Keywords
arthroplasty - hip fractures - intraoperative complications - periprosthetic fractures
Introduction
Total hip arthroplasty (THA) is a common and successful surgical procedure in the
treatment of patients with osteoarthritis or hip fractures, whose main objective is
to provide pain relief and restore joint biomechanics, substantially improving the
quality of life of the patient.[1]
Among the complications resulting from THA, periprosthetic fracture is mentioned.
It is considered a rare complication that, when not diagnosed during surgery, requires
complex, clinically relevant treatment and is generally associated with an unfavorable
outcome, such as the need for early revision arthroplasty and poor functional recovery.[2] For these reasons, it is important to identify which predisposing factors are present
in the cases to be operated on so that we can reduce the risks of this complication.[3]
Intraoperative fractures in primary arthroplasties can be observed during surgery
and are related to the type of implant, uncemented prostheses, incompatibility between
the rasp and the implant, joint interface, gender and/or defects created during the
osteotomy or milling of the femoral canal.[4] The fracture diagnosis can be made through direct visualization of the line, a change
in sound during the impaction of reamers, a sudden alteration in resistance during
the implantation of the definitive femoral component, or through the immediate postoperative
x-ray.[5]
After diagnosis, it is important to classify the fractures as this will guide the
treatment strategy. To this end, we used the Vancouver classification for intraoperative
femoral fractures, which takes into account the stability of the implant. Type A occurs
in the proximal metaphysis, type B occurs in the diaphysis, and type C occurs distal
to the shaft. They can further be subdivided into type I (perforation), type II (non-displaced
fracture line), and type III (unstable or displaced).[4]
Dorr et al.[6] classified the conformation of the proximal femur into types A, B, and C based on
the thickness of the cortices and the shape of the spinal canal. Patients with type
C femoral pain had a 6.5 times higher risk of postoperative periprosthetic fracture
compared to those with type A, while female patients had a 1.3 times higher risk of
experiencing this complication.[7] Considering intraoperative fractures, 86% occur during the use of cementless stems
in patients with good bone quality, Dorr A or B.[8]
The consequences of intraoperative periprosthetic fractures are varied resulting in
higher readmission rates, financial burden, inferior functional outcome, increased
revision rates, and increased mortality. Given this scenario, the importance of diagnosing
these fractures intraoperatively is highlighted, since the treatment is highly effective
and low complexity.[9]
In this scenario, the present study aims to determine the incidence of periprosthetic
femur fractures during the operative period in primary hip arthroplasties and correlate
it with the risk factors inherent to patients, implants, and the diagnosis of coxoarthrosis
and/or fracture of the femoral neck.
Materials and Methods
This is a retrospective cross-sectional study, based on medical records and imaging
exams from the archives of 2 tertiary hospitals from 01/01/2014 to 12/31/2019. All
patients undergoing primary total hip arthroplasty via posterolateral access during
the study period and with a period of at least 1 year of postoperative follow-up were
included. This research was submitted for analysis by the Research Ethics Committee
of Hospital Universitário Cajuru- Curitiba/PR and Hospital Marcelino Champagnat- Curitiba/PR.
Patients undergoing revision arthroplasty, those with previous surgeries on the affected
hip, resection arthroplasty, patients who missed follow-ups, and cases related to
pathological fracture were excluded.
The variables analyzed followed those suggested in the literature and were age, sex,
Dorr index, surgical indication, Vancouver fracture classification, location, type
of femoral fixation (cemented or uncemented), implant model used, intra or postoperative
diagnosis and corresponding treatment approach ([Table 1]). All cases were performed by a surgical team comprising 5 members, 4 of whom were
experienced senior surgeons with over 6 years of expertise, and 1 junior surgeon with
2 years of experience at the beginning of the study.
