Keywords hip - hip fractures - fracture fixation - device removal
Introduction
Life expectancy has increased worldwide, mainly due to improved social determinants
of health. As a result, the higher number of elderly people proportionally increased
the rate of chronic non-communicable diseases, including osteoporosis, which stands
out as a global public health problem. Osteoporosis mainly affects the elderly population,
especially female, postmenopausal patients. It is characterized by bone mineral density
reduction, leading to a lower bone mechanical strength. It has an important socioeconomic
impact due to the high incidence of proximal femoral fractures resulting from falls
and low-energy traumas.[1 ]
[2 ]
These fractures are approached in a manner as to provide patients with conditions
to resume normal activities as early as possible. Therefore, most cases are surgically
treated with implants, such as proximal femoral nails (PFNs), cannulated screws (CSs),
dynamic hip screws (DHSs), or even joint replacement (arthroplasty).[3 ]
Some complications associated with the surgical treatment of proximal femoral fractures
may require implant removal. Synthesis material removal is indicated mainly in cases
of persistent hip, gluteus, or thigh pain, and implant failure or infection.[4 ]
[5 ]
[6 ] Implant removal may predispose to femoral neck or intertrochanteric fractures, especially
in elderly patients with low bone quality.[7 ]
Due to the various dimensions and positions of implants in proximal femoral fractures,
we need to understand the biomechanical implications resulting from their removal
to raising surgeons' awareness of the safety and consequences of performing such procedure.[8 ]
[9 ]
As such, this study aims to identify the required energy (in Joules) to cause a fracture
in a synthetic proximal femur after removing three implant types: CSs, DHSs, and PFNs.
Materials and Methods
Twenty-five synthetic femurs (c1010 model manufactured by Nacional Ossos, Jaú, SP,
Brazil), composed of cortical and spongy bone, with 10 pounds per cubic foot and a
12-mm spinal canal, were used. These femurs were divided into four groups: control
group (CG), cannulated screw group (CSG), dynamic hip screw group (DHSG), and proximal
femoral nail group (PFNG).
The CG was formed by 10 intact femurs ([Figure 1 ]). For the CSG, 5 intact synthetic femurs were submitted to the placement of 3 7.5-mm
cannulated screws configured in an inverted triangle. For the DHSG and PFNG, each
group consisted of five synthetic femurs submitted to implant fixation using the Arbeitsgemeinschaft für Osteosynthesefragen (AO) technique shown in [Figure 2 ]. The sliding screws had 12 mm in diameter for the DHSG and 10.5 mm for the PFNG.
Eventually, all implants were removed, and the bones were sent to the biomechanical
analysis laboratory.
Fig. 1 Control group (CG) model.
Fig. 2 Cannulated screw (CSG), dynamic hip screw (DHSG) and proximal femoral nail (PFNG)
group samples after implant placement.
Tests were performed in static flexion using a servo-hydraulic machine (MTS 810 model,
FlexTest 40, MTS Sistemas do Brasil Ltda., São Paulo, SP, Brazil) with a 100 kilonewtons
power. Each femur was attached to the test device leaving 150 mm of its length outside
the machine, towards the hydraulic piston at its base with a horizontal inclination
of 10° and 15° in internal rotation according to a digital goniometer. The greater
trochanter was supported by a silicone disk with 8 × 2-cm in diameter ([Figure 3 ]). A preload of 40 Newtons was applied at a speed of 2 mm/s, followed by load applied
to the femoral head until fracture ([Figure 4 ]); the energy was determined in Joules (J).
Fig. 3 Experimental model on the biomechanical test platform.
Fig. 4 Experimental model after fracture.
The results were obtained through an inferential analysis using selected parameters
data and submitted to a one-way analysis of variance (ANOVA) to detect a potential
significant difference among the groups. Significance was set at 5%. The statistical
analysis was performed using the IBM Statistical Package for the Social Sciences (SPSS)
Statistics, version 20.0 (IBM SPSS Statistics, Armonk, NY, USA).
Results
All samples presented basicervical fractures.
The energy required for fracture was 7.1 J, 6.6 J, 6 J and 6.7 J for the CG, CSG,
DHSG, and PFNG, respectively, as shown in [Table 1 ].
Table 1
Variable
n
Mean value
95% CI for mean value
Minimum value
Maximum value
p -value∗
Energy (J)
CG
10
7.1
5.5 - 8.6
4.4
10.4
CSG
5
6.6
4.3 - 8.9
4
10
GDHS
5
6
4.9 - 7.1
4
7
GPFN
5
6.7
6.1 - 7.3
6.2
7.9
0.78
A one-way ANOVA revealed that there was no statistically significant difference in
the energy required for fracture (p = 0.78) among the study groups.
Discussion
Proximal femoral implant removal can result in local biomechanical changes. For instance,
DHS removal can generate bone defects in the subtrochanteric area due to its position,
while PNF removal causes a major bone defect in the greater trochanter. Therefore,
before implant removal from the proximal femur, the surgeon must consider the biomechanical
changes and the potential complications resulting from the procedure.[7 ]
[8 ]
[9 ]
In this study, synthetic bones were chosen to standardize the biomechanical properties
between samples and to minimize bone-inherent differences (bone density, length, biochemical
composition, age, diameter).[10 ] The simulated fracture mechanism, which was the fall over the greater trochanter,
is accepted as the most common in this type of injury, especially in the elderly population.[11 ]
All fractures in our study were basicervical injuries. The literature suggests that,
after implant removal, a bone failure aggravated by the low bone density in elderly
patients may contribute to the weakening of the femoral neck region, making it more
susceptible to stress and fracture.[12 ]
[13 ]
[14 ] Other studies have suggested that pain after fracture consolidation may have been
misinterpreted, consisting in a clinical sign of stress injury at the femoral neck,
which would contribute to fracture after implant removal.[15 ]
In addition, our results show a regular trend towards lower maximum energy in the
CSG, DHSG, and PFNG when compared to the CG, even though there were no statistically
significant differences. Yang et al., in a similar biomechanical study using 15 cadaveric
femurs, also failed to demonstrate a significant difference in the maximum energy
required for proximal femoral fractures after PFN and DHS removal.[6 ]
[9 ]
Other studies have tested femoral reinforcement with bone cement as a technique to
protect osteoporotic proximal femurs from fractures after synthesis material removal.
One of these studies used synthetic femurs divided into two groups, with or without
bone cement reinforcement after DHS removal, and performed biomechanical tests to
determine the maximum energy required for fracture. Interestingly, no statistical
difference was found in the maximum energy required for fracture, suggesting that
cementation after implant removal has no benefit.[8 ]
[16 ]
As limitations of our study, we realized that the load applied to the models was essentially
a pure lateral compression force, although other variables, including rotational and
axial forces, may play a role in vivo. Another important limitation was the use of
a synthetic bone model. We know that it does not reproduce the true biomechanics of
human bones, especially in the elderly population, which presents low bone mineral
density and is most susceptible to proximal femoral fractures. In addition, morphological
changes inherent to fracture healing, such as callus formation, remodeling and malunion,
were not evaluated. Synthetic models do not allow for ethnicity, age, metabolic conditions,
and lifestyle habits assessment. However, the iatrogenic aggression necessary to remove
the synthetic material cannot be evaluated. Last, the sample was restricted to 25
bones, which is limited, due to the high cost of the models.
Conclusion
None of the evaluated bone models presented significant differences in the energy
required for fracture when compared to the control group. Further studies are needed
to corroborate our results, preferably with bone models that biomechanically resemble
those of the population with a higher incidence of proximal femur fracture.
