The Journal of Arthroplasty Vol. 27 No.

7 2012

Does Osteoporosis Increase Early Subsidence of

Cementless Double-Tapered Femoral Stem in
Hip Arthroplasty?
Kee Hyung Rhyu, MD,* Se Min Lee, MD,y Young Soo Chun, MD,*
Kang Il Kim, MD,* Yoon Je Cho, MD,y and Myung Chul Yoo, MD*

Abstract: Among 320 hip arthroplasties performed between January 2007 and March 2008,
patients younger than 50 years old and patients older than 70 with a T-score at the proximal femur
less than −2.5 made up the control and study group, respectively. There were 40 patients in each
group. We measured stem subsidence, both digital and manual methods. Measurements were
made from radiographs taken serially from 2 weeks to 1 year after surgery. The amount of mean
subsidence for each group was not different, and all stems showed stable fixation in the final
radiographs. Our study suggests that even in osteoporotic proximal femurs, press-fit fixation of
double-tapered stems for hip arthroplasty can be safe and effective without excessive early
subsidence. Keywords: osteoporosis, subsidence, tapered stem, total hip arthroplasty.
© 2012 Elsevier Inc. All rights reserved.

As the elderly population has increased, so have the ages about the medical consequences during cementing,
of patients undergoing hip arthroplasty including bipolar including sudden death [6] and increased fatal cardio-
hemiarthroplasty (BHA) [1]. Hip arthroplasty for elderly pulmonary complications [7], have recently led to
patients may have several theoretical problems. Among reconsideration of cementless fixation in elderly pa-
them, the decreased bone quality of the elderly would be tients. Unfortunately, press-fitting a wedge-shaped stem
a major concern. Although our knowledge of osteopo- to an osteoporotic femur is difficult. Moreover, we do
rosis has grown enormously, there is still much left to not know the most appropriate rehabilitation program
learn. We know that osteoporosis can increase the risk of for these osteoporotic patients, especially in terms of the
fracture, is painful when it is severe and can increase timing and the amount of weight-bearing after surgery.
mortality indirectly by increasing morbidity in fractured Bottner et al [8] suggested that early weight-bearing
patients [2]. However, we still do not have thorough after a cementless THA showed no adverse effect in
understandings of its effects on fracture healing and young and healthy patients. They also recommended
bone ingrowth to the implants [2,3]. against early weight-bearing in osteoporotic elderly
Historically, the choice for implant fixation in total hip patients. However, few studies have dealt with this
arthroplasty (THA) to an osteoporotic femur has been issue. We performed the current study to investigate the
cemented fixation. However, because of the ease of effects of osteoporosis on the early stabilities of double-
surgical procedure and increasing number of predict- tapered cementless stems in hip arthroplasty, under the
ably, good results have made the cementless THA more hypothesis that the amount of early subsidence would
popular recently [4,5]. Moreover, increasing concerns not increase in an osteoporotic patient even during and
after early tolerable weight-bearing.

Patients and Methods

From the *Department of Orthopaedic Surgery, Kyung Hee University Patients
hospital at Gangdong, Seoul, South Korea; and yDepartment of Orthopaedic
Surgery, Kyung Hee University Hospital, Seoul, South Korea. A total of 320 hip arthroplasties were performed
Submitted February 21, 2011; accepted October 24, 2011. between January 2007 and March 2008 at our institute.
The Conflict of Interest statement associated with this article can be Among them, the number of total hip resurfacing
found at doi:10.1016/j.arth.2011.10.026.
Reprint requests: Kee Hyung Rhyu, MD, Department of Orthopae- arthroplasty was 143. In the remaining 177 arthroplas-
dic Surgery, Kyung Hee University Hospital at Gangdong, 149 Sangil- ties, 132 cases were THA, and 45 were BHA. Among
dong, Gangdong-Ku, Seoul, South Korea. 132 THA, 52 cases were younger than 50 years old, 9
© 2012 Elsevier Inc. All rights reserved.
0883-5403/2707-0008$36.00/0 were 70 years or older, and 71 cases were in between. In
doi:10.1016/j.arth.2011.10.026 cases of BHA, 33 cases were 70 years or older, and 1 case

