28 / 72
healing [6]. As the aging population is expected to double by 2050 [7]
and the occurrence of osteoporotic fractures rise in the near future,
impairment in osteoporotic fracture healing is becoming an emerging
public health concern. Moreover, it has previously been reported that
the risk of non-union increases with age [8,9]; and that osteoporotic
fracture is associated high morbidity, mortality rate [10,11] and
increased healthcare costs.
As the pathophysiology of both post-menopausal estrogen defi-
ciency (type I) and senile (type II) account for the major causes of
osteoporosis and subsequently osteoporotic fractures, this paper is
intended to review our current understanding on fracture healing in
osteoporotic bone in both types and to discuss a number of key
determining factors that are impaired during osteoporotic fracture
healing. These factors include the recruitment, proliferation and
differentiation of progenitor cells; the revascularization of callus; and
also the role of mechanical sensitivity in the healing osteoporotic bone.
These factors are of high potential as therapeutic targets in future
research. Some experiences in animal studies on diaphyseal osteopor-
otic fracture are summarized in this paper; nonetheless, a general
direction of future development in metaphyseal osteoporotic fracture
model is suggested in order to improve our research work in terms of
clinical relevance and translational applicability.
Mechanical sensitivity in estrogen deficiency-induced osteoporotic
fracture (type I) and the role of estrogen receptors
A number of reports revealed the differences of mechano-biology
between osteoporotic and normal bones [12] and osteoporotic fracture
healing was impaired in both early [13] and late phases with decrease
in callus cross-sectional area, bone mineral density (BMD) and
mechanical properties [14]. The mechanism of impaired osteoporotic
fracture healing is multi-factorial and some reports indicated that low
sensitivity of osteoblasts to mechanical signals [15,16], reduced
angiogenesis [17,18], and decreased mesenchymal stem cells [19]
might be the causes. To enhance fracture healing, mechanical
stimulation by means of weight bearing is the current commonest
clinical approach. However, previous finding showed that osteoblasts
from osteoporotic donors were less responsive to 1% cyclic strain
stretching in terms of proliferation and TGF
β
release, as compared
with younger normal donors [15]. Therefore, this is generally believed
that osteoporotic bone is less responsive to mechanical stimulation;
however, therewere some opposite reports, e.g. Leppänen et al showed
that osteoporosis was not attributable to impaired mechano-
responsiveness of aging skeleton [20]; also, male adult rats with
lower estrogen level demonstrated better mechanical responses than
females [21]. Hence, mechanical sensitivity of osteoporotic bone
remains obscure.
To compare the responses of normal and osteoporotic fractured
bones to mechanical signals, fracture healing of nine-month-old
normal (Sham) and ovariectomy (OVX)-induced osteoporotic SD rats
in response to cyclic vibration (35 Hz, 0.3 g where
g
=gravitational
acceleration; 20 min/day and 5 days/week) were assessed using
radiography, microCT, histomorphometry and four-point bending
mechanical test at 2, 4, and 8 weeks post-treatment. Results showed
that fracture healing in OVX animals responded to cyclic vibration very
well, as reflected in all the assessment outcomes, particularly in the
early phases of healing [22]. Callus formation, mineralization and
remodeling were enhanced by 25
–
30%, while energy to failure was
increased by 70% as compared to corresponding OVX control. The
outcomes were comparable to those of age-matched normal fracture
healing in Sham group. These findings also revealed that both
intramembranous and endochondral ossification were enhanced well
in osteoporotic fracture healing augmented by cyclic vibration. In the
meantime, these osteogenesis findings were further substantiated by
the angiogenesis data performed in another study using the same
experimental design and cyclic vibration treatment [17]. Significantly
increased blood flow velocity (+10
–
19%) and vascular volume
(+25
–
57%) than corresponding OVX control were demonstrated at
the fracture sites of OVX-induced osteoporotic rats at week 2 and 4
post-treatment, whereas its non-OVX counterpart showed +2.2
–
13.2%
increase of vascular volume (Sham treatment vs. Sham control) at
week 2
–
4 only. Also, similar findings were found when the mechanical
loading was changed to low intensity pulsed ultrasound (1.0 kHz,
30.0 mW/cm
2
spatial-averaged temporal-averaged intensity; 20 min/
day and 5 days/week) with the same study design [23], which again
showed comparable responses (similar increase of energy-to-failure of
OVX treatment over OVX control vs. Sham treatment over Sham control
at week 8) to acoustic loading between osteoporotic fractured bone and
age-matched normal one. Rubinacci et al. also verified that OVX non-
fractured rats treated with vibration treatment (30 Hz, 3 g) showed
significant increase in cortical and medullary areas, periosteal and
endosteal perimeters but not in Sham animals, illustrating that OVX
might sensitize cortical bone to mechanical stimulation [24]. All these
evidences confirm that osteoporotic bones respond effectively to
mechanical loading (regardless of physical or acoustic form), which
was not worse than normal ones.