Table 1
|
Age
|
Sex
|
Door
|
Indication
|
Vancouver
|
Location
|
Cemented stem
|
Model used
|
Diagnosis
|
Treatment
|
|
74
|
F
|
C
|
Fracture
|
A2
|
GT
|
No
|
Stryker- Accolade II (Portage, MI, USA)
|
Yes
|
Conversion + cerclage
|
|
56
|
M
|
B
|
Osteoarthritis
|
A2
|
GT
|
No
|
Zimmer-ML (Warsaw, IN, USA)
|
Yes
|
Cerclage
|
|
73
|
F
|
A
|
Osteoarthritis
|
A2
|
Calcar
|
No
|
Stryker- Accolade II (Portage, MI, USA)
|
Yes
|
Cerclage
|
|
74
|
F
|
B
|
Fracture
|
A2
|
Calcar
|
No
|
Jhonson- Corall (São Paulo, SP, BR)
|
Yes
|
Cerclage
|
|
65
|
F
|
B
|
Osteoarthritis
|
A2
|
Calcar
|
No
|
Jhonson- Corall (São Paulo, SP, BR)
|
Yes
|
Conversion + cerclage
|
|
78
|
F
|
A
|
Osteoarthritis
|
A2
|
GT
|
No
|
Aesculap- Bicontact (Tullingen, GER)
|
No
|
RATQ
|
|
67
|
F
|
B
|
Osteoarthritis
|
A2
|
Calcar
|
No
|
Jhonson- Corall (São Paulo, SP, BR)
|
Yes
|
Conversion + cerclage
|
|
47
|
M
|
A
|
Osteoarthritis
|
A2
|
Calcar
|
No
|
Aesculap- Bicontact (Tullingen, GER)
|
Yes
|
Cerclage
|
|
83
|
M
|
A
|
Osteoarthritis
|
A2
|
Calcar
|
No
|
Jhonson- Corall (São Paulo, SP, BR)
|
Yes
|
Conversion + cerclage
|
|
62
|
M
|
B
|
Osteoarthritis
|
B2
|
Diaphysis
|
Yes
|
Baumer-Alpha (Mogi Mirim, SP, BR)
|
Yes
|
Cerclage
|
|
32
|
F
|
A
|
Osteoarthritis
|
A2
|
Calcar
|
No
|
Aesculap- Bicontact (Tullingen, GER)
|
Yes
|
Cerclage
|
|
40
|
F
|
B
|
Osteoarthritis
|
A2
|
Calcar
|
No
|
Aesculap- Bicontact (Tullingen, GER)
|
Yes
|
Cerclage
|
The sample consisted of 2,217 arthroplasties performed on 2,154 patients, with 12
fractures (0.56%) identified. All radiographs from the immediate postoperative period,
as well as subsequent follow-up imaging examinations (15th day, 4 and 6 weeks, 3 months,
6 months, and yearly), were analyzed.
The data obtained were analyzed quantitatively using the Microsoft Excel Office 2010
software (Microsoft Corporation, Redmond WA, USA) for absolute and relative frequency
measurements. Comparisons between variables were carried out using the Student t test,
after checking the normality and variance of the data. Statistical significance was
p < 0.05.
Results
12 femoral fractures that occurred intraoperatively (0.56%) were detected in a total
of 2,117 hip arthroplasties in 2,054 patients over 6 years ([Table 2]).
Table 2
|
Year
|
Total number of surgeries
|
Total number of periprosthetic fractures
|
Percentage of periprosthetic fractures
|
|
2014
|
272
|
0
|
0%
|
|
2015
|
280
|
1
|
0.35%
|
|
2016
|
335
|
1
|
0.29%
|
|
2017
|
382
|
0
|
0%
|
|
2018
|
440
|
6
|
1.36%
|
|
2019
|
408
|
4
|
0.98%
|
The femoral component was cemented in 1,208 cases (57.06%) and uncemented in 909 cases
(42.93%), with an increase in the frequency of use of uncemented implants over the
years ([Table 3]).
Table 3
|
Year
|
Cemented
|
Uncemented
|
|
2014
|
156
|
116
|
|
2015
|
161
|
119
|
|
2016
|
176
|
159
|
|
2017
|
117
|
265
|
|
2018
|
179
|
261
|
|
2019
|
120
|
288
|
Eleven fractures were detected intraoperatively. In a case in which there was no diagnosis
during surgery, nor was it possible to show the fracture on the immediate control
x-ray, the patient developed disproportionate postoperative discomfort and pain, and,
on the 13th day, he presented to the emergency room reporting severe pain associated
with functional impotence without history of trauma, with the fracture being identified,
which occurred intraoperatively.
Regarding cementation of the femoral component, there was 1 case (0.11%) involving
the use of a cemented implant (n = 909, p = 0.004), and 11 cases (0.91%) involving the use of uncemented implants (n =1,208,
p = 0.005). Of the 12 reported cases, 8 occurred in women (66.6%), and the age of the
patients ranged between 32 and 83 years, with an average of 62.5 years.