1305
1306 The Journal of Arthroplasty Vol. 27 No. 7 August 2012

was younger than 50 years. This left 42 patients who use of crutches or a walker. All patients were able to
were 70 years old or older and 53 cases who were start tolerable weight-bearing at least 3 days after the
younger than 50 years old. The study group was defined surgery. Regardless of the degree of tolerable weight-
that the patient who were 70 years or older with T-score bearing, we recommended using crutches or walker
for the proximal femur was −2.5 or less. The control until 6 weeks after the surgery.
group included patients younger than 50 years old who
Method of Assessment
had no systemic disease known to affect the bone
Radiographic assessments were performed by 2 differ-
quality. All but 1 patient who were 70 years or older fit
ent orthopedic doctors who did not participate in the
the inclusion criteria. Among the patients younger than
surgeries. Measurements were made electrically using a
50 years old, 5 cases of rheumatoid arthritis, 4
built-in instrument in a digital imaging system (Infinitt,
ankylosing spondylitis, and 3 sequelae of childhood hip
Seoul, Korea) and manually using digital calipers (Mitu-
diseases were excluded. One patient in each group was
toyo, Kawasaki, Japan) with a 0.005-mm error limit. With
excluded because they were not ambulatory or were not
the 2 methods, the distance between the most proximal
able to follow the rehabilitation protocol for any reason.
point of the greater trochanter and the proximal apex of
Thus, after more than a 1-year follow-up, complete
the shoulder of the prosthesis along the axis of the
radiographs were available in 40 patients in each group.
proximal part of the femoral stem on anteroposterior
The average age of the patients in the study group was
images of the hip joint was measured. To avoid possible
79.1 years (range, 70-95 years) and was 38.6 years in the
errors caused by magnification, we used a template of the
control group (range, 23-49 years). There were 13 men
acetabular cup. We adjusted the anteroposterior radio-
and 27 women in the study group and 22 men and
graph images in the digital imaging system until the outer
18 women in the control group (P = .07). The average
diameter of the cup exactly fit the known size of the
T-score of the patients in the study group was −3.2
template. After the adjustment, the magnification was
(−2.5 to −4.8). Examination for bone mineral density
corrected. Measurements were conducted using images
in the control group was incomplete because the group
taken 2 and 6 weeks, 3 and 6 months, and 1 year after
consisted of young patients without a likelihood of
surgery. We measured the amount of subsidence during
osteoporosis. The average follow-up periods were 26
each period. The measurements were then repeated 1
months (17-39 months) for the study group and 30
month after initial data collection to obtain intraobserver
months (16-40 months) for the control group. In the
and interobserver errors. We compared them between the
study group, the diagnoses leading to hip arthroplasty
study and control groups, between THA and bipolar
included femoral neck fracture (n = 31), osteonecrosis
groups, according to the proximal femoral geometry
of the femoral head (n = 6), secondary osteoarthritis
classified according to the Dorr index [9] and, finally,
(n = 2), and cutout of the fixative device after fixation
according to the presence or absence of cerclage wire.
of intertrochanteric fracture (n = 1), whereas in the
Clinical analysis was also carried out using the patients'
control group, they were osteoarthritis by any reason
Harris Hip Scores [10].
other than osteonecrosis (n = 23), osteonecrosis of the
femoral head (n = 16), and high-energy femoral neck Statistics
fracture (n = 1). The types of arthroplasty performed in For statistical analyses, we used the Student t test for
the study group were THA (n = 11) and BHA (n = 29), the comparison of the data between the groups, the
whereas those in the control group were THA (n = 39) paired t test for the comparison of the data of the
and hemiarthroplasty (n = 1). Surgery was performed individuals according to the observing period, analysis of
by 2 orthopedic specialists using the same femoral stem variance for the comparison of the data from the patients
made with the beta titanium alloy (Accolade TMZF; of the 3 groups according to the proximal femoral
Stryker, Mahwah, NJ) for all patients in both groups. In geometry, and Fisher exact test for the comparison of the
all cases, the surgery was done via posterolateral categorical data. To investigate the errors between and
approach with lateral position. The femoral canals within observers, the interclass correlations were calcu-
were all prepared with sequential manual rasping lated with a 95% confidence interval. After confirming
with broaches. After the operation, tolerable weight- the significant interobserver and intraobserver agree-
bearing was permitted as soon as the pain was ments, we used the median of the 4 measured data as a
alleviated. The same rehabilitation protocol was used single representative value for each measurement.
in both groups. The average sizes of the femoral stems Because the number of values for each single radiograph
were number 3 in the study group and number 2 in the was only 8 (twice by 2 rater with 2 methods) and
control group. Intraoperative radiographs were checked because not all values were found to follow the normal
to confirm the size and alignment of the implants. The distribution, we choose median rather than mean or
patients were allowed to sit up at the day after surgery, average to determine the representative value. All
to move in a wheelchair at 1 or 2 days postoperatively, statistical assessments were made using bundled soft-
and, thereafter, to bear tolerable body weight with the ware (SPSS version 11.0; SPSS Inc, Chicago, Ill).
 Cementless Double-Tapered Femoral Stem in Hip Arthroplasty  Rhyu et al 1307