As the immediate effects of estrogen depletion is sensed and
relayed by estrogen receptors (ERs), as well as ERs was known to
function as mechanical signal transduction through its ligand-
independent function [25], this is not surprising to postulate the
quantity of ERs may play a role in determining bone formation during
fracture healing. Furthermore, ERs have been reported to localize in
fracture callus [26] that indicates the potential roles of ERs in fracture
healing. When comparing the gene expression of ERs at fracture callus
between 9-month-old Sham and OVX closed fractured rats, it was
found that ERs expressions were significantly higher in Sham group at
week 2 but later significantly lower at week 8 than OVX group, while
the OVX group demonstrated an opposite trend [27]. Meanwhile,
moderate correlations were found between ER-
α
and BMP-2 (
r
= 0.545,
p
= 0.003), between ER-
α
:ER-
β
ratio and BMP-2 (
r
= 0.601,
p
= 0.001),
between BMP-2 and callus width/callus area (
r
= 0.709,
p
= 0.000/
r
= 0.588,
p
= 0.001). These gene expression data were also validated
by immunohistochemistry at protein level. These findings depict that
impaired healing of OVX-induced osteoporotic fracture may be
associated with delayed expression of ERs.
As delayed expression of ERs may be the cause of impaired
osteoporotic fracture healing, this is interesting to look into the
changes of ERs expression in osteoporotic fracture healing augmented
by mechanical stimulation. In the study, the fractured rats were
randomly assigned to 4 groups
–
Sham control (SHAM), OVX-induced
osteoporotic control (OVX), OVX vibration treated at 35 Hz, 0.3 g for
20 min/day and 5 days/week (OVX-VT) and OVX vibration supplemen-
ted by daily 1.5 mg/kg/day ICI182,780 (Fulvestrant, a complete ER
antagonist) (OVX-VT-ICI). The results demonstrated that ER-
α
expres-
sion level was higher in SHAM and OVX-VT groups at week 2 and
gradually decreased at week 4 and week 8, while that of OVX group
showed lower expression at week 2 and later surged at week 8 [28].
Also, ER-
α
gene expression levels were similar between SHAM and
OVX-VT groups with no significant difference between two groups.
This indicated that cyclic vibration could induce the increase of ER-
α
level in osteoporotic fractured bone close to SHAM normal level.
Interestingly, in OVX-VT-ICI group, the ER-
α
expression was suppres-
sed to a significantly lower level. Similarly, the osteogenesis gene
expressions (Col-1 and BMP-2) and callus morphometry parameters
(callus width, callus area) echoed the ER-
α
data with the highest levels
in SHAM and OVX-VT groups from week 2
–
4, while the group of OVX-
VT-ICI was the lowest. This further substantiates the fractured bone
’
s
ability to transmit mechanical strain to stimulate callus formation. Both
gene expression data and fracture outcomes suggested that the
presence of ER-
α
was essential for mechanical transduction and
responsible for the enhancement effects induced by cyclic loading.
The induced increase of ER-
α
level at fracture callus may be sourced
W. H. Cheung et al. / Injury, Int. J. Care Injured 47S2 (2016) S21
–
S26
S22