UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A rheological study of treatments for osteoarthritis Chernos, Michael Benjamin 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2016_september_chernos_michael.pdf [ 4.86MB ]
JSON: 24-1.0305129.json
JSON-LD: 24-1.0305129-ld.json
RDF/XML (Pretty): 24-1.0305129-rdf.xml
RDF/JSON: 24-1.0305129-rdf.json
Turtle: 24-1.0305129-turtle.txt
N-Triples: 24-1.0305129-rdf-ntriples.txt
Original Record: 24-1.0305129-source.json
Full Text

Full Text

A RHEOLOGICAL STUDY OF TREATMENTSFOR OSTEOARTHRITISbyMICHAEL BENJAMIN CHERNOSBASc, Queen’s University, 2014A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR THE DEGREE OFMASTER OF APPLIED SCIENCEinTHE FACULTY OF GRADUATE AND POSTDOCTORALSTUDIES(Biomedical Engineering)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)June 2016©Michael Benjamin Chernos, 2016AbstractOsteoarthritis is a common joint disease causing pain and inflammation that limits mobilityand functionality. Osteoarthritis is a widespread disease, and despite it’s prevalence thereis no cure. The disease progresses by degrading joint cartilage and synovial fluid.The current work aims to contribute to existing knowledge regarding osteoarthritisthrough three research projects including a rheological study of novel anti-inflammatoryhyaluronic acid derivatives, a case study on the e↵ects of oral glucosamine supplementationon synovial fluid viscosity, and development of a microfluidic rheometer that may be usedto study low viscosity fluids such as hyaluronic acid and synovial fluid at high shear rates.The anti-inflammatory hyaluronic acid derivative demonstrated shear thinning and vis-coelastic behaviour as expected. This behaviour is common among viscosupplements andis therefore promising for it’s potential use as a viscosupplement, however, the viscosityand viscoelasticity were significantly lower than commercial viscosupplements. It is rec-ommended to investigate modulating viscosity and viscoelasticity with techniques such ascross-linking and increasing hyaluronic acid concentration.In a patient discontinuing oral glucosamine supplementation the viscosity was found to begreater prior to stopping treatment at low shear rates and greater after stopping treatmentat high shear rates. In a second patient no significant change was observed in viscosity.Future study should investigate the e↵ect of glucosamine on viscoelastic behaviour, and tostudy the e↵ect of glucosamine on synovial fluid viscosity over an extended period of time.A PDMS microfluidic rheometer was developed using a soft-lithography process. Therheometer was validated with water at room temperature and was found to predict fluidviscosity with a maximum of 6% error at shear rates as high as 30,000 s1. It is recommendedto investigate possible channel deformation as a cause for decreased accuracy at higher shearrates and resultant operating pressures.iiPrefaceFor all chapters, all research, experimental work and analysis was completed by me exceptas specified below.In Chapter 3, all hyaluronic acid derivatives were prepared and freeze dried at Queen’sUniversity by Dr. Tassos Anastassiades and Dr. Karen Reese-Milton. The work in Chapter3 will be submitted for publication: Chernos, Michael; Grecov, Dana; Kwok, Ezra; Bebe,Siziwe; Babasola, Oladunni; Anastassiades, Tassos, “Rheological Study of Hyaluronic AcidDerivatives”, 2016.The synovial fluid samples used in Chapter 4 were aspirated by Dr. Ezra Kwok.WOMAC surveys were also administered by Dr. Ezra Kwok. This work was approved by theUniversity of British Columbia Research Ethics Board (H08-02272)(H10-00146). These ethi-cal approvals were originally drafted and submitted by Anwar Madkhali who has since gradu-ated.iiiTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ixList of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 Thesis organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.1 Knee anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Synovial fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.2.1 Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.2.2 Rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.3 Osteoarthritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.3.1 Clinical presentation and diagnosis . . . . . . . . . . . . . . . . . . . . 10iv2.3.2 Osteoarthritis classification . . . . . . . . . . . . . . . . . . . . . . . . 122.3.3 E↵ect on synovial fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . 122.3.4 Longterm joint health . . . . . . . . . . . . . . . . . . . . . . . . . . . 132.4 Treatment for osteoarthritis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132.4.1 Non-pharmacologic treatments . . . . . . . . . . . . . . . . . . . . . . 142.4.2 Pharmacologic treatments . . . . . . . . . . . . . . . . . . . . . . . . . 142.4.3 Viscosupplementation . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4.4 Glucosamine supplementation . . . . . . . . . . . . . . . . . . . . . . . 172.5 Rheometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192.5.1 Rotational rheometers . . . . . . . . . . . . . . . . . . . . . . . . . . . 192.5.2 Rheometry of synovial fluid . . . . . . . . . . . . . . . . . . . . . . . . 222.5.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.6 Microfluidics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.6.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.7 Microfluidics and rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.7.1 Rheological measurements . . . . . . . . . . . . . . . . . . . . . . . . . 252.7.2 Extensional rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262.8 Device fabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262.8.1 Master fabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272.8.2 Replica moulding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272.8.3 Sealing the channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272.8.4 Connecting to the chip . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Rheological Evaluation of a Novel Viscosupplement . . . . . . . . . . . . . 293.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.2 Methods and materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303.2.1 Hyaluronic acid derivatives . . . . . . . . . . . . . . . . . . . . . . . . 303.2.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313.2.3 Shear rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31v3.2.4 Viscoelastic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 323.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.3.1 Shear rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.3.2 Viscoelastic properties . . . . . . . . . . . . . . . . . . . . . . . . . . . 353.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383.4.1 Shear rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383.4.2 Viscoselastic properties . . . . . . . . . . . . . . . . . . . . . . . . . . 393.4.3 Suitability of anti-inflammatory hyaluronic acid as a viscosupplement 403.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414 Case Study: Glucosamine Supplementation and Synovial Fluid Rheology 424.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424.2 Methods and materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444.2.1 Ethics approvals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444.2.2 Patient recruitment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444.2.3 Synovial fluid collection . . . . . . . . . . . . . . . . . . . . . . . . . . 444.2.4 Rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454.2.5 WOMAC questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . 464.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464.3.1 Shear viscosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464.3.2 WOMAC questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . 484.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504.4.1 Shear rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504.4.2 WOMAC questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . 514.5 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515 Development of a Microfluidic Rheometer . . . . . . . . . . . . . . . . . . . 535.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535.2 Methods and materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55vi5.2.1 Device fabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555.2.2 Working principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595.2.3 Experimentation and validation . . . . . . . . . . . . . . . . . . . . . . 615.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645.3.1 Device cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695.4.1 Device fabrication methods . . . . . . . . . . . . . . . . . . . . . . . . 695.4.2 Microrheometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 705.4.3 Device improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 726 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746.2 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766.4 Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90A Rheological Evaluation of a Novel Viscosupplement . . . . . . . . . . . . . 91B Case Study: Glucosamine Supplementation and Synovial Fluid Viscosity 95C Development of a Microfluidic Rheometer . . . . . . . . . . . . . . . . . . . 99viiList of Tables3.1 Hyaluronic acid derivative sample labelling . . . . . . . . . . . . . . . . . . . 303.2 Carreau-Yassuda model parameters for hyaluronic acid and derivatives . . . . 343.3 Hyaluronic acid derivative dynamic moduli and crossover frequency . . . . . . 383.4 Hyaluronic acid and derivative molecular weight and concentration . . . . . . 394.1 WOMAC category improvements . . . . . . . . . . . . . . . . . . . . . . . . . 495.1 Microfluidic device costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68A.1 Shear rheology data for hyaluronic acid and derivatives . . . . . . . . . . . . . 92A.2 Sample 1 dynamic moduli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93A.3 Sample 2 dynamic moduli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93A.4 Sample 3 dynamic moduli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93A.5 Sample 4 dynamic moduli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94A.6 Sample 5 dynamic moduli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94B.1 Patient A shear rheology data . . . . . . . . . . . . . . . . . . . . . . . . . . . 96B.2 Patient B shear rheology data . . . . . . . . . . . . . . . . . . . . . . . . . . . 97B.3 Patient A WOMAC scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98B.4 Patient B WOMAC scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98C.1 Pressure drop in microrheometer . . . . . . . . . . . . . . . . . . . . . . . . . 100C.2 Viscosity estimated from microrheometer data . . . . . . . . . . . . . . . . . . 100viiiList of Figures2.1 Synovial joint schematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Synovial fluid rate dependent viscosity. . . . . . . . . . . . . . . . . . . . . . . 72.3 Protein aggregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.4 Radiograph of an osteoarthritic joint . . . . . . . . . . . . . . . . . . . . . . . 112.5 Rheometer geometries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193.1 Hyaluronic acid derivative amplitude sweep . . . . . . . . . . . . . . . . . . . 323.2 Hyaluronic acid a derivative shear viscosity . . . . . . . . . . . . . . . . . . . 333.3 Hyaluronic acid and derivative dynamic moduli . . . . . . . . . . . . . . . . . 374.1 Synovial fluid viscosity: Patient A . . . . . . . . . . . . . . . . . . . . . . . . 474.2 Synovial fluid viscosity: Patient B . . . . . . . . . . . . . . . . . . . . . . . . 484.3 WOMAC scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495.1 PDMS microrheometer fabrication procedure . . . . . . . . . . . . . . . . . . 555.2 Photomask used in master development . . . . . . . . . . . . . . . . . . . . . 565.3 Microfluidic device connection . . . . . . . . . . . . . . . . . . . . . . . . . . . 595.4 Microrheometer experimental setup schematic . . . . . . . . . . . . . . . . . . 625.5 Experimental setup photograph . . . . . . . . . . . . . . . . . . . . . . . . . . 635.6 Microrheometer pressure drop as a function of flow rate . . . . . . . . . . . . 645.7 Microrheometer estimated shear viscosity . . . . . . . . . . . . . . . . . . . . 655.8 % Error as a function of pressure drop. Error magnitude increases withpressure in the channel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665.9 Pressure drop corrected for PDMS channel deformation . . . . . . . . . . . . 675.10 Viscosity corrected for PDMS channel deformation . . . . . . . . . . . . . . . 67ixList of SymbolsDH hydraulic diameter.G” Loss Modulus.G0 Storage Modulus.KC entrance pressure drop estimating factor.KH entrance pressure drop estimating factor.L length.Le entrance length.T0 torque amplitude.⌦ rotational speed.⇥0 cone angle. phase shift.˙ shear rate.˙w wall shear rate.⌘ shear viscosity.⌘0 zero shear viscosity.⌘1 infinite shear viscosity.0 strain amplitude.x relaxation time.! frequency.0 angular amplitude.⇢ density.⌧ shear stress.⌧0 stress amplitude.⌧w wall shear stress.a Yassuda exponent.c cross constant.d depth.n power law index.p cross exponent.v fluid velocity.w width.R rheometer geometry radius.T torque.xiAcknowledgementsI would like to thank my supervisors, Dr. Dana Grecov and Dr. Ezra Kwok for theirguidance and expertise throughout my degree. Their research experience, both experimentaland clinical was instrumental in my progress and achievements throughout this degree.I would like to thank Dr. Tassos Anastassiades for his e↵orts in collaboration on ourresearch regarding the hyaluronic acid derivatives.I would like to thank the committee members Dr. Mark Martinez and Dr. Boris Stoeberfor their time spent reviewing my work and providing advice and recommendations.I would like to thank all of the following individuals, without whom this research wouldnot have been possible. Nick Yeh for his patience in training me on the rheometer. Dr.Mario Beaudoin for his time and advice spent training me in the cleanroom. SamanthaGrist and Sahan Ranamukhaarachchi for all of their advice in working with PDMS anddeveloping microfluidic devices. Masoud Daneshi for his help in working with microfluidicpressure sensors and in developing a microfluidic device. Arian Amirkeyvan for her helpand collaboration in ordering supplies for my microfluidics project and in trouble shootingthe Kinexus rheometer. To all of the above mentioned fellow graduate students and Dr.Beaudoin, I am extremely grateful for the time and advice you have volunteered. Thisresearch would not have been possible without you.Finally, I must thank Anwar Madkhali for all of the hard work he put in to his master’sdegree as I have referred to his thesis on a frequent basis.xiiChapter 1Introduction1.1 MotivationOsteoarthritis is a common degenerative joint disease that causes pain, inflammation and aloss of mobility. Osteoarthritis is widespread, a↵ecting approximately 12% of all adults [51].In 2010 the United States saw 36 million outpatient visits with total accrued costs of $ 81billion [63]. As significant weight bearing joints, osteoarthritis most commonly a↵ects theknees [51]. According to the Centre for Disease Control, as many as 16% of adults over theage of 45 show signs of osteoarthritis in the knees. Mobility impairment is common in latestage osteoarthritis, further necessitating the need for improved osteoarthritis treatments.Osteoarthritis symptoms are the result of cartilage and synovial fluid breakdown withinthe joint, which limits protection of articulating surfaces. Caused by excessive mechanicalstress, osteoarthritis tends to develop later in life but does not exclusively a↵ect elderly.Several factors will predispose an individual to developing osteoarthritis such as obesity,age, joint trauma, joint malalignment, and hereditary predisposition to osteoarthritis [51].Obesity is the greatest risk factor for developing osteoarthritis, while mechanical factors arethe greatest determinants for disease progression [51].Many treatment options for osteoarthritis prioritize symptom relief over altering theprogression of the disease. Consequently, most cases of osteoarthritis increase in severityover time, causing an increase in the incidence of mobility impairment [3]. Viscosupple-mentation is an emerging treatment for osteoarthritis with promise of alleviating symptoms1and impeding its progression. Viscosupplements are designed to be similar in propertiesto synovial fluid and are injected into the joint cavity in order to protect articular carti-lage and supplement the diseased synovial fluid. Despite these advancements, osteoarthritisis largely untreatable and in many cases progresses to the point of disability. There is aneed to continue studying osteoarthritis and methods for slowing and preventing diseaseprogression.1.2 ObjectivesThe overarching purpose of this research is to contribute to the current body of scientificliterature concerning treatment of osteoarthritis and methods for evaluating treatment ef-fectiveness. This was accomplished by first identifying gaps in the current literature andsubsequently carrying out three research projects.1. In certain cases, viscosupplementation has been shown to relieve symptoms of os-teoarthritis for up to 6 months following an injection, however, one aspect of thedisease which has not yet been addressed by current viscosupplements is the incidenceof flare up.In collaboration with Dr. Tassos Anastassiades, a faculty member of the School ofMedicine and department of Biomedical and Molecular Sciences at Queen’s Univer-sity, the rheology of a low molecular weight derivative of hyaluronic acid with anti-inflammatory properties was studied with the purpose of evaluating suitability as anovel viscosupplement.The objectives of this study are to first investigate how the synthesis and processingof the derivative compound is a↵ecting the fluid rheology and second to compare therheology of the derivative compound with commercial viscosupplements in order toassess it’s suitability for use in treating osteoarthritis.2. Oral glucosamine supplements are commonly administered as a treatment for mild andmoderate cases of osteoarthritis. A large number of clinical studies have been pub-2lished with conflicting results on the clinical ecacy of the treatment. These studiespredominantly rely on patient feedback surveys and radiographic evidence of diseasemodification. The current body of literature does not investigate how glucosaminesupplementation may a↵ect synovial fluid rheology and its ability to lubricate andprotect a synovial joint such as the knees.The framework for a case study on the e↵ect of oral glucosamine supplements onsynovial fluid rheology was developed. Ethical approvals were obtained and prelimi-nary data was collected for one patient beginning glucosamine supplementation and asecond patient discontinuing treatment.The objectives of this study are to establish a study protocol that can be appliedto a larger scale study and to identify changes to synovial fluid viscosity following adiscontinuation or commencement of oral glucosamine supplementation.3. Rheological studies of hyaluronic acid and synovial fluid are often limited in the rangeof shear rates they are able to investigate by inherent limitations of conventionalrheometry. Consequently, the shear viscosity of hyaluronic acid and synovial fluid hasnot yet been studied at high shear rates.A microfluidic rheometer was developed and validated in order to use the device inthe future to obtain high shear measurements of low viscosity fluids such as hyaluronicacid and synovial fluid.The objectives of this study were to develop a device capable of measuring viscosityat shear rates orders of magnitude higher than would be feasible with a conventionalrheometer and to validate the device with a known low viscosity fluid. The device