Results cases in the control group and in 11 cases in the study

Radiographic Results group. Among those cases, intraoperative femoral
Mean intraobserver and interobserver correlations fracture occurred in 3 cases in the control group and in
for measurements using both methods were summa- 5 in the study group (P = .71). No difference in average
rized in Table 1. Intermethod correlation was 0.85 for subsidence was observed between the wired and
rater 1 and 0.78 for rater 2. All correlations met nonwired groups (P = .95). Final follow-up radiography
significant agreement. showed stable osseous fixation in all femoral stems.
The average subsidence at 1 year was 1.19 ± 1.18 mm
Clinical Results
(0.17-4.83 mm) for the control group and 0.94 ± 0.74
The patients' Harris Hip Scores increased from 51
mm (0.10-2.82 mm) for the study group, with no
preoperatively to 87 postoperatively in the study group
significant difference between the groups (P = .30).
and from 53 to 90 in the control group. The preoperative
Three hips in the control group and none in the study
Harris Hip Scores were not significantly different (P =
group showed migrations of more than 3 mm. No
.78), and the scores significantly improved postopera-
statistical difference was detected between the groups
tively in both groups (P b .01 in both groups), with no
in terms of the occurrence of migration more than 3 mm
difference in the scores between the groups at the final
(P = .24). The average subsidence during each period in
follow-up (P = .55). There were no cases of loosening of
the control and study groups were 0.40 ± 0.48 mm
the prosthesis, severe pain, or surgical site infection.
(0.00-2.02 mm) and 0.35 ± 0.29 mm (0.00-0.93 mm)
There was no patient-complained thigh pain at the final
(at 2-6 weeks postoperatively), 0.19 ± 0.39 mm (0.00-
follow-up. Other than incomplete intraoperative femo-
1.81 mm) and 0.17 ± 0.23 mm (0.00-0.79) (at 6 weeks to
ral fractures, 2 complications related to surgery were
3 months), 0.10 ± 0.24 mm (0.00-0.92 mm) and 0.22 ±
noted in the study group. There were a dislocation that
0.32 mm (0.00-1.21 mm) (at 3-6 months postopera-
occurred 20 months after surgery, which was treated by
tively), and 0.11 ± 0.20 mm (0.00-0.57 mm) and 0.15 ±
closed reduction, and a periprosthetic fracture that
0.18 mm (0.00-0.71 mm) (at 6 months to 1 year
occurred 29 months after the operation, which was
postoperatively). Except for the difference in sub-
treated with open reduction and internal fixation. The
sidence at 6 weeks after the operation in the control
male patient with periprosthetic fracture initially re-
group (P = .01), no differences were noted in the average
ceived BHA because of femoral neck fracture. He had
migrations during each period within each group or
