C hapter 8

Osseointegration of Implants
Christer Dahlin and Carina B. Johansson

8.1 General Overview – approaches (Steinemann, 1998) as well as the

Evaluating Osseointegration clinical observations; the latter one is probably
of most relevance since it describes “clinical
Osseointegration was coincidently discovered measures” (Zarb and Albrektsson, 1991). The
more than 40 years ago when Brånemark and definition of osseointegration in Dorland’s Medi-
colleagues were about to retrieve a titanium cal Dictionary today is as follows: “the formation
implant, which had been used for evaluating of a direct interface between an orthopedic or
microcirculation in bone, and experienced a very dental implant and bone without intervening
firm bone anchorage of the screw, i.e., the soft tissue” (Dorland’s Medical Dictionary, 2007),
implant was osseointegrated. Still, there are no which is quite similar to the original light micro-
“exact numbers” or “exact values” that describe scopic level definition in 1981.
the osseointegration phenomena (Brånemark et Nevertheless, from time to time, the question
al., 1969). The term osseointegration was coined arises as to when is an implant osseointegrated
and for the first time defined in 1981 (Albrekts- and to what extent should it be integrated in
son et al., 1981). The original definition was order to call it osseointegrated? To the best of our
based on observations at a light microscopic knowledge this is still not known. Bone is a living
level, although it did not indicate what percent- and dynamic tissue being constantly remodeled.
age of bone needed to be in contact with the There are no exact measures and no exact
implant in order to call it osseointegration. Over requirements set up in terms of how much bone
the years osseointegration has been redefined to needs to be in contact with the implant in order
span a range of criteria from the original light to fulfill the criteria for osseointegration. The
microscopic level to more precise engineering major focus today is on the implant surface mor-

103
Osteology Guidelines for Oral and Maxillofacial Regeneration

phology (including all surface parameters, i.e., 8.2 The Rat Model
roughness, chemistry, oxide type, oxide thick-
ness, etc.) and we have no exact knowledge of 8.2.1 Aim
which surface structural parameter, or the combi- The use of rat models for studying implant inter-
nation of parameters, is the most important one faces is a controversial issue in the literature.
to exactly describe and guide osseointegration Although the rat is one of the most commonly
(Coelho et al., 2009; Dohan Ehrenfest et al., used species in medical research, it has not been
2010). This is even more complicated today since extensively used for implant research mainly due
the targets of research are not only at the micro to concerns regarding limitation in size and dis-
cellular level but also at the nano-protein and similarities between rat and human bone (ISO
gene expression levels. Hence for the latter levels [International Organization for Standardization]
of investigations of osseointegration we are only 10993–6, 1994; Pearce et al., 2007). At the
in the beginning stages, and more tools and tech- Department of Biomaterials in Gothenburg, we
niques are emerging rapidly which may well lead are of a different opinion and see several advan-
to further redefinitions of osseointegration in the tages of using a rat model for screening of novel
near future. The research around osseointegra- implant surfaces and detailed analysis of tissue
tion is as fascinating now as in the early days! dynamics. The use of microimplants in the rat
The development of an optimal interface offers unique opportunities to combine state of
between bone tissue and an orthopedic or den- the art histology, biomechanical torsion tests
tal implants is an ongoing process. In order to and analysis of gene expression in the same ani-
determine whether a newly developed implant mal model (Omar et al., 2009; 2010a,b). The rat
material conforms to the requirements of bio- model is also suitable for studying osseointegra-
compatbility, mechanical stability, and safety, it tion in compromised bone situations, i.e., radia-
must undergo rigorous testing both in vitro and tion damage (Nyberg et al., 2010). Last but not
in vivo. As per the state of the art in biomaterial least, these types of study can also be reasona-
research today, characterization of bone-con- bly affordable due to the relatively low cost of
tacting materials is initiated by means of in vitro animal and housing.
testing. The advantages of such procedures
include a rapid response with regard to toxicity 8.2.2 A
 dvantages/Disadvantages
and cytocompatbility. Other reasons for the of the Rat Model
increasing popularity of in vitro testing within
the medical field are the principles of animal Advantages:
reduction. It is accepted that in vitro testing is • Low cost animal and housing.
used as a first stage for acute toxicity and cyto- • Easy surgical access. Few complications.
compatibility to avoid the unnecessary use of • Sufficient number of animals in groups.
animals. • Access to wide range of antibodies etc. for
In order to evaluate dental (and orthopedic) sophisticated analysis.
implants, it is essential to have a range of animal • Possibility to combine histology, biomechan-
models suitable for various analysis. The present ical testing, and gene expression in same ani-
chapter will describe suitable animal models for mal model.
dental implant research on three different levels
as well as discuss the very special circumstances Disadvantages:
that exist when performing histological analysis • Limitation in size = customized implants
involving hard and soft tissue in combination • Dissimilarities between rat and human bone
with an implanted biomaterial. in tissue dynamics.

104
 8.2 The Rat Model 8

8.2.3 Timing
Important time points from the project start up
to the collection of specimens and data analysis
(graphical): ordering of animals, quarantine sur-
gery, collection of specimen and preparation of
sections.
• 2 to 3 weeks usually for ordering of animals
(with huge individual variations between coun-
tries).
• Quarantine 1 week.
• Surgery x days.
• Healing period 0–x weeks.
• Specimen collection and preparation of sec- Fig 8-1   Installation of two microimplants in the rat tibia
tions: 2 to 3 months. under clean conditions.

8.2.4 Surgical Procedure

After shaving and cleaning with chlorhexidine
5 mg/mL in 70% ethanol, the medial aspect of
either the distal femoral epiphysis or tibial meta- and Coleman, Hull, UK) subcutaneously prior to
physis is exposed via an anteromedial skin inci- the implantation and daily postoperatively. Yet
sion. This is followed by skin and periosteum another method of rat anesthesia is to use a
reflection with blunt instruments (Fig 8-1). In the freshly prepared cocktail of fentanyl + mida-
present model, specially manufactured implants zolam (2 parts sterile water, 1 part fentanyl/flu-
2 mm in diameter and 2.3 mm in length are used anizone [Hypnorm® Vet, Saunderton, UK] and 1
(Fig 8-2). The implant sites are prepared by using part midazolam, 5mg/mL, Dormicum®, Roche,
1.4 and 1.8 mm spiral burs, respectively. All drill- France). The cocktail is administered by intra-
ing is done under profuse irrigation with 0.9% peritoneal injections to each animal initially
NaCl. Implants are installed manually. The surgi- with dose of 2.7 mL/kg (Flecknell, 1996). Addi-
cal wounds are closed in layers. The subcutane- tional anesthesia is given when needed.
ous layer is closed with resorbable polyglactin
sutures (5-0 Vicryl®, Ethicon, Johnson and John- 8.2.6 Detailed Methodology
son, Brussels, Belgium). See Section 8.5.