willthen be available for future study of hyaluronic acid and synovial fluid at high shearrates.31.3 Thesis organizationChapter 2 is a literature review on basic relevant anatomy, osteoarthritis and existing treat-ments for osteoarthritis. Additionally, basic principles of rheometry are discussed. Finally,a literature review of microfluidics and its application to rheology is discussed.Chapter 3 describes in detail the evaluation of the novel hyaluronic acid derivative sam-ples.Chapter 4 describes the framework developed for a case study on the e↵ect of oralglucosamine supplements on synovial fluid rheology. Preliminary data is also presented.Chapter 5 describes the methods and materials used to develop a microfluidic rheometer.Measurements are presented to validate the rheometer’s operation, and a discussion of futureapplications is presented.Chapter 6 presents the overall conclusions from the work presented as well as recom-mendations for future study.4Chapter 2Literature Review2.1 Knee anatomyThe knee joint is a synovial joint located between the femur and tibia [2]. In synovial jointsthe articulating surfaces do not come into direct contact and are separated by a gap referredto as the synovial cavity [2]. The terminal ends of bone are covered with a protective layer ofarticular cartilage that serves as the articulating surface [2]. The synovial cavity is filled withsynovial fluid, which helps prevent bone on bone contact and acts a lubricant to facilitatejoint movement [2][94]. In symptomatic cases of osteoarthritis, a combination of degradingarticular cartilage and synovial fluid contribute to joint pain, inflammation and a loss ofmobility. A schematic of a typical synovial joint is shown in Figure 2.1.Figure 2.1: Synovial joint schematic. Articulating bone, articular cartilage, and synovialfluid are labelled. Image reprinted with permission of Wikimedia Commons.52.2 Synovial fluid2.2.1 CompositionSynovial fluid is secreted into the joint cavity by the synovial membrane and is comprisedprimarily of hyaluronic acid, lubricin, proteins such as -globulin and albumin, lipids, choles-terol and water [34]. Hyaluronic acid is the primary component in synovial fluid and iswidely thought to dictate the fluid rheology [98][34][27]. Other constituents present in muchsmaller concentrations include cytokines, collagen, enzymes, proteoglycans, and fibronectin.Leukocytes are also present in small concentrations in healthy synovial fluid. [34][8].In healthy synovial fluid the concentration of hyaluronic acid can range from 0.35 to 3.65mg/mL, while protein concentration has been reported to range from 10.4 to 21.3 mg/mL[34][8]. Phospholipids and cholestoerol are present in lower concentrations. The averagemolecular weight of hyaluronic acid is reported in the range of 6.3 to 7.6 MDa [34][8].2.2.2 RheologyShear rheologySynovial fluid is a non-Newtonian, shear-thinning fluid [40][34]. As applied shear forcesincrease, the viscosity of synovial fluid decreases. This property has practical implicationsthat allow the fluid to become thinner during periods of elevated shear such as running andto increase in viscosity during periods of low shear such as standing. In the former case,the thinning property facilitates articulating surface lubrication, while in the latter casea thicker synovial fluid will prevent articulating surfaces from coming into contact whenstationary and bearing weight [8][34]. A viscosity curve for synovial fluid as a function ofshear rate is presented in Figure 2.2.6Figure 2.2: Synovial fluid viscosity as a function of shear rate. Shear thinning behaviour ofsynovial fluid is demonstrated as the applied shear rate increases.Modelling simple shear behaviourExperimental shear thinning behaviour of synovial fluid is often modelled with an empiri-cal relationship such as the power law, Cross model, simplified Cross model and Carreau-Yassuda model, which are presented in Equation 2.1, Equation 2.2, Equation 2.3, and Equa-tion 2.4 respectively [1].⌧ = ˙⌘n (2.1)⌘  ⌘1⌘0  ⌘1 =11 + c˙p(2.2)⌘⌘0=11 + c˙p(2.3)7⌘  ⌘1⌘0  ⌘1 = [1 + (˙)a]1na (2.4)Where• ⌧ is the shear stress• ˙ is the shear rate• ⌘ is the shear viscosity• n is the power law index related to the degree of shear thinning• ⌘1 is the viscosity at high shear• ⌘0 is the zero shear viscosity• c is the Cross constant• p is the Cross exponent• 1/ is the critical shear rate at which shear thinning behaviour is observed• a is the Yassuda exponent and is proportional to the extent over which shear thinningbehaviour is observedThe simple power law describes shear thinning behaviour only, while the Cross modeldescribes a Newtonian plateau at low shear rates and shear thinning behaviour at highershear rates. The simplified Cross model assumes that ⌘1 is much smaller than ⌘0. TheCarreau-Yassuda model accounts for a Newtonian plateau at both low and very high shearrates [1]. Given sample behaviour and the investigated range of shear rates, some of theabove models may be more appropriate than others for modelling experimental data of aparticular fluid.8ViscoelasticitySynovial fluid is viscoelastic, exhibiting both viscous and elastic behaviour [8][34]. Thisresponse is generally characterized with the dynamic moduli, G0 and G”, where G0 is thestorage modulus, while G” is the loss modulus, corresponding to the elastic and viscousresponses respectively. At low frequencies G”>G0 and the fluid behaviour is dominated bythe viscous response [34]. At higher frequencies G”<G0 and the fluid behaviour is domi-nated by the elastic response. The frequency at which G”=G0 is referred to as the crossoverfrequency, and is often used as an estimate of the point at which fluid behaviour transi-tions from viscous to elastic behaviour. For healthy synovial fluid, the crossover frequencygenerally occurs at a frequency of approximately 0.02 Hz [80].RheopexySynovial fluid has also been observed to demonstrate rheopectic behaviour at low shear rates[72]. Rheopectic behaviour is a useful synovial fluid property in the knees that results inincreased viscosity during periods of low shear such as standing, providing a more elasticbu↵er between the femur and tibia. Protein aggregation has been proposed as a possiblemechanism that would explain the observed rheopectic behaviour of synovial fluid. At lowshear rates proteins in synovial fluid aggregate and form loose structures around the muchlonger hyaluronic acid polymers. Over time these structures grow, causing the observedincrease in viscosity [72]. Protein aggregation around hyaluronic acid chains is depictedschematically in Figure 2.3.9Figure 2.3: Protein aggregation. Protein aggregation around hyaluronic acid chains is aproposed mechanism for rheopectic behaviour [72]. Image reprinted with the permission ofthe Journal of the Royal Society interface.The Role of proteins in synovial fluidThe role of proteins in synovial fluid has been proposed to explain instances of observedbulk yielding, suggesting that proteins tend to adsorb to the shear surface, forming a proteinaggregate on the shear surface that inhibits induced flow at low shear rates [96][69]. Proteinadsorption to the shear interface has led recent research to propose the formation of aninterfacial layer with distinct rheology, not representative of the bulk rheology [96].Thismay explain the large variation in rheological measurements when correlated to proteinconcentration, however further research is required to validate this claim. The formation ofa locally viscous and sti↵ interfacial layer has also been proposed as a mechanism for jointsti↵ness after long periods of inactivity [96].2.3 Osteoarthritis2.3.1 Clinical presentation and diagnosisDue to the broad spectrum of severity in osteoarthritis cases, there exists a large amount ofvariability in clinical presentation. Typically osteoarthritis will a↵ect specific joints ratherthan manifesting as general joint pain throughout the body [87]. Osteoarthritis is character-10ized by usage related pain and sti↵ness after periods of inactivity, joint e↵usion, and reducedrange of motion [87]. Bony enlargement and coarse crepitus are also signs of osteoarthritis[49]. In severe cases, patients may experience resting pain [87].It is important to assess suspected cases of osteoarthritis with medical imaging to ensureappropriate treatments are prescribed and to understand how the disease will progress. MRIpermits visualization of most intra-articular structures, however, combining MRI with X-ray and radiographs provides a more comprehensive view of an arthritic joint [18]. Whenusing medical imaging to diagnose osteoarthritis, clinicians will generally look for threekey features; osteophyte formation along the edges of the articulating surfaces, joint spacenarrowing, and cartilage breakdown as shown in Figure 2.4 [87] [18]. Osteophyte formationcauses added joint friction during articulation and can impinge on cartilage when standing[18]. Joint space narrowing is usually attributed to a degradation of the synovial fluid causedby a decrease in fluid elasticity and ability to maintain adequate separation of the articularsurfaces [98].Figure 2.4: Radiograph of an osteoarthritic joint. Joint space narrowing (1) and osteophyteformation (2) in the knee are indications of osteoarthritis severity [87]. Image reprinted withpermission of American family physician.112.3.2 Osteoarthritis classificationSeveral grading scales have been developed to improve osteoarthritis characterization and tostandardize treatments. In general, the grading scales are based on a pain and functionalityindex survey completed by the patient or on a radiological based index completed by theclinician.The Kellgren-Lawrence grading scale was developed to classify osteoarthritis based onosteophyte formation and severity of joint space narrowing [18]. There are four severities ofosteoarthritis on the Kellgren-Lawrence grading scale, the least severe being Grade 1 withdoubtful joint space narrowing and possible formation of small osteophytes, while Grade 4severity would correspond to significant joint space narrowing, large osteophytes and severesclerosis [18].The WOMAC questionnaire was developed as a method for assessing pain, sti↵ness, andphysical function in patients with hip and or knee osteoarthritis. The Likert Scale can beused to classify responses as none, mild, moderate, severe and extreme corresponding toscores of 0-4 respectively. Maximum scores in subcategories are 20, 8, and 68 for pain,sti↵ness and physical function respectively [53]. The overall WOMAC score was calculatedby summing the scores from the three subcategories with a maximum score of 96The Osteoarthritis Research Society International (OARSI) has recently developed a newjoint pain index specifically for cases of osteoarthritis in the knees [94]. The survey uses avariety of questions to develop an understanding on how osteoarthritis has impacted patientdaily activities, the frequency and severity of pain, and patient level of concern. Severalother indexes exist to aid in osteoarthritis classification, many of which are nonspecific tothe joint in question.2.3.3 E↵ect on synovial fluidThe changes to synovial fluid caused by osteoarthritis vary widely depending on the individ-ual and the severity of the disease. In general the average molecular weight and concentrationof hyaluronic acid decrease, while the concentration of proteins, lipids, and cells increase12[34][8][27]. These changes may correlate with increased joint inflammation as low molecularweight hyaluronic acid has been demonstrated to correlate with inflammatory response [6].Additionally, with the onset of osteoarthritis, the white blood cell count in synovial fluid isobserved to increase which also correlates with joint inflammation [27][23].The rheological e↵ects of the above changes result in a decrease in viscosity and degreeof shear thinning. Zero shear viscosity of healthy synovial fluid has been reported between 6and 175 Pas while the zero shear viscosity of osteoarthritic fluid has been reported between0.1 and 1 Pas. For inflammatory fluid this di↵erence is even greater [34][27]. Changes arealso observed in the viscoelastic properties, which include a shift of the cross-over frequencythat results in a transition from liquid to gel behaviour at much higher frequencies, and adecrease in both the elastic and viscous moduli [34].With the onset of these rheological changes, diseased synovial fluid is no longer able toe↵ectively lubricate and protect the joint and may therefore lead to increased cartilage wear,pain and additional inflammation [59][61].2.3.4 Longterm joint healthThe progression of osteoarthritis is dicult to predict. Recent research has been directedat determining which patients are at a greatest risk for early progression of osteoarthritis.Kashevarov et. Al have recently identified risk factors and indicators for early progressionof knee osteoarthritis. Among those identified were high levels of pain, elevated body massindex and bone density in the lumbar [55].As the patient ages, osteoarthritis symptoms will become more pronounced. It is commonfor osteoarthritis in the knees to eventually require total knee arthroplasty to alleviatesymptoms [49].2.4 Treatment for osteoarthritisDespite the prevalence of knee osteoarthritis, relatively few treatment options exist fornon-severe cases. Most treatment options available for osteoarthritis have had limited or13inconclusive clinical ecacy.Despite e↵orts made by OARSI there is no commonly accepted treatment plan for kneeosteoarthritis and decisions are made on a case-by-case basis [94]. Consequently, manypatients live with pain and mobility impairments until their symptoms progress to the pointof requiring knee arthroplasty [49].In general, osteoarthritis treatments can be divided into four categories; non-pharmacologic,pharmacologic, alternative, and surgical [87].2.4.1 Non-pharmacologic treatmentsObesity is one of the leading causes for development of osteoarthritis in the knees [49][43][7].Therefore, many surgeons will prescribe exercise and leg muscle strengthening as treatmentfor early cases of osteoarthritis. Weight loss has been demonstrated to partially relieve jointstresses, reducing pain and wear [67]. Leg muscle strengthening has been proven to reducepain and stress in the knee by improving joint support and reducing the impact of jointmalalignment [55]. Unfortunately, in the majority of osteoarthritis cases neither weight lossnor leg muscle strengthening has significantly inhibited the progression of osteoarthritis [85].2.4.2 Pharmacologic treatmentsAcetaminophenAcetaminophen is often prescribed in conjunction with exercise and weight loss as a primarytreatment for mild cases of osteoarthritis [90]. Acetaminophen is inexpensive, safe and hasbeen proven to alleviate pain when taken daily [87]. Acetaminophen is administered orallyand has relatively few side e↵ects [87]. Acetaminophen is commonly found in other over thecounter and prescription medications and as a result, daily intake limits may prohibit theuse of acetaminophen to treat mild cases of osteoarthritis [87]. Acetaminophen does notdelay the progression of osteoarthritis and is prescribed to reduce inflammation and pain[90].14Non-steroidal anti-inflammatory drugs (NSAIDs)NSAID therapy is typically administered for moderate to severe cases of osteoarthritis,although it is occasionally prescribed in place of acetaminophen [87]. Generally, NSAIDsare reserved for more debilitating cases of osteoarthritis due to the severity of the side e↵ects,including intestinal bleeding, renal dysfunction, and elevated blood pressure [87][70].Cortico-steroidsCortico-steroid injections are administered in an e↵ort to boost synovial fluid properties.Prior to injecting the cortico-steroid, a small volume of synovial fluid must be removed toavoid pressure build-up in the synovial cavity caused by an excess of fluid. Cortico-steroidinjections have been demonstrated to reduce acute pain in cases of knee osteoarthritis forup to three weeks [52]. Cortico-steroid injections have not proven to improve joint mobilityor function [52]. Physicians frequently recommend that patients receive no more than fourcortico-steroid injections per year per weight bearing joint due to a risk of osteoporosis andelevated rates of infection during subsequent knee arthroplasty [52] [12].2.4.3 ViscosupplementationViscosupplementation was proposed as a method to supplement degrading synovial fluidin order to provide adequate protection to the articulating surfaces in the knees. Visco-supplementation e↵ectiveness is reliant on the clinician accurately locating and injectingthe viscosupplement into the intra-articular region of the joint [87]. Short term outcomesfrom viscosupplementation have proven less e↵ective than cortico-steroid injections, oftenproviding no symptom relief in the first two to four weeks [93][87]. However, following thefour week point, viscosupplementation has been reported to alleviate pain, and to improvejoint mobility for up to six months [98] [93]. In contrast to NSAIDs, viscosupplementationside e↵ects are typically restricted to the a↵ected joint, resulting in minor concerns suchas joint inflammation, irritation and pain [98]. A major drawback for patients wishing toundergo viscosupplementation is the cost, which is not covered by Canadian healthcare.15Recent improvements to viscosupplementation have focused on enhancing the injectabil-ity of the supplement without compromising the viscoelastic properties. It has recentlybeen shown that by adding nanoparticles to a hyaluronic acid solution, the viscosity can bereduced (and injectability improved) without reducing the elasticity of the supplement[33].HYADD4, a relatively new viscosupplement has improved injectability due to its partiallyhydrophobic hyaluronic acid based hydrogel structure [36]. In the particular study, HYADD4was found to be the only gel-like viscosupplement that underwent full recovery of viscoelasticproperties following a high shock such as that delivered by injection or large compressiveforces [33].A separate, novel approach to viscosupplementation investigated the potential of au-rophilically cross-linked polyethylene glycol (PEG) based hydrogels in mimicking healthysynovial fluid. The gels were found to have good injectability and were not found to haveadverse e↵ects on cell growth when tested in vitro [21].Further research by Bausani et al. has focused on extending the lifetime of rheologicalchanges in synovial fluid caused by viscosupplementation. Synolis-Va, an innovative vis-cosupplement combining hyaluronic acid and sorbitol employs sorbitol as an oxygen freeradical (OFR) scavenger, thereby reducing the oxidative stress on chondrocytes and synovi-cytes found in synovial fluid while delaying the degradation of hyaluronic acid [11]. Simi-larly, Conrozier et al. have studied the benefits of employing mannitol as a reactive oxygenspecies (ROS) scavenger in order to preserve the viscoelastic properties of hyaluronic acid.Results from the study demonstrated that hyaluronic acid solutions containing mannitolbetter maintained their viscoelastic properties when exposed to oxygen peroxide comparedto pure hyaluronic solutions, suggesting that addition of mannitol may improve the lifetimeof viscosupplements [28].Despite recent advances in viscosupplement formulation, clinical trials reveal mixed re-sults and inconclusive improvements to the viscoelastic properties of synovial fluid overextended time periods [41][36][98]. Meta-analysis has also raised questions regarding thee↵ectiveness of the treatment [78].162.4.4 Glucosamine supplementationAlternative supplements such as glucosamine, chondroitin, and s-adenosylmethionine havebeen proposed for pain relief in mild cases of osteoarthritis [87]. Results for supplementsadministered independently and in conjunction have not been conclusively proven to havea significant clinical benefit. The rheological