Secondary Logo

Journal Logo


Compliant Compression Reconstruction of the Proximal Femur Is Durable Despite Minimal Bone Formation in the Compression Segment

Christ, Alexander B. MD1; Fujiwara, Tomohiro MD, PhD1; Fabbri, Nicola MD1; Healey, John H. MD1

Author Information
Clinical Orthopaedics and Related Research: July 2021 - Volume 479 - Issue 7 - p 1577-1585
doi: 10.1097/CORR.0000000000001663



Durable fixation of proximal femoral replacement and preservation of bone stock is particularly important for the 60% to 70% of young bone sarcoma patients who have a long life expectancy [9]. The early 5- and 10-year survivorship of cemented or uncemented proximal femoral replacement is good and may exceed that of distal femoral replacement, but it precipitously declines to 56% at 20 years [3]. Alternative fixation methods are needed to improve prosthetic survivorship.

One such technology is compliant compression fixation, which effectively achieves durable fixation of distal femoral replacements and is FDA approved for this indication; it also preserves bone stock in the segment of bone under compression [8] [10]. The constant self-adjusting compression promotes hypertrophy of the underlying segment, limiting the stress bypass that occurs around long stiff conventional stems. In the distal femur, compressive forces of 181 to 363 kg promote osteointegration and the implant-bone interface and are associated with cortical hypertrophy across the compression segment [5]. Although chemotherapy curtails this process, it does not seem to affect implant survival [6]. The biomechanics of the proximal femur at the bone-implant interface differ from those of distal femur in that it is further displaced and off axis from the mechanical axis of the lower limb. Theoretically, this is strong rationale for the use of compression fixation to promote increased viable bone and ingrowth at the prosthetic interface [5, 14, 17, 18]. A recent report suggests that compliant compression reconstructions of the proximal femur may have even better results than those of the distal femur [13], but to date, only small series with short-term follow-up have been evaluated, and no study, to our knowledge, has evaluated the surrogate outcome of bone formation in the bone segment under compression. This is the focus of our study.

We therefore asked: (1) How much bone formation occurs across the compression segment in patients treated with a proximal femoral replacement implant using compliant compression fixation? (2) What were the Musculoskeletal Tumor Society (MSTS) scores at a minimum 24-month follow-up of patients who received this reconstruction? (3) What is the implant survivorship free from removal or revision for any reason at final follow-up?

Patients and Methods

Study Design and Setting

From June 2006 to January 2018, we performed 213 proximal femoral replacements as part of our orthopaedic oncology practice at a tertiary referral cancer center (Fig. 1). Of those, 6% (12 of 213) were performed using a compliant compressive fixation prosthesis (Compress® Device, Zimmer Biomet) that we retrospectively reviewed (Fig. 2). During this period, we considered the use of a compliant compressive fixation device in patients younger than 65 years of age with primary bone tumors of the proximal femur without metastases, not having prior or planned local radiotherapy, with adequate bone stock (defined as ≥ 2.5 mm of cortical thickness circumferentially intraoperatively). Thus, we selected patients for whom the benefit of possibly improved implant durability weighed favorably against the 12-week protected weightbearing regimen. Contraindications to this prosthesis were inadequate cortical thickness (< 2.5 mm), pre- or postoperative bone irradiation, the potential need for postoperative radiation, metastatic disease, or the inability to comply with therapy restrictions.

Fig. 1:
This STROBE flow diagram shows the study inclusion criteria.
Fig. 2:
This illustration shows the Compress® (Zimmer Biomet) proximal femur replacement.

Of 213 patients who had proximal femoral replacements, 72 patients underwent proximal femoral replacement for primary bone sarcomas, 23 for soft tissue sarcomas adjacent to the proximal femur, and 118 for metastatic or hematologic malignancies. Other exclusions were those who received or might need radiotherapy, such as those with Ewing sarcoma, and patients > 65 years of age. Of those who were treated with this approach, 8% (one patient) died before their second year of follow-up. None were lost to follow-up, leaving 11 available for follow-up at a minimum of 2 years (median [range] 5.9 years [2 to 10 years]) (Fig. 1).


We evaluated the 12 patients who received this implant. Seven patients had high-grade sarcomas, five of which were osteosarcomas. Five patients had intermediate-grade or low-grade sarcomas, including one low-grade osteosarcoma and four chondrosarcomas. During the study period, all patients with high-grade sarcomas were treated with chemotherapy per active Children’s Oncology Group protocols. None was on an experimental protocol. All resections and reconstructions were performed at our institution. The median (range) follow-up duration was 71 months (13 to 144 months). Four patients died of their disease at 13, 30, 70, and 93 months. Eight patients were alive after 37, 45, 66, 85, 104, 117, 133, and 144 months of follow-up. The median (range) patient age was 32 years (13 to 62 years). There were seven males and five females. The median femoral resection was 38% (22% to 59%). One patient developed local recurrence at a contaminated biopsy tract, and 6 of 12 patients developed distant metastases.