8.2.5 Preparation – Anesthesia 8.2.7 Postoperative Care

Sprague-Dawley rats (200 to 250 g) are recom- For pain relief, see above. After surgery, the ani-
mended for the model. The animals can be fed mals are allowed free postoperative movement
on standard laboratory diet with pellets and with food and water ad libitum.
water ad libitum.
Anesthesia is performed using a Univentor Retrieval Procedure
400 anesthesia unit (Univentor Ltd, Zejtun, Sacrifice by an intraperitoneal overdose of
Malta) under isoflurane (Isoba® Vet, Schering- sodium pentobarbital (60 mg/mL; ATL Apoteket
Plough, Uxbridge, UK) ventilation. Anesthesia Production & Laboratories, Kungens Kurva,
should be maintained by continous administra- Sweden) under anesthesia with a 0.5 mL mix-
tion of isoflurane via a mask. Each rat receives ture of pentobarbital (60 mg/mL), sodium chlo-
an analgesic (Temgesic®, 0.03 mg/kg, Reckitt ride, and diazepam (1:1:2).

105
Osteology Guidelines for Oral and Maxillofacial Regeneration

a b

c d
Fig 8-2   Radiographic image and histologic sections of the implant insertion sites in rat bone; one implant placed in the
femur condyle region and two implants in the tibia region. The implant diameter is 2.0 mm. The proximal implant (c) has
been removal torque tested prior to histological sectioning.

Healing Periods 8.3 The Rabbit Model

This model is suitable for early analysis as well
as more long-term studies. In brief, this means 8.3.1 Aim
that healing dynamics can be studied from day The rabbit is one of the most commonly used
1. Solid osseointegration can be expected even animals for dental implant research. It is used in
around 4 weeks postoperatively. approximately 35% of all musculoskeletal
research studies (Neyt et al., 1998). The rabbit
8.2.8 Endpoint Measurements attains skeletal maturity at around 6 to 8 months
See Section 8.5. of age (Gilsanz et al., 1988). This can be studied
by radiographs of the metaphysis (Fig 8-3).
8.2.9 Statistical Analysis Plan The most commonly used experimental sites
See Chapter 4. are the femoral diaphyseal bone and the tibial
bone. The former is considered more ideal for
8.2.10 Materials, Consumables, Equipment studying bone dynamics and the interaction
See Section 8.5. between surfaces and adjacent tissues. The

106
 8.3 The Rabbit Model 8

a b
Fig 8-3   Radiographs of rabbit bone. (a) The rabbit is 7 months old. Note the open growth zones. (b) The rabbit is 10
months old with closed epiphyseal lines.

Fig 8-4   Installation of three implants in the rabbit tibia. Fig 8-5   A specimen of a rabbit hindleg with one implant
inserted in the femur condyle region and three implants
placed in the tuberositas tibia region. The mean implant
diameter is 3.7 mm. Illustration courtesy of Dr Young-Taeg Sul.

tibial bone is also suitable for some mechanical sized implants such as a bone harvest chamber
testing including torque removal tests and (BHC) can also succesfully be used. The BHC is
push-out tests. These methods will be described an implant with a penetrating canal that per-
in more depth in Section 8.5. There are certain mits bone ingrowth. The ingrown tissue may be
size limitations associated with the use of rabbit harvested after removal of a titanium lid with-
models. The international standard for the bio- out disturbing the anchorage of the implant.
logic evaluation of medical devices recom- This device is approximately 6 mm in dimen-
mends a maximum of six implants (three test sion. Hence, only one device per site can be
and three control implants) (ISO 10993–6, used (Albrektsson et al., 1981).
1994). In order to avoid pathologic fractures of
the test sites, the same ISO report suggests that 8.3.2 Advantages/Disadvantages
cylindrical implants placed into the rabbit tibial of the Rabbit Model
and femoral diaphyseal bone should be no The femoral and tibial bone of the rabbit has
larger than 2 mm in diameter and 6 mm in several advantages. The dimensions and the
length (Fig 8-4 and Fig 8-5). However, larger- anatomy of the bone corresponds fairly well

107
Osteology Guidelines for Oral and Maxillofacial Regeneration

with the edentulous jaw in humans. It is easily 8.3.6 Detailed Methodology

accessible and seems to generate a minimum of See Section 8.5.
morbidity for the animals.
Prior to surgery the animals should be kept 8.3.7 Postoperative Care
in a purpose-designed room and be fed ad libi-
tum with water, a standard laboratory animal Postoperative Antibiotics and Analgesics
diet, and carrots. The animals should be given antibiotics (Inten-
penicillin 2,250,000 IU/5 mL 0.1 mL/kg body
8.3.3 Timing weight, LEO, Helsingborg, Sweden) for 3 days.
• Important time points from the project start Temgesic® 0.05 mg/kg (Reckitt and Colman,
up to the collection of specimens and data NJ, USA) is given as single intramuscular injec-
analysis (graphical): tion for pain relief for 3 days.
• Ordering of animals 3 months in advance
(huge variations in different countries 8.3.8 Endpoint Measurements
• Quarantine approx 2 weeks See Section 8.5.
• Surgery, collection of specimen, x = weeks
• Preparation of sections 2 to 3 months 8.3.9 Statistical Analysis Plan
See Chapter 6.
8.3.4 Surgical Procedure
After shaving and appropriate cleaning with, 8.3.11 Materials, Consumables, Equipment
e.g., Panadyme®, the tibial or femoral bone is See Section 8.5.
exposed via a skin incision followed by a careful
subperiosteal incision. It is advisable to slightly
overlap the skin–periosteal flap in order to 8.4 The Canine Model
secure an optimal primary closure. The perio-
steal flap and the skin should be closed in sepa- 8.4.1 Aim
rate layers with slow resorbable sutures (4-0 or The dog is one of the more frequently used
5-0) (see Fig 8-8 and Fig 8-9 below). large animal species for musculoskeletal and
dental research. It has been found that there is
8.3.5 Preparation – Anesthesia most similarity in bone composition (ash
• General anesthesia is given by an intramuscular weight, hydroxyproline, extractable proteins
injection of fluanison and fentanyl (Hypnorm®, and insulin growth factor-1 content) between
Janssen Pharmaceutica, Brussels, Belgium) the dog and humans (Aerssens et al., 1998). In
0.2 mg/ kg body weight and intraperitoneal terms of bone density the dog most closely rep-
injection of diazepam (Stesolid®, Dumex, Copen- resent the human situation (Pearce et al.,
hagen, Denmark) 1.5 mg/kg body weight. 2007). Depending on the size and the breed of
• Additional Hypnorm® should be adminis- dog, there may be some discrepancy in the size
tered when needed. and shape of the canine bones. However, com-
mercially dental implants can be utilized in par-
Local Anesthesia ticular in the mandible. It should be mentioned
Local anesthesia should always be given at the that there are increasing ethical issues relating
surgical site. Both for hemostasis as well as to the use of dogs in medical research due to
postoperative pain prophylaxis (1 mL 2.0% lido- their status as companion animals. This model
caine/epinephrine solution, Astra AB, is ideally suited for studies on functional load-
Södertälje, Sweden). ing of implants.

