9621_Approaches To Fracture Healing Under Inflammatory Conditions Infection And Diabetes

Yale University
Yale University
EliScholar – A Digital Platform for Scholarly Publishing at Yale
EliScholar – A Digital Platform for Scholarly Publishing at Yale
Yale Medicine Thesis Digital Library
School of Medicine
January 2020
Approaches To Fracture Healing Under Inflammatory Conditions:
Approaches To Fracture Healing Under Inflammatory Conditions:
Infection And Diabetes
Infection And Diabetes
Sean Vincent Cahill
Follow this and additional works at: https://elischolar.library.yale.edu/ymtdl
Recommended Citation
Recommended Citation
Cahill, Sean Vincent, “Approaches To Fracture Healing Under Inflammatory Conditions: Infection And
Diabetes” (2020). Yale Medicine Thesis Digital Library. 3887.
https://elischolar.library.yale.edu/ymtdl/3887
This Open Access Thesis is brought to you for free and open access by the School of Medicine at EliScholar – A
Digital Platform for Scholarly Publishing at Yale. It has been accepted for inclusion in Yale Medicine Thesis Digital
Library by an authorized administrator of EliScholar – A Digital Platform for Scholarly Publishing at Yale. For more
information, please contact elischolar@yale.edu.

Approaches to Fracture Healing Under Inflammatory Conditions: Infection and Diabetes

A Thesis Submitted to the
Yale University School of Medicine
in Partial Fulfillment of the Requirements for the
Degree of Doctor of Medicine

Sean Vincent Cahill

2020

2
Abstract

Non-union is a devastating complication of fracture and can be precipitated by abnormal
inflammatory states including infection and diabetes.
This thesis focuses on four related research problems that are addressed through original
scientific investigation and literature review. In addressing these questions, this dissertation
presents evidence for the following conclusions through in vivo animal models and using methods
including bacterial cell culture and counting, histology, radiography, and micro-computed
tomography:
1. Rifampin-loaded hydrogels decrease bacterial load and improve fracture healing in
a MRSA-infected open fracture model.
2. MRSA-infected nonunion is characterized by impaired chondrocyte maturation and
is associated with IL-1 and NF-KB activation.
3. Local teriparatide improves radiographic fracture healing in a type 2 diabetic mouse
model, but is inferior to systemic treatment.
4. Systemic administration of teriparatide, along with systemic antibiotics, improves
fracture healing in a diabetic, MRSA-infected mouse tibia fracture model.
This current work is not without limitation, and many aspects of this work are still in
progress. Nevertheless, the author hopes that this dissertation will serve as providing meaningful,
foundational data for future laboratory and clinical studies to improve our understanding of
inflammatory fracture healing and arrive at new therapies to advance the practice of fracture care.

3
Acknowledgements

The author would like to acknowledge the following for their mentorship, intellectual
contributions, technical assistance, and financial support of this thesis:

Francis Lee, MD, PhD for exceptional guidance, support, and encouragement to pursue
challenging and rewarding research; the Yale Department of Orthopaedic Surgery, especially
Jonathan Grauer, MD, Lisa Lattanza, MD, Gary Friedlaender, MD, Dieter Lindskog, MD,
Adrienne Socci, MD, Daniel Cooperman, MD, and Andrea Halim, MD; Lee lab members,
without whom this work would have been impossible, including Jungho Back, PhD, Hyuk-Kwon
Kwon, PhD, Yeon-Ho Chung, PhD, MD, Minh-Nam Nguyen, PhD, Kareme Alder, BS, Zichen
Hao, MS, Kristin Yu, BS, Christopher Dussik, BS, Inkyu Lee, and Saelim Lee; members of the
Tompkins Orthopaedics Research Department including: Mark Horowitz, PhD, Steven
Tommasini, PhD, Nancy Troiano, MS, and Jackie Fretz, PhD.

With much gratitude to his previous Yale Orthopaedics research mentors for their teaching and
encouragement: Cordelia Carter, MD, and Melinda Sharkey, MD.

Special thanks to the Office of Student Research including John Forrest, MD, Kelly-Jo Carlson,
Donna Carranzo, and Reagin Carney for their outstanding support and guidance.

Finally, a sincere thank you to all the faculty and residents of the Yale Department of
Orthopaedics and Rehabilitation for helping me start on my orthopaedic surgery career.

