007, 61, 236–239.

(148) Kaufman, M.; Meaike, J.; Izaddoost, S. Orthopedic Prosthetic Infections:

Diagnosis and Orthopedic Salvage. Semin Plast Surg 2016, 30, 66–72.

(149) Zhang, Y. G.; Guo, X. L.; Song, Y.; Miao, C. F.; Zhang, C.; Chen, N. H. Diagnosis

and Treatment of Vascular Surgery Related Infection. Open Biomed Eng J 2015, 9,

250–255.

(150) Cyphert, E. L.; Lu, C.; Marques, D. W.; Learn, G. D.; von Recum, H. A.

Combination Antibiotic PMMA Composites Provide Sustained Broad-Spectrum

Antimicrobial Activity and Allow for Post-Implantation Refilling.

267

Biomacromolecules 2020, 21 (2), 854–866.

https://doi.org/10.1021/acs.biomac.9b01523.

(151) Gogia, J. S.; Meehan, J. P.; Di Cesare, P. E.; Jamali, A. A. Local Antibiotic

Therapy in Osteomyelitis. Semin Plast Surg 2009, 23 (2), 100–107.

(152) Hernigou, P.; Flouzat-Lachianette, C.-H.; Jalil, R.; Batista, S.; Guissou, I.;

Poignard, A. Treatment of Infected Hip Arthroplasty. Open Orthop J 2010, 4,

126–131.

(153) Martinez-Moreno, J.; Merino, V.; Nacher, A.; Rodrrigo, J.; Climente, M.; Merino-

Sanjuan, M. Antibiotic-Loaded Bone Cement as Prophylaxis in Total Joint

Replacement. Orthop. Surg 2017, 9, 331–341.

(154) Neut, D.; van de Belt, H.; Stokroos, L.; van Horn, J.; van der Mei, H.; Busscher, H.

Biomaterial-Associated Infection of Gentamicin-Loaded PMMA Beads in

Orthopaedic Revision Surgery. J Antimicrob Chemother 2001, 47, 885–891.

(155) van Vugt, T.; Arts, J.; Geurts, J. Antibiotic-Loaded Polymethylmethacrylate Beads

and Spacers in Treatment of Orthopedic Infections and the Role of Biofilm

Formation. Front. Microbiol. 2019, 10, 1–11.

(156) Wendling, A.; Mar, D.; Wischmeier, N.; Anderson, D.; McIff, T. Combination of

Modified Mixing Technique and Low Frequency Ultrasound to Control the Elution

Profile of Vancomycin-Loaded Acrylic Bone Cement. Bone Jt. Res. 2016, 5, 26–

268

32.

(157) Symonds, T.; Parkinson, B.; Hazratwala, K.; McEwen, P.; Wilkinson, M.; Grant,

A. Use of Regional Administration of Prophylactic Antibiotics in Total Knee

Arthroplasty. ANZ J Surg 2018, 88, 848–853.

(158) Moehring, D.; Gravel, C.; Chapman, M.; Olson, S. Comparison of Antibiotic

Beads and Intravenous Antibiotics in Open Fractures. Clin Orthop Relat Res 2000,

372, 254–261.

(159) Strandberg, G.; Larsson, A.; Lipcsey, M.; Michalek, J.; Eriksson, M. Intraosseous

and Intravenous Administration of Antibiotics Yields Comparable Plasma

Concentrations during Experimental Septic Shock. Acta Anaesthesiol Scand 2015,

59, 346–353.

(160) Neuhaus, D. Intraosseous Infusion in Elective and Emergency Pediatric

Anesthesia: When Should We Use It? Curr Opin Anaesthesiol 2014, 27, 282–287.

(161) Johnson, M.; Inaba, K.; Byerly, S.; Falsgraf, E.; Lam, L.; Benjamin, E.;

Strumwasser, A.; David, J. S.; Demetriades, D. Intraosseous Infusion as a Bridge

to Definitive Access. Am Surg 2016, 82, 876–880.