108
 8.4 The Canine Model 8

Fig 8-6   Placement of standard Brånemark Implants® in Fig 8-7   Radiographs demonstrating implants placed in
edentulous mandible in conjunction with bone augmenta- mandibular bone of dog.
tion study.

8.4.2 A
 dvantages/Disadvantages of the Canine Implant Selection
Model According to ISO recommendations, cylindrical
implants with a 4 mm diameter and up to 12 mm
Advantages: in length (Fig 8-6, Fig 8-7, Fig 8-8, and Fig 8-9)
• Bone density and quality most closely resem- should be used. The breed of animal used in the
ble humans study must of course be considered when choos-
• Standard size implants ing the exact implant dimensions. It is extremely
• Study of functional load. important that control implants are included in
the study design. These implants should prefer-
Disadvantages: ably be of a material already in clinical use (ISO
• Expensive (including housing) 10993–6, 1994).
• Ethical concerns. The chosen implant design will determine
the experimental techniques used to evaluate
8.4.3 Timing the material. Common mechanical testing used
Important time points from the project start up to on tissues harvested from in vivo studies include
the collection of specimens and data analysis torque removal tests, and pull-out and push-
(graphical): out tests. These tests are used to evaluate the
• Ordering (3 months) strength of the bone–implant surface contact.
• Quarantine (3 weeks) A high force encountered is considered indirect
• Surgery (day 0) evidence of a solid osseointegration. Regarding
• Collection of specimen (2 months to × months) the histologic analysis, see below.
• Processing (3 to 4 months)
8.4.5 Preparation
8.4.4 Surgical Procedures – Baseline Surgery The following description is based on a sequence
Since an edentulous space is required for related to implant placement in the mandible
implant installation, so-called baseline surgery of canines.
has to be performed, which includes bilateral Adult mongrel dogs, with a body weight of a
surgical extractions of all the four mandibular minimum of 20 kg, are usually recommended.
premolars and the first molar. A healing period In compliance with guidelines for the care and
of 3 to 4 months is usually recommended prior use of laboratory animals the animals should be
to the placement of oral implants. vaccinated, receiving anti-vermin drugs, and

109
Osteology Guidelines for Oral and Maxillofacial Regeneration

Fig 8-8   Tension-free suturing following surgery in Fig 8-9   Uneventful healing after suture removal (10 days)
edentulous mandible of dogs. Note the multiple single in the mandible of dogs.
sutures in combination with horizontal mattress sutures.
Nonresorbable Teflon 4-0 suture has been used.

put into quarantine for clinical observation 8.4.7 Postoperative Care

prior to the experiment. Chlorhexidine (0.2%) rinse should be done
twice a week during the duration of the study.
Anesthesia The sutures are removed when adequate
Local variations with regards to access to drugs healing is observed, usually around 7 to 10 days
etc. will apply in different countries. However, postoperatively.
the following protocol is well established within The dogs are maintained on a soft diet for
this type of study: the duration of the study.
• Pre-anesthesia with xilazine (Ronpum ®,
20 mg/kg intramuscularly and Ketamine 1 g 8.4.8 Endpoint Measurements
(Dopalen® 0.8 g/kg intramuscularly). See Section 8.5.
• Anaesthetized with thionembutal 1 g (Tio-
pental®, 20 mg/kg intravenously). 8.4.9 Statistical Analysis Plan
See Chapter 6.
The animal should be kept on intravenous infu-
sion of saline during surgery. 8.4.10 Materials, Consumables, Equipment
See Section 8.5.
Analgesic
Anti-inflammatory/analgesic drug postopera-
tively (Banamine®). 8.5 Evaluating Osseointegration
Antibiotics 8.5.1 Aim
Postsurgical infection control includes 1.0 mg What is osseointegration, how can we evaluate
dexamethasone the day following surgery, and measure it and how it is measured today –
Amoxicillin 500 to 750 mg BID or Pentabiotic® these are some questions that will be elabo-
for 10 days. rated on in this section. Below we present some
of our in-house materials and methods related
8.4.6 Detailed Methodology to the techniques used for preparing samples
See Section 8.5. as well as final tools and techniques used for
evaluation of osseointegration. We start with a

110
 8.5 Evaluating Osseointegration 8

Fig 8-10   The preparation of cut and ground sections is illustrated here. (1) The resin-embedded implant is divided in the
long axis of the implant. (2) The sample block is attached onto a supporting glass. (3) The surface of the divided implant is
ground parallel on wet grinding paper in water-cooled grinding equipment. (4) A PlexiglasTM of known thickness is glued
onto the plan sample. (5) A section of thickness of about 200 µm is prepared. (6) The thick section is further ground to a
thickness of about 15 µm. (7) The section is histologically stained. (8) After drying, the samples are cover-slipped. (9)
Qualitative inspection under the light microscope.

brief explanation of the preparation of cut and niques that are frequently used in research
ground sections used for light microscopy. The studies. Some other novel methods are also
methods presented are related to nondecalci- presented.
fied bone tissue with implants in situ. Almost all
steps in the handling and treatment of samples 8.5.2 U
 ndecalcified Cut and Ground Sections
will render some artifacts. This includes every- With the Biomaterial In Situ
thing from fixation artifacts to evaluation arti- The concept of preparing undecalcified cut and
facts. Human-made and technical artifacts are ground sections was already implemented by
rather easy to identify but technical unexpected one of the authors (CBJ) in 1983 when collabo-
and unaccounted artifacts can arise and thus ration was started with Professor Karl Donath,
clear and standardized protocols should be oral pathologist at the Department of Pathol-
used. ogy, University of Hamburg, Germany (Donath
Our routine quantification techniques of tis- and Breuner, 1982; Donath, 1985) (Fig 8-10).
sue reactions to biomaterials will be explained. Throughout the years the laboratories have
Biomechanical tests will briefly be mentioned in used the so-called Donath technique involving
the end of the section since these are tech- the machine part related to the Exakt cutting