Lee lab members, spring 2019. From left: Hyuk-Kwon Kwon, PhD; Jungho Back, PhD; Zichen
Hao, MS; Minh-Nam Nguyen, PhD; Francis Lee, MD, PhD; Sean Cahill, BA; Kareme Alder,
BS; Kristin Yu, BS; Yeon-Ho Cheung, PhD.

4

Table of Contents

2. Introduction………………………………………………………………………………..7
1. Overview: Fracture healing is essential to human health……………………….…7
2. Bone quality in health and disease……………………………………………..….8
3. Normal fracture healing depends on controlled inflammation………………..…23
4. Infection and osteomyelitis: mechanisms and the inflammatory response………26
5. Consequences infected fracture: case presentation and treatment approaches..…29
6. Open fracture: minimizing infection risk with systemic and local strategies…….36
7. Diabetes is a pro-inflammatory condition that increases fracture risk………..…37
8. Diabetic fracture healing and the need for new treatment approaches……….….44
9. The role of murine models to study fracture healing and musculoskeletal
disease……………………………………………………………………………………………………..52
3. Purpose and specific aims……………………………………………………..…………53
4. Methods……………………………………………………………………………………55
1. Summary of experimental designs……………………………………………….55
i. MRSA infection and antibiotic hydrogels……………………………….55
ii. Diabetic fracture healing with local and systemic PTH…………………56
iii. Diabetic fracture healing under infected conditions……………………..56
2. Detailed methods
i. Animals…………………………………………………………………..57
ii. Type 2 diabetic mouse model and metabolic testing………………….…57
iii. Hydrogel preparation…………………………………………………….58
iv. Surgical open fracture model…………………………………………….59

5
v. Bacterial colony-forming unit analysis…………………………………..61
vi. Radiographic and histologic analysis…………………………………….61
vii. Immunohistochemistry…………………………………………………..63
viii. Biomechanical testing……………………………………………………64
ix. Statistics………………………………………………………………….65
5. Results……………………………………………………………………………………67
1. Rifampin-loaded hydrogels decrease bacterial load and improve fracture healing
in a MRSA-infected open fracture model………………………………………..67
2. MRSA-infected nonunion is characterized by impaired chondrocyte maturation and
is associated with IL-1 and NF-KB activation……………………………………74
3. A high-fat, high-sugar diet induces a type 2 diabetic phenotype characterized by
obesity, impaired glucose metabolism, increased infection burden, and poor
fracture healing characteristic of type 2 diabetes…………………………………80
4. Systemic and local PTH improves fracture healing in a type 2 diabetic mouse
model, but more data collection is required to fully evaluate this hypothesis……..87
5. Systemic administration of parathyroid hormone, along with systemic antibiotics,
improves fracture healing under infected conditions…………………………….89
6. Discussion……………………………………………………………………………………….93
1. Rifampin-loaded hydrogels reduce bacteria load and improve fracture healing in a
MRSA-infected, open fracture mouse model……………………………………93
2. MRSA-infected fracture is marked by poor chondrocyte proliferation and
maturation as well as IL-1 and NFKB inflammatory signaling………………….96
3. High-fat, high-sugar diet induces a mouse model of type 2 diabetes…………..99

6
4. Fracture healing is improved by systemic and hydrogel-delivered teriparatide
treatment in diabetic mice………………………………………………………102
5. Use of teriparatide to improve fracture healing in a MRSA-infected open fracture
model in diabetic and normal mice……………………………………………..103
6. Inflammatory fracture healing: summary, conclusions and future directions….107
7. References……………………………………………………………………………………..109

