Keywords
hemiarthroplasty - hip fractures - arthroplasty, replacement, hip - hip prosthesis
Introduction
The elderly population with osteoporosis has a high incidence of implant failure and
poor functional results in terms of limb shortening with external rotational deformity
following osteosynthesis with dynamic hip screw or intramedullary hip screw fixation
in case of unstable intertrochanteric fractures.[1] These fears restrict weight bearing, resulting in prolonged bed rest and subsequent
morbidity and mortality.[2]
[3] The requirement of a second surgery in these cases is an important consideration
before choosing the appropriate procedure and implant.
Hemiarthroplasty can be posed as a solution to these problems of osteosynthesis. It
allows for immediate weight bearing and allays the fears of nonunion or malunion at
the fracture site.[4] Few authors suggest reconstruction of the femoral calcar to avoid early subsidence
of the implant and maintain the initial stability.[5]
[6] However, calcar reconstruction increases the complexity of the procedure and, therefore,
the overall complication rates.[7]
[8] Hence, calcar replacement is an option to avoid complications such as calcar nonunion,
especially in octogenarians.[9]
Zha et al.,[10] in their retrospective study, advocate the use of distal fixing long cementless
stem to decrease the surgical complications. However, due to osteoporosis and wide
femoral canals, there are high chances of implant failure with uncemented stems in
the elderly population.[11] There are studies that reported complications with the use of cement in hemiarthroplasty,[12]
[13] but others have reported good outcomes.[14]
[15]
We are not aware of any study comparing the results of treatment with cemented calcar
replacing with long stem cemented hemiarthroplasty. We hypothesized that calcar replacing
cemented hemiarthroplasty would be associated with better hip function and better
health-related quality of life in comparison with long stem cemented hemiarthroplasty
in cognitively intact octogenarians with an unstable intertrochanteric fracture.
Materials and Methods
Study Design
The present single center, prospective randomized controlled trail was conducted according
to the guidelines of good practice and the Consolidated Standards of Reporting Trials
(CONSORT) statement.[16] Between March 2013 and November 2018, 140 cases with unstable intertrochanteric
femoral fractures (AO/OTA type 31-A2, A3) were enrolled in the study, which was approved
by the institutional review board (AIMC/CT02092013). Informed consent was taken from
all the participants in the trial.
Participants
All patients with unstable intertrochanteric fracture were screened for participation
in the study. The inclusion criteria were unstable intertrochanteric femoral fractures
(AO/OTA type 31-A2, A3), age ≥ 80 years old, independent walking with or without help
of walking aid prior to the injury and intact cognitive function with a short portable
mental status questionnaire (SPMSQ) score of 8 to 10 points.[17] Patients with pathological fracture, rheumatoid arthritis in the involved hip, polytraumatic
injuries, simultaneous fractures of the ipsilateral extremity and patients who refused
participation and were unfit for operation were excluded from the study ([Fig. 1]).
Fig. 1 CONSORT flow diagram of the participants in the study. The surgeon chooses to perform
total hip arthroplasty in seven cases due to acetabular erosions encountered during
surgery. One patient was taken up for internal fixation due to suspected urinary tract
infection.
Randomization and Blinding
The patients were block randomized with a block size of 10 in a ratio of 1:1 to undergo
an operative procedure with either calcar replacing or long stem femoral stem hemiarthroplasty.
The envelopes were sealed and stratified by gender to ensure similar gender distribution
in both groups. The patients were blinded to the treatment while the surgeon and three
nursing assistants were not. The nursing assistants were instructed not to reveal
the allocation to the patients. The postoperative care protocol did not differ between
the two groups. The patients were not shown their radiographs and were asked whether
they knew about the treatment at the time of the last follow-up.
Data Collection
Three nursing assistants involved in the research interviewed the patients and obtained
baseline data of the patients 1 week prior to the fracture. The patients were followed-up
at 4 months, 1 year and 2 years. The functional outcome was self-reported by the patients.