Out of the 10 elective arthroplasties performed, 8 were due to arthrosis, while only
2 were prompted by femoral neck fractures. Despite being carried out by different
surgeons, all procedures employed the same posterolateral access approach. The surgical
technique for all femoral stems involved a sequential milling of the canal until achieving
implant stability, and there were no fractures during test reductions. In 4 instances,
the decision was made to replace the femoral stem with a cemented one and secure the
fracture with cerclage, utilizing a 1.5-mm steel wire positioned above the lesser
trochanter or, if necessary, involving the greater trochanter. In the remaining cases,
the stem was retained, and treatment was administered using cerclage with a 1.5-mm
steel wire, as shown in [Fig. 1].
Fig. 1 (A-B) Preoperative radiograph, C) Immediate control radiograph, maintenance of the nail and cerclage above the lesser
trochanter. Source: researcher's collection.
In the case in which the fracture was not identified intraoperatively but upon the
patient's postoperative return, the radiography detected a fracture starting in the
greater trochanter and extending to the metaphyseal region (Vancouver B2) in addition
to signs compatible with implant loosening. In this case, a revision arthroplasty
was performed with a distal fixation femoral stem ([Fig. 2]).
Fig. 2 (A) Immediate postoperative X-ray, (B) X-ray on the 13th day of evolution, (C) Control X-ray after review. Source: researcher's collection.
Visual and tactile stability was achieved in all cases, and weight bearing was permitted
by the patient as tolerated.
Discussion
Intraoperative periprosthetic fracture of the femur is a complication frequently described
in the literature and lacks studies with a broad sample. When not detected immediately,
it leads to early implant failure, increased hospital stay, increased morbidity, and
costs inherent to the procedure.[10] This has a variable incidence in the literature, being approximately 0.62% for cemented
implants and 0.87% for non-cemented implants in a sample of 793,823 THAs.[11] There is currently an increase in the incidence of this event, which is related
to the aging of the population and the increased use of uncemented stems.[12]
[13]
Over the course of 6 years, we obtained an incidence of 0.11% using cemented stems
and 0.91% using uncemented stems, which is compatible with the global literature.
The risk approximately doubles in women.[14] Our data point to an increase in the prevalence of fractures with increasing age
in women, which does not occur in men. This difference may exist because women are
more affected by postmenopausal osteoporosis, which leads to a decrease in bone resistance.[15]
Increasing age has already been associated with an increase in intraoperative fractures,[14] but young patients may have a greater risk of suffering calcar fractures due to
the proximal femoral canal being narrower and requiring longer reaming with greater
force transmitted between the bone and the instrument.
The analysis in [Table 2] shows the progressive increase in the use of uncemented femoral stems instead of
cemented ones, a change that has occurred over the years due to a global trend in
changing the technique. Uncemented stems, when properly indicated and executed, can
reduce surgical time and also the chances of pulmonary embolism resulting from the
cementation process.[16] Another possible explanation for this migration is lower revision rates due to aseptic
loosening.[17]
The cementless stem model is the most common in THAs performed in the United States,
although studies have demonstrated excellent results over the years using different
femoral fixation options, a scenario that points to an increased risk of intraoperative
fracture when using uncemented stems,[14] as evidenced in our data. Correct planning of the prosthesis with radiographs and
transparencies is essential to reduce this risk, since there is the possibility that
specific radiographic parameters indicate a greater risk of periprosthetic fracture,
for example, those with Dorr type A and type B conformation. In our sample, 42% of
fractures occurred in type B canals and 50% in type A canals.