7
Introduction

I.
Overview: Fracture healing is essential to human health
Unlike repair mechanisms of nearly every other human tissue, bone fracture
healing has the potential to restore the original structure and physical properties without leaving
functional deficits, scar, or other evidence of previous injury [1]. The biologic process of fracture
healing is complex and requires mechanical stability, growth factors, stem cells, and other factors
in order to restore structure and function [2].
Successful fracture healing is essential to human health, as fracture is one the most common
traumatic injuries to humans [1,3]. Fracture nonunion and delayed union results in pain and
disability, and can be devastating for patient’s quality of life [4-5]. Specifically, in a 2013 study,
tibia shaft non-union resulted in a negative effect on mental and physical health that was worse
than congestive heart failure and equivalent to end-stage hip arthrosis [4]. In a similar study,
Schottel et al found that femoral fracture nonunion demonstrated a reduced quality of life similar
to type 1 diabetes, stroke, and acquired immunodeficiency syndrome [6]. Forearm and clavicle
nonunion resulted in the greatest degree of impairment, compared to femur, tibia, fibula, and
humerus fracture [6].
Fracture non-union also poses a major burden to our healthcare and economic systems. An
estimated 100,000 fractures result in non-union in the United States every year [7]. In the US,
additional healthcare costs due to tibia fracture nonunion range from $11,333 to $13,870 [8-9].
Indirect costs of nonunion, most notably productivity loss, account for the majority of the
economic burden resulting from fracture nonunion. Among Canadian and European healthcare
systems, these indirect costs make up for an estimated 67-79% and 82-93% of total costs burden,
respectively [7].

8
The overall fracture nonunion rate is cited to be approximately 5-10% in the orthopaedic
literature [10-11]. In a 2016 study of open long bone fractures, 17% progressed to nonunion and
an additional 8% demonstrated delayed union [12]. Non-union risk is variable and depends on
injury factors such as site, mechanism, and severity; and patient factors such as age, sex, and
comorbidities [13-14]. The incidence of non-union and delayed union is proposed to have
increased over the past decades due to improved patient survival and advances in medical and
surgical care following major injuries [15].
Many approaches to improving fracture healing have been investigated, from biologic and
surgical approaches to traditional medicine practices [16]. This dissertation will discuss
translational science approaches to improve fracture healing in altered inflammatory environments
including diabetes and infection. It will discuss the use of a locally-applied hydrogel to deliver
antibiotics and teriparatide under inflammatory conditions, using mouse models of infected
nonunion and diabetes. It will also identify key cellular processes and potential avenues for
targeted therapies. It is the author’s hope that these findings will enhance our understanding of
fracture non-union and move the field of orthopaedic surgery forward by providing a basis for
future clinical investigations.

II.
Bone quality in health and disease1

Successful fracture healing and underlying bone quality are closely related. Mesenchymal stem
cells, chondrocytes, osteoblasts, osteocytes, and osteoclasts form a tightly-regulated cellular
network that performs in the tasks of building and maintaining bone as well as fracture healing.

1 Based on: SC, Lee, FY. “Orthopaedic Tissues,” Orhtopaedic Knowledge Update 13, AAOS
2020. All text and figures in this thesis, including hand drawings, are original and were prepared
by the author, unless explicitly noted. The author acknowledges Dr. Lee’s guidance in preparing
and revising this portion of the text.

9
Hormonal regulation is essential, with mesenchymal progenitor cells playing major signaling roles.
This section will investigate the normal workings of this cellular network of orthopaedic tissues
and how it can fail in diseased states such as smoking and cancer. This section will present basic
components of bone biology that are relevant to fracture healing, diabetic bone disease,
methodologies, and findings presented in this dissertation.

Figure 1. Major transcription factors and regulators of bone cell differentiation

Mesenchymal stem cells differentiate via a stepwise progression into chondrocytes, adipocytes,
osteoblasts, osteocytes, tenocytes, and myocytes. Osteoclast arise from the monocyte lineage of
hematopoietic stem cells. A host of transcription factors, genes, and growth factors regulate

10
differentiation. Activation via RANK-L and inhibition by OPG, both expressed by pre-osteoclasts,
are major regulators for osteoclastogenesis.