benefits of such treatments are unproven andconsequently, their e↵ectiveness in delaying and inhibiting the progression of osteoarthritiscannot be confirmed.In general, glucosamine studies are restricted to patients with mild to moderate cases ofosteoarthritis, however it is occasionally used in more severe cases [98]. Glucosamine is alsogenerally accepted to have very few, if any, adverse side e↵ects [98]. A recent study has in-vestigated the possibility that glucosamine supplementation may a↵ect glucose metabolism.Results from the study suggest that glucosamine dosages would have to be significantlyhigher than currently administered in order to a↵ect glucose metabolism [81].The e↵ects of glucosamine have been studied in both animal and human trials. In gen-eral, most trials find that pain symptoms, as evaluated by WOMAC and Lequesne indexestend to improve, even in trials as short as 3 months [88][56][53]. Long term administration ofglucosamine has demonstrated that it may inhibit progression of osteoarthritis with radio-graphic evidence of slowing joint space narrowing and osteophyte formation [53][88]. Littlework has been done on evaluating the rheological e↵ects of glucosamine supplements, how-ever, a recent study has suggested that glucosamine does not cause any rheological changeswithin a period of 3 months [62]. It has also been demonstrated that glucosamine has ledto improved range of motion and overall joint functionality [62][53][88].Sadeghi et al. have attempted to reconcile conflicting results from glucosamine trials byinvestigating variation in dosage [79]. By studying 3 groups; placebo, full dose and 1/3 dose,Sedeghi et al. conclude that functionality and pain scale improvement are dependent ondosage. Consequently, they have attributed some variation in published glucosamine trialsto variation in dosage.A meta-analysis conducted by Erikson et Al attempted to determine the underlying17reason for the large degree of variation in results from glucosamine clinical trials. Thestudy concluded that a considerable proportion of the clinical trials were at risk of a bias,contributing at least in part to the discrepancy between results. Additionally, the studyattributes the variation in results to variation in glucosamine formula between brands [30].The mechanism by which glucosamine improves the condition of an osteoarthritic jointis often considered to be the promotion of cartilage regeneration [32]. However, this pro-cess is slow acting and is unlikely to explain the immediate pain relief reported by certainglucosamine clinical trials [32]. McCarty proposes, contrary to the findings later proposedby Matsuno, that glucosamine may stimulate the endogenous production of high molecularweight hyaluronic acid in addition to stimulating cartilage building [65]. This e↵ect meritsfurther study, and can be verified by testing changes in synovial fluid properties within thefirst several weeks of glucosamine treatment.As is the case for viscosupplementation, glucosamine trials have achieved mixed results,some claiming to alleviate painful symptoms of osteoarthritis and to alter disease patho-genesis by impeding joint space narrowing while others refute these findings, suggestingany benefits are not statistically and clinically significant [38][48][53][79][81]. The large de-gree of ambiguity in the literature may be attributable to inadequate dosage, variation ofglucosamine brands, bias, and the subjective nature of osteoarthritis pain scales such asWOMAC and Lequesne [48][30][17].The potential of glucosamine as a disease altering treatment for osteoarthritis meritsfurther study as it may provide an e↵ective, low cost, non-invasive and minimal side-e↵ecttreatment option for altering the progression of osteoarthritis. In addition to the standardmetrics used to evaluate the e↵ectiveness of glucosamine treatment such as pain scale andjoint space narrowing, metrics such as rheological changes to the synovial fluid and changesin white blood cell counts should be used to evaluate treatment e↵ectiveness.182.5 RheometryRheology is defined as the science of deformation and flow of matter. Rheometry is the studyof rheological material functions. In practice rheometry can be carried out in a variety ofexperimental setups, however, rheometry is most commonly carried out with a rheometer.Rheometers can be designed to study rheological parameters under many di↵erent types offlow, including simple shear flow, oscillatory flow, and capillary flow [68].2.5.1 Rotational rheometersRotational rheometers are commonly used in the measurement of shear flow and smallamplitutde oscillatory flow. These rheometers measure rheological properties by applying ashear force to a small volume of sample and measuring the corresponding torque requiredto apply the shear force. In general, this is completed by loading the sample in a small gapbetween two surfaces. The geometry of the two surfaces may vary, but frequently includeparallel plates, a cone and plate, or a cup and bob configuration as shown in Figure 2.5.Figure 2.5: Rheometer geometries. Schematic of bob and cup, cone and plate and parallelplate rheometer geometries.For low viscosity samples it is common to use larger plates or a cup and bob geometry togenerate more torque. For highly viscous samples, smaller geometries are used. Additionally,the angle of the cone geometry can also be changed.19The working equations for these geometries are commonly found in literature. As thecone and plate geometry was used for this research the working equations for steady shearare presented in Equation 2.5 through Equation 2.7.⌧ =3T2⇡R3(2.5)˙ =⌦⇥0(2.6)⌘ =⌧˙=3T⇥02⇡R3⌦(2.7)Where• T is the measured torque• ⌦ is the rotational speed• R is the radius of the cone geometry• ⇥0 is the cone angleSmall amplitude oscillatory shear (SAOS) is often used to investigate dynamic rheologicalproperties, including the storage and loss modulus (G0, G00). When applying SAOS, thesample is subjected to a small sinusoidal oscillatory strain described by Equation 2.8(t) = 0 sin!t (2.8)20where• 0 is the strain amplitude• ! is the frequency• t is the timeAt low strain amplitudes, the sample will remain in the linear viscoelastic region, wherestress is independent of the strain amplitude. The resultant sample stress is then measuredas a function of time and can be written according to Equation 2.9.⌧(t) = ⌧0 sin!t+  (2.9)Where• ⌧0 is stress amplitude•  is the phase shiftThe measured stress amplitude, imposed strain amplitude, and phase shift can then beused to calculate the dynamic moduli according to Equation 2.10 and Equation 2.11.G0 =⌧00cos  (2.10)G00 =⌧00sin  (2.11)For a cone and plate geometry, the working equations for dynamic moduli are describedby Equation 2.12 and Equation 2.1321G0 =3⇥0T0 cos 2⇡R30(2.12)G00 =3⇥0T0 sin 2⇡R30(2.13)Where• T0 is the measured torque• 0 is the angular amplitude2.5.2 Rheometry of synovial fluidSeveral methods of joint lubrication have been proposed for synovial joints, including hy-drodynamic, elastohydrodynamic, and boundary lubrication [84]. It is likely that all threeregimes will contribute to knee joint lubrication and protection with varying degrees ofimportance under di↵erent types of motion and applied forces [72]. Under regimes of hy-drodynamic and elastohydrodynamic lubrication the bulk rheology of the lubricating fluidplays an important role in reducing friction [84]. It is therefore necessary to develop anunderstanding of synovial fluid and hyaluronic acid bulk rheology in order to evaluate theire↵ectiveness in joint lubrication.When measuring the viscosity and viscoelastic properties of synovial fluid, there are afew fluid specific challenges that should be considered. First, the viscosity of synovial fluidcan vary greatly between subjects. In the lowest viscosity samples, at low applied shearstrains and frequency of oscillation, it can be dicult to generate large amounts of torquewhich results in increased noise in experimental data. Typically this problem would bemitigated by using a larger diameter plate or switching from a cone and plate geometry toa cup and bob geometry, however synovial fluid samples are often very low in volume which22limits the size of geometry that can be used.Second, synovial fluid is susceptible to evaporation, in particular when it is measured atphysiological temperature. Therefore, care should be taken to reduce the time of experimentsand limit the extent of evaporation. Additional measures may be taken to reduce evaporationby humidifying the enclosure around the sample or using a base plate with a water trap.Finally, synovial fluid contains components such as proteins and lipids which may becomesurface active and form an interfacial layer and the plate surfaces [72]. This e↵ect may alsocontribute to variability in rheology measurements.2.5.3 LimitationsAlthough rotational rheometers are the standard method used for measuring rheologicalbehaviour, they are limited in several ways. First, most rheometers can apply shear rateson the order of 102 to 103 s1 [75]. For synovial fluid, it is estimated that shear rates inthe knee may reach as high as 105 s1 [50]. The maximum shear rates that can be studiedwith a rotational rheometer will depend on the fluid and the geometry used, however it isgenerally limited by the presence of turbulence or inertial e↵ects [10].A second limitation of conventional rheometry is the required sample volume. In orderto generate sucient torque large geometries must be used for low viscosity samples. Theselarge geometries often require several millilitres of sample to be correctly loaded. For syn-ovial fluid, sample volumes can be very small and completing replicate runs with a geometryrequiring several millilitres of sample per run may be unfeasible.2.6 Microfluidics2.6.1 BackgroundG. Whiteside defines microfluidics as the study of fluid flow through channels with dimen-sions on the order of 100 microns or less [92]. The field of microfluidics has developed awide range of applications ranging from high resolution separations, chemical reactions, and23molecule detection in research fields including chemistry, biochemistry, microbiology andengineering [73]. These applications are often combined into a single, continuously flowingsystem to create what is referred to as a lab on a chip in order to carry out multiple op-erations on a single microfluidic chip [73]. Microfluidic devices have also been used widelyin the field of micro-electro-mechanical systems (MEMS), which itself is applied in manydi↵erent scientific fields. More recently, microfluidic devices have been used for rheologicalmeasurements of both synthetic and biological, and Newtonian and non-Newotnian fluidswhere conventional rheometer measurements may be limited in some way [74][76].Microfluidic devices are small in scale and therefore require little lab space to operate.Fabrication is in general, inexpensive and can be completed over a short period of time[74][92]. The small size of the device is also an advantage when sample volume is limitedas the devices generally require less volume than conventional methods (such as a conven-tional rotational or extensional capillary breakup rheometer). A less immediately apparentadvantage of microfluidic devices over tradational rheometry is the ability to achieve highshear rates while maintaining laminar flow due to the small scale of the channel dimensions[74][76].2.7 Microfluidics and rheologySeveral types of microfluidic rheometers have been developed to study steady shear, un-steady behaviour, and extensional rheology. These devices have been used for both Newto-nian and non-Newtonian fluids and results have shown general agreement with conventionalrheometry measurements [46][9][97].Microfluidic rheometers have started to see use in the field of biofluids for measuringviscosity of dilute solutions of hyaluronic acid representative of synovial fluid [45]. This isin large part due to the limited sample volumes of blood and synovial fluid available foranalysis that make use of a conventional rheometer dicult. The low viscosity of thesebiofluids is also a motivator for use of microfluidic devices, where high shear measurementscan be obtained without creating turbulent flow that would lead to error with a conventional24rheometer [45][47].Pressure and flow rate measurement instruments used to obtain viscosity estimates formicrofluidic devices are generally compatible microrheometers designed to measure shear,extensional or dynamic rheological properties, whereas conventional rheometry requires dif-ferent machines in order to measure shear and extensional properties [76][75]. The use of asingle experimental setup to achieve di↵erent types of measurement is also an advantage interms of conserving laboratory bench space and costs.2.7.1 Rheological measurementsIn general, obtaining viscosity measurements from a microfluidics device requires eithercontrolling the pressure drop across the device and measuring the flow rate, or by con-trolling the flow rate and measuring the resultant pressure drop [76]. Various approachesexist to control these parameters, however the most common is to control either flow orpressure with a syringe pump while measuring the other. Pressure measurements can beobtained with a variety of pressure sensors, including small chips built into the channel orin line pressure sensors [75][54]. Flow measurements can be obtained through in-line flowmeasurement devices or from micro-PIV measurements [76][9][45]. There are various advan-tages to operating a micro-rheometer under each set of conditions which include cost, accessto laboratory equipment such as a high resolution microscope and micro-PIV software orappropriate pressure sensors [92][76].Channel geometryTo measure simple shear viscosity it is most common to use a straight channel and toobtain various shear rate measurements by manipulating the flow rate. These channels canbe capillary channels, rectangular channels, or square channels [76][75][54]. The use of a Tjunction has also been proposed to obtain a range of shear rates in a single measurementby creating a range of shear rates as fluid is forced around the corner in the T junction[9]. A serpentine channel has also been developed for measurement of relaxation times inviscoelastic fluids [97].252.7.2 Extensional rheologyIn addition to shear rheology, microfluidic devices have also been used to obtain measure-ments of extensional viscosity for non-Newtonian fluids. Distinct geometries are used forextensional microrheometers compared to shear microrheometers. These geometries typ-ically cause extensional strains by creating a stagnation point where polymer moleculesbecome entrapped. The most common geometry is a cross-slot, which has been extensivelystudied by Haward et al. [46][45][47]. This geometry has been used to study both synthetic,dilute polymer solutions as well as biofluids.In the extensional microrheometeres developed by Haward et al. the flow kinematics aremonitored with micro particle image velocimetry (micro-PIV), while sample birefringence isused to evaluate sample deformation. Briefly, birefringence is the optical property of a ma-terial having a refractive index that depends on the polarization and propagation directionof incident light. In dilute polymer solutions, the amount of birefringence is proportional tothe extensional strain and can thereby be used to indirectly measure extensional viscosity[45]. In the setup used by Haward et al. circularly polarized monochromatic light is passedthrough the sample, then through a liquid crystal compensator and then finally captured bya CCD array. Complex data processing algorithms are then required to combine multipleimage frames into a single full field map of light retardation and orientation angle [45].2.8 Device fabricationMicrofluidics devices are commonly fabricated with poly(dimethylsiloxane) (referred to asPDMS) with a soft lithography process [83][76][5]. This process is widely practised forit’s simplicity, minimal use of specialized facilities such as a cleanroom, and for the speedat which prototypes can be replicated [5][22]. PDMS also has excellent optical propertiesthat make it a suitable material for experimentation requiring visualization of flow, such asmicro-PIV [66]. Other techniques for manufacture are also employed to a lesser degree, suchas etching of silicon or glass or precision machining [5][9].262.8.1 Master fabricationMasters are a smooth surface embossed with the negative of the channel geometry. Em-bossing can be achieved through a variety of techniques including deep reactive ion etching(DRIE) of silicon, wet etching of silicon, and photolithography. Depending on the techniqueemployed, quality of pattern transfer, surface roughness and cost may all be a↵ected andthe particular technique used should be chosen based on the application [31].The current work employs the use of SU-8 photolithography. SU-8 is a photo-curableepoxy that is deposited onto a smooth substrate such as a silicon wafer. The desired channelgeometry is designed with a CAD software and then sent to a photomask producer that willprint the mask. SU-8 is spin coated to a desired thickness on a silicon wafer. The mask isthen held over the SU-8 during UV exposure to selectively cure the epoxy to the substrate[5]. Once cured, the substrate and SU-8 are rinsed with the SU-8 developer to wash awayuncured SU-8, leaving behind only the embossed geometry. This master can then be usedas a mould for many PDMS replicas provided that it is kept clean [31].2.8.2 Replica mouldingPDMS replica moulding is a simple, inexpensive, and fast process that enables fabricationof microfluidic devices with relatively little laboratory equipment or resources [5][66].PDMS base and curing agent are mixed thoroughly and degassed to remove any airbubbles introduced during the mixing process. The PDMS is then poured over the embossedmaster and cured. PDMS can be cured at room temperature over the course of several days,or it can be cured in hours if cured in an oven. Once cured, the PDMS mould can thenbe removed from the master. Removal of the PDMS should not damage the master, whichallows reuse of the master to produce many PDMS replicates.2.8.3 Sealing the channelOnce removed from the master, the PDMS must be bonded to another surface to seal themicrofluidic channel. This can be accomplished by bonding to a variety of materials, and is27commonly completed by bonding to glass or another PDMS layer [77][57][66].When bonding PDMS to a glass slide, several techniques can be used depending onavailable laboratory equipment. Treatment of adjoining PDMS and glass surfaces with airplasma prior to bringing the two surfaces into contact has been demonstrated to create apermanent bond with shear strength as high as 70 psi (480 kPa) [13]. Air plasma treatmenthas also been used for PDMS-PDMS bonding and was demonstrated to achieve a bondstrength of 500 kPa [29].2.8.4 Connecting to the chipConnecting the macro scale fluid reservoir with the micro scale of the microfluidics channelis a non trivial step in the development of the device. A typical setup will include a syringeas the fluid reservoir. The syringe can then be connected to a needle which is subsequentlyconnected to tubing. The second crucial junction will occur at the chip interface. Here itis common to again use a needle to connect tubing to the microfluidic channel. There maybe additional connections between the syringe and the chip for pressure sensors or otherinstrumentation [39].For the above method, entrance points to the chip must be cored out at an appropriatediameter to accommodate the gauge of the connecting needle. Punching tools can be used tocore the entrance and exit points to the chip prior to sealing the channel. Once the setup iscomplete, it is common to reinforce all seals by curing PDMS or another compatible sealantaround all connections. For low pressure applications this