Operative Technique
Cemented Calcar Replacement Hemiarthroplasty
The injured hip was approached posterolaterally under spinal anesthesia. We used different
lengths of calcar extensions on a case-to-case basis in order to provide a stable
platform for the hip prosthesis to rest. The length of the stem varied between 170
and 220 mm with increments of 15 mm. The fractured greater trochanter was attached
with help of #5 Ethibond (Ethicon, Somerville, NJ, USA) cerclage wires to the lateral
aspect of the femoral stem ([Fig. 2a-b]).
Fig. 2 (A-B) Radiographic image showing unstable intertrochanteric fracture treated with calcar
replacing hemiarthroplasty and greater trochanter attached to the proximal and lateral
aspect for femur.
Cemented Long Stem Hemiarthroplasty
The surgical approach was similar as that of in cases of calcar replacement. The fractured
fragments were removed from the proximal femur. The femoral canal was prepared using
broach and a trial modular femoral stem was inserted followed by hip reduction. Preoperative
planning helped in deciding intraoperatively the appropriate length of the stem. Bone
cement was used to build up the posteromedial defect in proximal femur ([Fig. 3a]). The length of the stem varied between 170 and 300 mm. The stability of the prosthetic
hip joint was ascertained and, subsequently, an appropriately sized head was used.
Fig. 3 (A-C) Build-up of calcar with cement in a case of unstable fracture treated with long
stem bipolar hemiarthroplasty.
Postoperative Protocol
The patients were mobilized and allowed to bear weight as tolerated with the help
of crutches. All patients in both groups were administered aspirin 75 mg daily starting
on the 1st postoperative day and was continued until discharge from the hospital at the 10th postoperative day. The radiographic evaluation was done by an independent observer
(Lakhani AK) ([Fig. 3 b-c]).
Primary End Points
The primary end points were the hip function as assessed by the Harris Hip Score (HHS).[18] The HHS is a validated tool for reporting outcomes in pertrochanteric fractures.[19] Harris hip scores were graded < 70 as poor, 70 to 79 as fair, 80 to 89 as good,
and 90 to 100 as excellent.
Secondary End Points
The secondary end points were hip function as assessed by the Barthel Index (BI)[20] and the EuroQol-5 (EQ-5D).[21] The BI includes 10 activities of daily living such as feeding, grooming, bathing,
dressing, bowel care, bladder care, toilet use, ambulation, transfers, and stair climbing.
The total range of the BI is from zero to 20. The EQ-5D is a generic instrument to
measure health-related quality of life. Other parameters studied were hip-related
complications, pain in the involved hip, surgical time, intraoperative blood loss,
mortality, and ability to regain previous walking function.
Sample Size
The sample calculation was based on two variables: the HHS and the EQ-5D. Based on
a previous study,[22] we assumed that a difference of 10 ± 15 (mean ± standard deviation [SD]) would be
the smallest clinically relevant change in the HHS. So, to achieve a power of 80%,
a minimum of 60 patients should be enrolled in the study. This calculation also allowed
to prove noninferiority in the case of the EQ-5D, with an assumption of mean values
of 0.73 ± 0.18 (0.1 as limit of noninferiority). The statistical significance level
was set at 0.025 to include multiplicity encountered by the calculation of 2 sample
sizes. The final study included 140 patients, with 70 patients in each group to account
for attrition.
Statistical Method
The analyses of outcomes were done on the intent to treat principle, and all patients
were analyzed in the group allocated regardless of the treatment administered. Analysis
according to protocol was also performed. The chi-squared test was used to test correlations
in ordinal data, and the student t-test was used to compare the HHS, the BI, and the EQ-5D between the two groups. Analysis
of covariance (ANCOVA) of the primary end points was used to decrease variance with
inclusion of exposure variables and stratification. Statistical analysis was performed
using IBM SPPS Statistics for Windows, Version 22 (IBM Corp., Armonk, NY, USA).
Results
A total of 432 patients were screened for eligibility, and 292 were excluded for not
meeting the inclusion criteria. A total of 140 patients were randomized into 2 groups
with 70 cases in each group. In one group, the patients were treated with calcar replacing
cemented hemiarthroplasty (calcar replacement stem; Xlo, New Delhi, India); in the
other group, the patients were treated with long stem cemented hemiarthroplasty (Xlo,
New Delhi, India).