111
Osteology Guidelines for Oral and Maxillofacial Regeneration

and grinding system (Exakt Apparatebau, Nor- in formaldehyde solution for a long time renders
derstedt, Germany). All changes made by Pro- artifacts such as difficulties in the staining abil-
fessor Donath in collaboration with the author ity. If enzyme- and immunohistochemical meth-
have been implemented and this is an ongoing ods are the target then it is a must to use a fixa-
issue. Despite being a well-known technique, tive that will preserve the proteins. (Johansson
it must be regarded as a “rough” technique et al., 1999; Röser et al., 2000). In order to suc-
and it does not allow serial sectioning; how- ceed with such samples it is recommended to
ever, it is considered the state of the art tech- follow the protocol strictly, i.e., freshly made
nique when using biomaterials in situ. As with solutions with exact pH adjustment right
all special instruments and techniques it before usage and exact timing of the steps
requires specially trained technicians. One involved in the methodology. The initial fixa-
major drawback of the equipment is that it is tive steps are the most important and crucial
time consuming and costly. Having said this, ones for a positive result. Recently we have
one positive aspect is that samples, e.g., small used a commercially available fixative contain-
cochlea wall implants up to human knee and ing zinc and formalin (4% zinc formaldehyde
hip samples, can be prepared using the very [Zn-F], Mallinckrodt Baker B.V., Holland). The
same equipment. Note that all steps during rat bones were immersed for a few days in the
the tissue handling both before and during Zn-F fixative followed by routine dehydration.
cutting and grinding should be performed Some specimens were decalcified using EDTA
strictly observing appropriate laboratory with PVP (ethylenediaminetetraacetic acid dis-
safety conditions. This involves avoidance of odium salt dehydrate with polyvinylpyrro-
skin contact with the solutions and usage of lidone), which is a mild decalcification solution
lab coats and gloves as well as working in a that preserves proteins. The latter samples
hood during some of the steps. were embedded in paraffin followed by routine
sectioning. Such sections were tested with von
8.5.3 Preparation of Samples for Willbrand factor (factor VIII) rendering positive
Undecalcified Sections results of bone marrow cells (megacariocytes)
(Nyberg et al., 2010).
Fixation of Samples
In research animal studies it is quite common to Dehydration
apply perfusion fixation with, e.g., glutaralde- All samples must be fully dehydrated before
hyde. This is strongly recommended if samples being embedded in supporting material. Among
are going to be evaluated at the electron micro- dehydration solutions, ethanol is the most com-
scopic level. However, for routine morphology monly used one. After appropriate fixation time
of retrieved human and animal samples it is the routine immerse-fixed samples are rinsed in
common to immerse samples in fixative. Four tap water. Dehydration starts in 70% ethanol
percent neutral buffered formaldehyde is a followed by 80%, 96%, and absolute ethanol. It
suitable solution for fixation via immersion. is recommended to change each solution at
However, for proper fixation it is important to least twice during the dehydration process.
trim and decrease the sample size as much as Hence, this depends on the sample size as well
possible. Moreover, it is important to use a gen- as the amount of solution used. The exact dehy-
erous amount of fixative including stirring and dration time depends on the sample size, the
vacuum treatment. number of samples and the size of the jar. The
When the samples are fully fixed it is recom- latter means that a generous amount of solu-
mended to store them in 70% ethanol. Storage tion must be used.

112
 8.5 Evaluating Osseointegration 8

Preinfiltration and Infiltration sample sizes and between batches (older

The first step in sample treatment with resin glue is less viscous and results in a thicker
solution involves a dilution of the pure resin. glue compared to new).
Usually we use three dilutions and each step is 3. A section of approximately 150 to 200 μm is
conducted under stirring and vacuum condi- prepared in the bandsaw. According to own
tions. Following this, we use two batches of experience, preparation of sections around
pure resin before the samples are embedded in 100 μm with an implant in situ is not recom-
pure resin. The routine resin in the laboratories mended since this more often results in
is the Technovit 7200 VLC (Kulzer, Germany). implant detachment. However, if no implant
However, for enzyme- and immunohistochemi- is in situ one may well prepare thinner sec-
cal samples one must use a resin that can be tions and by using a thinner sawing-band the
dissolved and for such we use Technovit 9100 actual loss of material is reduced.
NEW (Kulzer).
Grinding
Embedding Following this the grinding starts by using wet
Some extra planning in this last embedding step grinding papers of rough (800) to fine grains
will save time when preparing the final cut and (1200). Due to the so-called water-planing effect,
ground sections. This means that the preferred resulting in “shadow effects” in the interface
cutting direction should be known beforehand (which will be visualized in the microscope as an
so that the samples can be embedded “smart”. interfacial artifact) it is advisable to avoid using
It is advisable to let the sample have a few extra too much weight during grinding. Moreover, too
days of “surface drying” before handling and smooth grinding papers will also cause similar
cutting and grinding starts. artifacts and should therefore be avoided. A
sample thickness of 10 to 15 μm is optimal for
Sawing histomorphometry. And it must be noted that
The embedded sample is divided for example in the thicker the section the greater is the over­
the long axis of the implant in the bandsaw. Note estimation of the bony contact (Johansson and
that all samples in the same study must be sec- Morberg, 1995b). However, if enzyme and
tioned in the same “anatomical manner” in immunohistochemical studies are conducted
order to avoid comparisons of apples to pears. one must prepare sections below 10 μm. Such
The histomorphometric outcome is dependent sections are not prepared on Plexiglas but on
on section direction (Johansson and Morberg, precoated glass slides. This is due to the fact that
1995a). Implants retrieved with a trephine must the sections must be de-plastified before com-
be marked so that all samples will be divided in mencement of the subsequent “immune-proto-
the same manner. col” (Johansson et al., 1999; Röser et al., 2000).
1. The sample surface is ground parallel, and
the thickness (1) of this is registered, before Staining
a supporting Plexiglas of known thickness (2) Almost all histologic stainings (with some modi-
is mounted on the sample. fications) can be used on cut and ground sec-
2. Measure the sandwich, i.e., the mounted tions. It is recommended to use a routine stain-
sample with the Plexiglas (3). Knowing (1) ing that differentiates between various bone,
and (2) and deducting this from (3) will result i.e., old- and new bone as well as soft tissue.
in the glue thickness between the sample Our in-house routine staining technique is
and the Plexiglas. This is important to know referred to as the toluidine blue mixed with
since the glue thickness varies both between pyronin G (Johansson, 1991). Bone stains in