Description of Surgery

Reconstructions of the proximal femur were performed according to previously outlined procedures [6]. Wide resection of the proximal femur tumor was performed first, followed by Compress reconstruction. Briefly, this entailed securing an anchor plug with cross pins in the retained bone, inserting the spindle on the convexly milled bone interface, and applying 400 to 800 pounds of force to the bone-implant interface. The length of proximal femur resected was estimated, and a corresponding-length implant was chosen. Modular segments were assembled with a standard hip hemiarthroplasty. This was then reduced into the native acetabulum. The capsule was closed with a heavy nonabsorbable purse-string suture and any remaining gluteus medius was reattached with two or three double rows of interlocking sutures to the implant.


When indicated per active Children’s Oncology Group protocols, chemotherapy was resumed routinely 2 to 3 weeks postoperatively. All patients were given abduction and posterior hip precautions for the first 3 months postoperatively. Weightbearing was limited to 5 to 9 kg for 6 weeks, followed by 50% weightbearing for 6 weeks. Patients were subsequently allowed to bear weight as tolerated, and they were encouraged to wean from crutches to the maximum level of achievable ambulation. Abduction strengthening was started after 3 months.

Description of the Follow-up Routine

The follow-up frequency was based on the sarcoma grade. Patients with high-grade tumors were followed every 2 to 3 months in the first year, every 3 to 4 months in the second year, every 6 months in the third and fourth years, and then annually. Patients with low-grade tumors were followed at 3-month intervals for the first 2 years postoperatively and semiannually until year 5, then annually. Function was evaluated with manual abductor strength testing, the use of assistive devices, characterization of the patient’s limp, and MSTS score at the last follow-up visit [7].

Variables, Outcome Measures, Data Sources, and Bias

The baseline area of bone in the compression segment was quantified on AP and lateral radiographs in the immediate postoperative period, and measured again at 3, 6, 9, 12, 18, and 24 months during scheduled follow-up visits. A radiographic image analysis was performed with the NIH-developed ImageJ software, Version 1.52 d (United States National Institutes of Health;, a simple and reliable open-source software developed for use with multidimensional scientific images. Radiographic measurements were performed by measuring the area of bone between the spindle and the closest pin in orthogonal planes, using the known centering sleeve diameter for calibration of all radiographic measurements, as previously described [6] (Fig. 3). The reported measurements were the average measurements performed by two blinded, experienced observers (ABC, TF) familiar with the technique. Prosthetic failure was defined as implant removal resulting from mechanical failure or aseptic loosening.

Fig. 3:
A-B These (A) initial and (B) final AP radiographs show an example of new bone formation of the femur (in mm2) between the anchor plug and spindle, quantified using ImageJ open-source software on edge reformats of original radiographs. The measured surface area was corrected for the known diameter of the spindle. A color image accompanies the online version of this article.

Primary and Secondary Study Outcomes

The primary outcome of this study was to measure the amount of bone hypertrophy in the compression segment as defined by the change in bone area on orthogonal radiographs in that segment over time.

The secondary outcomes of this study were to describe functional outcomes at a minimum of 24 months postoperatively using the MSTS scoring system and to determine the survivorship of this reconstruction. MSTS questionnaires were completed for patients at each visit, and the MSTS score at their most recent follow-up visit was defined as their final functional outcome. Manual muscle testing was performed and documented by the examining surgeon using a standard 5-point scale. The use of walking aids was also documented by the examining surgeon at each time point, and the need for a walking aid at the most recent follow-up visit was defined as the patient’s final outcome.

Ethical Approval

Ethical approval for this study was obtained from Memorial Sloan Kettering (IRB 12-287).

Statistical Analysis

All statistical analyses were performed using SPSS Version 25.0 (IBM Corp). We used an independent t-test to compare the two-dimensional surrogates of volumetric bone formation, traction bar distance, cortical thickness, and differences between groups regarding chemotherapy.

Prosthesis survival was analyzed using competing risk analysis. We defined implant survival as the time from surgical implantation to either surgical revision or the final follow-up of the patient, and the competing risk as death of the patient.