All cases were operated on by a single surgeon (Bansal D.) within an interval of 4
days after the initial injury following management of comorbidities. Both groups were
comparable in terms of age, gender, and osteoporosis index as defined by Singh et
al.[23] ([Table 1]).
Table 1
|
Group A (n = 70) *
|
Group B (n = 70)
|
|
Age (years old)#
|
83 ± 3.7
|
82 ± 3.4
|
|
Gender
|
|
|
|
Male
|
48
|
48
|
|
Female
|
12
|
12
|
|
Body Mass Index (kg/m2)#
|
21 ± 3.2
|
22 ± 2.4
|
|
Singh osteoporosis Index
|
2.7 ± 0.6
|
2.6 ± 0.7
|
|
Charnley functional classification (number of patients)
|
|
|
|
A
|
58
|
56
|
|
B
|
7
|
11
|
|
C
|
5
|
3
|
|
ASA classification (number of patients)
|
|
|
|
1–2
|
43
|
38
|
|
3–4
|
27
|
32
|
Primary End Points
In the intent to treat analysis, the functional outcome scales (HHS, EQ-5D, and BI)
deteriorated over time and revealed no significant difference between cemented calcar
replacing and long stem hemiarthroplasty ([Table 2]; [Figs. 4] and [5]). The findings remained similar after per protocol and ANCOVA analysis of end points.
The American Society of Anesthesiologists (ASA) classification at baseline and higher
walking ability prior to injury did not affected the primary end point.
Fig. 4 Line graph showing mean Harris Hip Score during the study period.
Fig. 5 Line graph showing mean EQ-5D index scores (a health-related quality of life measure)
during the study period.
Table 2
|
Intent to Treat
|
Per Protocol
|
|
Outcome measure
|
Calcar replacement hemiarthroplasty (n = 70)
|
Long stem hemiarthroplasty (n = 70)
|
Mean difference or relative risk (95%CI)#
|
Calcar replacement hemiarthroplasty (n = 67)
|
Long stem hemiarthroplasty (n = 65)
|
Mean difference or relative risk (95%CI)#
|
|
Harris Hip Score
|
|
|
|
|
|
|
|
Baseline
|
83.4 ± 8.6[*] (n = 69)
|
82.4 ± 11.2 (n = 70)
|
1.0 (−5.7373–3.7373)
|
83.4 ± 8.6 (n = 69)
|
82.4 ± 11.2 (n = 65)
|
1.0 (−5.7373–3.7373)
|
|
4 months
|
80.4 ± 9.4 (n = 64)
|
77.5 ± 8.1 (n = 62)
|
2.9 (7.0953–1.2953)
|
80.4 ± 9.4 (n = 64)
|
77.5 ± 8.3 (n = 58)
|
3.9 (8.0953–1.2953)
|
|
12 months
|
78.4 ± 8.7 (n = 59)
|
75.4 ± 9.6 (n = 60)
|
3.0 (−7.3414–1.3414)
|
78.4 ± 8.7 (n = 59)
|
75.4 ± 9.6 (n = 55)
|
3.0 (−7.3414–1.3414)
|
|
24 months
|
74.5 ± 9.6 (n = 53)
|
73.8 ± 7.8 (n = 55)
|
0.7 (−6.78–2.348)
|
74.5 ± 9.6 (n = 53)
|
73.5 ± 7.8 (n = 51)