deformed femoral shaft because of preexisting malunion
between groups.
of femoral fracture. The fracture occurred at the apex of
The average subsidence was 0.91 ± 0.79 mm (0.10-
the deformity after fall from a height.
2.84 mm) in the BHA group and was 1.15 ± 1.10 mm
(0.12-4.85 mm) in the THA group, with no significant Discussion
difference between the groups (P = .33). Regarding the Cemented stems may be preferred for femoral fixation
proximal femoral geometry based on the Dorr index, the in hip arthroplasty of elderly patients because the
numbers of proximal femurs in types A, B, and C were cement can provide strong and immediate stability in
12 (30%), 22 (55%), and 6 (15%), respectively, in the the operating room and requires no additional time for
control group. Those of the study group were 2 (5%), biologic fixation, enabling patients to be ambulatory
6 (15%), and 32 (80%), respectively. The distribution immediately after surgery. Because of these advantages,
was significantly different (P = .01). Mean subsidence at the cemented stem has been used with good results and
1 year was 1.23 ± 1.09 mm (0.08-4.21 mm) in type A with no particular long-term problems [11,12]. As the
femur, 0.93 ± 1.04 mm (0.17-4.83 mm) in type B, age of the patients who need arthroplasty has increased,
and 1.14 ± 0.97 mm (0.10-3.22 mm) in type C. No however, more problems related to the use of cement
difference was observed in the subsidence of the femoral have been reported. Parvizi et al [6] investigated cases of
stem according to the geometry of the proximal femur in-hospital mortality in patients who underwent arthro-
(P = .82). Prophylactic cerclage wires were used in 6 plasty and found that mortality primarily occurred
during the application of cement in patients with
femoral neck fractures, highlighting the risks related to
Table 1. Interclass Correlations (Ranges in 95% Confidence
interval) According to Raters and Methods the use of cement in such procedures. Issak et al [7]
reported that the reduced cardiovascular reserve of
Embedded Intermethod
Caliper Digital Calipers Correlation
elderly patients could make them unable to adapt to or
overcome the problems related with the use of cement
Rater 1 0.99 (0.98-0.99) * 0.98 (0.97-0.99) * 0.85 (0.75-0.91)
Rater 2 0.97 (0.96-0.99) * 0.97 (0.96-0.99) * 0.78 (0.65-0.87) in arthroplasty, possibly leading to serious complica-
Interrater 0.70 (0.52-0.82) 0.66 (0.45-0.80) tions. Because of these concerns as well as technical
correlation simplicity and recent good tract records, cementless hip
All values were statistically significant (P b .01). arthroplasty has recently attracted more attention from
* Intrarater correlations for each given method. orthopedic surgeons and patients [13-17].
1308 The Journal of Arthroplasty Vol. 27 No. 7 August 2012