step may be unnecessary. Forhigh pressure applications, needle to tubing adapters may be required to ensure a strongseal is obtained [39].Should flow properties be measured indirectly by measuring pressure drop across thechannel, special consideration should be taken to avoid unnecessary pressure drop in tubingjunctions and at the entrance and exit of the device. Pressure drop can be reduced byavoiding needles with elbows, using a larger diameter needle, and minimizing the numberof junctions. Additionally, the pressure sensor should be placed as close to the device aspossible to minimize the pressure drop in tubing before flow reaches the device.28Chapter 3Rheological Evaluation of a NovelViscosupplement3.1 IntroductionAs no standard viscosupplement formulation has been widely accepted as an ideal treatmentfor osteoarthritis, there is a need for further improvement upon existing supplements. Aspreviously discussed, many novel improvements to viscosupplements have been investigatedsuch as controlling rheological properties with nano particles [33] and increasing treatmente↵ect lifetime [28]. One area which has not yet been extensively studied is the possibility ofdeveloping a viscosupplement with anti-inflammatory properties.Recently, it has been demonstrated that while high molecular weight hyaluronic acid isprotective in joints, low molecular weight hyaluronic acid is pro-inflammatory in a numberof systems [6][19][20]. The hyaluronic acid derivatives in this study were developed with thepurpose of investigating pro-inflammatory response of low molecular weight hyaluronic acidsamples and to synthesize a hyaluronic acid derivative with anti-inflammatory properties [6].Reacylation with butyryic anhydride resulted in partially substituted low molecular weighthyaluronic acid which had anti-inflammatory properties. This was not the case with otherN-acylations [6].The objectives of this study are to investigate the rheological properties of these low29molecular weight hyaluronic acid derivatives with the purpose of better understanding thee↵ects of the chemical preparation on the final solution rheology and to evaluate the suit-ability of the anti-inflammatory derivative as a potential treatment for mild to moderateosteoarthritis.3.2 Methods and materials3.2.1 Hyaluronic acid derivativesParent (intact) hyaluronic acid derived from streptococcus equi along with four, low molec-ular weight derivatives were studied. The samples were prepared according to the proceduredescribed by Babsola et al. [6] at Queen’s University, Kingston, ON and were shipped to theUniversity of British Columbia, Vancouver, BC in dehydrated form. Briefly, the derivativesamples were formed by first deacetylating the parent compound. One sample was left assuch, while two other samples were reacetylated. The final sample was butyrylated andwas found to have anti-inflammatory properties [6]. The hyaluronic acid derivatives wereprepared at Queen’s University according to the procedure described by Babasola et al. [6].The hyaluronic acid derivatives, preparation method and sample number are presented inTable 3.1.Table 3.1: Hyaluronic acid derivatives and the corresponding sample number used for ref-erence throughout this report.Compound Preparation SampleHyaluronic acid (parent compound) - Sample 1DHA Deacetylated hyaluronic acid1 Sample 2AHA-1 Reacetylated DHA1 Sample 3AHA-2 Reacetylated DHA2 Sample 4BHA Butyrylated DHA1 Sample 51Deacetylated by hydrazinolysis2Deacetylated with NaOHIt should be noted that the deacetylation step in the formulation of all derivatives in30this study causes bond cleavage, resulting in a decrease in molecular weight from parent toderivative compounds [6].In preparation for rheological testing each compound was diluted to a concentrationof 5 mg/mL in phosphate bu↵er solution (PBS) 7.2 pH at room temperature and wassubsequently mixed for 3 hours using a magnetic stirrer. Rheological tests were conductedimmediately following mixing.3.2.2 MethodsAll measurements were conducted using a Malvern Kinexus Ultra rheometer with a 50 mm,1 degree cone and plate geometry. A sample volume of 0.58 mL was required for loading therheometer with the given geometry. All tests were conducted at 37 °C. To reduce sampleevaporation, a base plate with a water trap was used. The sample was then covered withaluminium and plastic covers respectively.Three replicate runs were completed for each test. The final results presented are theaverage of all three runs completed.3.2.3 Shear rheologyAll samples were pre-sheared at 0.1 Pa for 1 minute followed by a zero shear rest period of 2minutes to remove any loading e↵ects and shear history. Shear viscosity was then evaluatedover the shear range of 0.01 to 1000 s1. It is important to note that this shear range doesnot describe the entire spectrum of shear rates observed in knee joints, however viscosity atshear rates outside this range cannot be measured accurately with the given apparatus dueto low sample viscosity, sample inertia, and inherent limitations in conventional rheometry[42]. This shear range does describe the range of shears typically observed during walkingand the transition to running [37].The shear viscosity data for all samples was fitted with the Carreau-Yassuda model.The Carreau-Yassuda model has previously been used to model both synovial fluid andviscosupplement viscosity, and describes a zero shear plateau, a shear thinning region andan infinite shear plateau [1][15]. The Carreau-Yassuda model is given in Equation 2.4.31Estimates for ⌘0 and ⌘1 are determined experimentally while the remaining parame-ter estimates for the Carreau-Yassuda were obtained using the curve fitting tool cftool inMATLAB.3.2.4 Viscoelastic propertiesPrior to commencing oscillatory measurements all samples were pre-sheared at 0.1 Pa for 1minute followed by a zero shear rest period for 2 minutes to remove any shear history fromthe shear measurements.Oscillatory flow was used to investigate the viscoelastic properties of the hyaluronic acidderivatives. Strain amplitude sweep measurements were first conducted at 1 Hz to identifythe linear viscoelastic region for each sample. 5 % strain was determined to fall withinthe linear viscoelastic region for all compounds investigated and was therefore used for allfrequency sweep measurements. A typical amplitude sweep curve is shown in Figure 3.1.Figure 3.1: Sample 2 amplitude sweep. The linear viscoelastic region is clearly shown up to10% strain.Following strain amplitude sweep measurements a pre-shear at 0.1 Pa for 1 minuteand a zero shear rest period of 2 minutes was used to remove any history e↵ects prior to32commencing frequency sweep measurements.Frequency sweeps were completed over an oscillation frequency range of 1 to 10 Hzin order to obtain estimates for the storage and loss moduli (G0 and G00). For the lowerviscosity samples accurate measurements could only be obtained over a smaller range offrequencies as was the case for Samples 4 and 5.Averages from three replicate runs were used to plot the data. Where applicable, theaveraged data sets were used to determine the crossover frequency.3.3 Results3.3.1 Shear rheologyThe shear viscosity is plotted as a function of shear rate for all samples in Figure 3.2. Thesolid lines represent the Carreau-Yassuda model fits.Figure 3.2: Viscosity as a function of shear rate. A Newtonian plateau is observed at lowshear rates for Sample 1 with shear thinning behaviour at higher shear rates. Samples 2through 5 were not observed to demonstrate a zero-shear plateau in the investigated range.Solid lines represent the Carreau-Yassuda fit.Non-Newtonian shear thinning behaviour is observed for all samples, however Sample 133(parent hyaluronic acid) is observed to exhibit distinct behaviour from the four derivativesamples. At low shear rates, Sample 1 is found to exhibit a Newtonian plateau while athigher shear rates exhibits moderate shear thinning behaviour.The four derivative samples exhibited shear thinning behaviour at low shear with ahigh shear Newtonian plateau. The slope of the shear thinning behaviour is similar forall derivative samples. The major di↵erence among derivative samples was in the viscosityat low shear rates with Sample 2 (DHA) being most viscous. Of particular interest wasSample 5 (BHA, anti-inflammatory), which displayed along with Sample 4 (AHA-2) thelowest viscosities over the investigated range. The Newtonian plateau at high shear rates wassimilar for all samples. Parameter estimates for the Carreau-Yassuda model are presentedin Table 3.2.Table 3.2: Carreau-Yassuda model parameters for hyaluronic acid shear viscosity data. Allparameters were found to be significant at the 95% confidence level.a n  ⌘0 ⌘1Sample 1 1.02 0.530 0.03 0.08 0.003Sample 2 55.00 -0.101 99.70 32.00 0.002Sample 3 20.00 -0.044 8.36 0.70 0.002Sample 4 10.62 0.026 110.60 0.50 0.003Sample 5 4.68 -0.030 122.20 1.20 0.002Values for ⌘0 and ⌘1 are estimated from experimental data, while the remaining pa-rameters are estimated with the MATLAB curve fitting tool cftool. All parameters werefound to be significant at the 95% confidence level suggesting the models are a good fit forthe data over the observed shear rates. For the derivative samples, no low shear Newtonianplateau was observed over the investigated range of shear rates, therefore an estimate wasmade for ⌘0 based on the trend of the observed data. Similarly, for the parent hyaluronicacid solution an estimate was made for ⌘1. Although using this procedure allowed us toobtain model fits with a good representation of the data over the investigated shear rangethe parameter estimates may lack physical significance due to the uncertainty introducedby estimating values for ⌘0 and ⌘1.The Sisko model may provide a better fit to the derivative samples as it describes shear34thinning behaviour at low shear rates with a Newtonian plateau at high shear rates [1]. TheEllis model may provide a more reliable fit to the parent compound data as it describesa low shear Newtonian plateau with shear thinning behaviour at elevated shear rates [1].These models were not applied to the present work as to the best of our knowledge, theyhave not previously been used to describe viscosity behaviour of hyaluronic acid or synovialfluid solutions and therefore provide no basis for comparison with values elsewhere in theliterature.3.3.2 Viscoelastic propertiesThe dynamic moduli for hyaluronic acid and the four derivatives are plotted as functions ofoscillation frequency in Figure 3.3a through Figure 3.3e.(a) Storage and loss moduli plotted as functions of oscillation frequency for Sample 1. A crossoverof storage and loss moduli is observed at approximately 10 Hz.35(b) Storage and loss moduli plotted as functions of oscillation frequency for Sample 2. Gel likebehaviour is observed over the investigated frequency range.(c) Storage and loss moduli plotted as functions of oscillation frequency for Sample 3. Gel likebehaviour is observed over the investigated frequency range.36(d) Storage and loss moduli plotted as functions of oscillation frequency for Sample 4. A crossoverfrequency of approximately 4.2 Hz is observed.(e) Storage and loss moduli plotted as functions of oscillation frequency for Sample 5. A crossoverfrequency of approximately 2.8 Hz is observed.Figure 3.3As expected, Sample 1 exhibited clear viscoelastic behaviour, with a crossover frequencyat approximately 10 Hz. Samples 2 and 3 both exhibited purely gel behaviour in the investi-gated range of frequencies while Sample 4 exhibited a crossover frequency at approximately4.2 Hz. Sample 5 exhibited a crossover frequency at approximately 2.8 Hz. A summary ofthe viscoelastic behaviour of all samples along with those for three commercial viscosupple-37ments is provided in Table 3.3 [15]. Note that dynamic moduli are commonly reported at2.5 Hz as it roughly corresponds to the frequency of running [64].Table 3.3: Crossover frequency and modulus, dynamic moduli at 2.5 Hz, and solutionbehaviour. Note that no crossover frequency was observed for Sample 2 and Sample 3.The viscoelastic properties of three commercial viscosupplements are also presented forcomparison [15].fc (Hz) Gc (Pa) G02.5 (Pa) G”2.5 (Pa) Solution BehaviourSample 1 10.0 2.3 0.19 1.04 Liquid and gelSample 2 - - 0.58 0.17 GelSample 3 - - 1.18 0.24 GelSample 4 4.5 0.20 0.129 0.15 Liquid and gelSample 5 2.8 0.06 0.013 0.063 Liquid and gelOrthovisc® 0.398 - 111.2 61.48 Liquid and GelSuplasyn® - - 3.36 10.78 GelSynvisc® - - 118.1 22.46 Gel3.4 Discussion3.4.1 Shear rheologyAs observed in Figure 3.2, the derivative samples were found to have lower viscosity thanthe parent hyaluronic acid. It is believed that the bond cleavage caused by the deacetylationprocess is primarily responsible for the decrease in viscosity and the disappearance of theNewtonian plateau at low shear rates. Sample 2 was found to have a greater viscositythan Samples 3, 4 and 5, suggesting that the reacetylation and butyrylation processes mayalso have a thinning e↵ect on the compounds. Sample 3 was found to have a greaterviscosity than Sample 4, suggesting that deacetylation via hydrazinolysis may cause a lesssignificant decrease in viscosity than when NaOH is used to perform the deacetylation. Asshown in Table 3.2 and Figure 3.2, the infinite shear viscosity of all derivative samples arenearly identical, suggesting the processing steps involved in each compound formulationhave minimal e↵ect on the shear viscosity at high shear rates.The shear viscosity of the hyaluronic acid derivatives was found to be significantly lowerthan those reported for commercial viscosupplements [15][35][37]. This di↵erence is primar-38ily attributed to the low molecular weight of the hyaluronic acid derivatives and the relativelylow concentration at which the samples were studied [35][37]. The molecular weight andconcentration of several commercial viscosupplements are compared to those values for thesamples in this study in Table 3.4.Table 3.4: The molecular weight and concentration of hyaluronic acid derivatives and severalcommercial viscosupplements [6][37][14][71][4].Viscosupplement Molecular weight (kDa) Concentration (mg/mL)HA derivs1 30-214 5Hyalgan® 500-700 10Hyalubrix® 1500 15Durolane® 1000 20Synvisc® 6000-7000 8Suplasyn® 730 20Orthovisc® 1000-2900 151Hyaluronic acid derivatives from this study3.4.2 Viscoselastic propertiesThe deacetylation process appears to have reduced the viscoelastic properties of the deriva-tive samples as Sample 1 was found to have greater dynamic moduli than Samples 2 through5 over the investigated range of frequencies. Samples 2 and 3 were found to exhibit purelygel-like behaviour in the range of studied frequencies while the remaining samples displayeda cross-over frequency and transition from liquid to gel behaviour. This may indicate thatdeacetylation via hydrazinolysis causes gel behaviour while deacetylation by NaOH does not.This may also suggest that the butyrylation process causes a return towards combined liq-uid and gel behaviour. This may suggest that moieties such as N-acetyl and butyryl groupshave an e↵ect on the viscoelastic behaviour of the hyaluronic acid derivatives. Sample 5 wasobserved to exhibit the smallest dynamic moduli, suggesting that butyrylation may furtherreduce sample viscoelasticity beyond the initial reduction caused by deacetylation.As was observed with the shear viscosity, the oscillatory behaviour of the investigatedsamples was found to be significantly less viscoelastic than commercially available viscosup-39plements and other hyaluronic acid solutions [15][37][28][24]. This result is attributed to thelow molecular weight and concentration of hyaluronic acid in the derivative samples studiedhere.Similarly to shear viscosity measurements, the viscoelastic measurements demonstratethat sample rheology is influenced by method of deacetylation and butyrylation. Thissuggests that rheology of hyaluronic acid solutions is notexclusively dependent on averagemolecular weight, concentration, and degree of cross-linking.3.4.3 Suitability of anti-inflammatory hyaluronic acid as a visco-supplementSample 5 (butyrylated, anti-inflammatory) was found to exhibit shear thinning behaviourand viscoelastic behaviour, traits present in both healthy synovial fluid and commercialviscosupplements [35][15][37]. This is a promising result as prior studies have shown thatviscosity and viscoelasticity can be modulated by adjusting the solution concentration anddegree of cross-linking [35][24]. Therefore, even though at the investigated concentrationof 5 mg/mL the viscosity and viscoelastic properties were significantly lower than thosefor three commercial viscosupplements, it is expected that Sample 5 could be modifiedto enhance the viscosity and viscoelastic properties to fall within the range for a typicalviscosupplement[24][86][35][33].In addition to potentially suitable viscous and viscoelastic properties, Sample 5 alsopossesses anti-inflammatory properties which may reduce the incidence of flare up [6]. Thee↵ect of flare up in patients with knee osteoarthritis has been observed to significantlya↵ect synovial fluid properties such as hyaluronic acid molecular weight, concentration, andprotein concentration [27].Incidents of flare up may be detrimental long-term to joint health by reducing the vis-cosity and viscoelasticity as a result of a decrease in concentration and molecular weightof hyaluronic acid [35][27]. It is therefore expected that a viscosupplement that adequatelylubricates and protects an osteoarthritic knee while also providing anti-inflammatory prop-erties would be advantageous in the treatment of mild to moderate osteoarthritis.403.5 ConclusionsThe shear viscosity of four low molecular weight hyaluronic acid derivative samples wasinvestigated. All samples were found to exhibit shear thinning behaviour which was suc-cessfully fitted with the Carreau-Yassuda model. Viscosity of the derivative samples wasfound to be lower than for the parent compound. The decrease in viscosity from parent toderivatives is largely attributed to the bond cleavage and subsequent decrease in molecularweight that occurs during the deacetylation step of sample preparation. Additionally, Sam-ple 5 was found to have a lower viscosity than Sample 2, suggesting that the butyrylationstep further decreases the viscosity after deacetylation has taken place.The oscillatory behaviour of the samples was also investigated. Viscoelastic behaviourwas observed for all samples. Samples 2 and 3 were found to exhibit purely gel behaviourover the investigated range. This result suggests that the method of deacetylation and thebutyrylation step may a↵ect the viscoelastic behaviour.Variation in rheology across derivative compounds demonstrates that the rheology ofhyaluronic acid solutions is not only dependent on polymer concentration, average molecularweight and degree of cross-linking but may also be a↵ected by method of deacetylation andbutyrylation.The viscosity and viscoelasticity of all derivative samples were found to be lower than val-ues reported elsewhere for commercially available viscosupplements. It is believed that thisresult is primarily caused by a decrease in hyaluronic acid molecular weight and concentra-tion. It is expected that by adjusting the hyaluronic acid concentration and degree of crosslinking, the viscosity and viscoelasticity of the investigated samples could be modulated tomore closely resemble properties of commercial viscosupplements.By adjusting the concentration and degree of cross-linking, it is believed that Sample5 could be used as an e↵ective viscosupplement. The e↵ect of flare up in cases of kneeosteoarthritis has been proven to be detrimental, and it would therefore be highly advanta-geous to use a viscosupplement which is anti-inflammatory.41Chapter 4Case Study: Glucosamine Sup-plementation and Synovial FluidRheology4.1 IntroductionIt has been suggested that