|
1.0 (−6.78–2.348)
|
|
Barthel Index
|
|
|
|
|
|
|
|
Baseline
|
18 (51%)
|
17 (49%)
|
1.04 (0.7888–1.3734)
|
18 (51%)
|
17 (49%)
|
1.04 (0.876–1.4573)
|
|
4 months
|
16 (46%)
|
16 (46%)
|
0.9409 (0.6974–1.2695)
|
16 (46%)
|
14 (43%)
|
0.8409 (0.7074–1.256)
|
|
12 months
|
15 (43%)
|
14 (41%)
|
1.0238 (0.7416–1.4134)
|
15 (44%)
|
13 (42%)
|
1.0238 (0.7416–1.4134)
|
|
24 months
|
12 (38%)
|
13 (39%)
|
1.3214 (0.7653–1.3675)
|
12 (38%)
|
11 (36%)
|
1.1214 (0.8765–1.4675)
|
|
EQ-5D
|
|
|
|
|
|
|
|
Baseline
|
0.63 ± 0.29 (n = 69)
|
0.62 ± 0.31 (n = 70)
|
0.01 (−0.18–0.02)
|
0.63 ± 0.29 (n = 69)
|
0.62 ± 0.31 (n = 65)
|
0.01 (−0.18–0.02)
|
|
4 months
|
0.63 ± 0.27 (n = 64)
|
0.62 ± 0.25 (n = 62)
|
0.01 (−0.06–0.11)
|
0.63 ± 0.27 (n = 64)
|
0.59 ± 0.25 (n = 58)
|
0.04 (−0.08–0.12)
|
|
12 months
|
0.61 ± 0.25 (n = 59)
|
0.63 ± 0.23 (n = 60)
|
0.02 (−0.18–0.02)
|
0.61 ± 0.25 (n = 59)
|
0.62 ± 0.23 (n = 55)
|
0.01 (−0.18–0.02)
|
|
24 months
|
0.54 ± 0.26 (n = 53)
|
0.57 ± 0.30 (n = 55)
|
0.03 (−0.18–0.02)
|
0.54 ± 0.26 (n = 53)
|
0.56 ± 0.30 (n = 51)
|
0.02 (−0.16–0.01)
|
|
Pain Rating Scale
|
|
|
|
|
|
|
|
Baseline
|
0.5 ± 1.3 (n = 69)
|
0.43 ± 1.7 (n = 70)
|
0.02 (0.4–0.6)
|
0.5 ± 1.3 (n = 69)
|
0.41 ± 1.7 (n = 65)
|
0.19 (0.3–0.6)
|
|
4 months
|
2.1 ± 1.8 (n = 64)
|
2.0 ± 1.5 (n = 62)
|
0.1 (0.4–1)
|
2.1 ± 1.8 (n = 64)
|
2.0 ± 1.5 (n = 58)
|
0.1 (0.4–1)
|
|
12 months
|
1.5 ± 1.6 (n = 59)
|
1.3 ± 1.2 (n = 60)
|
0.02 (0.6–0.9)
|
1.5 ± 1.6 (n = 59)
|
1.4 ± 1.2 (n = 55)
|
0.1 (0.6–0.9)
|
|
24 months
|
1.4 ± 1.9 (n = 53)
|
1.5 ± 1.6 (n = 55)
|
0.2 (0.8–0.6)
|
1.5 ± 1.9 (n = 53)
|
1.2 ± 1.6 (n = 51)
|
0.3 (0.8–0.6)
|
|
Ability to walk with aid
|
|
|
|
|
|
|
|
Baseline
|
33 (94%)
|
32 (88%)
|
1.0682 (0.9785–1.1661)
|
31 (91%)
|
32 (88%)
|
1.2422 (0.7895–1.4325)
|
|
4 months
|
27 (78%)
|
29 (80%)
|
0.9750 (0.8451–1.1248)
|
27 (78%)
|
29 (80%)
|
0.8897 (0.8456–1.1248)
|
|
12 months
|
30 (85%)
|
30 (84%)
|
0.9882 (0.8776–1.1128)
|
30 (85%)
|
27 (81%)
|
0.7654 (0.8965–1.5643)
|
|
24 months
|
28 (79%)
|
27 (78%)
|
0.8977 (0.7894–1.2314)
|
26 (74%)
|
25 (75%)
|
0.9875 (0.6754–1.3345)
|
Secondary End Points
The intraoperative parameters are listed in [Table 3]. There was no significant difference in pain scores and activities of daily living
in the two groups. Both parameters deteriorated with time ([Table 2]).