Bone Formation Across the Compression Segment

Throughout the follow-up, bone formation across the compression segment was scant. The amount of cortical bone decreased slightly at 3 months and stabilized throughout follow-up (Fig. 4). At the time of surgery, the median (range) cross-sectional area of bone was 2605 mm2 (1982 to 3029 mm2). At 12 months, the median cross-sectional area was 2664 mm2 (1852 to 3163 mm2), for a median increase of 60 mm2 (2.3%). At 24 months, the median (range) cross-sectional area was 2690 mm2 (1833 to 3300 mm2) for a total median increase of 85 mm2 (3.2%). At the final follow-up visit, the median (range) percentage of cortical bone formation was only 4 (-7 to 14) above baseline. When cortical thickening was observed, it was at the medial or posterior cortex in all but one instance of lateral cortical hypertrophy (Table 1). The median (range) radiographic projection of traction bar shortening and cortical segment shortening was 0.4 mm (0 to 1.4 mm), demonstrating stability of the implant. Cortical thickness increased by a mean ± SD of only 0.1 ± 1.3 mm at the bone-spindle junction. One patient had a traumatic periprosthetic fracture and eventually underwent implant revision, precluding accurate follow-up measurements.

Fig. 4:
This graph shows the change in bone area over time for all patients.
Table 1. - Cortical thickening by location
Location Number
Medial 6 of 12
Posterior 4 of 12
Lateral 1 of 12
Anterior 0 of 12
None 4 of 12

None of the biological, clinical, or mechanical factors analyzed were related to bone formation given the number of patients studied. Chemotherapy did not influence the bone area at any timepoint (Fig. 5). However, this study was underpowered to detect a difference based on this parameter given the small number of patients in the study and scant bone formation observed.

Fig. 5:
This graph shows the change in bone area over time based on whether patients received chemotherapy.

Functional Outcomes After Compliant Compression Reconstruction of the Proximal Femur

All patients achieved independent ambulation postoperatively, and only two needed walking aids in the form of a cane. Six of 12 patients had a Trendelenburg gait at the final follow-up visit. The median (range) manual strength grade for abduction across the group was 4 of 5 (3 to 5). The median MSTS functional outcome score was 27 (19 to 30), with no difference between those treated with chemotherapy and those who were not, given the small patient numbers.

During the study period, one patient developed a recurrence at the site of an unresected needle biopsy tract. The recurrent disease and modular proximal femur were excised, and a new proximal femur was implanted while retaining the intact distal femoral compression fixation, which has been maintained for 144 months. Four patients died of disease, and four are alive with disease.

Implant Survivorship of Proximal Femur Compression Reconstructions Free from Implant Removal or Revision

Using competing risk analysis, the 5-year cumulative incidence of implant failure was 8.3% (95% CI 0.4% to 32.4%) (Table 2).The surgical revision occurred after a traumatic periprosthetic fracture that caused fracture of the implant, but the compression mechanism remained intact (Fig. 6).

Table 2. - Competing risk analysis demonstrating 5-year cumulative incidence of implant failure and death
Time in months Risk, n Events, n Implant failure Death
Incidence 95% CI Incidence 95% CI
13 12 1 0.00 0.00-0.00 0.08 0.01-0.32
20 11 1 0.08 0.01-0.32 0.08 0.01-0.32
30 10 1 0.08 0.014-0.32 0.17 0.02-0.43
70 7 1 0.08 0.01-0.32 0.27 0.05-0.56
93 5 1 0.08 0.01-0.32 0.40 0.01-0.70

Fig. 6:
Competing risk analysis survival curve plotting implant failure versus death.


Compliant, self-adjusting compression fixation has proven to be a successful, durable reconstructive method after tumor resections in the distal femur [5, 14]. At the distal femur, bony hypertrophy is seen across the compression segment over time. This has led to the adoption of this fixation technology at other sites, including the proximal femur. However, the success at the proximal femur has not been established. This investigation sought to quantify bone hypertrophy across the compression segment, patient functional outcomes, and implant survivorship free from revision or removal of compliant compression fixation of the proximal femur. The 2- to 12-year clinical outcome of proximal femur replacement with this fixation technology was good. Radiographic bone formation across the compression segment was scant. Implant survivorship was 91% at an average follow-up of almost 6 years, demonstrating good mid-term results.