oral glucosamine may slow or prevent the progression of os-teoarthritis by stimulating the production of proteoglycan-4 and articular cartilage [91][62].Glucosamine is partially absorbed in the small intestine and it is possible that some of the ab-sorbed glucosamine can reach osteoarthritic joints to stimulate production of proteoglycan-4and articular cartilage. A separate hypothesis which may explain symptom relief caused byglucosamine supplementation proposes that oral glucosamine supplementation may stimu-late enhanced production of hyaluronic acid as glucosamine is a rate limiting pre-cursor forsynthesis of hyaluronic acid [65]. This theory provides a clear motivation for studying thee↵ect of oral glucosamine supplementation on synovial fluid viscosity.Glucosamine supplementation has been studied extensively as a treatment for osteoarthri-tis. Most common measures of treatment ecacy are treatment scales that assess the treat-ment’s e↵ect on quality of life through impact on daily activities and patient feedback onpain levels [25][26][38]. Additional studies have assessed the e↵ectiveness of glucosamine42supplementation through radiographic evidence of joint space narrowing, osteophyte forma-tion and cartilage breakdown. These studies have produced conflicting results on the ecacyof glucosamine treatment, the required dose, and the required duration of treatment beforeany clinical benefits can be observed [17][87][79][88].The e↵ect of glucosamine supplementation on synovial fluid rheology has not been stud-ied comprehensively. In a study by Matsuno et al. the viscosity of synovial fluid was studiedat a single shear rate of 40 s1 rather than fully characterizing the shear thinning behaviourbefore and after, while viscoelastic properties were not studied at all [62]. Despite the linkbetween hyaluronic acid, synovial fluid, and cartilage and the hypothesized e↵ect of glu-cosamine on cartilage formation, no further studies have been presented on the e↵ects ofglucosamine on hyaluronic acid or synovial fluid. McCarty hypothesized that immediatesymptom relief in high dosage cases of synovial fluid was caused by an increase in highmolecular weight production of hyaluronic acid [65], however this theory has not yet beenvalidated.As the e↵ect of osteoarthritis on synovial fluid is a significant factor in disease progressionand joint health, there exists a need to further study the e↵ects of glucosamine supplemen-tation on synovial fluid rheology. Should a significant e↵ect exist, it may provide insightinto some of the confounding factors that have led to conflicting results in previous clinicaltrials, and may elucidate a mechanism by which reported symptom relief functions.The present work is intended to establish a framework for future, more exhaustive studyof the e↵ects of oral glucosamine supplementation on synovial fluid rheology. The presentwork will describe appropriate collection and handling of synovial fluid, and a methodol-ogy used to obtain shear rheology measurements. Preliminary results are discussed for onepatient beginning glucosamine supplementation and a second patient discontinuing treat-ment. Changes in the synovial fluid will be identified as areas requiring future study. Ascomplimentary results WOMAC surveys are also discussed. Not included in this study is acharacterization of oscillatory and extensional rheology of synovial fluid. As patients werevoluntarily changing their treatment regimen, blinding could also not be incorporated.434.2 Methods and materials4.2.1 Ethics approvalsPrior to commencing study, a valid ethics approval was acquired from the University ofBritish Columbia Board of Ethics and the Vancouver Coastal Health Institute for aspirationof synovial fluid from individuals with mild to moderate osteoarthritis. The ethics approvalis valid for subjects currently taking glucosamine supplements and willing to discontinuetreatment, and for participants not taking a glucosamine supplement but willing to begintreatment. A second ethics approval was obtained from the University of British ColumbiaBoard of Ethics and the Vancouver Coastal Health Institute for studying the rheology ofaspirated synovial fluid samples.4.2.2 Patient recruitmentPatients were recruited from the family practice clinic of Dr. Ezra Kwok. Informed consentforms were signed by all eligible participants prior to enrolment. Eligible participants werebetween the ages of 30 and 85 years, clinically diagnosed with moderate to severe kneeosteoarthritis, and are recommended for synovial fluid aspiration to relieve joint discomfortcased by e↵usion. Subjects will be excluded from participation who are clinically diagnosedwith inflammatory arthritis, have had previous joint surgery in the study knee or havereceived an injection to the study knee within the past 12 months. Only mild to moderatecases of osteoarthritis will be included in the study. Patient treatment was modified byeither discontinuing glucosamine supplementation or starting supplementation immediatelyfollowing the first visit.4.2.3 Synovial fluid collectionAt the baseline visit, after informed consent has been obtained and patients have beenassessed against the inclusion and exclusion criteria, patient demographics were recordedalong with the degree of joint e↵usion and a WOMAC questionnaire was completed. After44noting the degree of joint e↵usion, an experienced physician aspirated a sample of synovialfluid later used for viscosity and viscoelasticity analysis.Follow up visits were scheduled for 4 weeks after the initial visit. At the follow upvisit, a second WOMAC questionnaire was completed and the degree of joint e↵usion wasreported prior to aspirating a synovial fluid sample from the knee under study. The syn-ovial fluid sample was then used to compare viscosity and viscoelasticity with the baselinemeasurement.4.2.4 RheologyAll measurements were conducted using a Malvern Kinexus Ultra rheometer with a 50 mm,1 degree cone and plate geometry. A sample volume of 0.58 mL was required for loading therheometer with the given geometry. All tests were conducted at 37 °C. To reduce sampleevaporation, a base plate with a water trap was used. The sample was then covered withaluminium and plastic covers respectively to further prevent evaporation. All samples weremeasured on the same day as collection. Three replicate runs were completed for each test.The final results presented are the average of all three runs completed. The variation inreplicate runs ranged between 20 % at low shear rates and 1% at high shear rates.Data is presented prior to and post treatment change for two patients. Patient A discon-tinues glucosamine supplementation while Patient B begins glucosamine supplementation.All samples were pre-sheared at 0.1 Pa for 1 minute followed by a zero shear rest period of2 minutes to remove any loading e↵ects and shear history. Shear viscosity was then evaluatedover the shear range of 0.1 to 1000 s1. It is important to note that this shear range doesnot describe the entire spectrum of shear rates observed in knee joints, however viscosity atshear rates outside this range cannot be measured accurately with the given apparatus dueto low sample viscosity, sample inertia, and inherent limitations in conventional rheometry[42]. This shear range describes the range of shears typically observed during walking andthe transition to running [37].454.2.5 WOMAC questionnaireThe WOMAC questionnaire was completed at each patient visit and used as complimen-tary data to the synovial fluid rheology. The WOMAC questionnaire was developed as amethod for assessing pain, sti↵ness, and physical function in patients with hip and or kneeosteoarthritis. The WOMAC questionnaire was administered by Dr. Ezra Kwok at thebeginning of each patient visit. The questionnaire was completed orally. The Likert Scalewas used to classify responses as none, mild, moderate, severe and extreme correspondingto scores of 0-4 respectively. Maximum scores in subcategories are 20, 8, and 68 for pain,sti↵ness and physical function respectively [53]. For this study, the overall WOMAC scorewas calculated by summing the scores from the three subcategories with a maximum scoreof 96.4.3 Results4.3.1 Shear viscosityThe shear viscosity as a function of shear rate of synovial fluid from Patient A is presentedin Figure 4.1.46Figure 4.1: Patient A viscosity as a function of shear rate. Shear thinning behaviour isobserved both prior to and following glucosamine discontinuation.Shear thinning behaviour is observed both prior to and following discontinuation ofglucosamine supplementation. Variability is observed to be greatest at low shear rates forboth samples. At shear rates of approximately 1 s1 and greater the post treatment sampleis observed to have a greater viscosity although at shear rates of approximately 100 s1 andhigher there is overlap in sample confidence intervals (defined as the measured value ± thestandard deviation).The shear viscosity as a function of shear rate of synovial fluid from Patient B is presentedin Figure 4.2.47Figure 4.2: Patient B Viscosity as a function of shear rate. Shear thinning behaviouris observed Prior to taking glucosamine and after taking glucosamine supplements. Nostatistically significant di↵erence is observed between treatment and post treatment samples.Shear thinning behaviour is observed both prior to and following commencement ofglucosamine supplementation. The viscosity is found to decrease after one month of glu-cosamine supplementation, however the confidence intervals, as defined above, overlap overthe entire range of investigated shear rates, suggesting the di↵erence in viscosity pre treat-ment and during treatment is not statistically significant.4.3.2 WOMAC questionnaireTotal WOMAC scores at initial and follow up visit are presented in Figure 4.3.48Figure 4.3: WOMAC scores at initial and follow up visit for Patients A and B. A decreasein WOMAC score is observed for both patients.A decrease in WOMAC score is observed for both patients. A greater improvement wasobserved for Patient A than for Patient B. Relative improvements in WOMAC subcategoryscores are presented in Table 4.1. The relative improvement is defined as the change insubcategory score divided by the maximum score attainable in the relevant subcategory.Table 4.1: Patient A and B relative improvements per WOMAC subcategory. The greatestimprovement is observed for both patients in the Sti↵ness category.Pain Sti↵ness Physical FunctionPatient A 35% 63% 35%Patient B 10% 50% 9%Sti↵ness was observed to have the greatest improvement in subcategory score for bothpatients. No categories scores were found to increase. Pain and physical function improve-ments were equal for Patient A and within 1% for Patient B.494.4 Discussion4.4.1 Shear rheologyFor Patient A, the non-treatment sample was found to have a greater shear viscosity athigh shear rates, while the glucosamine treated sample was found to have a greater viscosityat low shear rates. No significant di↵erence was observed for Patient B. These findingsimply that the a greater extent of shear thinning is observed in the treatment sample thanin the non-treatment samples. As proposed by McCarty, it is possible that this changeis caused by an increase in high molecular weight hyaluronic acid production, howeverwith just two samples it is impossible to assert whether these findings are statisticallysignificant. It is also suspected that the observed changes over the one month period ofstudy may become more pronounced at higher doses of glucosamine and over longer periodsof treatment administration.There exist a number of confounders that may influence the outcome of viscosity mea-surements, in addition to the treatment change. No studies have been published on thechanges in synovial fluid viscosity over time, therefore it is unknown if the one month pe-riod of this study is a sucient amount of time to have any e↵ect on synovial fluid viscosity.The incidence of flare-up has been previously demonstrated to alter the composition andcharacteristics of synovial fluid [27]. Patient flare up was not recorded during the course ofthis study, and therefore, it is possible that a flare up incident for one or both patients couldhave influenced the synovial fluid rheology at follow up visit. Glucosamine dosage may alsohave an impact as individuals of di↵erent body mass may require di↵erent glucosamine dosesin order for an e↵ect on synovial fluid rheology to be observed. Finally, it is possible thatthe four week study duration was an insucient amount of time for an e↵ect on synovialfluid viscosity to become apparent.504.4.2 WOMAC questionnaireThe WOMAC scores were found to improve substantially for both patients, regardless oftreatment change. In particular the greatest improvement was observed in the sti↵nesscategory. It is suspected that improvement in this score may be influenced by a reductionin synovial fluid volume due to aspiration. Pain and physical function improvements weresimilar for both patients. It is also possible that improvements in these subcategories arecaused by a relief of pressure in the joint due to synovial fluid aspiration.A major concern with the WOMAC questionnaire results is the lack of blinding to treat-ment change. As both patients were found to have significant improvements in WOMACscore regardless of treatment group, it is possible that the placebo e↵ect is a↵ecting ques-tionnaire results.4.5 RecommendationsBased on the preliminary results and findings in the present work, several recommendationsare made for future analysis of the e↵ect of oral glucosamine supplementation on synovialfluid rheology. In addition to the shear viscosity, it is recommended to investigate changes todynamic rheological properties and extensional rheology as these properties play an impor-tant role in joint protection and lubrication. Incidence of flareup should be well documentedas it’s occurrence close to a synovial fluid aspiration may have significant e↵ects on the sam-ple rheology. It is also recommended to extend the period of study past one month, in orderto ensure any slow acting changes caused by change in glucosamine treatment are detectedby the study.4.6 ConclusionsThe present work presents a framework for aspirating, handling, and measuring synovialfluid shear rheology. Valid human ethical approvals have been obtained for the aspirationand study of osteoarthritic synovial fluid which may be renewed in subsequent years for51future, more exhaustive study on the e↵ects of oral glucosamine supplementation on synovialfluid rheology. In Patient A the synovial fluid viscosity was found to be greater at lowshear rates for the glucosamine treatment samples. At high shear rates the viscosity wasfound to be greater in the non-glucosamine treatment samples. No statistically significantdi↵erence in synovial fluid viscosity was observed for Patient B. Despite these findings, thelack of additional samples make it dicult to draw conclusions regarding the e↵ect of oralglucosamine on osteoarthritic synovial fluid rheology. Additionally, the results are subjectto confounding variables that may influence results, including possible incidence of flareupand insucient time for treatment to have an e↵ect on synovial fluid rheology.Several recommendations have been made for a future, more exhaustive study, includinginvestigating changes to dynamic and extensional rheological parameters, documenting anyincidence of flareup, and extending the period of study past one month to allow for studyof potentially slow acting changes caused by a change in treatment.52Chapter 5Development of a Microfluidic Rheome-ter5.1 IntroductionMicroscale devices have recently seen increasing use for measuring rheological properties offluids [75][46]. The primary advantages of these devices are the relatively low cost of fabri-cation, and the ability to study fluids at deformation rates significantly higher than thoseof conventional rheometry (> 103s1). This is possible because laminar flow is maintainedin such small scale devices even at high shear rates. This is demonstrated by the Reynoldsnumber, Re = ⇢ULµ , which is describes the ratio of inertial to viscous forces. In the case ofthe microrheometer, the characteristic length L is suciently small that even at large fluidvelocities U the Reynolds number will remain low, indicating that flow will be dominatedby the viscous rather than inertial forces and therefore remain laminar [76][74].Various geometries exist for measuring di↵erent rheological parameters, the simplest ofwhich being a microscale capillary rheometer. In this configuration, fluid flows through asquare or rectangular channel at either a controlled pressure or controlled volumetric flowrate. In a pressure controlled device the shear viscosity can be determined by measuring thevolumetric flow rate. In a flow rate controlled device the shear viscosity can be determinedby measuring the pressure drop. Measuring the flow rate in a pressure controlled set up53is typically more resource intensive and more complex than the flow rate controlled setup,requiring the use of micro particle image velocimetry (micro-PIV) software and a high res-olution microscope with appropriate light filters [9][76]. In contrast, controlling volumetricflow rate with a syringe pump and measuring the pressure drop across the channel by plac-ing an in-line pressure sensor is inexpensive and simple [54][74]. A drawback to measuringthe pressure drop in order to estimate the shear viscosity is that no information on the flowdynamics at entrance and exit points can be obtained.Polydimethylsiloxane (PDMS), is often used as a material of construction for these chan-nels for its ease of use, compatibility with various chemicals and good optical properties thatallow flow visualization [5][66][57]. PDMS is used in a soft lithography process in order toreplicate a master containing a negative of the desired channel geometry. This process canyield results with high fidelity to the original master even on features at the nanometer scale,making PDMS an ideal material for microfluidic device construction [5][95][58]. A limitationin the use of PDMS in microfluidic devices is the material deformation that occurs at highpressures. PDMS has previously been demonstrated to deform at pressures as low as 15 kPadepending on thickness of the channel walls [44]. Therefore, careful consideration should begiven to channel dimensions, wall thickness, and intended operating pressures when usingPDMS to create a microfluidic device.The present work describes a process for developing a microfluidic rheometer used formeasuring the shear viscosity of fluids by estimating the pressure drop across the deviceas a function of imposed flow rate. SU-8 epoxy is used to produce a silicon master viaphotolithography. The master is then used for PDMS replica moulding. The PDMS replicateis bonded to a glass substrate via oxygen plasma bonding. Finally, the rheometer is validatedby estimating the viscosity of water at room temperature. This study is intended to establisha framework for rapid and inexpensive microrheometer development for study of fluidsundergoing high rates of deformation. The process described here may also be adapted tocreate new geometries with applications in extensional or dynamic rheology.545.2 Methods and materials5.2.1 Device fabricationThe device fabrication process described here employs soft lithography replica moulding ofan SU-8 epoxy master. The process has been broken down into the following sections:• Master fabrication• Replica moulding• Sealing and connecting to the chipAn overview of the fabrication process is summarized in Figure 5.1.Figure 5.1: Microfluidic device fabrication process overview. (A) SU-8 epoxy is spin coatedon a silicon wafer. (B) The SU-8 is exposed to UV light and developed, forming the SU-8master. (C) PDMS is cured onto the master. (D) The cured PDMS is removed from themaster. (E) The PDMS layer is then bonded to a glass substrate to seal the channel. Note,a similar figure and explanation is provided in an article published by Cooper et. al. [66].Briefly, SU-8 epoxy is spin coated onto a silicon wafer to a desired thickness (A). Aphotomask is then held over the layer of epoxy during a UV exposure to selectively curethe desired channel geometry. An SU-8 developer is then used to wash away uncured SU-8,leaving behind only a negative of the desired final channel geometry (B). The product at55step B is referred to as the master. PDMS is then cured onto the master (C). When fullycured, the PDMS layer is peeled o↵ of the master (D), and then bonded to a glass substrateto seal the channel (E).Master fabricationMaterials required for master fabrication include a 100 mm test grade silicon wafer (Univer-sity Wafer), SU-8 photoresist and developer (Microchem), a quartz plate required for SU-8UV exposure (Fisher Scientific) and a UV photomask (Outputcity).Photomask geometry was designed using CleWin 4 software. The CAD file was thensent to Outputcity for printing. The photomask used is shown in Figure 5.2.Figure 5.2: Photomask design. 