Orthopaedic tissues are derived from pluripotent stem cells which become increasingly
more specialized (figure 1). Bone is unique in that regulation of cellular and metabolic processes
occur primarily at the level of the stem cell. Osteoblasts, chondrocytes, adipose cells, fibroblasts,
and myocytes share the mesenchymal stem cell as the common precursor, while osteoclasts are
derived from the macrophage/monocyte lineage of hematopoietic stem cells. Altered development
and function of these precursor lineages underly many of the processes that alter bone quality and
fracture healing potential.
Variable expression of transcription factors facilitates stem cell differentiation into
terminal lineages to form orthopaedic tissues as cellular migration and ossification take place [17].
Runx2 and osterix are essential for differentiation of the osteoblast lineage. Sox5, 6, and 9 are
markers of chondrocyte development, with Sox9 having been identified as an essential regulator
[18] (Figure 1). These signaling pathways involved in mesenchymal stem cell differentiation have
important consequences for fracture healing under infected conditions (pages 75-80, 97-100). The
Wnt/-catenin pathway is one of the most important signaling pathway for regulating bone
formation, leading MSCs towards osteoblastic differentiation and suppressing adipose
development, and is altered in diabetes (page 42).
Ossification is a foundational principle that underlies both skeletal development and
fracture healing. During human development, contact between mesenchymal cells and epithelial
cells triggers pre-osteoblastic differentiation and intramembranous ossification, during which
mesenchymal cells differentiate directly into periosteum and osteoblasts [19]. During

11
endochondral ossification (Figure 2), mesenchymal tissue develops into bone from a cartilage
template [20]. Chondrocytes proliferate and undergo hypertrophy and apoptosis, and the remaining
matrix is mineralized and invaded by vasculature. Systemic factors, such as growth hormone and
thyroid hormone, and local factors such as Indian hedgehog and PTHrP, promote and regulate
these processes (Figure 2). Woven bone, secreted by osteoblasts, is eventually replaced by
lamellar bone. After a rudimentary skeleton is formed, osteoblasts and chondrocytes undertake
skeletal modeling to shape the skeleton and improve its strength and resilience. More information
about the ossification process as it relates to fracture healing can be found on pages 23-26.
Secondary ossification widens bones, with peripheral growth from the apophysis. In
contrast to primary ossification, which begins in the embryonic stage and continues through
adolescence, secondary ossification only begins during the post-natal period.
Bone Cellular Biology. Bone is a rich, biologically active tissue. Osteoblasts, osteocytes,
and osteoclasts maintain and renew the bony matrix and are involved in systemic processes such
as mineral metabolism. An understanding of bone cellular biology is essential for understanding
the mechanisms behind fracture healing and diabetic bone disease.
Mature osteoblasts contain abundant rough endoplasmic reticulum for collagen synthesis,
as well as an extensive Golgi apparatus. Osteoblasts synthesize bone through type I collagen
secretion and production of osteoid (unmineralized matrix). Parathyroid hormone stimulation and
Runx2 expression induce the expression of alkaline phosphatase, type 1 collagen, and bone
sialoprotein II in the preosteoblast stage [21] (Figure 1). Transcriptional activation of RUNX2 and
osterix result in osteoblast differentiation, allowing for matrix mineralization and expression of
other proteins such as osteocalcin to occur. Experimental evidence for the importance of RUNX2

12
in fracture healing is given on page 75-81. Osteoblasts create a basic environment with alkaline
phosphate that helps catalyze calcium-phosphate crystal deposition.

Figure 2. Endochondral Ossification and Cartilage Differentiation
Endochondral ossification occurs at the physis, during which chondrocytes undergo proliferation,
hypertrophy, and apoptosis. The matrix left behind is mineralized and invaded by blood vessels.
Growth factors, including growth hormone, thyroid hormone, and FGF3, promote osteogenesis.
Indian hedgehog and PTHrP create a feedback loop to modulate and regulate chondrocyte
proliferation and hypertrophy. Fracture healing, as discussed in this thesis, relies on this process.

13
Osteocytes, which comprise over 90% of all bone cells in adults, are differentiated from
osteoblasts [22]. Our understanding of osteocytes has dramatically increased over the past decade,
especially our understanding of the mechanisms of response to mechanical loading and their role
in bone metabolism. The process of the osteoblast becoming embedded in bone lacunae induces
changes genetic expression to induce osteocyte differentiation, including participation in
mineralization regulation and development of dendritic processes [22]. The expression of
membrane type 1 matrix metalloproteinase is necessary for canaliculi formation, through which
osteocytes form an extensive intercellular network of dendritic processes that directly
communicate via gap junctions [23]. The lacunar-canalicular system contains circulating fluid that
provides osteocytes with oxygen and nutrients and allows osteocytes to sense acute deformation
of the bone matrix, inducing release of anabolic factors to increase bone mass in response to strain.
These lacunar networks are significant in the setting of osteomyelitis, as it will be shown (Page
75) that the lacunae are the main harboring site of bacteria following MRSA-infected fracture.
Osteocytes are important regulators of bone metabolism. In addition to sensing fluid
microcurrents and responding to bone matrix strain, osteocytes produce sclerostin and Dkk1,
which are potent Wnt inhibitors and therefore key negative regulators of osteoblastogenesis [22].
Sclerostin and Dkk1 monoclonal antibodies are under investigation as a potential means of
increasing bone mass and improving fracture healing, especially in diabetes [24]. A more thorough
discussion of the roles of sclerostin and Dkk1 in the altered inflammatory environment of diabetic
bone disease is provided on page 42.
Osteoclasts, derived from hematopoietic stem cells, appear as large, multinucleated cells
housed in Howship’s Lacunae, microscopic grooves on the bone surface. The ruffled border seals