(162) Charette, R.; Melnic, C. Two-Stage Revision Arthroplasty for the Treatment of

Prosthetic Joint Infection. Curr Rev Musculoskelet Med 2018, 11, 332–340.

(163) Di Benedetto, P.; Di Benedetto, E.; Buttironi, M.; De Franceschi, D.; Beltrame, A.; 269

Gisonni, R.; Cainero, V.; Causero, A. Two-Stage Revision after Total Knee

Arthroplasty. Acta Biomed 2017, 88, 92–97.

(164) Izakovicova, P.; Borens, O.; Trampuz, A. Periprosthetic Joint Infection: Current

Concepts and Outlook. EFORT Open Rev 2019, 4, 482–494.

(165) Sigmund, I.; McNally, M. Diagnosis of Bone and Joint Infections. Orthop. Trauma

2019, 33, 144–152.

(166) Zimmerli, W. CHAPTER 52: Bone and Joint Infections. In Antibiotic and

Chemotherapy; 2010; pp 659–666.

(167) Khatoon, Z.; McTiernan, C.; Suuronen, E.; Mah, T.-F.; Alarcon, E. Bacterial

Biofilm Formation on Implantable Devices and Approaches to Its Treatment and

Prevention. Heliyon 2018, 4, e01067.

(168) Caplin, J.; Garcia, A. Implantable Antimicrobial Biomaterials for Local Drug

Delivery in Bone Infection Models. Acta Biomater. 2019, 93, 2–11.

(169) Arciola, C.; Campoccia, D.; Montanaro, L. Implant Infections: Adhesion, Biofilm

Formation and Immune Evasion. Nat. Rev. Microbiol. 2018, 16, 397–409.

(170) Teughels, W.; Van Assche, N.; Sliepen, I.; Quirynen, M. Effect of Material

Characteristics and/or Surface Topography on Biofilm Development. Clin. Oral

Implants Res. 2006, 17, 68–81.

(171) Ammar, Y.; Swailes, D.; Bridgens, B.; Chen, J. Influence of Surface Roughness on 270

the Initial Formation of Biofilm. Surf. Coat. Technol. 2015, 284, 410–416.

(172) Chatterjee, S.; Biswas, N.; Datta, A.; Maiti, P. Periodicities in the Roughness and

Biofilm Growth on Glass Substrate with Etching Time: Hydrofluoric Acid

Etchant. PLoS One 2019, 14, e0214192.

(173) Ulrich, S. D.; Seyler, T. M.; Bennett, D.; Delanois, R. E.; Saleh, K. J.;

Thongtrangan, I.; Kuskowski, M.; Cheng, E. Y.; Sharkey, P. F.; Parvizi, J.; et al.

Total Hip Arthroplasties: What Are the Reasons for Revision? Int. Orthop. 2008,

32 (5), 597–604.

(174) Pithankuakul, K.; Samranvedhya, W.; Vistutipol, B.; Rojviroj, S. The Effects of

Different Mixing Speeds on the Elution and Strength of High-Dose Antibiotic-

Loaded Bone Cement Created With the Hand-Mixed Technique. J Arthroplast.

2015, 30, 858–863.

(175) Hoey, D.; Taylor, D. Quantitative Analysis of the Effect of Porosity on the Fatigue

Strength of Bone Cement. Acta Biomater. 2009, 5, 719–726.

(176) Jelecevic, J.; Maidanjuk, S.; Leithner, A.; Loewe, K.; Kuehn, K.-D. Methyl

Methacrylate Levels in Orthopedic Surgery: Comparison of Two Conventional

Vacuum Mixing Systems. Ann. Occup. Hyg. 2014, 58 (4), 493–500.

(177) Zivic, F.; Babic, M.; Grujovic, N.; Mitrovic, S.; Favaro, G.; Caunii, M. Effect of

Vacuum-Treatment on Deformation Properties of PMMA Bone Cement. J Mech.

271

Behav. Biomed. Mater. 2012, 5, 129–138.

(178) Scott, J.; Tidball, A.; Uitvlugt, J. M.; Lucia, M.; Vander Griend, D. A.; Louters, L.