6 parallel channels with a diameter of 100 µm and a lengthof 6 cm connect inlet and outlet reservoirs with a diameter of 4 mm.The photomask has 6 parallel channels each with a width of 100 µm and a length of6 cm. Inlet and outlet reservoirs positioned at the distal ends of each channel are circularwith a diameter of 4 mm. Reservoirs were made large compared to the channel geometriesto facilitate punching an access point to the chip. The channel length was made as large aspossible given the silicon wafer size in order to reduce the relative importance of entrance56and exit e↵ects (increasing the ratio channel length to entrance length). A channel widthof 100 µm was chosen to ensure the pressure drop was within the operating range of thepressure sensors used, and to ensure channel features could be easily defined by the SU-8epoxy UV exposure.Master fabrication was completed in the University of British Columbia Cleanroom,located in AMPEL 446. The silicon substrate was cleaned with isopropyl alcohol, methylalcohol, and distilled water successively, and was then dried with clean air in a fume hood.The silicon wafer was further dried by placing it on a hot plate at 95 °C for 15 minutes oruntil no moisture could be observed. Once dry, approximately 10 mL of SU-8 was pouredonto the silicon substrate and was spin coated at a final speed of 1750 rpm to achieve athickness of 50 µm. The SU-8 is baked by placing the silicon substrate on a hotplate at 65°C for 2 minutes followed by a second hotplate at 95 °C for 6 minutes. The photoresist isthen exposed to 150 mJ/cm2 with a Newtronix UV light source with an intensity of 8.26mW/cm2. A post exposure bake is then completed for 1 minute at 65 °C followed by 6minutes at 95 °C. The photoresist is then developed by immersing the silicon substrate inSU-8 developer and agitating for 5 - 7 minutes. The master is then rinsed with isopropylalcohol and dried with air. Finally a hard bake is performed at 150 °C for 15 minutes. Toconfirm SU-8 epoxy structure, a Dektak XT stylus profiler (Bruker) was used to measurechannel width and thickness at 7 locations along the length of each channel. The DektakXT has a measurement repeatability of 4 A˚. For the channel used in validation testing, theaverage SU-8 depth was found to be 45.2 µm with a standard deviation of 1.9 µm. Theaverage channel width was found to be 100.5 µm with a standard deviation of 1.4 µm.Variation in SU-8 thickness is likely resultant from small di↵erences in UV light intensitydelivered to di↵erent locations on the silicon wafer.Replica mouldingPDMS base and curing agent (Sylgard 184) were mixed by hand in a 10:1 weight to weightratio for approximately 1 minute. The mixture was then placed in a mixer and degasser,and further mixed for 2 minutes, followed by 8 minutes of degassing. The mixture was then57gently poured onto the master placed in a petri dish until it filled the dish to a depth of1 cm above the master surface. If air bubbles were formed by pouring the mixture, themaster was placed under vacuum for 30 minutes or until all bubbles were removed. Themaster was then placed in an oven at 60 °C for 4 hours to cure the PDMS. After removingthe master from the oven, the PDMS mould was gently removed from the master. Entryand exit points to the channel are then punched into the entry and exit reservoirs at thedistal ends of the channels. A 1 mm diameter coring tool (UniCore) was used to punchthe chip access points. PDMS has previously been demonstrate to have excellent fidelitywhen moulding micro structures [95][58]. As the density of SU-8 structures was low and thestructure aspect ratio was also low, it was assumed that any discrepancies between masterand replica structure would be suciently small as to have no significant e↵ect on results.Sealing and connecting to the chipA 4 by 5 inch glass slide(Ted Pella), is rinsed with methyl alcohol and distilled water andthen dried with air. The glass slide and PDMS replica are then both placed into A PDC-32G115 V Basic Plasma Cleaner (Harrick Plasma). Once a vacuum is established, the plasmacleaner is turned on to full power for 1 minute and 15 seconds. The power is turned o↵,and the channel is slowly brought back to atmospheric pressure. The PDMS mould andglass substrate are then quickly removed from the chamber and brought into contact. Lightpressure may be applied to aid in the bonding process and to remove air bubbles betweenthe PDMS and glass surfaces. The PDMS and glass are then placed in an oven at 60 °C forone hour to solidify the bond. The PDMS channel is now sealed.To connect to the chip 15 gauge microfluidic needles (Nordson EFD) are force fit into thepreviously cored channel entry and exit points. Quick-turn tubing couplings (McMaster-Carr) can then be screwed onto the open side of the needle. These couplings can then serveas a connection point for 1/16” Tygon tubing (McMaster-Carr). Needle connection to themicrofluidic device is presented in Figure 5.3. To improve the needle-PDMS connection, asmall amount of PDMS can be applied around the base of the needle and cured in an ovenat 60 °C.58Figure 5.3: Microfluidic device connection. Flow enters the chip via a 15 gauge needlefitted with tube coupling on the upstream side. On the downstream side, a 15 gauge needleconnects to tubing at the outlet of the channel.5.2.2 Working principlesFor the given apparatus, we assume flow to be laminar. This assumption is validated bycalculating the Reynolds number at the intended operating flow rates. The Reynolds numberis defined according to Equation 5.1.Re =⇢vDH⌘(5.1)Where ⇢ is the fluid density, v is the fluid velocity, and DH is the hydraulic diameter,defined according to Equation 5.2 for a rectangular channel.DH =4wd2(w + d)(5.2)Where w and d are the channel width and depth and w > d. For the given channelgeometry and a flow rate of 100 µL/min the Reynolds number for water at room temperature59is approximately 6.5, which is well within the regime for laminar flow.It is also assumed that flow is fully developed in the channel. This assumption is validfor channels with length much greater than the length of the entrance region (L>>Le). Theentrance length can be estimated according to Equation 5.3 [16].Le = 0.035DHRe (5.3)At a flow rate of 100 µL/min, the entrance length is approximately 50 µm, which issignificantly less than the channel length of 6 cm. Pressure drop at the entrance region istypically estimated using a Bagley plot, which requires plotting the full channel pressuredrop as a function of L/DH , while changing the channel length, L. The intercept of thiscurve is taken to be the inlet pressure drop. This approach is not possible for the givenapparatus as the channel length is fixed. Therefore, the empirical correlation developedby Van Wazer et al. for inlet pressure drop for Newtonian fluids was used, as shown inEquation 5.4 [60][89].pe12⇢v2= KH +KcRe(5.4)Where pe is in the inlet pressure drop, and KH and KC are functions of the ratio ofreservoir to channel width (b = w/r), as defined below.KH = 2.32⇣1 b2⌘(5.5)KC = 159⇣1 b2⌘(5.6)No wall slip is expected to occur as the working fluid is water, and such a phenomenon istypically only observed in suspensions, gels and polymer solutions [60]. Additional assump-tions include incompressible flow, steady state flow achieved at each imposed flow rate, andunidirectional flow. With these assumptions, the equation of motion may be simplified torelate the wall shear stress, ⌧w to the channel pressure drop according to Equation 5.7 [75].60⌧w =wdP2L(w + d)(5.7)For a Newtonian fluid of constant viscosity, the wall shear rate, ˙w, is a linear functionof the imposed flow rate, as shown in Equation 5.8 [75].˙w =6Qwd2(5.8)Recall that for a rectangular channel, w > d. The viscosity can then be estimated bymeasuring the pressure drop in a microfluidic rheometer according to Equation 5.9.⌘ =⌧w˙w=w2d312LQ(w + d)P (5.9)For fluids with a shear rate dependent viscosity, the viscosity must be calculated usingthe Weissenberg-Rabinowitsch-Mooney equation to estimate the true shear rate as follows[75].˙a =6Qwd2(5.10)˙true =˙a3h2 +dln(˙a)dln(⌧w)i(5.11)The expression of dln(˙a)dln(⌧w) can be evaluated graphically by plotting ln ˙a as a function ofexperimental data for ⌧w and evaluating the slope of a fitted curve to the data.5.2.3 Experimentation and validationTo validate the use of the microrheometer described here, a fluid with known viscositymust first be used to confirm measurement accuracy and reliability. Due to the inherentlylarge pressure drop across the present device caused by the small width and depth of thechannel (w, d) and the comparatively large channel length (L), a low viscosity fluid must beused. Therefore, to validate pressure drop measurements can be used reliably to estimate61viscosity, water will be used as a test fluid at room temperature. The viscosity of water atroom temperature is 1 cP or 0.001 kPa.A KD210 micro-syringe pump (KD Scientific) with an accuracy of ± 1 % was usedto control the volumetric flow rate. A 5 mL BD plastic syringe was used as the waterreservoir. Tygon tubing with an inner diameter of 1/16” was used to connect the syringeto the pressure sensor and microrheometer. An inline, microfluidic pressure sensor (MPS)from Elveflow was used to measure the channel pressure. The maximum pressure readingfor the given sensor is 2000 mbar, while the accuracy is ± 4 mbar. The pressure sensors wereverified to work according to the manufacturer specifications by measuring the pressure ina pressure controlled flow channel. A schematic and photograph of the experimental setupare presented in Figure 5.4 and Figure 5.5 respectively.Figure 5.4: Schematic of experimental setup used to measure the pressure drop across themicrorheometer. (A) syringe and syringe pump controlling flow rate. (B) Pressure sensor,connected to PC. (C) Microfluidic needle and tubing coupling connecting tubing to themicro channel. (D) Microrheometer. (E) PC used for data acquisition.62Figure 5.5: Full experimental setup. (A) Syringe pump with syringe and needle. (B)Microrheometer, as shown in Figure 5.3. (C) Pressure sensor used to measure back pressureof flow through the channel. (D) Transmits and connects pressure sensor signal to PC.Prior to beginning measurements, the channel was filled with water to evacuate anyair bubbles. The channel was then subjected to flow rates ranging from 10 µL/min to100 µL/min at increases of 10 µL/min. At each flow rate, the channel pressure drop wasrecorded over time, and allowed to stabilize for 100 s before the final pressure drop wasrecorded.It is expected that for a Newtonian, constant viscosity fluid under isothermal conditions,pressure drop will scale linearly with increasing flow rate, as described by Equation 5.9.PDMS has previously been demonstrated to deform at low pressures for a top wall thick-ness of 1.5 and 3 mm [44]. Therefore, it is important to consider that channel deformationmay a↵ect observed pressure drop across the channel. Should the channel deform, it isexpected that the pressure drop across the channel will be less than predicted.635.3 ResultsThe measured pressure drop as a function of volumetric flow rate is plotted in Figure 5.6.Measured pressure drop data presented are an average of three replicate runs. The coef-ficient of variation, defined as the standard deviation divided by the average of the threemeasurements, ranged between 1 and 5 %. The Reynolds number was calculated over theinvestigated range of flow rates according to Equation 5.1, and was found to range between0.065 and 4.582.Figure 5.6: Measured and predicted pressure drop as a function of volumetric flow rate. Atlow flow rates, the estimated pressure drop is less than the observed, while at high shearrates, the estimated pressure drop is greater than the observed. Error bars are shown asthe % variation.At low shear rates the predicted pressure drop is less than the observed pressure drop,while at higher shear rates, the predicted pressure drop is greater than the observed pressuredrop. The measured pressure drop compared to the predicted pressure drop was within13% at all flow rates, and within 6% at flow rates between 20 and 60 µL/min, where the %di↵erence is defined as (measured  predicted)/predicted. Shear viscosity was calculatedfrom pressure drop data according to Equation 5.9. The shear rate was calculated fromchannel dimensions and flow rate according to Equation 5.8. The observed viscosity is64plotted as a function of shear rate in Figure 5.7.Figure 5.7: Shear viscosity is plotted as a function of shear rate in the microrheometer.Shear rates as high as 5 ⇥ 104s1 were applied. Viscosity is measured with the greatestaccuracy at low shear rates.The experimentally determined viscosity is observed to be greater than the true viscosityat low shear rates, and lower than the true value at high shear rates. The percent error inmeasured viscosity, defined as (measured  true)/true, is within 12 % at all shear rates,and within 6 % at shear rates between approximately 104 and 3 ⇥ 104s1. The % error isplotted as a function of pressure drop in Figure 5.8. A linear model is fitted to the data. Theequation and R2 value are presented, correlating pressure drop to relative error (p < 0.001).65Figure 5.8: % Error as a function of pressure drop. Error magnitude increases with pressurein the channel.With the linear model correlating % error as a function of channel pressure drop, theobserved pressure drops may be corrected to obtain a more accurate estimate of channelpressure drop and corresponding sample viscosity. Corrected pressure drop is plotted as afunction of flow rate along with the predicted pressure drop and the uncorrected pressuredrop in Figure 5.9 and the corrected viscosity is plotted as a function of shear rate alongwith the true viscosity and uncorrected viscosity in Figure 5.10.66Figure 5.9: Corrected, measured, and predicted pressure drop as a function of volumetricflow rate. At high flow rates, the corrected pressure drop is observed to have improvedaccuracy.Figure 5.10: Corrected, measured, and true viscosity as a function of shear rate. At highshear rates the corrected viscosity is observed to have improved accuracy.It is observed that the corrected pressure drop and corrected viscosity more accuratelypredict the true viscosity than the uncorrected data, in particular at high volumetric flowrates and shear rates. At low flow rates and shear rates, this correction has less e↵ect onaccurately predicting the true sample viscosity.675.3.1 Device costA major advantage of the microrheometer compared to a traditional rheometer are the lowcosts associated with fabricating the device and subsequent reuse. The raw materials costsfor the initial prototype are summarized in Table 5.1.Table 5.1: Raw materials used in fabricating the microrheometer and associated costs. Totalmaterials costs are less than $2,000.Item CostQuartz plate $600SU-8 and developer $840Photomask $100Silicon plates $190Glass plates $175Tygon Tubing $30Microfluidic Needles $25Tubing Couplings $20Sylgard 184 and curing agent $50Total costs $1,980Although the cost of the initial prototype is approximately $2,000, many of the materialspurchased in the above table would be reusable for several subsequent prototypes or havebeen purchased in bulk. Specifically, the SU-8 is purchased at a volume of 500 mL. Thephotomask only needs to be replaced if damaged or if a new geometry is desired. Thereare 10 silicon plates ordered. There are 50 glass plates. There is 30 feet of Tygon tubing.There are 20 microfluidic needles per order. There are 10 tubing couplings per order. Itis estimated that after the initial master is produced, the cost of materials per subsequentprototype would be approximately $ 20. This cost includes the per unit costs of a siliconwafer, SU-8 epoxy, microfluidic needles, couplings, and tubing. Additional project costsinclude cleanroom access fees at a rate of $45 per hour. For the present work, pressuresensors and a micro syringe pump were borrowed from a partner research group and werenot purchased. The cost of these devices can be considerable depending on the precision ofthe instruments.685.4 Discussion5.4.1 Device fabrication methodsThe purpose of the present work is to present a low cost and simple method of fabrication todevelop a microrheometer capable of measuring shear viscosity at high shear rates, therefore,a discussion on the advantages and shortcomings of the fabrication procedure should bepresented.Overall, the device fabrication is simple and easy to implement. In particular, SU-8 is a relatively safe epoxy to work with in the cleanroom and does not require the useof dangerous solvents. As the given channel geometry does not have a high density ofstructures and the desired height of the features is not exceedingly small or large, the UVexposure is uncomplicated, and can be completed in a single step. The resolution of theDektak XT stylus was a highly accurate method for confirming channel geometry. The maindisadvantage to the current master fabrication method is the cost to use the cleanroom.PDMS replica moulding is a very easy process yielding high fidelity in channel dimensionsto the master geometry. PDMS and the curing agent are quite safe and do not require speciallaboratory safety precautions. PDMS can be cured at a variety of temperatures, includingroom temperature depending on the facilities available. One challenge in working withPDMS is degassing the solution after it has been mixed. If a vacuum chamber is available,this step is straightforward.The most unreliable part of the process was the plasma bonding to seal the channel. Itis fairly common for the bond not to work. In such cases it is usually possible to repeat theplasma cleaning and achieve an adequate bond on the second attempt, however occasionallythe plasma bond may work only in select sections of the PDMS - glass interface. In thesesituations, the PDMS cannot be removed from the glass without damaging the channel.Considerable literature discusses various PDMS - glass bonding techniques, including astamp and stick method [83], and a PDMS adhesive bonding technique [82], which onlyrequire the use of a spin coater. These techniques may provide more reliable and consistent69results, however, the plasma bonding method was easy to implement given the availabilityof the necessary equipment and when bonded, formed a strong enough seal for the givenapplication.The fabrication method described here requires an initial investment in equipment andmaterials, however, subsequent prototyping is very inexpensive. The present fabricationmethod is ideally suited to an iterative approach to developing an ideal channel geometryand configuration.5.4.2 MicrorheometryMeasurements were found to