Table 3
|
Parameter
|
Calcar replaced (n = 67)
|
Long Stem (n = 65)
|
Mean Difference
|
95%CI
|
p-value
|
|
Operative time (minutes)
|
109 ± 20[*]
|
104 ± 23
|
− 5.000
|
−15.2166–5.2166
|
0.3323
|
|
Blood loss (ml)
|
730 ± 112
|
745 ± 109
|
15.000
|
−37.3229–67.3229
|
0.5692
|
|
Blood
transfused (units)
|
2.3 ± 0.6
|
2.5 ± 0.7
|
0.200
|
−0.1091–0.5091
|
0.2010
|
The complications encountered in the postoperative period are presented in [Table 4]. There were two cases of hip dislocation in the calcar replacing hemiarthroplasty
group A. One patient suffered dislocation while sitting on a traditional woven bed,
while another had fallen in the washroom. In both cases, the dislocation was reported
after discharge from hospital. Prosthetic hip dislocation was reduced by open reduction
in both cases. Hip abduction braces were advised to be worn for 6 weeks. There were
no subsequent recurrences.
Table 4
|
Calcar replacement (n = 67) group
|
Long stem (n = 65) group
|
|
Immediate postoperative
|
|
|
|
a. Cardiovascular events
|
4
|
2
|
|
b. Urinary tract infection
|
2
|
0
|
|
c. Neurologic complications
|
0
|
1
|
|
d. Deep vein thrombosis
|
1
|
0
|
|
e. Respiratory complications
|
2
|
3
|
|
f. Superficial infection
|
3
|
4
|
|
g. Deep infection
|
1
|
0
|
|
h. Mortality in hospital
|
1
|
0
|
|
i. Nerve injury
|
0
|
1
|
|
Late postoperative complications
|
|
|
|
a. Mortality within one year
|
3
|
2
|
|
b. Mortality after one year
|
8
|
6
|
|
c. Dislocation
|
2
|
0
|
|
d. Trochanteric nonunion
|
0
|
2
|
Two patients had trochanteric nonunion in group B ([Fig. 6]), and one of them had symptomatic lurch. The patient who had symptomatic lurch refused
any further intervention. Three patients in group A and four patients in group B had
superficial infection, which was treated with antibiotics and regular dressings ([Fig. 7]). One patient in group A developed deep seated infection, and debridement was performed,
but no organism was obtained on culture sensitivity reports. The patient developed
discharging sinus 3 weeks after debridement. He was advised repeat debridement with
removal of the prosthesis, but refused any further intervention at our institution.
Fig. 6 Radiographic image showing trochanteric nonunion.
Fig. 7 Clinical Image showing bleb formation in case of superficial infection treated with
debridement.
Success of Blinding
The success of blinding was tested in the 105 patients available at the time of the
last follow-up, that is, 2 years after the surgical intervention. A total of 28 patients
correctly identified their allocation, 24 were did not guess it correctly, and 53
did not know about the allocation ([Table 5]). In the test for blinding, the difference was not significant between the two groups
(p = 0.63). Moreover, the difference in the outcomes was not statistically significant
between the patients who had correctly guessed the allocation and those did not ([Fig. 8]).
Fig. 8 Test for blinding at the time of last follow up.
Table 5
|
Actual number of patients underwent allocation
|
|
Calcar replaced (n = 48) hemiarthroplasty
|
Long stem (n = 57) hemiarthroplasty
|
|
In which you were allocated?
|
|
|
Calcar replacing hemiarthroplasty
|
12 (25%)
|
16 (28%)
|
|
Long stem hemiarthroplasty
|
13 (27%)
|
11 (19%)
|
|
Don't know
|
23 (48%)
|
30 (53%)
|
Participants Who Refused Participation
There were 89 patients who declined participation in the study; they were similar
in terms of gender (p = 0.56), age (p = 0.53), and ASA classification (0.23).