This study has several limitations. First, its small size limits a more in-depth analysis and is likely responsible for several inconclusive findings. Small sample size predisposes our study to substantial risk of Type II errors, where true differences between groups are obscured by small numbers. To see a difference in bone formation in patients who did not receive chemotherapy versus patients who did, given the scant amount of bone formation described here, our study would require several hundred patients. Unfortunately, due to the rare nature of diseases seen and reconstructions performed in orthopaedic oncology practices, it is unlikely that any study is going to include that number of patients. However, dissemination of this information is important to orthopaedic oncologists and their patients, which is why we believe this study is important, despite being underpowered.

Second, this implant was used in only a small proportion of the total number of proximal femoral replacements performed at our institution during the study period. Although this substantial selection bias would normally make it difficult to generalize these results, we argue that in this case, it actually reinforces them. In a group of patients thought to be suited for this type of reconstruction, cortical hypertrophy was still scant, indicating that hypertrophy is likely the result of biomechanical properties of the anatomic location rather than patient-related factors. Third, chemotherapy was used as a dichotomous variable. Although this fails to highlight important differences in regimens and dosing, it would not likely have changed the results of this study [2, 19]. Finally, we used a previously published method with ImageJ software to measure cortical hypertrophy across the compression segment [6]. This method uses additive two-dimensional measurements to estimate three-dimensional structures. The method has been validated, but it is inherently imperfect. This method also assumes that hypertrophy is an appropriate surrogate for stable fixation and possibly even osteointegration. Previous authors have correlated the geometry of new bone formation with stable fixation [1], and the limited retrieval analyses available seem to suggest this as well [5, 14]. However, there is no radiographic technique to assess in vivo osteointegration, and this should be kept in mind when interpreting these results.

Bone Formation Across the Compression Segment

We found minimal bone formation across the compression segment in these reconstructions, which is different from the results of compliant compression reconstructions of the distal femur. Using the same radiographic measurement methods and software in the setting of distal femoral replacement for primary bone tumors, we found that bone formation in the first year after distal femoral replacement resulted in an increase in cortical bone area of 296.4 mm2 in patients who did not receive chemotherapy and 209.3 mm2 in those who received chemotherapy. We did not see this amount of increased cortical bone in the setting of proximal femoral replacement, and we found no difference between patients with chemotherapy and those without [15]. Despite the negligible bone formation in our study, the implants showed no evidence of instability, and no patient demonstrated radiographic or clinical evidence of aseptic loosening. Thus, bone formation in the compression segment appears not to be required for stability of the proximal femoral replacement, and it is not an early surrogate marker for long-term implant durability. Consequently, we suggest that compression forces in this range may have different effects based on the relationship with the weightbearing axis. This factor raises several additional questions: (1) Should the postoperative weightbearing regimen be predicated on early bone formation after proximal femoral replacement? (2) How do the biomechanical differences in stress or strain transfer between the implant and host bone differ between distal and proximal femoral replacements [17, 18]? (3) Will compression fixation of a proximal femur replacement reduce the stress bypass effect and actually preserve useful bone for later prosthetic revision?

Functional Outcomes After Compliant Compression Reconstruction of the Proximal Femur

The median (range) MSTS score at final follow-up in this study was 27 (19 to 30), which is consistent with the best scores seen in a recent systematic review of proximal femoral replacement for oncologic indications [11]. All patients maintained abductor function, with a median manual strength testing of 4 of 5. Although half of patients had a Trendelenburg gait at final follow-up, only two required walking aids in the form of a cane. The authors believe these high functional scores are indicative of stable fixation and osteointegration of the implant.

Implant Survivorship of Proximal Femur Compression Reconstructions Free from Implant Removal or Revision

At a mean of 71 months (range 13-144 months), we found the survivorship of this implant was 91%. These data are similar to those of another report regarding the proximal femur [3] and consistent with intermediate-term data on the survivorship of compliant compression fixation in the distal femur [6, 8, 20]. Interestingly, the single revision in this study occurred early after a traumatic fall, without disruption of the compression fixation. There were no failures in the five patients who were followed from 5 to 10 years, and no failures in the two patients followed for more than 10 years. This suggests that if early complications can be avoided during the osteointegration phase, we speculate that this reconstruction method might be more durable than conventional cemented or uncemented prostheses [11]. However, these are mid-term results, and the low end of the confidence interval described in this study is similar to quoted survivorship rates at 10 to 20 years in previous studies of oncologic proximal femoral replacements. Long-term studies are needed to test this hypothesis, and a multicenter trial will be necessary because proximal femoral replacement is infrequently indicated.