have good repeatability as % variation remained below 5% forall flow rates. As expected, the pressure was found to increase approximately linearly withincreasing flow rate. Correspondingly, the estimated viscosity was found to be constant. Atflow rates above 60 µL/min the pressure measurements begin to deviate from linearity, andpredicted values were found to exceed to measured values. A clear decrease in % error wasobserved with increasing flow rate, that is at low flow rates the viscosity estimate was greaterthan the true value, while at high flow rates, it was less than the true value. This resultsuggests that at low flow rates, additional frictional losses not accounted for are causingan increase in channel pressure drop. However, as flow rate is increased, and as a resultthe channel pressure drop, the gap between predicted and observed pressure drop decreaseseventually leading to the case where the observed pressure drop is less than the expected.This trend suggests that with increasing pressure, the PDMS channel may begin to deform.This result is consistent with previous studies that have demonstrated deformation of PDMSchannels at pressures as low as 15 kPa [44]. For pressure drops as great as 110 kPa (1100mbar), corresponding to flow rates as high as 60 µL/min and to shear rates of approximately3 ⇥ 104s1, deformation was found to have little e↵ect on the accuracy of the device, asestimated viscosity was within 6 % or less of the true value for water at room temperature.At greater pressure drops, the accuracy of the device was within 9 to 13 %. By plotting the% error as a function of pressure drop across the channel, a correction factor was obtainedto improve viscosity estimates by accounting for channel deformation. With the corrected70estimates, the measured viscosity was within 7 % of the true value at pressure drops as highas 180 kPa (mbar).These results demonstrate that the microrheometer developed here is a valid means forestimating the viscosity of Newtonian fluids at shear rates in the range of 104 to 3⇥104 s1.Additionally, the assumptions of fully developed flow and no wall slip are substantiated as themeasured viscosity was a reasonable estimate of the true value. Additionally, the Reynoldsnumber was found to be 4.652 or less, confirming the assumption of laminar flow. Thepressure drop correction used to accommodate for channel deformation further improvedthe accuracy of the viscosity estimates. As this correction factor is simply a function ofchannel pressure drop, it would be possible to apply this correction to a variety of fluidswith unknown viscosityExperimental sources of error include PDMS deformation, which is observed to have ane↵ect at flow rates above 60 µL/min. Additionally, channel irregularities may contributeto an additional pressure drop not accounted for in pressure drop estimation calculations.This variability is likely caused by spacial variations in UV light intensity during the SU-8UV exposure. Finally, it is also expected that change in channel geometry at the exit willresult in an additional pressure drop not estimated here.5.4.3 Device improvementsA major shortcoming of the present device is the pressure limitation. PDMS deformationis believed to cause a decrease in channel pressure drop at high flow rates. This e↵ect maybe reduced somewhat by increasing the thickness of the PDMS channel walls, however it isunlikely that the e↵ect could be removed altogether. Consequently, to measure viscosity atshear rates on the order of 105 or higher, an alternative material of construction with in-creased sti↵ness would be required. It may also be possible to increase the channel sti↵nessby fixing the PDMS channel between two glass plates for added rigidity. A second poten-tial improvement would be to estimate the Bagley Correction factor experimentally ratherthan empirically, as was done in the present work. This could be easily implemented bysuccessively shortening each of the 6 channels on the photomask. In doing so, a final device71with 6 channels of equal width and depth but di↵erent lengths could be obtained. Then bymeasuring pressure drop in each channel at an identical flow rate the inlet pressure dropcould be estimated graphically by plotting pressure drop as a function of the L/DH ratio[60]. It is also recommended to increase the channel width, w, in order to accommodateviscosity measurements of fluids with greater viscosity. The resultant pressure drop, evenat low flow rates, for the present geometry would result in significant channel deformationsthat would make viscosity estimates unreliable or inaccurate.Although the entrance and exit pressure drops were observed to have minimal e↵ect onthe overall channel pressure drop, it is possible that these e↵ects become more significantat higher flow rates, with a higher viscosity fluid, or with a less compliant material ofconstruction. Therefore, it is recommended to use an channel design similar to Kang et al.to design channels with a progressive narrowing from the fluid reservoirs to the body of thechannels in order to reduce inlet pressure drop [54].This device also requires validation with a non-Newtonian fluid. This should be com-pleted with a fluid of known viscosity, either previously measured with a conventionalrheometer at high shear rates, or with a viscoelastic standard.5.5 ConclusionsThe present work describes in detail the procedure for designing and fabricating a low costmicrofluidic rheometer. The fabrication procedure requires minimal resources and lab spaceand after an initial master is produced, prototypes can be replicated quickly and at verylow cost. Measurement repeatability was good, with % variation in measurement less than5% at all imposed flow rates. Measurement error was less than 6% at flow rates between20 and 60 µL/min with (corresponding to channel pressure as high as 110 kPa), however athigher flow rates and resultant channel pressure, the channel is suspected to deform, causingan underestimate of sample viscosity. A correction factor was employed to account for theerror at higher pressures, and was found to result in viscosity estimates that were within7% of the true water viscosity at all flow rates investigated.72The current device is limited in maximum flow rate and shear rate by PDMS defor-mation. Investigating reducing channel deformation by increasing the channel wall widthor by modifying the material of fabrication should be considered. Modifying the channellengths should also be considered in order to investigate the inlet pressure drop via Bagleycorrection factor. Finally, this device must be validated for use with a non-Newtonian fluid.73Chapter 6Conclusions6.1 SummaryA butyrylated, low molecular weight, derivative of hyaluronic acid with anti-inflammatoryproperties has been demonstrated to have shear thinning and viscoelastic properties. Whencompared to the parent compound, the derivative was found to have lower viscosity and dy-namic moduli. This di↵erence is largely attributed to the decrease in molecular weight thatoccurs during the derivative synthesis. When compared to commercial viscosupplements,the derivative sample was found to have significantly lower viscosity and dynamic moduli. Itis believed that by increasing the hyaluronic acid concentration the derivative sample rheol-ogy could be modulated to increase viscosity and viscoelasticity. The occurrence of flare-uphas been previously demonstrated to be detrimental to knee joint health, and therefore, thebutyrylated hyaluronic acid derivative presents an opportunity for improvement to currentviscosupplementation treatment.Preliminary results from a case study on the e↵ects of oral glucosamine supplementationon synovial fluid rheology are presented. In both the treatment and treatment discontinu-ation participants, it was observed that while administered glucosamine compared to whilenot taking glucosamine, synovial fluid had greater viscosity at low shear rates and lowerviscosity at high shear rates. This trend is strong motivation for future study. WOMACscores were found to increase in both participants. This results may be caused by a lack ofblinding, or by joint pain relief caused by a decrease in joint pressure due to synovial fluid74aspiration.A rapid, low cost method was used to develop a microfluidic rheometer. The rheometerwas subsequently validated with water, and was found to estimate the viscosity of waterwithin 6 % at flowrates between 20 and 60 µL/min. The present work demonstrates a simpleand easily repeatable process for fabricating a microrheometer. Simply by designing a newphotomask, alternate geometries can be fabricated that would allow studying shear viscosityof high viscosity fluids, dynamic rheological parameters, and extensional viscosity. PDMSwas demonstrated to deform at higher channel pressures. This result provides an estimate ofthe range of operating pressures for a PDMS channel of similar thickness. Finally, a pressurecorrection correlation was estimated based on the expected versus observed pressure data.With the correction factor, the microrheometer was used to predict the viscosity of waterat room temperature at flow rates between 10 and 100 µL/min within 7 %.6.2 ContributionsThe present work contributes to current literature concerning osteoarthritis, treatments forthe disease, and evaluation methods of those treatments in the following ways.The present work demonstrates that the anti-inflammatory hyaluronic acid derivativedisplays shear thinning and viscoelastic behaviour similar to some commercial viscosupple-ments. It is proposed that by increasing the concentration of the derivative in solution,greater viscosity and dynamic rheological properties could be achieved. In doing so, ad-vancement could be made in viscosupplementation treatment by developing a hyaluronicacid solution that can supplement synovial fluid joint protection while also reducing theincidence of flare up.In a preliminary study the viscosity of synovial fluid was found to increase at high shearrates after glucosamine discontinuation, and was found to be greater at low shear rates whentaking glucosamine supplements. This trend merits further study that may be completedusing the present study methods and ethical approvals. By providing a clear motivation forfuture study on the e↵ect of glucosamine on synovial fluid rheology progress is made towards75a more complete understanding on the use and ecacy of glucosamine supplementation fortreatment of osteoarthritis.A simple and low cost microrheometer has been developed that will facilitate shear rhe-ology measurements of low viscosity solutions. This rheometer can be used to inexpensivelyand quickly evaluate the shear rheology of hyaluronic acid and synovial fluid. Addition-ally, the rapid prototyping procedure described can be easily modified to produce alternategeometries for study of dynamic and extensional rheological properties.6.3 LimitationsThe hyaluronic acid derivatives investigated in this study were measured at a concentrationof 5 mg/mL. Viscosupplements typically have a hyaluronic acid concentration much higher.It is hypothesized that the viscoelastic properties of the derivatives investigated would scaleto similar values to those of commercial viscosupplements, however this was not completedin the current study. Additionally, the variation in dynamic rheological properties was muchhigher than those reported elsewhere in the literature.As only two patients were studied, it is impossible to make any statistical inferences aboutthe e↵ect of glucosamine on synovial fluid rheology. The study did not include dynamicrheological data. The study was limited in time of glucosamine supplementation to fourweeks and two synovial fluid aspirations. The WOMAC scales are subject to bias as blindingcould not be incorporated to the present study. It is also possible that undocumentedincidents of flareup may have a↵ected results.The current microrheometer design is limited to fluids with very low viscosity because asmall channel geometry was selected and the device was fabricated with PDMS, a compliantmaterial. The PDMS compliance also limits the maximum attainable shear rates in thechannel. The present device does not accommodate estimation of a Bagley Correctionfactor.766.4 Future directionsRheological characterization of the anti inflammatory hyaluronic acid derivative at higherconcentrations is required to further validate its use as a viscosupplement. It is believed thatby studying the solution at higher concentrations, more precise and repeatable measurementswill be obtainable with the given apparatus. Beyond that, industrial partnership withCanCog technologies will be required in order to validate the treatment in canine modelsbefore it can be used with humans.A full scale clinical study of the e↵ects of glucosamine on synovial fluid rheology isrequired in order to statistically substantiate the findings presented here. Additionally, ameans for evaluating the clinical significance of the rheological changes must be established.Finally, the study should be set up to facilitate treatment blinding in order to reduce thepotential for bias in WOMAC survey scores.The microrheometer channel geometry should be enlarged to accommodate higher vis-cosity fluids. Other geometries, including a serpentine channel and a cross slot geometry,should be developed to facilitate measurement of dynamic and extensional rheological prop-erties. Finally, the present rheometer should be used to validate the use of micro-PIVmeasurements for estimating fluid viscosity as a means to establish an appropriate protocolfor birefringence measurements used in extensional rheological measurements.77Bibliography[1] The Rheology Handbook. Coatings Compendia, 2006.[2] Principles of Anatomy and Physiology, 14th edition. Wiley, 2014.[3] A Abhishek and Michael Doherty. Diagnosis and clinical presentation of osteoarthritis.Rheumatic Disease Clinics of North America, 39(1):45–66, 2013.[4] Anita Aggarwal and Ian P Sempowski. Hyaluronic acid injections for knee osteoarthri-tis. systematic review of the literature. Canadian Family Physician, 50(2):249–256,2004.[5] Janelle R Anderson, Daniel T Chiu, Hongkai Wu, OJ Schueller, and George M White-sides. Fabrication of microfluidic systems in poly (dimethylsiloxane). Electrophoresis,21(1):27–40, 2000.[6] Oladunni Babasola, Karen J Rees-Milton, Siziwe Bebe, Jiaxi Wang, and Tassos P Anas-tassiades. Chemically modified n-acylated hyaluronan fragments modulate proinflam-matory cytokine production by stimulated human macrophages. Journal of BiologicalChemistry, 289(36):24779–24791, 2014.[7] Kristin R Baker, Miriam E Nelson, David T Felson, Jennifer E Layne, Robert Sarno,and Ronenn Roubeno↵. The ecacy of home based progressive strength training inolder adults with knee osteoarthritis: a randomized controlled trial. The Journal ofrheumatology, 28(7):1655–1665, 2001.[8] Endre A. Balazs. The physicial proeperties of synovial fluid and the special role ofhyaluronic acid. Disorders of the Knee, 1974.78[9] H C Hemaka Bandalusena, William B Zimmerman, and Julia M Rees. Microfluidicrheometry of a polymer solution by micron resolution particle image velocimetry: amodel validation study. Measurement Science and Technology, 20(11):115404, 2009.[10] C Baravian and D Quemada. Correction of instrumental inertia e↵ects in controlledstress rheometry. The European Physical Journal Applied Physics, 2(02):189–195, 1998.[11] M. Bausani. Assessing the ecacy of a viscosupplement combining hyaluronic acidand sorbitol (synolis-va) in patients with high grades knee osteoarthritis for whomcorticotherapy is contraindicated. Osteoarthritis and Cartilage, 21, Supplement(0):S270–, 2013.[12] Jean-Marie Berthelot, Benoˆıt Le Go↵, and Yves Maugars. Side e↵ects of corticosteroidinjections: What’s new? Joint Bone Spine, 80(4):363–367, 2013.[13] Shantanu Bhattacharya, Arindom Datta, Jordan M Berg, and Shubhra Gangopad-hyay. Studies on surface wettability of poly (dimethyl) siloxane (pdms) and glass underoxygen-plasma treatment and correlation with bond strength. MicroelectromechanicalSystems, Journal of, 14(3):590–597, 2005.[14] Petcharatana Bhuanantanondh. Rheology of synovial fluid with and without viscosup-plements in patients with osteoarthritis: A pilot study. Master’s thesis, University ofBritish Columbia, 2009.[15] Petcharatana Bhuanantanondh, Dana Grecov, and Ezra Kwok. Rheological study ofviscosupplements and synovial fluid in patients with osteoarthritis. J. Med. Biol. Eng,32(1):12–16, 2012.[16] R Bird, W Stewart, and E Lightfoot. Transport phenomena–second edition, 2002.[17] J.A. Block, T.R. Oegema, J.D. Sandy, and A. Plaas. The e↵ects of oral glucosamineon joint health: is a change in research approach needed? Osteoarthritis and Cartilage,18(1):5 – 11, 2010.79[18] Hillary J Braun and Garry E Gold. Diagnosis of osteoarthritis: imaging. Bone,51(2):278–288, 2012.[19] Giuseppe M Campo, Angela Avenoso, Salvatore Campo, Angela D’Ascola, Gian-carlo Nastasi, and Alberto Calatroni. Molecular size hyaluronan di↵erently modulatestoll-like receptor-4 in lps-induced inflammation in mouse chondrocytes. Biochimie,92(2):204–215, 2010.[20] Giuseppe M Campo, Angela Avenoso, Salvatore Campo, Angela DAscola, GiancarloNastasi, and Alberto Calatroni. Small hyaluronan oligosaccharides induce inflammationby engaging both toll-like-4 and cd44 receptors in human chondrocytes. Biochemicalpharmacology, 80(4):480–490, 2010.[21] Pablo Casuso, Adrian Perez-San Vicente, Haizea Iribar, Araika Gutierrez-Rivera, AnderIzeta, Iraida Loinaz, German Cabanero, Hans-Jurgen Grande, Ibon Odriozola, andDamien Dupin. Aurophilically cross-linked ”dynamic” hydrogels mimicking healthysynovial fluid properties. Chem. Commun., 50:15199–15201, 2014.[22] Mary B. Chan-Park, Jun Zhang, Yehai Yan, and C.Y. Yue. Fabrication of large su-8mold with high aspect ratio microchannels by {UV} exposure dose reduction. Sensorsand Actuators B: Chemical, 101(12):175 – 182, 2004.[23] Yin-Quan Chen, Pei-lun Chou, Chen-Yu Cheng, Chia-Chun Chiang, Ming-Tzo Wei,Chin-Ting Chuang, Yi-Lin Sophia Chen, and Arthur Chiou. Microrheology of humansynovial fluid of arthritis patients studied by di↵using wave spectroscopy. Journal ofBiophotonics, 5(10):777–784, 2012.[24] Sung Chul Choi, Mi Ae Yoo, Su Yeon Lee, Hyun Ji Lee, Dong Hoon Son, Jessica Jung,Insup Noh, and Chan-Wha Kim. Modulation of biomechanical properties of hyaluronicacid hydrogels by crosslinking agents. Journal of Biomedical Materials Research PartA, pages n/a–n/a, 2015.[25] Jolanda Cibere, Jacek A. Kopec, Anona Thorne, Joel Singer, Janice Canvin, David B.Robinson, Janet Pope, Paul Hong, Eric Grant, and John M. Esdaile. Randomized,80double-blind, placebo-controlled glucosamine discontinuation trial in knee osteoarthri-tis. Arthritis Care and Research, 51(5):738–745, 2004.[26] Daniel O. Clegg, Domenic J. Reda, Crystal L. Harris, Marguerite A. Klein, James R.O’Dell, Michele M. Hooper, John D. Bradley, Clifton O. Bingham, Michael H. Weisman,Christopher G. Jackson, Nancy E. Lane, John J. Cush, Larry W. Moreland, H. RalphSchumacher, Chester V. Oddis, Frederick Wolfe, Jerry A. Molitor, David E. Yocum,Thomas J. Schnitzer, Daniel E. Furst, Allen D. Sawitzke, Helen Shi, Kenneth D. Brandt,Roland W. Moskowitz, and H. James Williams. Glucosamine, chondroitin sulfate,and the two in combination for painful knee osteoarthritis. New England Journal ofMedicine, 354(8):795–808, 2006. PMID: 16495392.[27] T Conrozier, J-C Balblanc, M Piperno, and M Rinaudo. Di↵erences in the osteoarthriticsynovial fluid composition and rheology between patients with or without flare-up. apilot study. Osteoarthritis and Cartilage, 20:S244, 2012.[28] Thierry Conrozier, Pierre Mathieu, and Marguerite Rinaudo. Mannitol preserves theviscoelastic properties of hyaluronic acid in an in vitro model of oxidative stress.Rheumatology and Therapy, 1(1):45–54, 2014.