Discussion
Hip fractures are among the most frequent causes of morbidity and mortality in the
elderly population.[24] Intertrochanteric fractures account for 50% of all hip fractures. The majority of
these fractures is unstable, with loss of posteromedial support and displaced lesser
trochanteric fragment.[25] Although surgical techniques and patient care have improved a lot in recent years,
the ideal treatment for intertrochanteric fractures is still under debate.[25]
Currently, the treatment options for unstable intertrochanteric fracture are internal
fixation or hemiarthroplasty with either cemented or cementless stems. High union
rates have been reported with internal fixation of stable intertrochanteric fractures,
but failure rates of up to 55% have been recorded in unstable, osteoporotic fracture
patients.[26] The poor bone quality in the elderly population does not provide firm purchase in
many cases, leading to failure in terms of screw cut-out, varus collapse, and retroversion
of femoral head.[27]
Bipolar hemiarthroplasty is proposed as a treatment modality for managing unstable
intertrochanteric fractures due to its advantages such as early mobilization, good
functional results, and lower implant failure rates. Although uncemented femoral stems
have lower chances of cement implantation syndrome, there is always risk of femoral
stem loosening with rates varying up to 18%.[28] Cobden et al.[28] reported a higher rate of femoral stem loosening, especially in the elderly population.
Using the HHS, Socci et al.[1] reported good to excellent results in ∼ 80% of the patients; there was no case of
dislocation or of loosening. Zha et al.,[10] in their study of patients > 75 years old treated with hemiarthroplasty, reported
good results in 83% of the subjects. Many authors believe that following surgery,
the patient should be mobilized as early as possible to prevent complications and
reduce mortality.[3]
[27] Our patients were permitted weight bearing using a walker on day one of the surgery.
Our study demonstrated that elderly patients with unstable intertrochanteric fractures
treated with either cemented calcar replacement or long stem hemiarthroplasty had
successful clinical results. The cost of the implants (calcar replacing or long stem
hemiarthroplasty) used in either treatment group are comparable but are more expensive
than the implants routinely used in the treatment of unstable intertrochanteric fractures.
To obtain good implant stability, a few authors suggested reconstructing the calcar
femorale.[29]
[30]
[31] However, others argue that calcar replacement with prosthesis augments the deficit
in the proximal medial part of the femur. To reconstruct the calcar femorale, it would
require further wound exposure for fixation of the lesser trochanteric fragment. An
extensive soft tissue stripping and reconstruction technique is likely to result in
longer operative time and increased blood loss in already cardiovascularly compromised
patients. Abdelkhalek et al.[32] reported an average operative time of 140 minutes using cemented hemiarthroplasty
with reconstruction of the calcar femorale in elderly patients. In contrast, a shorter
operative time of 109 ± 20 minutes in calcar replacement and of 104 ± 23 minutes in
the long stem group was noted in our study. However, there was a higher number of
cases of limb length discrepancy noted with use of calcar replacing prosthesis than
with calcar reconstruction. Knutson et al.[33] reported that a limb length discrepancy < 1 cm does not lead to any major function
disability. In our study, the limb length discrepancy was between 4 and 8 mm, so the
patients had good functional outcomes.
Strength of the Study
The strengths of the present study were its prospective nature, blinding, randomization,
inclusion of intention treat and per protocol analyses, randomization stratified by
gender to attain equal gender distribution in both groups, and adherence to a predefined
hypothesis and outcome measurements assessment. Other significant strength factors
were inclusion of methods like check for success of blinding and analysis of patients
who had declined to participate in the study. To the best of our knowledge, this is
the first randomized controlled trail (RCT) comparing calcar replacing with long stem
hemiarthroplasty.
Nonetheless, the present study has a few limitations. First, it did not include a
large number of patients. Second, the follow-up of patients was relatively short;
therefore, long-term complications such as acetabular erosion, stem failure, late
dislocation or infection could not be reported. However, a long-term follow-up is
difficult due to the typical senile age of the patients and its questionable clinical
relevance in view of their remaining life expectancy. Third, the cases that required
greater trochanteric fixation might have different outcomes in comparison with those
with an intact greater trochanter.
Conclusion
With the increase in the number of geriatric populations worldwide, the choice of
implant with better and long-term results is taking a center stage. Our study did
not find any significant difference between the two groups in terms of functional
outcome and complication rates. Long-term studies may be required to identify any
long-term advantage of calcar replacing versus long stem hemiarthroplasty. Hence,
surgeons have a choice to opt for calcar replacement or long hemiarthroplasty in the
management of unstable intertrochanteric fractures in the elderly population.