L. Methyl-b-Cyclodextrin Directly Binds Methylene Blue and Blocks Both Its Cell

Staining and Glucose Uptake Stimulatory Effects. Biochimie 2009, 91, 271–276.

(179) Zhang, Z. Introduction to Machine Learning: K-Nearest Neighbors. Ann Transl

Med 2016, 4 (11), 218.

(180) Kuhn, K.-D. Bone Cements: Up-to-Date Comparison of Physical and Chemical

Properties of Commercial Materials; Springer-Verlag Berlin and Heidelberg

GmbH & Co. KG: Berlin, Germany, 2011.

(181) Rachid, Z.; Djafar, A.; Abderahmane, S.; Abdelmadjid, M.; Smail, B. Numerical

Simulation of a Crack Emanating from a Micro-Cavity in the Orthopedic Cement

by Technical Sub Modeling of Total Hip Prosthesis. Ratt. ed Integrrita Strutt.

2019, 49, 586–598.

(182) Ensing, G. T.; Hendriks, J. G.; Jongsma, J. E.; van Horn, J. R.; van der Mei, H. C.;

Busscher, H. J. The Influence of Ultrasound on the Release of Gentamicin from

Antibiotic-Loaded Acrylic Beads and Bone Cements. J Biomed Mater Res B Appl

Biomater 2005, 75, 1–5.

(183) Kuechle, D. K.; Landon, G. C.; Musher, D. M.; Noble, P. C. Elution of

Vancomycin, Daptomycin, and Amikacin from Acrylic Bone Cement. Clin Orthop

272

Relat Res 1991, 264, 302–308.

(184) Rivera-Delgado, E.; Xin, A.; von Recum, H. A. Using QSAR for Predictions in

Drug Delivery. BioRxiv 2019, 727172.

(185) Khoo, B.; Grenci, G.; Lim, J.; Lim, Y.; Fong, J.; Yeap, W.; Lim, S.; Chua, S.;

Wong, S.; Yap, Y.-S.; et al. Low-Dose Anti-Inflammatory Combinatorial Therapy

Reduced Cancer Stem Cell Formation in Patient-Derived Preclinical Models for

Tumour Relapse Prevention. Br. J Cancer 2019, 120, 407–423.

(186) Oak, P.; Kopp, F.; Thakur, C.; Ellwart, J.; Rapp, U.; Ullrich, A.; Wagner, E.;

Knyazev, P.; Riodl, A. Combination Treatment of Mammospheres with

Trastuzumab and Salinomycin Efficiently Targets HER2-Positive Cancer Cells

and Cancer Stem Cells. Cancer Cell Biol. 2012, 131, 2808–2819.

(187) Weisberg, E.; Liu, Q.; Nelson, E.; Kung, A.; Christie, A.; Bronson, R.; Sattler, M.;

Sanda, T.; Zhao, Z.; Hur, W.; et al. Using Combination Therapy to Override

Stromal-Mediated Chemoresistance in Mutant FLT3-Positive AML: Synergism

between FLT3 Inhibitors, Desatinib/Multi-Targeted Inhibitors, and JAK Inhibitors.

Leukemia 2012, 26, 2233–2244.

(188) Al-Lazikani, B.; Banerji, U.; Workman, P. Combinatorial Drug Therapy for

Cancer in the Post-Genomic Era. Nat. Biotechnol. 2012, 30, 679–692.

(189) Chaudhry, S.; Dunlop, D. Bone Cement in Arthroplasty. Orthop. Trauma 2012,

273

26, 391–396.

(190) Wimhurst, J.; Brooks, R.; Rushton, N. The Effects of Particulate Bone Cements at

the Bone-Implant Interface. J Bone Jt. Surg 2001, 83-B, 588–592.

(191) Provenzano, M.; Murphy, K.; Riley, L. Bone Cements: Review of Their

Physicochemical and Biochemical Properties in Percutaneous Vertebroplasty. Am

J Neuroradiol. 2004, 25, 1286–1290.