113
Osteology Guidelines for Oral and Maxillofacial Regeneration

various purple grades (old bone light purple The image analysis tools and techniques are
and new bone dark purple). The cement lines an issue that is sometimes frustrating. This is
in the bone are sharply demarcated being due to the lack of consensus in what and how
darker stained. Osteoid is stained blue-grey to perform measurements. How to perform
and osteoblasts are most often clearly measurements: manually, automatically, or
observed on osteoid. The trapped osteoblasts semi-automatically? It is our opinion that one
in the osteoid, i.e., the osteocytes, are clearly cannot rely on automatic programs. A trained
visible in its various stages. Osteoclasts are and experienced eye is required to judge
also recognized and sometimes the ruffled whether something is an artifact or not. More-
boarders are clearly distinguished. Muscle and over in order to understand the biologic proc-
other soft tissue stains blue. One can clearly esses we must rely not only on figures and
discriminate between various cell types such data. The qualitative description is a must but
as macrophages, multinucleated giant cells, unfortunately it seems that it has less priority
plasma cells and other inflammatory cells. in the analysis part. The methods and equip-
Mast cells are also clearly depicted using this ments used for histomorphometrical quantifi-
staining protocol. Note that it is very impor- cations vary among laboratories and it is diffi-
tant that the sections are not too thick because cult to compare in-house data with out-house
cellular details cannot be observed on thick results and from one study to another. How-
sections. This is probably a reason why several ever, in our laboratories we emphasize that
scientific papers are lacking information the same equipment and the same person
regarding the cellular findings on cut and perform all measurements in the same study.
ground sections. In fact, it is quite common to Our in-house routine quantifications are
find scientific papers where thick sections described below.
(from 50 μm up to a few 100 μm) have been
used for evaluating osseointegration as well as 8.5.5 Assessing the Bone Quality
other features (Ohashi et al., 2000; Becker et
al., 2006; Cordaro et al., 2008; Iwaniec et al., Bone-to-implant Contact
2008). Such studies are methodologically inap- The undecalcified cut and ground sections are
propriate since the thicker the sample the histologically stained with toluidine blue mixed
greater is the overestimation of the bone-to- in pyronin G prior to observations under the
implant contact. We have shown that, cut and light microscope. This is our in-house standard
ground, a section of a thickness above 30 μm and rule (Fig 8-11).
up to 100 μm renders overestimations of at The bone-to-implant contact measurement
least 30% (Johansson and Morberg, 1995a). requires a trained person to judge the interface.
Learning time and practice involves measure-
8.5.4 Histomorphometry ments performed by an experienced person
“The quantitative measurement and character- and the nonexperienced person together. Dis-
ization of microscopic images using a compu- cussions and test measurements are conducted
ter; manual or automated digital image analysis and individual as well as cross-comparisons are
typically involves measurements and compari- performed. The laboratories have internal crite-
sons of selected geometric areas, perimeters, ria set up for judging whether or not an artifact
length angle or orientation, form factors, center is of relevance for the measurement, and all
or gravity coordinates, as well as image analysis and judgments follow the internal
enhancement.” (www.mondofacto.com/facts/ standards. Such artifacts are, for example,
dictionary?histomorphometry) related to shrinking and staining errors.

114
 8.5 Evaluating Osseointegration 8

The laboratories started with PC-based histo-

morphometric analyses about 25 years ago. The
first PhD project reported using this tool was in
1991 (Johansson, 1991). The original semiauto-
matic Microvid program coupled to a PC and a
Leitz Aristoplan light microscope is still in use
(Leitz Wetzlar, Germany). In fact most of the PhD
theses from the laboratories related to histo-
a
morphometric comparisons have used this pro-
gram (i.e., Wennerberg et al., 1996; Ivanoff et al.,
2001). We find it very useful, not only as a
research and training tool, but also since the
researchers can study the section with various
magnifications and judge the interface in very
high magnifications before performing the actual
measurements on the preselected magnifica-
tion. The advantage with this system is that one
conducts histomorphometric analysis directly in b
the microscope. This means that all measure-
ments are performed directly with the eyepiece
of the microscope. Unfortunately this piece of
equipment is no longer available on the market.
The enlargement used in the majority of his-
tomorphometric studies involves one region of
interest (ROI) at a time. For screw-shaped
implants, our in-house rule is that one implant
thread at a time is in focus. The implants curva-
ture is outlined followed by marking nonbone c
lengths. The program automatically depicts the
bone length measured as well as bone and non- Fig 8-11   (a) Bone-to-implant contact (BIC) is outlined in
the interface region. (b) Bone area (BA) measurements are
bony contacts. The percentage of bone-to- performed in the inner threads. (c) The area of the inside
implant contact is also shown. The laboratories threads is out-folded depicting the mirror image (MI) area.
are equipped with other image analysis tools (These figures illustrate a toluidine blue stained section
but they will not be discussed here since they with subsequent quantitative histomorphometric methods
that are used routinely. Note that all measurements are
do not render any additional information performed with only one thread visible in the eyepiece of
related to this issue. To give one example of the microscope.)
what has been observed in terms of the
osseointegration process when measured as each ROI along the surface. A different bone
bone-to-implant contact, a unique clinical remodeling pattern could be observed and thus
material related to 275 retrieved Brånemark the osseointegration varied both with time of
implants was published by Bolind et al. (2005). insertion as well as within the implant geome-
Selected samples, both unloaded and loaded try (Fig 8-12). Hence, the specific quality of the
with various times of follow-up, were investi- bone tissue could not be commented on and to
gated along various regions of the implant, i.e., the best of our knowledge this is seldom
four different areas/regions were selected in reported in scientific papers.

115
Osteology Guidelines for Oral and Maxillofacial Regeneration

% 100
90
80
70
60
50
40
30
20
10
0
Unloaded Loaded Loaded Loaded
1-2.5 years 2.5–5 years >5 years

Top
Co

l
ica
ron

Ap
al

Bottom

Fig 8-12   This figure illustrates bone-to-implant contact (BIC) measurements performed on retrieved human samples after
being either unloaded or loaded with various times of follow-up. The first gray bar to the left in each group demonstrates the
mean BIC of the entire implant. As can be seen the longer the loading time the greater is the BIC. The four bars to the right
represent mean values of BIC when performed in selected regions, i.e., the coronal, bottom, apical, and top regions of the
implant surface. The unloaded implants demonstrated the lowest BIC in the thread peaks (top) while with increasing time of
loading the greatest BIC values varied along the four regions. After a loading time of more than 5 years the greatest BIC
values were observed in the bottom part of the implant surface. Note the left side of the enlarged implant is facing the upper,
coronal part of the implant while the right side depicts the apical region of the implant.