[29] Mark A Eddings, Michael A Johnson, and Bruce K Gale. Determining the optimalpdmspdms bonding technique for microfluidic devices. Journal of Micromechanics andMicroengineering, 18(6):067001, 2008.[30] Patrick Eriksen, Else M. Bartels, Roy D. Altman, Henning Bliddal, Carsten Juhl, andRobin Christensen. Risk of bias and brand explain the observed inconsistency in trialson glucosamine for symptomatic relief of osteoarthritis: A meta-analysis of placebo-controlled trials. Arthritis Care and Research, 66(12):1844–1855, 2014.[31] Mandy B. Esch, Sahil Kapur, Gizaida Irizarry, and Vincent Genova. Influence ofmaster fabrication techniques on the characteristics of embossed microfluidic channels.Lab Chip, 3:121–127, 2003.81[32] Richy F, Bruyere O, Ethgen O, Cucherat M, Henrotin Y, and Reginster J. Structuraland symptomatic ecacy of glucosamine and chondroitin in knee osteoarthritis: Acomprehensive meta-analysis. Archives of Internal Medicine, 163(13):1514–1522, 2003.[33] Amir Fakhari, Quang Phan, Santosh V. Thakkar, C. Russell Middaugh, and CoryBerkland. Hyaluronic acid nanoparticles titrate the viscoelastic properties of viscosup-plements. Langmuir, 29(17):5123–5131, 2013.[34] H Fam, JT Bryant, and M Kontopoulou. Rheological properties of synovial fluids.Biorheology, 44(2):59–74, 2006.[35] H Fam, M Kontopoulou, and JT Bryant. E↵ect of concentration and molecular weighton the rheology of hyaluronic acid/bovine calf serum solutions. Biorheology, 46(1):31–43, 2009.[36] Ivana Finelli, Ester Chiessi, Devis Galesso, Davide Renier, and Gaio Paradossi. A newviscosupplement based on partially hydrophobic hyaluronic acid: a comparative study.Biorheology, 48(5):263 – 275, 2011.[37] Ivana Finelli, Ester Chiessi, Devis Galesso, Davide Renier, and Gaio Paradossi. A newviscosupplement based on partially hydrophobic hyaluronic acid: a comparative study.Biorheology, 48(5-6):263–275, 2011.[38] Marlene Fransen, Maria Agaliotis, Lillias Nairn, Milana Votrubec, Lisa Bridgett, SteveSu, Stephen Jan, Lyn March, John Edmonds, Robyn Norton, Mark Woodward, RichardDay, and on behalf of the LEGS study collaborative group. Glucosamine and chon-droitin for knee osteoarthritis: a double-blind randomised placebo-controlled clinicaltrial evaluating single and combination regimens. Annals of the Rheumatic Diseases,74(5):851–858, 2015.[39] Carl K Fredrickson and Z Hugh Fan. Macro-to-micro interfaces for microfluidic devices.Lab on a Chip, 4(6):526–533, 2004.[40] ThomasB. Goudoulas, EleftheriosG. Kastrinakis, StavrosG. Nychas, LysimachosG. Pa-pazoglou, GeorgeM. Kazakos, and PanagiotisV. Kosmas. Rheological study of synovial82fluid obtained from dogs: Healthy, pathological, and post-surgery, after spontaneousrupture of cranial cruciate ligament. Annals of Biomedical Engineering, 38(1):57–65,2010.[41] Bagga H, Burkhardt D, Sambrook P, and March L. Longterm e↵ects of intraarticularhyaluronan on synovial fluid in osteoarthritis of the knee. Journal of Rheumatology,2006.[42] M Kontopoulou H. Fam and J/T. Bryant. E↵ect of concentration and molecular weighton the rheology of hyaluronic acid/bovine calf serum solutions. Biorheology, 2009.[43] MY Han, JJ Dai, Y Zhang, Q Lin, M Jiang, XY Xu, and Q Liu. Identification of os-teoarthritis biomarkers by proteomic analysis of synovial fluid. Journal of InternationalMedical Research, 40(6):2243–2250, 2012.[44] Brian S Hardy, Kawika Uechi, Janet Zhen, and H Pirouz Kavehpour. The deformationof flexible pdms microchannels under a pressure driven flow. Lab on a Chip, 9(7):935–938, 2009.[45] Simon J. Haward, A Jaishankar, MSN Oliveira, MA Alves, and GH McKinley. Exten-sional flow of hyaluronic acid solutions in an optimized microfluidic cross-slot devicea).Biomicrofluidics, 7(4):044108, 2013.[46] Simon J. Haward, Mo´nica S. N. Oliveira, Manuel A. Alves, and Gareth H. McKinley.Optimized cross-slot flow geometry for microfluidic extensional rheometry. Phys. Rev.Lett., 109:128301, Sep 2012.[47] Simon J. Haward, Vivek Sharma, and Je↵rey A Odell. Extensional opto-rheometrywith biofluids and ultra-dilute polymer solutions. Soft Matter, 7(21):9908–9921, 2011.[48] Yves Henrotin, Ali Mobasheri, and Marc Marty. Is there any scientific evidence for theuse of glucosamine in the management of human osteoarthritis? Arthritis Res Ther.,2012.83[49] David J. Hunter. Osteoarthritis. Best Practice Research Clinical Rheumatology,25(6):801 – 814, 2011. Musculoskeletal Science.[50] ZM Jin, D Dowson, and J Fisher. Analysis of fluid film lubrication in artificial hipjoint replacements with surfaces of high elastic modulus. Proceedings of the Institutionof Mechanical Engineers, Part H: Journal of Engineering in Medicine, 211(3):247–256,1997.[51] Victoria L. Johnson and David J. Hunter. The epidemiology of osteoarthritis. BestPractice Research Clinical Rheumatology, 28(1):5 – 15, 2014. Osteoarthritis: Movingfrom Evidence to Practice.[52] Ty Jones, Gary Kelsberg, and Sarah Safranek. Fpin’s clinical inquiries: Intra-articularcorticosteroid injections for osteoarthritis of the knee. American family physician,90(2):115, 2014.[53] Pavelk K, Gatterov J, Olejarov M, Machacek S, Giacovelli G, and Rovati LC. Glu-cosamine sulfate use and delay of progression of knee osteoarthritis: A 3-year,randomized, placebo-controlled, double-blind study. Archives of Internal Medicine,162(18):2113–2123, 2002.[54] Kai Kang, L James Lee, and Kurt W Koelling. High shear microfluidics and its appli-cation in rheological measurement. Experiments in fluids, 38(2):222–232, 2005.[55] NG Kashevarova, EM Zaitseva, AV Smirnov, NV Demin, and LI Alekseeva. Ab0601risk factors for early osteoarthritis progression. Annals of the Rheumatic Diseases,72(Suppl 3):A974–A974, 2013.[56] C. Kent Kwoh, Frank W. Roemer, Michael J. Hannon, Carolyn E. Moore, John M.Jakicic, Ali Guermazi, Stephanie M. Green, Rhobert W. Evans, and Robert Boudreau.E↵ect of oral glucosamine on joint structure in individuals with chronic knee pain: Arandomized, placebo-controlled clinical trial. Arthritis and Rheumatology, 66(4):930–939, 2014.84[57] Kevin S. Lee and Rajeev J. Ram. Plastic-pdms bonding for high pressure hydrolyticallystable active microfluidics. Lab Chip, 9:1618–1624, 2009.[58] T-W Lee, Oleg Mitrofanov, and Julia WP Hsu. Pattern-transfer fidelity in soft lithog-raphy: The role of pattern density and aspect ratio. Advanced Functional Materials,15(10):1683–1688, 2005.[59] Taryn E. Ludwig, Jenelle R. McAllister, Victor Lun, J. Preston Wiley, and Tannin A.Schmidt. Diminished cartilage-lubricating ability of human osteoarthritic synovialfluid deficient in proteoglycan 4: Restoration through proteoglycan 4 supplementation.Arthritis and Rheumatism, 64(12):3963–3971, 2012.[60] CW Macosko and Principles Rheology. Measurements and applications. VCH, NewYork, 1994.[61] Anwar Madkhali. Osteoarthritic synovial fluid rheology and correlation with proteinconcentration. Master’s thesis, University of British Columbia, 2013.[62] Hiroaki Matsuno, Hiroshi Nakamura, Kou Katayama, Seigaku Hayashi, Syogo Kano,Kazuo Yudoh, and Yoshinobu Kiso. E↵ects of an oral administration of glucosamine-chondroitin-quercetin glucoside on the synovial fluid properties in patients with os-teoarthritis and rheumatoid arthritis. Bioscience, Biotechnology, and Biochemistry,73(2):288–292, 2009.[63] Gloria L Matthews and David J Hunter. Emerging drugs for osteoarthritis. Expertopinion on emerging drugs, 16(3):479–491, 2011.[64] Dan Mazzucco, Gareth McKinley, Richard D Scott, and Myron Spector. Rheologyof joint fluid in total knee arthroplasty patients. Journal of Orthopaedic Research,20(6):1157–1163, 2002.[65] M.F. McCarty. Enhanced synovial production of hyaluronic acid may explain rapid clin-ical response to high-dose glucosamine in osteoarthritis. Medical Hypotheses, 50(6):507– 510, 1998.85[66] J. Cooper McDonald, , and George M. Whitesides*. Poly(dimethylsiloxane) as a mate-rial for fabricating microfluidic devices. Accounts of Chemical Research, 35(7):491–499,2002. PMID: 12118988.[67] Stephen P Messier, David J Gutekunst, Cralen Davis, and Paul DeVita. Weight lossreduces knee-joint loads in overweight and obese older adults with knee osteoarthritis.Arthritis & Rheumatism, 52(7):2026–2032, 2005.[68] Faith A Morrison. Understanding rheology. Oxford University Press, USA, 2001.[69] Connor William Myant and Philippa Cann. The e↵ect of transient conditions onsynovial fluid protein aggregation lubrication. Journal of the Mechanical Behaviorof Biomedical Materials, 34(0):349 – 357, 2014.[70] M Namikawa et al. Bleeding gastric ulcer: case report. Reactions, 1393:17, 2012.[71] David Neustadt, Jacques Caldwell, Mary Bell, John Wade, and Joseph Gimbel. Clinicale↵ects of intraarticular injection of high molecular weight hyaluronan (orthovisc) inosteoarthritis of the knee: a randomized, controlled, multicenter trial. The Journal ofrheumatology, 32(10):1928–1936, 2005.[72] Katherine M.N Oates, Wendy E Krause, Ronald L Jones, and Ralph H Colby. Rheopexyof synovial fluid and protein aggregation. Journal of The Royal Society Interface,3(6):167–174, 2006.[73] Ken-ichi Ohno, Kaoru Tachikawa, and Andreas Manz. Microfluidics: Applica-tions for analytical purposes in chemistry and biochemistry. ELECTROPHORESIS,29(22):4443–4453, 2008.[74] Lichao Pan and PauloE. Arratia. A high-shear, low reynolds number microfluidicrheometer. Microfluidics and Nanofluidics, 14(5):885–894, 2013.[75] Christopher J Pipe, Trushant S Majmudar, and Gareth H McKinley. High shear rateviscometry. Rheologica Acta, 47(5-6):621–642, 2008.86[76] Christopher J. Pipe and Gareth H. McKinley. Microfluidic rheometry. MechanicsResearch Communications, 36(1):110 – 120, 2009. Recent Advances in Microfluidics.[77] Adrien Plecis and Yong Chen. Fabrication of microfluidic devices based on glasspdms-glass technology. Microelectronic Engineering, 84(58):1265 – 1269, 2007. Proceedingsof the 32nd International Conference on Micro- and Nano-Engineering.[78] Anne W.S. Rutjes, Peter Jni, Bruno R. da Costa, Sven Trelle, Eveline Nesch, andStephan Reichenbach. Viscosupplementation for osteoarthritis of the kneea systematicreview and meta-analysis. Annals of Internal Medicine, 157(3):180–191, 2012.[79] Alireza Sadeghi, Nooshin Jalili, and Masoud Asadi-Khiavi. Low does glucosamine andchondroitin sulfate use in knee osteoarthritis. Journal of Chemical and PharmaceuticalResearch, 2014.[80] M Safari, A Bjelle, M Gudmundsson, C Hoegfors, and H Granhed. Clinical assessmentof rheumatic diseases using viscoelastic parameters for synovial fluid. Biorheology,27(5):659–674, 1989.[81] Juan Salazar, Luis Bello, Mervin Chvez, Roberto Aez, Joselyn Rojas, and ValmoreBermdez. Glucosamine for osteoarthritis: Biological e↵ects, clinical ecacy, and safetyon glucose metabolism,. Arthritis, 2014.[82] Bjo¨rn Samel, M Kamruzzaman Chowdhury, and Go¨ran Stemme. The fabrication ofmicrofluidic structures by means of full-wafer adhesive bonding using a poly (dimethyl-siloxane) catalyst. Journal of Micromechanics and Microengineering, 17(8):1710, 2007.[83] Srinath Satyanarayana, R.N. Karnik, and A. Majumdar. Stamp-and-stick room-temperature bonding technique for microdevices. Microelectromechanical Systems,Journal of, 14(2):392–399, April 2005.[84] Daniel J Schneck and Joseph D Bronzino. Biomechanics: principles and applications.CRC press, 2002.87[85] Leena Sharma, Dorothy D Dunlop, Jing Song, and Karen W Hayes. Quadricepsstrength and osteoarthritis progression in malaligned and lax knees. Annals of internalmedicine, 138(8):613–619, 2003.[86] Andra Arruda Martins Shimojo, Aline Mara Barbosa Pires, Rafael Lichy, Ana Am-lia Rodrigues, and Maria Helena Andrade Santana. The crosslinking degree controlsthe mechanical, rheological, and swelling properties of hyaluronic acid microparticles.Journal of Biomedical Materials Research Part A, 103(2):730–737, 2015.[87] Keith Sinusas. Osteoarthritis: diagnosis and treatment. American family physician,85(1), 2012.[88] Shin Taniguchi, Junnosuke Ryu, Masayuki Seki, Takanobu Sumino, Yasuaki Tokuhashi,and Mariko Esumi. Long-term oral administration of glucosamine or chondroitin sul-fate reduces destruction of cartilage and up-regulation of mmp-3 mrna in a model ofspontaneous osteoarthritis in hartley guinea pigs. Journal of Orthopaedic Research,30(5):673–678, 2012.[89] John R Van Wazer. Viscosity and flow measurement: a laboratory handbook of rheology.Interscience Publishers, 1963.[90] Saskia PJ Verkleij, Pim AJ Luijsterburg, Bart W Koes, Arthur M Bohnen, and Sita MABierma-Zeinstra. E↵ectiveness of diclofenac versus acetaminophen in primary carepatients with knee osteoarthritis:[ntr1485], dipa-trial: design of a randomized clinicaltrial. BMC musculoskeletal disorders, 11(1):1, 2010.[91] Simon Wandel, Peter Ju¨ni, Britta Tendal, Eveline Nu¨esch, Peter M Villiger, Nicky JWelton, Stephan Reichenbach, and Sven Trelle. E↵ects of glucosamine, chondroitin, orplacebo in patients with osteoarthritis of hip or knee: network meta-analysis. BMJ,341, 2010.[92] George M Whitesides. The origins and the future of mcirofluidics. Nature, 2006.[93] A Yayon. A novel autologous plasma based viscosupplementation. Osteoarthritis andCartilage, (22):S430, 2014.88[94] WMRNGea Zhang, RW Moskowitz, George Nuki, S Abramson, RD Altman, N Arden,S Bierma-Zeinstra, KD Brandt, P Croft, M Doherty, et al. Oarsi recommendations forthe management of hip and knee osteoarthritis, part ii: Oarsi evidence-based, expertconsensus guidelines. Osteoarthritis and cartilage, 16(2):137–162, 2008.[95] Ying Zhang, Chi-Wei Lo, J Ashley Taylor, and Shu Yang. Replica molding of high-aspect-ratio polymeric nanopillar arrays with high fidelity. Langmuir, 22(20):8595–8601,2006.[96] Z Zhang, S Barman, and GF Christopher. The role of protein content on the steadyand oscillatory shear rheology of model synovial fluids. Soft matter, 10(32):5965–5973,2014.[97] Josephine Zilz, Christof Scha¨fer, Christian Wagner, Robert J Poole, Manuel A Alves,and Anke Lindner. Serpentine channels: micro-rheometers for fluid relaxation times.Lab on a Chip, 14(2):351–358, 2014.[98] Michael E Zychowicz. Viscosupplementation for knee osteoarthritis. Journal of theAmerican Association of Nurse Practitioners, 26(3):163–175, 2014.89Appendices90Appendix ARheological Evaluation of a NovelViscosupplement91Table A.1: Shear rate and shear viscosity (Pas) for all five investigated samples.˙(s1) Sample 1 Sample 2 Sample 3 Sample 4 Sample 51.00E-01 1.11E-01 9.31E-01 4.79E-02 2.55E+00 9.17E-021.39E-01 1.00E-01 6.60E-01 3.49E-02 1.82E+00 6.46E-021.93E-01 9.25E-02 4.67E-01 2.63E-02 1.31E+00 4.60E-022.68E-01 8.69E-02 3.33E-01 2.00E-02 9.54E-01 3.44E-023.73E-01 8.30E-02 2.34E-01 1.25E-02 7.11E-01 2.43E-025.18E-01 8.05E-02 1.68E-01 1.07E-02 5.10E-01 1.75E-027.20E-01 7.84E-02 1.21E-01 1.00E-02 3.62E-01 1.30E-021.00E+00 7.67E-02 8.67E-02 8.20E-03 2.63E-01 1.13E-021.39E+00 7.59E-02 6.37E-02 7.40E-03 1.91E-01 8.60E-031.93E+00 7.51E-02 4.92E-02 6.80E-03 1.39E-01 6.90E-032.68E+00 7.41E-02 3.83E-02 5.50E-03 9.69E-02 5.30E-033.73E+00 7.28E-02 3.01E-02 4.80E-03 7.28E-02 4.30E-035.18E+00 7.17E-02 2.10E-02 4.40E-03 5.34E-02 3.50E-037.20E+00 6.99E-02 1.69E-02 4.20E-03 3.66E-02 3.10E-031.00E+01 6.79E-02 1.26E-02 4.00E-03 2.57E-02 2.80E-031.39E+01 6.54E-02 9.53E-03 3.90E-03 1.86E-02 2.60E-031.93E+01 6.25E-02 7.11E-03 3.80E-03 1.22E-02 2.40E-032.68E+01 5.91E-02 5.78E-03 3.70E-03 8.30E-03 2.30E-033.73E+01 5.53E-02 4.69E-03 3.60E-03 6.50E-03 2.30E-035.18E+01 5.11E-02 4.02E-03 3.60E-03 5.00E-03 2.20E-037.20E+01 4.67E-02 3.44E-03 3.60E-03 4.10E-03 2.20E-031.00E+02 4.22E-02 3.08E-03 3.60E-03 3.40E-03 2.20E-031.39E+02 3.78E-02 2.71E-03 3.50E-03 3.00E-03 2.20E-031.93E+02 3.35E-02 2.53E-03 3.50E-03 2.70E-03 2.20E-032.68E+02 2.95E-02 2.42E-03 3.40E-03 2.60E-03 2.20E-033.73E+02 2.57E-02 2.34E-03 3.40E-03 2.40E-03 2.20E-035.18E+02 2.32E-02 2.29E-03 3.40E-03 2.40E-03 2.20E-037.20E+02 2.00E-02 2.25E-03 3.40E-03 2.30E-03 2.20E-031.00E+03 1.73E-02 2.25E-03 3.40E-03 2.30E-03 2.20E-0392Table A.2: Dynamic moduli for Sample 1 as a function of oscillation frequency.f(Hz) G’ (Pa) G” (Pa)1.26 0.04 0.641.59 0.06 0.772.00 0.12 0.862.51 0.19 1.043.16 0.32 1.213.98 0.57 1.365.01 0.87 1.586.31 1.27 1.857.94 1.79 2.1010.00 2.34 2.2812.59 3.21 2.61Table A.3: Dynamic moduli for Sample 2 as a function of oscillation frequency.f(Hz) G’ (Pa) G” (Pa)1.00 1.20 0.181.26 1.18 0.201.59 1.20 0.212.00 1.19 0.222.51 1.18 0.243.16 1.26 0.263.98 1.39 0.275.01 1.55 0.306.31 1.81 0.347.94 2.19 0.37Table A.4: Dynamic moduli for Sample 3 as a function of oscillation frequency.f(Hz) G’ (Pa) G” (Pa)3.16 0.05 0.153.98 0.14 0.185.01 0.29 0.196.31 0.52 0.217.94 0.83 0.2110.00 1.10 0.2393Table A.5: Dynamic moduli for Sample 4 as a function of oscillation frequency.f(Hz) G’ (Pa) G” (Pa)1.00 0.64 0.111.26 0.62 0.151.59 0.56 0.152.00 0.53 0.152.51 0.58 0.173.16 0.68 0.153.98 0.80 0.175.01 0.92 0.176.31 1.17 0.217.94 1.50 0.2010.00 1.82 0.20Table A.6: Dynamic moduli for Sample 5 as a function of oscillation frequency.f(Hz) G’ (Pa) G” (Pa)2.51 0.01 0.063.16 0.10 0.063.98 0.19 0.105.01 0.34 0.086.31 0.54 0.127.94 0.85 0.1310.00 1.32 0.1694Appendix BCase Study: Glucosamine Sup-plementation and Synovial FluidViscosity95Table B.1: Shear rheology data for Patient A both before and after glucosamine supplemen-tation was discontinued.˙(s1) Viscosity prior to discontinuation (Pas) Viscosity after discontinuation (Pas)0.010 10.612 0.2300.013 7.552 0.2310.018 4.735 0.2040.024 3.026 0.1830.034 1.773 0.1640.047 1.097 0.1500.065 0.833 0.1360.091 0.604 0.1270.126 0.422 0.1200.175 0.277 0.1130.243 0.212 0.1090.338 0.167 0.1070.470 0.152 0.1030.653 0.113 0.0950.907 0.089 0.0901.260 0.085 0.0841.751 0.070 0.0792.432 0.055 0.0743.380 0.049 0.0704.696 0.044 0.0646.525 0.041 0.0599.066 0.034 0.05412.600 0.031 0.04917.507 0.027 0.04424.323 0.024 0.03933.797 0.022 0.03546.960 0.020 0.03165.247 0.018 0.02790.657 0.016 0.024126.000 0.015 0.021175.067 0.013 0.019243.233 0.012 0.016337.967 0.011 0.014469.600 0.010 0.013652.467 0.009 0.011906.567 0.008 0.01096Table B.2: Shear rheology data for Patient B both before and after glucosamine supplemen-tation was started.˙(s1) Viscosity prior to glucosamine (Pas) Viscosity after glucosamine (Pas)0.010 0.612 0.9750.014 0.588 0.8640.019 0.534 0.8070.027 0.445 0.6430.037 0.350 0.5090.052 0.278 0.4020.072 0.231 0.3110.100 0.195 0.2580.139 0.171 0.2050.193 0.155 0.1760.268 0.142 0.1570.373 0.129 0.1420.518 0.121 0.1170.720 0.113 0.1041.000 0.104 0.0911.390 0.096 0.0801.931 0.088 0.0712.683 0.081 0.0643.728 0.074 0.0595.180 0.068 0.0547.197 0.061 0.04810.000 0.056 0.04313.900 0.050 0.03919.310 0.045 0.03526.830 0.040 0.03237.280 0.035 0.02851.800 0.031 0.02571.970 0.027 0.022100.000 0.024 0.020139.000 0.021 0.017193.100 0.019 0.015268.300 0.016 0.014372.800 0.014 0.012518.000 0.013 0.011719.700 0.011 0.0101000.000 0.010 0.00897Table B.3: WOMAC scores at first and second aspiration for Patient A. An improvementis observed in all categories.Pain (Section 1) Sti↵ness (Section 2) Function (Section 3) TOTAL1st Score 13 7 54 742nd Score 6 2 30 38Table B.4: WOMAC scores at first and second aspiration for Patient B. An improvement isobserved in all categories.Pain (Section 1) Sti↵ness (Section 2) Function (Section 3) TOTAL1st Score 5 6 22 332nd Score 3 2 16 2198Appendix CDevelopment of a Microfluidic Rheome-ter99Table C.1: The predicted, observed, and corrected pressure drop across the PDMS mi-crorheoemeter.Q (µL/min) Ppredicted (mbar) Pobserved (mbar) Pcorrected (mbar)10 188.7 208.9 194.620 377.5 386.1 364.230 566.2 567.2 544.740 755.0 740.0 723.050 943.9 905.7 900.060 1132.8 1066.0 1077.470 1321.7 1198.5 1229.280 1510.6 1368.2 1428.990 1699.5 1511.1 1603.8100 1888.5 1648.6 1777.5Table C.2: Estimated viscosity and corrected viscosity estimates for the microrheometer forwater at room temperature.˙(s1) ⌘observed (cP) ⌘corrected (cP)4.88E+03 1.110 1.0349.77E+03 1.026 0.9681.47E+04 1.005 0.9651.95E+04 0.983 0.9602.44E+04 0.962 0.9562.93E+04 0.944 0.9543.42E+04 0.910 0.9333.91E+04 0.909 0.9494.40E+04 0.892 0.9474.88E+04 0.876 0.944100


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items