14
the bone surface, creating a closed microenvironment in which degradation products such as acid
(produced by carbonic anhydrase and H+ ATPase) and cathepsin K degrade the bone matrix [25].
Osteoclast development and differentiation is tightly linked to the osteoblastic lineage, and
dysregulation leads to bone pathology (Figure 1). Commitment to the osteoclastic lineage from
hematopoietic precursors is induced by macrophage colony stimulating factor, c-fos transcription
factor, and RANK expression [26]. RANK-ligand, expressed by osteoblasts, is required for pre-
osteoclast differentiation into osteoclasts. Osteoprotegerin (OPG), produced by cells of the
osteoblast lineage as well as some hematopoietic cells, serves as a decoy receptor for RANK-L
and competes with osteoclast RANK receptor; increased OPG decreases osteoclast differentiation
and activation. Systemic factors help to regulate RANK-L and OPG production, including TNF-
, a catabolic factor, and PTH and estrogen, anabolic factors. Factors that contribute to osteoclast
activation are found in the setting of infected fracture (pages 27-29).

Although osteoclast differentiation is tightly regulated by the osteoblastic lineage,
osteoclast activity is also influenced by hormones. Resorptive activity is increased by vitamin D,
PTH, PTHrP, and prolactin, and decreased by estrogen, calcitonin, and TGF-. Cytokines also
regulate osteoclast activity. IL-17 has been shown to decrease bone resorption; IL-6 and TNF-
increase resorption [26]. Parathyroid hormone is the main anabolic agent studied in this
dissertation, and is used to improve fracture healing in normal and diabetic mice. More detail about
the action of parathyroid hormone can be found on pages 49-50.

Marrow adipocytes have long considered inert place holders in the marrow space of long
bone, ribs, sternum, and vertebrae. However, marrow fat has been increasingly recognized as an
important, active element of the bone cell milieu and is considered a distinct type of adipocyte (as
opposed to white, brown, or beige fat found elsewhere in the body). As discussed in a 2017 study,

15
it has been shown that marrow adipocytes increase with age and are tightly correlated with reduced
bone mass [26]. These regulatory effects on bone mass are thought to be induced through
modulation of PPAR and RUNx2 proteins [27]. Thus, marrow adipocytes and associated
signaling pathways have become sought-after targets for bone disease therapies. In addition to
bone metabolism, it is thought that marrow adipocytes influence other cell populations within the
marrow and negatively affect hematopoiesis [26]. The adipocyte lineage is important to the
pathogenesis of diabetic bone disease, as discussed on page 44.
Matrix Composition. The cellular components produce and maintain a mineralized bone
matrix, which serves as the functional component of bone. The extracellular matrix is
heterogeneous and structured. Mineral and organic components together impart strength and
rigidity (Table 1a). The composition of bone varies with age, gender, and ethnicity. Health status
can also affect matrix composition, and alterations of the matrix underlie diseases such as
osteogenesis imperfecta and osteoporosis [28].
The mineral calcium hydroxyapatite, Ca10(PO4)6(OH)2, is the most abundant substance in
bone, 60-70% of its mineral composition. Other abundant minerals include sodium, magnesium,
and bicarbonate [28]. The organic component of bone, or osteoid, stabilizes the extracellular
matrix, facilitates calcification and mineralization, and provides tensile strength. Type I collagen
is the dominant organic substance of bone, comprising 90% of total protein and the second most
abundant substance following hydroxyapatite. The collagen triple helix, characterized by glycine-
X-Y repeating sequence, is highly crosslinked, providing elasticity. Fibronectin, another important
structural protein of the bony matrix, helps develop and maintain the structure of the collagen
network. Non-collagenous proteins, such as proteoglycans and osteocalcin, also play various roles
[29] (Table 1). The quality of bony matrix is determined by a number of factors, but one crucial

16
factor is accumulation of advanced glycosylation end products (AGEs) that occurs in the setting
of diabetes; a detailed discussion is provided on page 40-41 and figure 9.