(192) Qu, G.-X.; Ying, Z.-M.; Zhao, C.-C.; Yan, S.-G.; Cai, X.-Z. Mechanical Properties

and Porosity of Acrylic Cement Bone Loaded with Alendronate Powder. Int J Med

Sci 2018, 15, 1458–1465.

(193) Moran, E.; Byren, I.; Atkins, B. The Diagnosis and Management of Prosthetic

Joint Infections. J Antimicrob Chemother 2010, 65, iii45–iii54.

(194) Song, Z.; Borgwardt, L.; Hoiby, N.; Wu, H.; Sorensen, T.; Borgwardt, A.

Prosthesis Infections after Orthopedic Joint Replacement: The Possible Role of

Bacterial Biofilms. Orthop Rev 2013, 7, e14.

(195) Mazzaferro, E.; Ford, R. Section I- Emergency Care. In Kirk & Bistner’s

Handbook of Veterinary Procedures and Emergency Treatment; 2012; pp 1–294.

(196) Giunti, M.; Otto, C. Chapter 194-Intraosseous Catheterization. In Small Animal

Critical Care Medicine; 2015; pp 1009–1013.

(197) Cefalu, J.; Barrier, K.; Davis, A. Wound Infections in Critical Care. Crit. Care 274

Nurs. Clin. North Am. 2017, 29, 81–96.

(198) Pribaz, J.; Bernthal, N.; Billi, F.; Cho, J.; Ramos, R.; Guo, Y.; Cheung, A.;

Francis, K.; Miller, L. Mouse Model of Chronic Post-Arthroplasty Infection:

Noninvasive in Vivo Bioluminescence Imaging to Monitor Bacterial Burden for

Long-Term Study. J Orthop Res 2012, 30, 335–340.

(199) Bernthal, N.; Stavrakis, A.; Billi, F.; Cho, J.; Kremen, T.; Simon, S.; Cheung, A.;

Finerman, G.; Lieberman, J.; Adams, J.; et al. A Mouse Model of Post-

Arthroplasty Staphylococcus Aureus Joint Infection to Evaluate In Vivo the

Efficacy of Antimicrobial Implant Coatings. PLoS One 2010, 5, e12580.

(200) Samara, E.; Moriarty, T.; Decosterd, L.; Richards, R.; Gautier, E.; Wahl, P.

Antibiotic Stability over Six Weeks in Aqueous Solution at Body Temperature

with and without Heat Treatment That Mimics the Curing of Bone Cement. Bone

Jt. Res 2017, 6, 296–306.

(201) Carli, A.; Sethuraman, A.; Bhimani, S.; Ross, F.; Bostrom, M. Selected Heat-

Sensitive Antibiotics Are Not Inactivated During Polymethylmethacrylate Curing

and Can Be Used in Cement Spacers for Periprosthetic Joint Infection. J

Arthroplast. 2018, 33, 1930–1935.

(202) Karginov, V. Cyclodextrin Derivatives as Anti-Infectives. Curr Opin Pharmacol

2013, 13, 717–725.

275

(203) Capelezzo, A.; Mohr, L.; Dalcanton, F.; de Mello, J.; Fiori, M. B-Cyclodextrins as

Encapsulating Agents of Essential Oils. In Cyclodextrin-A Versatile Ingredient;

2018. https://doi.org/10.5772/intechopen.73568.

(204) Sun, H.-K.; Seshadri, M.; Lingard, S.; Monaghan, W.; Faoagali, J.; Chan, E.;

McDonald, H.; Houston, T.; King, M.; Peak, I.; et al. Antibacterial Activity of B-

Cyclodextrin and 2-Hydroxypropyl-B-Cyclodextrin Trimethoprim Complexes. Am

J Microbiol. 2011, 2, 1–8.

(205) Traub, W.; Leonhard, B. Heat Stability of the Antimicrobial Activity of Sixty-Two

Antibacterial Agents. J Antimicrob Chemother 1995, 35, 149–154.

(206) de Haro Moreno, A.; Salgado, H. Stability Study and Degradation Kinetics of

Ceftazidime in Pharmaceutical Preparations. Adv. Anal. Chem. 2012, 2, 1–5.