Bone Area compared with the newly formed bone in the

Another in-house method used is to quantify threads in the marrow cavity. With increasing
the bone area in a thread region or any pre­ time of follow-up it is more difficult to depict
selected ROI (Fig 8-11). While the former bone- new versus old bone. This is not so much elabo-
to-implant contact measurements may be more rated on in the presentation of results and to
related to osseointegration per se, the area the best of our knowledge, it is common to
measurements reflect the remodeling activity present data as bone area, i.e., old and newly
in the threaded region as a whole. The quality formed bone.
of bone can vary quite substantially in a ROI. For One example of comparing newly formed and
example in rabbit cortical bone the remodeling old bone inside threads was presented in combi-
pattern in the old cortical region is different nation with in vivo fluorescent labeling of bone

116
 8.5 Evaluating Osseointegration 8

(Carlsson et al., 2009). The occurrence of fluores- niques available as of today) compared with
cence was in agreement with the amount of earlier times when the implant chemistry was
newly formed bone; however, the amount and not in focus. Quite similar quantitative interfa-
type of fluorochrome (alizarin complexon and cial analysis of bone inside and outside the
calcein) varied between test and control inner threads (mirror image) has been con-
implants. Such data cannot be obtained based ducted by Jansen (2003), who refers to these
on newly formed bone alone when compared to area measurements as “bone density”. True or
the toluidine blue stained section. not, this is an effort from other biomaterials
research groups to investigate the bone tissue
Mirror Image Bone Area formation occurring around the implant in vari-
Yet another in-house standard measurement is ous regions, and thus not only focusing on the
related to the mirror images. Here one outlines interface. Whether or not we will find area
the inner area and folds it out, reflecting a mir- measurements of significant value for the
ror area (Johansson, 1991) (Fig 8-11). This type osseointegration process remains unknown for
of measurement may be of importance for the time being but at least we are reporting
studying various bone remodeling activities and data that can help and increase the understand-
reasons; if various biomaterials are used, one ing of the complex process of osseointegration.
may expect differences in the inner area com-
pared with the mirror images (Johansson et al., Bone-to-Implant Contact and Bone Volume
1998). Naturally this depends on the follow-up Measurement on 3D Material
time as well. Hypothetically one may expect a Micro-computed tomography (CT) techniques
greater bone area inside the threads if a more are emerging in the field of biomaterials
“tissue friendly material” has been used. A ratio research when studying integration of medical
of 1:1 between the inside area to mirror image devices in bone tissue. To the best of our knowl-
may depict similar tissue reactions or time- edge and with our own experience there is no
dependent healing. However, if there is a 1:2 such technique or resolution available with the
ratio between in and out this may indicate that ordinary micro-CT equipment. The interface is
there are material disturbances or the healing most often occupied by an artifact-layer of vari-
time is too short. A 2:1 ratio between in and out ous thickness (beam-hardening) resulting in the
on the other hand may indicate that the implant impossibility of judging osseointegration (Bar-
surface condition is very prone to attract bone rett and Keat, 2004). We have performed CT
tissue. Having said this, there are times when the tests at the Synchrotron Radiation facility (SR
bone-to-implant contact is very high although micro-CT) at the Institute for Materials
the bone areas in the inner threads are low. Research, GKSS Research Centre Geesthacht,
This may reflect chemical reactions and Hamburg, Germany (www.desy.de). This tech-
attachments, demonstrating that the bone-to- nique yields more accurate tomographic recon-
implant contact does not have to be “all over” structions due to the parallel beam acquisition
but the contact points that are between bone and higher resolution compared to standard
and implant are “strong enough” (Ellingsen et micro-CT and avoids the beam-hardening
al., 2004). These types of results and observa- effect. The aim of our SR micro-CT studies var-
tions need to be further elaborated on. Now­ ies. For example, the results from the two-
adays one may not have to have high bone-to- dimensional histologic sections are compared
implant contact but the contacts that are there to the three-dimensional material. The method
may involve much stronger forces (and forces for finding the two-dimensional slice in the
difficult to measure with the tools and tech- three-dimensional material (image registra-

117
Osteology Guidelines for Oral and Maxillofacial Regeneration

z’

BIC

100%

BTV z

0%

Φ
Fig 8-13   Computer tomography pictures are often shown as three-dimensional (3D) illustrations only. The material to these
figures is originally rendered from SR micro-CT (Synchrotron Radiation micro-tomography). The upper part of the figures
illustrates the entire bone implant contact (left) unfolded to a 2D-image (right). Black dashed lines show the approximate
location of the peaks of the threads. The vertical line indicates the corresponding angles in the two images. The lower part of
the figure illustrates the entire bone implant volume (left) unfolded to a 2D-image (right). This lower part illustrates the bone
tissue volumes. White dashed lines show the peaks of the threads. The vertical line indicates the corresponding angles in the
two images. Illustrations courtesy of Mr Hamid Sarve.

tion) is a challenge. This has been reported by understood. However, for research purposes
Sarve et al. (2009). One of our recent papers and eventually to perform studies related to,
describes a novel method related to bone tis- for example, loading conditions, we foresee
sue registration around implants. The bone- increased usage of equipment such as SR micro-
to-implant contact, bone area, and mirror CT. The drawback is the limited time available in
image were extracted from the three-dimen- the facilities and their location.
sional material and the results could be fol-
lowed simultaneously in a helix structure (Fig 8.5.6 Biomechanical Tests for Evaluating
8-11) (Sarve et al., 2010). Yet another novel Osseointegration
method is to “unfold” the tissue both as bone-
to-implant and bone tissue volumes (Fig 8-13) Removal Torque Tests (RTQ)
(Sarve et al., 2010). Whether this will impact Removal torque tests as well as other destruc-
how to evaluate osseointegration is not fully tive biomechanical tests such as push- and pull-