Bone metabolism: As impact and activity imparts microscopic damage, osteoclasts and
osteoblasts work in concert to remove old bone and replace it with new bone to maintain strength
and integrity. This process of remodeling is ongoing, working to replace approximately 10% of
the skeleton every year and replacing the entire bone mass every ten years [30].
Remodeling is a tightly regulated process, as osteogenesis is intimately coupled to
osteoclastogenesis. Remodeling imbalances can result in systemic disease such as osteoporosis as
well as local bone destruction as in cancer metastasis. Bone loading affects the rate of bone
remodeling as stated by Wolff’s Law, that mechanical stress results in greater bone density and
strength. Systemic and hormonal factors, such as parathyroid hormone and 1,25-dihydroxy vitamin
D, modulate remodeling by inhibiting bone resorption and promote differentiation of osteoblasts
and osteoclasts [31-32]. Direct communication between osteoblasts and osteoclasts is also
achieved through release of local factors from the bone matrix itself. As osteoblasts degrade the
bony matrix, proteins including TGF-, platelet-derived growth factor, and fibroblast growth
factor are released to stimulate osteoblasts and thus enhance bone formation. Micro-RNA
modulation has also been identified as a significant regulator of bone remodeling [33]. Bone
remodeling is essential not only in normal skeletal maintenance, but also in the later stages of
fracture healing; more detail can be found on pages 24-25.
Anatomy and Structure. Long bones allow for mechanical motion of the extremities and
are formed by endochondral ossification (Figure 2). The epiphysis, covered by articular cartilage,
forms joint surfaces (Figure 3). The physis, distal to the epiphysis, is the location of endochondral

17
ossification, and is located between the epiphysis and the metaphysis. The apophysis, a feature of
both long and flat bone, is an area of secondary ossification.
Periosteum is a thin, membranous tissue that surrounds both long and flat bones, and
endosteum is the corresponding inner surface. The periosteum contains a rich vasculature that
provides primary blood supply to bone, making it essential for fracture healing. It is also the site
of muscle and tendon attachment.
Cortical bone is dense, compact tissue which functions to provide mechanical rigidity.
Haversian canals run parallel to the diaphysis along the mechanical axis of bone. These spaces
house nerves and microvasculature that supply the bone tissue. Laminae are discreet, concentric
sheets of bone that surround the Haversian canal. Volkmann canals allow for communication
between periosteal vessels and Haversian system. Osteocytes are housed in lacunae and their
dendritic processes communicate via small canaliculi. Cutting cones comprise the remodeling unit
of cortical bone, with osteoclasts forming a canal along the longitudinal axis of bone and
osteoblasts following to close the gaps [32]. The mouse, an animal model used to fracture healing
under inflammatory conditions throughout this dissertation, lacks the haversian system found in
human bones.
Trabecular (or cancellous) bone is a lower-density tissue. Trabecular bone has a porous,
sponge-like structure which houses marrow elements, including hematopoietic stem cells and
marrow fat. Remodeling occurs directly on the surface of the trabeculae, with osteoclasts forming
lacunae which are subsequently filled by osteoblasts [32]. The trabecular network can be assessed
by micro-computed tomography, a major method used to assess the quality of fracture healing in
this dissertation; please see page 62 for more information.

18

.

Figure 3. Anatomy of Long Bone
The major regions of long bone include the epiphysis (nearest to the joint), physis, metaphysis,
and diaphysis. Blood supply via nutrient arteries is derived from the periosteum, which is also a
major source of stem cells in fracture healing. Endochondral ossification occurs at the physis.