(207) Lv, N.; Guo, T.; Liu, B.; Wang, C.; Singh, V.; Xu, X.; Li, X.; Chen, D.; Gref, R.;

Zhang, J. Improvement in Thermal Stablity of Sucralose by G-Cyclodextrin Metal-

Organic Frameworks. Pharm. Res. 2017, 34, 269–278.

(208) Shamsipur, M.; Pourmortazavi, S.; Beigi, A.; Heydari, R.; Khatibi, M. Thermal

Stability and Decomposition Kinetic Studies of Acyclovir and Zidovudine Drug

Compounds. AAPS PharmSciTech 2013, 14, 287–293.

(209) Bianchi, F.; Ginggen, A.; Tardy, Y. Stability and Compatibility of Drug Mixtures

in an Implantable Infusion System. Anaesthesia 2008, 63, 972–978.

276

(210) Vassallo, J.; Besinis, A.; Boden, R.; Handy, R. The Minimum Inhibitory

Concentration (MIC) Assay with Escherichia Coli: An Early Tier in the

Environmental Hazard Assessment of Nanomaterials? Ecotoxicol. Environ. Saf.

2018, 162, 633–646.

(211) Ozben, H.; Eralp, L.; Baysal, G.; Cort, A.; Sarkalkan, N.; Ozben, T. Cisplatin

Loaded PMMA: Mechanical Properties, Surface Analysis and Effects on Saos-2

Cell Culture. Acta Orthop Traumatol Turc 2013, 47, 184–192.

(212) Mestiri, M.; Puisieux, F.; Benoit, J. Preparation and Characterization of Cisplatin-

Loaded Polymethyl Methacrylate Microspheres. Int J Pharm. 1993, 89, 229–234.

(213) Mestiri, M.; Benoit, J.; Hernigou, P.; Devissaguet, J.; Puisieux, F. Cisplatin-

Loaded Poly(Methyl Methacrylate) Implants: A Sustained Drug Delivery System.

J Control. Release 1995, 33, 107–113.

(214) Miao, J.; Zhang, F.; Takieddin, M.; Mousa, S.; Linhardt, R. Adsorption of

Doxorubicin on Poly(Methyl Methacrylate)-Chitosan-Heparin Coated Activated

Carbon Beads. Langmuir 2012, 28, 4396–4403.

(215) Sivakumar, T. T.; Jeevadhas, P.; Raghavanpillai, V. B.; Sivaraj, S.; Joseph, A. P.;

Denny, S. E. Recurrent Osteosarcoma of Maxilla Presenting as Fungating Mass: A

Case Report. Int J Case Rep Images 2016, 7 (3), 149–153.

(216) Wang, Y.; Sun, L.; Mei, Z.; Zhang, F.; He, M.; Fletcher, C.; Wang, F.; Yang, J.;

277

Bi, D.; Jiang, Y.; et al. 3D Printed Biodegradable Implants as an Individualized

Drug Delivery System for Local Chemotherapy of Osteosarcoma. Mater. Des.

2020, 186, 108336.

(217) Wu, W.; Dai, Y.; Liu, H.; Cheng, R.; Ni, Q.; Ye, T.; Cui, W. Local Release of

Gemcitabine via in Situ UV-Crosslinked Lipid-Strengthened Hydrogel for

Inhibiting Osteosarcoma. Drug Deliv. 2018, 25, 1642–1651.

(218) Gu, W.; Wu, C.; Chen, J.; Xiao, Y. Nanotechnology in the Targeted Drug Delivery

for Bone Diseases and Bone Regeneration. Int J Nanomedicine 2013, 8, 2305–

2317.

(219) Daw, N.; Chou, A.; Jaffe, N.; Rao, B.; Billups, C.; Rodriguez-Galindo, C.; Meyers,

P.; Huh, W. Recurrent Osteosarcoma with a Single Pulmonary Metastasis: A

Multi-Institutional Review. Br. J Cancer 2015, 112, 278–282.

278