Fractures of the calcar or diaphysis tend to occur during canal preparation or implant
insertion, since most uncemented stems use the pressfit concept, which increases tension
in the bone cortex. Among the causes of fractures related to surgical conditions inherent
to the surgery that occur in the proximal region, we can mention the mismatch between
the drill used in the preparation and the definitive implant, or the excessive use
of force during surgical exposure and/or preparation of the femoral canal. Finally,
distal fractures generally occur due to collision of the straight tip of the implant
with the curved cortex of the femur. The type of fracture in which the lesser trochanter
is affected occurs secondary to the propagation of microfractures that may have occurred
during the primary injury, in the case of a fracture, or during osteotomy. The pattern
around the stem occurs due to the incompatibility of the metaphyseal-diaphyseal junction
of the stem and the canal during implantation in osteoporotic femurs.[18]
The durability of the cementless stem has been shown to be greater in young patients,
perhaps due to better bone quality, which reduces the risk of fractures.[19] However, the increased risk of intraoperative fracture and, consequently, its complications
must be considered in relation to the longer survival of the implant, especially in
patients in whom the proximal femur appears weak or in cases that require prolonged
preparation and greater use of force, as in the cases Dorr A.[11]
The evaluation of the nail design is something that must always be taken into consideration,
as the use of wedge nails reveals a 10 times greater risk of causing intraoperative
fracture when compared to first generation nails.[20] In two cases presented in the present study, simple wedge femoral implants with
the Fit-and-Fill concept (Accolade II – Stryker Corporation, Kalamazoo, MI, USA) were used; this design
is associated with a 3-fold increase in intraoperative fracture rates (p < 0.001) compared to with anatomical stems, fully coated and conical or rounded.[21] Four cases presented occurred using preservation rods such as Corall (São Paulo,
SP, BR) or Bicontact (Tullingmen, GER). In a study that corelates the rod model with
a greater risk of fracture, the rasp model in which metaphyseal bone impaction occurs,
resulting in bone preservation, would increase the tension around the canal at the
time of introduction of the definitive implant, thus increasing the chances of intraoperative
fractures.[22]
It is important to analyze whether the indication for arthroplasty, due to fracture
or osteoarthritis, interferes with the occurrence of intraoperative fractures. In
our study, only 2 cases (16.6%) were indicated for neck fracture and 10 cases (83.3%)
for osteoarthritis. The literature is controversial on this point. Those who advocate
for the use of cemented stems use as an argument the increased risk of intraoperative
fracture due to preexisting bone fragility that culminated in neck fracture; in parallel,
the use of cemented stems could increase mortality due to complications resulting
from femoral cementation and increased surgical time.[16]
Klein et al.,[23] in 2006, in a study that brought together 85 cases of neck fractures, with a mean
age of 78 years, treated with cementless stem, demonstrated that there was no significant
increase in the case of intraoperative fractures and that the patients showed good
functional recovery. Richardson et al.,[24] in 2020, with the analysis of 5,883 cases of total arthroplasty in neck fractures,
compared the mortality rates in both techniques and obtained as a result a reduction
in mortality rates when using cemented stems 30, 90, and 365 days after surgery. This
difference is mainly due to the occurrence of intraoperative fractures and, consequently,
reoperations when using non-cemented stems. Ekman et al.,[25] in 2019, studied the effects of cementation in 10,677 total and partial arthroplasties,
concluding that there was no difference in mortality. The literature is still very
controversial on this point and requires further studies.
Considering that the biggest complication of intraoperative fracture is precisely
not diagnosing it during surgery, it is important to pay attention to the signs of
this complication. It is defined as a fracture identified during the surgical procedure
or diagnosed on an immediate control radiograph; some signs, such as a sudden change
in resistance during insertion of the definitive component, stopping of the definitive
implant below the level of the osteotomy, audible noise, definitive implant with different
size than planned and/or loss of stability; suggest a possible fracture.[5]
In general, the treatment of acute periprosthetic fractures is related to a high rate
of complications (63%) and reoperations (23%), and intraoperative diagnosis results
in favorable outcome and low complexity of the procedure. The tactic consists of applying
a circumferential cerclage with steel wire above the lesser trochanter and using the
same implant after the cerclage with stabilization of the fracture.[9]
In this study, the failure of the intraoperative diagnosis generated the need for
a revision arthroplasty on the 13th postoperative day. The review of the exams showed
that it was a hidden fracture, which, after intervention, evolved with consolidation
of the fracture and full functional recovery. In three cases, the uncemented implant
was exchanged for a cemented stem and cerclage was associated at the trochanter; this
option being due to the greater femoral bone fragility, detected by the surgeon. All
cases evolved with bone consolidation, full functional recovery of the patient, and
without the implant loosening over the years.
When diagnosed intraoperatively and treated with the cerclage technique above the
lesser trochanter, periprosthetic fracture has a high-resolution rate, and a good
outcome is guaranteed. Late diagnosis leads to more complex treatment and greater
chances of complications, and in these cases we should use the Vancouver classification
for postoperative periprosthetic fractures, which will guide the most appropriate
treatment strategy.[26]
Thus, the development of skills by the surgeon, such as recognizing a specific femoral
morphology, can change surgical planning and even the selection of the femoral stem.[27]
This study provides information on the incidence, risks, and factors associated with
intraoperative periprosthetic femur fracture during total hip arthroplasty. Understanding
how it occurs and identifying possible factors that indicate such a complication is
essential so that we can reduce complications, as once diagnosed intraoperatively,
it has a less complex treatment and better results.
Conclusions
Data analysis allows us to state that the risk of intraoperative periprosthetic fracture
is greater in women and with the use of cementless stems. The occurrence of these
in the present study had a rate of less than 1%.