19

Cartilage, synovial membrane, and synovial fluid compose synovial joints. Together, they
provide lubrication and protection of joint surfaces and allow for movement along a virtually
frictionless surface. Inflammation and degradation of these tissues underlie arthritic processes and
result in debilitating pain and deformity. Interleukin 1 (IL-1), is a major inflammatory mediator
that has been demonstrated to cause cartilage destruction in osteoarthritis. A more complete
discussion of the proposed role of chondrocyte differentiation and IL-1 in fracture healing and
infected nonunion is provided on page 97.
The chondrocyte is the cellular component of cartilage, and is crucial to fracture healing.
A major portion of this dissertation is devoted to characterizing changes in chondrocytes and
cartilage in the setting of infected fracture (pages 75-80, 97-100). The chondrocyte comprises only
a small amount of total articular cartilage mass, approximately 5% of its dry weight. Cartilage is
composed primarily of water and cross-linked type II cartilage, secreted by chondrocytes (Table
1). Proteoglycans, most abundant in the deep layer, trap fluid and contribute to pressurization and
compressive tissue strength. SOX5 and SOX6 expression, along with interactions with
CEBP/p300, stimulate chondrocytes to secrete type II collagen and proteoglycans [34]. In addition
to secretion, chondrocytes regulate cartilage homeostasis by expression of metalloproteinases that
break down the cartilage matrix. Components of the cartilage extracellular matrix, especially type
XI collagen, have been identified in a 2017 study as playing a role in inducing chondrogenesis
[35].
Chondrocytes undergo terminal differentiation through the process of hypertrophy, marked
by cellular swelling, decreased proliferation, and eventual apoptosis. While chondrocyte

20
hypertrophy and death allow for bone formation in developing bone, a 2018 study reported that
hypertrophy has been shown to be an essential step in the pathogenesis of osteoarthritis and marks
the beginning of irreversible degradation to the cartilage matrix [34].

The quality of tissues including bone, cartilage, and ligaments plays a significant role in
prevention and treatment of skeletal disease. Physiologic causes, such as age, and pathologic states,
such as diabetes mellitus and metastatic bone disease, can decrease the integrity of these tissues,
resulting in increased fracture incidence and decreased surgical outcomes. A host of drugs, such
as corticosteroids, can also affect bone quality.
Bone health and bone density can be affected by a host of physiologic, pharmacologic, and
demographic factors that affect the balance of bone resorption and formation (Figure 4). While a
comprehensive review of causes of impaired skeletal health are out of the scope of this
introduction, a few key subjects, including aging, malignancy, and smoking are included here, as
they are common in the clinical setting affect fracture risk and healing potential. A thorough review
of the role of diabetes, a major focus of this dissertation, is provided on pages 37-51.
Aging: Aging bone is characterized by decreased bone mass due to progressively
dysregulated remodeling. With increasing age, the amount of new bone formed with each
remodeling cycle decreases slightly, while the amount of resorbed bone remains constant. This is
associated with increased osteocyte apoptosis and decreased number of precursor cells. Other
factors, such as decreased mechanical loading of bone and accumulation of reactive oxygen
species, may also contribute to decreased osteocyte function. In addition to decreased bone mass,
the bone quality changes with aging. As discussed in a 2017 study, collagen becomes increasingly
crosslinked, resulting in bone that is more brittle with a disrupted mineralized matrix [35]. Bones

21
become slenderer with decreased trabecular bone mass, thinning cortices, and changes in center of
gravity [36]. Overall, these changes contribute to greater risk of fracture in the elderly.

Figure 4. Bone Quality Determinants 2
Bone mass is determined by a balance between bone formation and resorption. Peak bone mass is
achieved in early adulthood, and slowly declines due to decreased bone synthesis during each
remodeling cycle.