Compliant, self-adjusting compression reconstruction of the proximal femur after oncologic resection demonstrated good survivorship and offers mid-term survivorship comparable to distal femur reconstructions, despite scant bone formation across the compression segment. Functional outcomes in terms of gait, walking aids, and MSTS scores are generally good and reflect stable fixation at the bone-prosthesis interface. However, these advantages must be weighed against an initial period of protected weightbearing to allow for osteointegration. Longer-term multicenter studies are needed for an appropriately powered investigation into variables that may affect bone formation in and survival of these implants.


We thank Jessica Massler MSW for her editorial assistance.


1. Avedian RS, Goldsby RE, Kramer MJ, O'Donnell RJ. Effect of chemotherapy on initial compressive osseointegration of tumor endoprostheses. Clin Orthop Relat Res. 2007;459:48-53.
2. Bath LF, Crofton PM, Evans AE, et al. Bone turnover and growth during and after chemotherapy in children with solid tumors. Pediatr Res. 2004;55:224-230.
3. Bernthal NM, Schwartz AJ, Oakes DA, Kabo JM, Eckardt JJ. How long do endoprosthetic reconstructions for proximal femoral tumors last? Clin Orthop Relat Res. 2010;468:2867-2874.
4. Capanna R, Scoccianti G, Frenos F, et al. What was the survival of megaprostheses in lower limb reconstructions after tumor resections? Clin Orthop Relat Res. 2015;473:820-830.
5. Christ AB, Baral E, Wright TM, Healey JH. Osteointegration in compliant self-adjusting compression fixation shown by backscatter electron microscopy: a case report. HSS J. 2019;15:302-306.
6. Elalfy MA, Boland PJ, Healey JH. Chemotherapy curtails bone formation from compliant compression fixation of distal femoral endoprostheses. Clin Orthop Relat Res. 2019;477:206-216.
7. Enneking WF, Dunham W, Gebhardt MC, Malawar M, Pritchard DJ. A system for the functional evaluation of reconstructive procedures after surgical treatment of tumors of the musculoskeletal system. Clin Orthop Relat Res. 1993;(286):241-246.
8. Farfalli GL, Boland PJ, Morris CD, Athanasian EA, Healey JH. Early equivalence of uncemented press-fit and Compress femoral fixation. Clin Orthop Relat Res. 2009;467:2792-2799.
9. Gorlick R, Janeway K, Lessnick S, Randall RL, Marina N. Children's Oncology Group's 2013 blueprint for research: bone tumors. Pediatr Blood Cancer. 2013;60:1009-1015.
10. Healey JH, Morris CD, Athanasian EA, Boland PJ. Compress knee arthroplasty has 80% 10-year survivorship and novel forms of bone failure. Clin Orthop Relat Res. 2013;471:774-783.
11. Janssen SJ, Langerhuizen DWG, Schwab JH, Bramer JAM. Outcome after reconstruction of proximal femoral tumors: a systematic review. J Surg Oncol. 2019;119:120-129.
12. Jeon DG, Kawai A, Boland P, Healey JH. Algorithm for the surgical treatment of malignant lesions of the proximal tibia. Clin Orthop Relat Res. 1999;(358):15-26.
13. Kagan R, Adams J, Schulman C, et al. What factors are associated with failure of compressive osseointegration fixation? Clin Orthop Relat Res. 2017;475:698-704.
14. Kramer MJ, Tanner BJ, Horvai AE, O'Donnell RJ. Compressive osseointegration promotes viable bone at the endoprosthetic interface: retrieval study of Compress implants. Int Orthop. 2008;32:567-571.
15. Monument MJ, Bernthal NM, Bowles AJ, Jones KB, Randall RL. What are the 5-year survivorship outcomes of compressive endoprosthetic osseointegration fixation of the femur? Clin Orthop Relat Res. 2015;473:883-890.
16. Pala E, Henderson ER, Calabro T, et al. Survival of current production tumor endoprostheses: complications, functional results, and a comparative statistical analysis. J Surg Oncol. 2013;108:403-408.
17. Taylor SJ, Perry JS, Meswania JM, et al. Telemetry of forces from proximal femoral replacements and relevance to fixation. J Biomech. 1997;30:225-234.
18. Taylor SJ, Walker PS. Forces and moments telemetered from two distal femoral replacements during various activities. J Biomech. 2001;34:839-848.
19. Wissing MD. Chemotherapy- and irradiation-induced bone loss in adults with solid tumors. Curr Osteoporos Rep. 2015;13:140-145.
20. Zimel MN, Farfalli GL, Zindman AM, et al. Revision distal femoral arthroplasty with the Compress® prosthesis has a low rate of mechanical failure at 10 years. Clin Orthop Relat Res. 2016;474:528-536.
© 2021 by the Association of Bone and Joint Surgeons