Considerations regarding the use of cementless The rate of subsidence during each period in the early
femoral stems in elderly patients can be divided into postoperative stage was not found to be different in
long- and short-term results. Short-term considerations the study group. In the control group, however, it was
include whether the weakened femoral cortical bone significantly higher during the first 6 weeks after the
can provide the strength required for the insertion of surgery than during the following time points. This is
wedge-shaped press-fit stems and whether the initial perhaps because the younger patients in the control
stability required for bone ingrowth can be maintained group began bearing their full weight (without assis-
if they are successfully fitted. Long-term considerations tance) earlier than those in the study group or because
include whether a severely osteoporotic femur can the activity level of the patients in the control group was
maintain stable bone ingrowth and sustainable stabil- higher after discharge, although the same rehabilitation
ity. These long-term considerations have been rela- program and time table were suggested. In addition to
tively well studied, with reports that the bone bone mineral density, we wanted to investigate possible
ingrowth capability in osteoporosis patients needed correlations between early subsidence and proximal
for stable fixation of the prosthesis was maintained femoral geometry, type of arthroplasty, and the occur-
[18] and that various stem designs have been clinically rence of incomplete femoral fracture during the oper-
successful [16,17]. Dutton and Rubash [5] reported ation. With an ordinary viewpoint, the amount of total
that cementless femoral stem fixation was successful in subsidence was expected to increase in the elderly study
elderly patients, and Bhalodiya and Singh [19] group, the bipolar group, the type C (lead pipe) femur
described the advantages of cementless wedge-shaped group, and the wired group. However, none of the
femoral stems by comparing them with those of differences were statistically significant. The results
cemented THA. Few studies, however, have examined suggest that, as long as the initial fit in the operating
the short-term considerations regarding the use of room is good, initial subsidence does not increase. In
cementless femoral stems in elderly patients. Although addition, it appears that tolerable weight-bearing in the
they are widely used because of their strong press-fit early postoperative stage does not create problems for
effect, wedge-shaped stems can cause slight initial the stability of this specific prosthesis even in osteopo-
subsidence with weight-bearing before obtaining bio- rotic patients. Because the early subsidence started
logic fixation [20]. Bottner et al [8] reported that theoretically just after the weight-bearing, one might
young patients with relatively good bone quality argue that the first measurement would have been a
showed good results even when they bear weight radiograph taken immediately after the surgery. How-
immediately after surgery, whereas patients who were ever, we found that the most of the radiographs taken
older or who had osteoporosis needed to exercise with portable devices during or just after the operation
caution. In a clinical setting, however, a patient's age were not standardized enough to measure the length.
or osteoporosis rarely affects the course of postopera- Thus, we decided the radiograph taken 2 weeks after
tive rehabilitation. Thus, studies on the short-term the surgery as our initial one.
considerations regarding the use of cementless femoral Although this study provided clinically meaningful
stems in elderly patients are important for optimizing information, it has several limitations. The first and
the course of postoperative rehabilitation for osteopo- probably the most important limitation is that there
rotic patients. can be a certain degree of error because we did not
In this study, our initial hypothesis was that there perform a radiosterometric analysis. Although we made
would be no statistical difference in early subsidence multiple measurements using 2 different methods,
between the groups. To reduce the possible errors verified them, and then used the median values, we
during the study, blinded measurements were con- could not tell that we completely eliminated all mea-
ducted by 2 unrelated orthopedic surgeons who did surement errors. The second limitation was the limited
not participate in the surgery and were unaware of the data for bone mineral density. Because our national
study design. Repeated measurements were performed medical insurance did not cover the bone densitometry
1 month after the initial measurements. After these for young and otherwise healthy patients, the data for
subsequent trials and statistical considerations, the the control group below the age 45 years were mostly
final results showed no differences between the groups. unavailable. To overcome this limitation, we added the
The intrarater correlations of both methods were proximal femoral geometry as another indicator to
excellent. Although the exact range for each degree of divide the groups. Second, from the viewpoint of reha-
agreement has not been clearly defined, the interrater bilitation, possible differences may exist between the
correlations of both methods were relatively low. How- rehabilitation courses of each patient after discharge
ever, all of the values of interclass correlations showed from the hospital, which could not be well controlled.
statistically significant agreement. Thus, we choose Moreover, we did not analyze the activity levels of
the median of the measured data as a representative patients according to the time elapsed. And the last,
value for each patient. the number of cases in each group was relatively small.
 Cementless Double-Tapered Femoral Stem in Hip Arthroplasty  Rhyu et al 1309