118
 8.5 Evaluating Osseointegration 8

out tests cannot be performed in the clinical essed to a cut and ground section. The section is
situation but in preclinical animal studies they stained and evaluated under the light micro-
are rapid tests revealing the actual osseointe- scope. Estimation of the bone length close to the
gration strength. The results depict the shear implant is performed, and by applying a geomet-
strength in a three-dimensional manner of the rical formula, the shear strength data can be cal-
force needed to loosen an implant from its culated (Derezende and Johansson, 1993; Sten-
bone bed. Such destructive biomechanical tests port and Johansson, 2008; Johansson et al.,
of implant integration in research animals are 2010). For more information regarding shear
performed with a variety of equipment. These strength calculations see Derezende and Johans-
range from manual hand-held Tonichi devices son (1993), Stenport and Johansson (2009), and
(Johansson and Albrektsson, 1987; Sennerby et Johansson et al. (2010).
al., 1992) to user-friendly and one-of-a-kind
electronic products, where the in-house 1992 Combination of Biomechanical Tests
products is still in use (Detektor, Göteborg, Yet another evaluation of osseointegration is to
Sweden). Several commercially available and combine biomechanical tests, i.e., removal
larger equipment such as the Instron equip- torque and pull-away tests. This was recently
ment are also frequently used for biomechani- performed with a novel medical device compo-
cal tests (Brånemark, 1996; Buser et al., 1998). nent in rabbit bone (Sul et al., 2010). With this
However, such equipment is not easily movable device it is possible to measure the removal
and tissue needs to be retrieved and trans- torque/shear strength on screws and the bone
ported to the test facilities, which may render bonding force between a disk-shaped implant
artifacts. In fact it is rather common that the tis- and bone. The device also renders the possibil-
sue is immersed in fixative on transportation to ity to investigate the qualitative and quantita-
the test facility, and this will affect the results tive bone tissue reactions (bone-to-implant
and the obtained data (Johansson CB, personal contact, bone area, mirror image) to implants
communication, 2009). Recently we have since one of the two screws attaching the
implemented and tested new user-friendly device to the bone is not RTQ tested but
removal torque equipment (Johansson et al., retrieved with surrounding bone and treated to
2010). The major advantages are the size of the cut and ground sections. For more information
instrument including the possibility of monitor- about this device see Sul et al. (2010).
ing the release torque over a selected time
span. The equipment has recently been verified Resonance Frequency Tests
for its accuracy and the results were as expected The nondestructive resonance frequency analy-
both when performed ex vivo on materials of ses (RFA) test is nowadays mostly performed
different hardness as well as in vivo tests of C.p. clinically and relates to the “implant stability as a
Ti and TiAlV implants in rabbit bone. function of the stiffness” (Meredith et al., 1996).
The test is referred to as “a bending test of the
Shear Strength Tests implant–bone complex where a transducer
The 3D removal torque tests roughly reflect the applies an extremely small bending force” (Sen-
shear strength. The removal torque results are nerby and Meredith, 2008). However, perform-
presented in Ncm. These results can be con- ing RFA tests in research animals is not a simple
verted to shear strength in N/mm2. For this pur- task and for more details related to this see Sen-
pose one has to leave the torque tested implant nerby et al. (2005). In a study by Sul et al. (2010),
in the bone bed. The bone block with the implant the RFA measurements were performed on six
in situ is retrieved, treated routinely and proc- groups of implants, with various topographies

119
Osteology Guidelines for Oral and Maxillofacial Regeneration

and surface chemical properties, at the time of methods and current trends of dental implant surfaces.
J Biomed Res Part B: Appl Biomater 88:579–596.
installation in rabbit bone and 6 weeks after. Sig- 11. Cordaro L, Bosshardt DD, Palattella P, Rao W, Serion G,
nificant differences were observed for bone Chiapasco M (2008). Maxillary sinus grafting with Bio-
measurements performed both in a longitudinal oss or Straumann bone ceramic: histomorphometric
results from a randomized controlled multicenter clini-
and in a perpendicular manner, comparing base- cal trial. Clin Oral Implants Res 19:796–803.
line to 6 weeks (Sul et al., 2010). Various RFA 12. Derezende MLR, Johansson CB (1993). Quantitative
studies were reviewed by Quesada-García et al. bone tissue response to commercially pure titanium
(2009) who reported “The studies reviewed implants. J Mater Sci Mater Med 4(3):233–239.
13. Dohan Ehrenfest DM, Coelho PG, Kang B-S, Sul Y-T,
demonstrate the usefulness of RFA as a non- Albrektsson T (2010). Classification of osseointegrated
invasive method to assess implant stability. Fur- implant surfaces: materials, chemistry and topography.
ther research is required to determine whether Trends Biotechnol 28:198–206.
14. Donath K (1985). The diagnostic value of the new
this system is also capable of measuring the method for the study of undecalcified bones and teeth
degree of dental implant osseointegration”. with attached soft tissue (Säge-Schliff (sawing and
grinding) technique). Pathol Res Pract 179:631–633.
15. Donath K, Breuner G (1982). A method for the study of
undecalcified bones and teeth with attached soft tis-
References sues. The Säge-Schliff (sawing and grinding) technique.
J Oral Pathol 11:318–326.
1. Aerssens J, Boonen S, Joly J, Dequeker J (1997). Varia- 16. Dorland´s Medical Dictionary (2007). Edinburgh: Elsevier.
tions in trabecular bone composition with anatomical 17. Ellingsen JE, Johansson CB, Wennerberg A, Holmen A
site and age: potential implications for bone quality (2004). Improved retention and bone-to implant con-
assessment. J Endocrinol 155:411–421. tact with fluoride-modified titanium implants. Int J Oral
2. Albrektsson T, Brånemark PI, Hansson HA, Lindström J Maxillofac Implants 19:659–666.
(1981). Osseointegrated titanium implants. Acta 18. Flecknell P (1996). Laboratory animal anaesthesia, 2nd
Orthop Scand 52:155–179. ed. Maryland Heights: Academic Press, Elsevier.
3. Barrett JF, Keat N (2004). Artifacts in CT: Recognition 19. Gilsanz V, Roe TF, Gibbens DT, Schulz EE, Carlson ME,
and avoidance. RadioGraphics 24:1679–1691. Gonzalez O, Boechat MI (1988). Effects of sex steroids
4. Becker J, Kirsch A, Schwarz F, Chatzinikolaidou M, Roth- on peak bone density of growing rabbits. Am J Physiol
amel D, Lekovic V et al. (2006). Bone apposition to tita- 255: E416-E421.
nium implants biocoated with recombinant human 20. ISO 10993-6 (1994). Biological evaluation of medical
bone morphogenetic protein-2 (rhBMP-2). A pilot devices. Part 6. Tests for local effects after implanta-
study in dogs. Clin Oral Invest 10:217–224. tion. Geneva: International Organization for Standardi-
5. Bolind P, Johansson CB, Balshi T, Langer B, Albrektsson T zation, pp 1–11.
(2005). A study of 275 retrieved Brånemark oral implants. 21. Ivanoff CJ, Hallgren C, Widmark G, Sennerby L, Wenner-
Int J Periodontics Restorative Dent 25:425–437. berg A (2001). Histologic evaluation of the bone inte-
6. Brånemark PI, Breiner U, Lindström J, Adell R, Hansson gration of TiO2 blasted and turned titanium microim-
BO, Ohlsson P (1969). Intraosseous anchorage of den- plants in humans. Clin Oral Implants Res 12:128–134.
tal prostheses. I. Experimental studies. Scand J Plast 22. Iwaniec UT, Wronski TJ Turner RT (2008). Histological
Reconstr Surg 3:81. analysis of bone. In: Alcohol: Methods and protocols
7. Brånemark R (1996). A biomechanical study of 325, methods in molecular biology. Nagy LE, editor.
osseointegration in-vivo measurements in rat, rabbit, No. 447. Toronto, New Jersey: Humana Press.
dog and men. PhD Thesis. University of Göteborg, 23. Jansen JA (2003). Histological analysis of bone-implant
Göteborg, Sweden. interface. In: Handbook of histology methods for bone
8. Buser D, Nydegger T, Hirt H, Cochran D, Nolte LP and cartilage. An YH, Martin KL, editors. Toronto, New
(1998). Removal torque of titanium implants in the Jersey: Humana Press, pp 353–360.
maxilla of miniature pigs. Int J Oral Maxillofac Implants 24. Johansson C, Albrektsson T (1987). Integration of screw
13:611–619. implants in the rabbit. A 1-year follow-up of removal torque
9. Carlsson C, Holmgren-Peterson K, Jönsson J, Johansson of Ti implants. Int J Oral Maxillofac Implants 2:69–75.
P, Albrektsson A, Hoffman M et al. (2009). Comparing 25. Johansson CB (1991). On tissue reactions to metal
light- and fluorescence microscopic data: A pilot study implants. PhD Thesis. Gothenburg, Sweden: Depart-
of titanium- and magnesium oxide implant integration ment of Handicap Research, Biomaterials, University of
in rabbit bone. Online. TITANIUM Int Sci J Dent Implants Gothenburg.
Biomater 1:61–70. 26. Johansson CB, Morberg P (1995a). Importance of
10. Coelho PG, Granjeiro JM, Romanos GE, Suzuki M, ground section thickness for reliable histomorphomet-
Silva NRF, Cardaropoli G et al. (2009). Basic research rical results. Biomaterials 2:91–95.