Osteoporosis is a bone disease distinct from aging that is marked by decreased bone
mineral density, especially in trabecular bone. The hallmark of osteoporosis is fracture

2 Original figure was prepared by Dr. Lee and modified by the author.

22
predisposition, including non-traumatic fractures and vertebral body fractures. While osteoporosis
is an age-associated disease, it is generally not considered a disease of aging, as osteoporosis can
(although rarely) affect the young and does not affect all elderly individuals. Treatment generally
is targeted against bone resorption, and including bisphosphonates, hormone therapy, and
calcitonin. Anabolic agents, such as teriparatide (recombinant parathyroid hormone), are also used.
Preventative measures, including weight-bearing exercise and vitamin D supplementation, is also
recommended for patients at risk of developing osteoporosis. Patients with diabetes can present
with osteoporosis, especially type 1 diabetics early in life. Please see pages 37-40 for more
information on diabetes-induced osteoporosis.
Malignancy: Bone destruction can be a devastating source of morbidity in cancer. Two
major causes of bone destruction will be covered here: radiotherapy and metastasis to bone.
Radiation therapy, common bone and soft tissue tumor treatment, can cause inflammation,
vascular fibrosis, and reduced tissue circulation. Bone necrosis can result from hypoxic and
hypovascular conditions, tissue breakdown, and disrupted wound healing [46]. These conditions
raise the risk for pathologic fracture and allow for development of chronic infections that are poorly
responsive to systemic antibiotics. A case presentation on fracture and infection following tumor
resection and radiation therapy can be found on pages 29-35.
Malignant bone metastases, an incurable progression of a primary tumor, can lead to pain,
fracture, and hypercalcemia due to dysregulated bone remodeling [47]. Lung, breast, and renal
cancer, as well as multiple myeloma, are best known for causing osteolysis. A host of pathways
contribute to tumor-driven bone destruction, including increased expression of RANK-L, matrix
metalloproteinases, and PTHrP. Tumor cells can also undergo osteoclastic mimicry by fusing with
osteoclast precursors, gaining the ability to participate in bone resorption. Furthermore, osteolytic

23
matrix destruction facilitates cancer progression via TGF-B release, increase calcium, and
hypoxia. Prostate cancer bone metastasis, on the other hand, is marked by pathologic production
of immature, woven bone. Tumor cell epithelial-to-mesenchymal transdifferentiation and
osteomimicry are the major osteoblastic pathways involved.
Smoking: Smoking is a leading cause of preventable morbidity and mortality worldwide,
and has significant, deleterious effects on bone health. Smoking generates reactive oxygen species,
impairs mitochondrial activity, impairs fibroblast migration, and reduces blood flow to sites of
injury [48]. Smoking is a risk factor for low bone density, with recent animal models demonstrating
increased osteoclast numbers and impaired bone growth in response to long-term cigarette smoke
exposure [49]. Smokers demonstrate greater time to union and impaired chondrogenesis following
fracture, as well as higher rates of spinal fusion failure and pseudarthrosis compared to non-
smokers [48,50]. In addition to impaired bone integrity, smokers carry significantly increased risk
of infection following trauma [51].

III.
Normal fracture healing is an inflammation-dependent process.3

Fracture healing is a complex process that results from tightly-ordered interactions of cells,
growth factors, and extracellular matrices [52].6 Fracture healing is broadly classified as secondary
or primary, delineations that refer to underlying healing mechanisms dependent on rigidity of the
fracture construct. Primary fracture healing occurs when bone fragments are in direct contact with
one another and are highly stable with minimal strain [52]. Haversian remodeling facilitates
healing of the fracture site, dependent on osteoclasts and osteoblasts to directly repair the bone
fragments. Secondary healing is more common, and is characterized by four steps: 1) the initial

3 The author acknowledges the assistance of Yeun-Ho Cheung, PhD, in revising the first two paragraphs of this
section.

24
inflammatory phase, 2) soft callus formation, 3) hard callus formation, and 4) remodeling [52-54] (Figure 5).

Figure 5. Phases of Normal Fracture Healing.
Schematic of secondary fracture healing, which progresses through step-wise phases to restore
original strength and shape to bone without leaving a scar. The inflammatory phase is crucial,
during which the presence of hematoma from disrupted blood vessels forms a clot, attracting
immune cells and mesenchymal progenitor cells via inflammatory signaling. Soft-callus, primarily
made of cartilage, bridges the bone gap. New bone formation follows the process of endochondral
ossification. The bone’s normal shape is returned by remodeling via osteoclasts osteoblasts.

The early inflammatory phase of secondary fracture healing, prior to soft callus formation, is
considered to be crucial to the entire healing process [55]. Blood vessel disruption in the bone and
surrounding soft tissue results in hematoma formation, which incites an inflammatory reaction.
Conversion of fibrinogen to fibrin occurs in the hematoma, allowing for trapping of innate immune