Post hoc power analysis revealed that if we set the effect 8. Bottner F, Zawadsky M, Su EP, et al. Implant migration
size as high, this study showed significant statistical after early weightbearing in cementless hip replacement.
power as 0.94. However, if the effect size is medium, the Clin Orthop Relat Res 2005;132.
power of this study falls 0.67. To meet the sufficient 9. Dorr LD, Faugere MC, Mackel AM, et al. Structural and
cellular assessment of bone quality of proximal femur.
power (≥0.80) at the medium effect size, the number of
Bone 1993;14:231.
patients in each group should be at least 65. Further
10. Harris WH. Traumatic arthritis of the hip after dislocation
studies with more patients and longer term follow-up and acetabular fractures: treatment by mold arthroplasty.
with strict control of activity would be helpful to derive a An end-result study using a new method of result
more definitive result. evaluation. J Bone Joint Surg Am 1969;51:737.
In conclusion, hip arthroplasty performed to osteopo- 11. Ogino D, Kawaji H, Konttinen L, et al. Total hip
rotic patients using this cementless, double-tapered replacement in patients eighty years of age and older.
wedge stem showed low subsidence until 1 year after J Bone Joint Surg Am 2008;90:1884.
the surgery with excellent fixation despite early toler- 12. Runkel M, Wenda K, Jaeger U, et al. 2- to 7-year results of
able weight-bearing. Thus, we believe that osteoporosis cement-free and cemented joint replacement in femoral
itself is not a contraindication for the use of this specific neck fractures and coxarthrosis. Aktuelle Traumatol 1994;
24:173.
cementless stem.
13. Nakoshi Y, Hasegawa M, Sudo A, et al. A long-term follow-
up study of the cementless THA with anatomic
stem/HGPII cup with 22-mm head. Int Orthop 2009;33:381.
14. Eingartner C, Volkmann R, Winter E, et al. Results of an
References uncemented straight femoral shaft prosthesis after 9 years
1. Singh JA, Vessely MB, Harmsen WS, et al. A population- of follow-up. J Arthroplasty 2000;15:440.
based study of trends in the use of total hip and total knee 15. Aldinger PR, Jung AW, Breusch SJ, et al. Survival of the
arthroplasty, 1969-2008. Mayo Clin Proc 2010;85:898. cementless Spotorno stem in the second decade. Clin
2. Campbell BJ. Osteoporosis: the basics and case-based Orthop Relat Res 2009;467:2297.
advanced treatment update for the orthopaedic surgeon. 16. Garcia-Cimbrelo E, Cruz-Pardos A, Madero R, et al. Total
Instr Course Lect 2008;57:595. hip arthroplasty with use of the cementless Zweymuller
3. Gehrig LM, Lane JM, O'Connor MI. Osteoporosis: man- Alloclassic system. A ten to thirteen-year follow-up study.
agement and treatment strategies for orthopaedic sur- J Bone Joint Surg Am 2003;85-A:296.
geons. Instr Course Lect 2009;58:817. 17. Moon KH, Kang JS, Cho KJ, et al. Hip hemiarthroplasty
4. Kim YH, Kim JS, Park JW, et al. Total hip replacement with using extensively porous coated femoral stem in femoral
a short metaphyseal-fitting anatomical cementless femoral neck fracture aged 70 years or older. J Korean Orthop
component in patients aged 70 years or older. J Bone Joint Assoc 2004;39:785.
Surg Br 2011;93:587. 18. Stenderup K, Justesen J, Eriksen EF, et al. Number
5. Dutton A, Rubash HE. Hot topics and controversies in and proliferative capacity of osteogenic stem cells are
arthroplasty: cementless femoral fixation in elderly pa- maintained during aging and in patients with osteoporosis.
tients. Instr Course Lect 2008;57:255. J Bone Miner Res 2001;16:1120.
6. Parvizi J, Holiday AD, Ereth MH, et al. The Frank 19. Bhalodiya HP, Singh SP. Results with the cementless
Stinchfield Award. Sudden death during primary hip Spotorno stem in total hip arthroplasty. J Arthroplasty
arthroplasty. Clin Orthop Relat Res 1999;39. 2009;24:1188.
7. Issack PS, Lauerman MH, Helfet DL, et al. Fat embolism 20. Rao RR, Sharkey PF, Hozack WJ, et al. Immediate
and respiratory distress associated with cemented femoral weightbearing after uncemented total hip arthroplasty.
arthroplasty. Am J Orthop (Belle Mead NJ) 2009;38:72. Clin Orthop Relat Res 1998;156.