120
 References 8

27. Johansson CB, Morberg P (1995b). Cutting directions of 38. Pearce AI, Richards RG, Milz S, Schneider E, Pearce SG
bone with biomaterials in situ does influence the out- (2007). Animal models for implant biomaterial research in
come of the histomorphometrical quantification. Bio- bone: A review. European Cells and Materials 13:1–10.
materials 13:1037–1039. 39. Quesada-García MP, Prados-Sánchez E, Olmedo-
28. Johansson CB, Han CH, Wennerberg A, Albrektsson T Gaya MV, Muñoz-Soto E, González-Rodríguez MP, Vall-
(1998). A quantitative comparison of commercially ecillo-Capilla M (2009). Measurement of dental implant
pure (c.p.) titanium and Titanium-6Aluminum-4Vana- stability by resonance frequency analysis: a review of
dium (Ti6Al4V) implants in rabbit bone. J Oral Maxillo- the literature. Medicina Oral Patologia Oral Cirugía
fac Implants 13:315–321. Bucal 14:e538–e546.
29. Johansson CB, Röser K, Bolind P, Donath K, Albrektsson 40. Röser K, Johansson CB, Donath K, Albrektsson T (2000).
T (1999). Bone-tissue formation and integration of tita- A new approach to demonstrate cellular activity in
nium implants: an evaluation with newly developed bone formation adjacent to implants. J Biomed Mater
enzyme and immunohistochemical techniques. Clin Res 51:280–291.
Implant Dent Relat Res 1:33–40. 41. Sarve H, Lindblad J, Johansson CB (2009). Quantifica-
30. Johansson CB, Jimbo R, Stefenson P (2010). Ex vivo and tion of bone remodeling in SRuCT images of implants.
in vivo biomechanical test of implant attachment to Lect Notes Comput Sci 5575:770–779.
various materials. Introduction of a new user-friendly 42. Sarve H, Lindblad J, Johansson CB, Borgefors G (2010).
removal torque equipment. Clin Implant Dent Relat Methods for visualization of bone tissue in the proxim-
Res. 17 July .[Epub ahead of print]. ity of implants. In: Computer vision and graphics. Bolc
31. Meredith N, Alleyne D, Cawley P (1996). Quantitative L, Tadeusiewicz R, Chmielewski LJ, editors. Proc. ICCVG.
determination of the stability of the implant interface 43. Sennerby L, Meredith N (2008). Implant stability meas-
using resonance frequency analysis. Clin Oral Implants urements using resonance frequency analysis: biologi-
Res 7:261–267. cal and biomechanical aspects and clinical implications.
32. Neyt JG, Buckwalter JA, Carrol NC (1988). Use of animal Periodontology 2000 47:51–65.
models in musculoskeletal research. Iowa Orthop J 18: 44. Sennerby L, Thomsen P, Ericson LE (1992). A morpho-
118–123. metric and biomechanic comparison of titanium
33. Nyberg J, Hertzman S, Svensson B, Johansson C, Gran- implants inserted in rabbit cortical and cancellous
ström G, Johansson CB (2010). Single-dose irradiation bone. Int J Oral Maxillofac Implants 7:62–71.
followed by implant insertion in rat bone. An investiga- 45. Sennerby L, Dasmah A, Larsson B, Iverhed M (2005).
tive study to find a critical level for osseointegration. Bone tissue responses to surface-modified zirconia
J Osseointegration 2:93–101. implants: A histomorphometric and removal torque
34. Ohashi H, Kobayashi A, Kadoya Y, Yamano Y, Oonishi H, study in the rabbit. Clin Implant Dent Relat Res 7 (Suppl
Iwaki H (2000). Effect of particles and interface condi- 1):S13–20.
tions on fibrous tissue interposition between bone and 46. Steinemann SG (1998). Titanium: the material of
implant. A particle challenge model in rabbit. J Mater choice? Periodontology 2000 17:7–21.
Sci Mater Med 11:255–259. 47. Stenport VF, Johansson CB (2008). Evaluation of the tis-
35. Omar O, Lennerås M, Svensson S, Suska F, Emanuelsson sue integration to pure titanium and titanium alloy
L, Hall J et al. (2010a). Integrin and chemokine receptor implants. Clin Implant Dent Relat Res 10:191–199.
gene expression in implant-adherent cells during early 48. Sul YT, Johansson CB, Albrektsson T (2010). A novel
osseointegration. J Mater Sci Mater Med 21:969–980. in vivo method for quantifying the interfacial biochemi-
36. Omar O, Svensson S, Zoric N, Lennerås M, Suska F, cal bond strength of bone implants. J R Soc Interface
Wigren S et al. (2010b). In vivo gene expression in 7:81–90.
response to anodically oxidized versus machined tita- 49. Wennerberg A, Albrektsson T, Lausmaa J (1996). Torque
nium implants. J Biomed Mater Res A 92:1552–1566. and histomorphometric evaluation of c.p. titanium
37. Omar O, Suska F, Lennerås M, Zoric N, Svensson S, screws blasted with 25- and 75- microns-sized-particles
Hall J et al. (2009). The influence of bone type on the of Al2O3. J Biomed Mater Res 30:251–260.
gene expression in normal bone and at the bone- 50. Zarb G, Albrektsson T (1991). Osseointegration – A req-
implant interface: experiment in animal model. Clin uiem for the periodontal ligament? An Editorial. Int J
Implant Dent Relat Res. May 7 [Epub ahead of print]. Periodontics Restorative Dent 11:88–91.

121