Due to improved oncological therapeutic procedures with longer survival times, the stabilization of osteolyses and pathological fractures is gaining importance. The proximal femur is often affected by metastases. As femoral stability can be compromised by such bone lesions, stabilization as a palliative measure is indicated to restore function and relieve pain. Besides intramedullary osteosynthesis and endoprosthetic reconstruction, compound osteosynthesis is an alternative method for stabilization of the proximal femur.
Between 1994 and 2004, 34 compound osteosyntheses were performed for a tumor-caused lesion compromising mechanical stability of the proximal femur. Of those cases, 22 double-plate compound osteosyntheses and 12 single-plate compound osteosyntheses were performed for 9 pathological fractures and 25 osteolyses. Both techniques provided good primary stability. The average survival time after compound osteosynthesis was 14.2 months (range, 0-72 months). Double-plate compound osteosyntheses showed a lower mechanical failure rate than single-plate compound osteosyntheses (14.3% vs 33.3%) and a higher survival probability after 5 years (76.4% vs 38.6%). No surgical revision was required due to perioperative complications in any case.
We conclude that reliable stabilization of extensive osteolyses and pathological fractures of the proximal femur can be achieved with compound osteosynthesis. Our data suggest that double-plate compound osteosyntheses is a more favorable technique than single-plate compound osteosyntheses based on a lower rate of mechanical failure and higher survival probability.
The femur is the second skeleton section most affected by metastases, the first being the spinal column.1 With three-fourths of femoral metastases, the proximal part is the most affected.2,3 Due to high bending and shear forces during normal gait, the proximal femur is susceptible to fracture when its primary stability is put at risk by osteolytic lesions.1-3 This rationale explains the high rate of pathological fractures in this skeletal region.4
As a result, any applied stabilization technique requires good primary stability, a low rate of complications, and, most importantly, long-term stability. The latter is difficult to achieve due to absent bone healing and a resulting bone defect after neoplastic infiltration.1,2,5 Therefore, any applied stabilization procedure must be capable of permanently withstanding high mechanical stresses.2,5 Surgical treatment of metastatic bone lesions represents a purely palliative measure mainly intended to provide pain relief and to preserve or restore the patient’s mobility to ensure a better quality of life.
Various therapeutic procedures have been proposed for the treatment of femoral metastases compromising stability. Conventional joint replacement with a known favorable outcome6 is only appropriate for proximal lesions, as pertrochanteric and subtrochanteric defects with a limited bone stock require more elaborate implants such as tumor endoprosthesis for reconstruction.7-9 An intramedullary stabilization with intramedullary nails and femoral neck components is increasingly recommended for metastases of the proximal femur.10-13 A technically demanding alternative method for the treatment of large osteolytic lesions of the proximal femur is plate compound osteosynthesis.2,5,14 With this technique, stabilization is achieved by implanting a condylar plate after removal of tumor tissue through a cortical window and filling the defect with bone cement (single-plate compound osteosyntheses). In addition, an intramedullary plate can be introduced that provides a medial load transfer and supports the condylar plate blade (double-plate compound osteosyntheses; Figure 1).5,14
The purpose of this study was to retrospectively analyze whether compound osteosynthesis for the therapy for metastases in the proximal femur achieved an immediate and long-term stability while preserving hip joint function.
Figure 1: Single- and double-plate compound osteosynthesis.
Materials and Methods
All patients treated at our institution with compound osteosynthesis of the proximal femur between 1994 and 2004 were retrospectively analyzed. The indication for compound osteosynthesis in each case was an osteolysis (n=25) with compromised mechanical stability or a pathological fracture (n=9) of the proximal femur based on metastases of various neoplastic diseases in 31 patients and 2 primary tumors (Table 1). Figure 2 shows tumor infiltration sites of the proximal femur. The prerequisites for compound osteosynthesis were the affection of only the pertrochanteric and subtrochanteric femur, an extensively intact ipsilateral hip joint, no previous surgery on the proximal femur, and an anticipated life expectancy of at least 3 months.
Figure 2: Localization of pathological fractures and osteolyses.
A total of 34 compound osteosyntheses were performed on 33 patients (20 women and 13 men). Average patient age was 61.2 years (range, 26-86 years), and average American Society of Anesthesiologists (ASA) score was 2.56 (range, 1-4).
A lateral exposure of the proximal femur was chosen as a standardized surgical approach. Through a ventral cortical bone window, the tumor tissue was removed and the defect was subsequently filled with bone cement. Osteosynthesis was performed as single-plate compound osteosynthesis (n=12) with a 95° condylar plate (Synthes GmbH, Solothurn, Switzerland) and in advanced instability as double-plate compound osteosynthesis (n=22) with an additional narrow large fragment plate (DCP; Synthes GmbH; Figure 1).
During the postoperative hospital stay, all patients with double-plate compound osteosynthesis and most patients with single-plate compound osteosynthesis (10 of 12 patients) were allowed full weight bearing. This was determined by the surgeon in each case (S.K., J.S., U.H.), and the retrospectively conducted study did not detect any objective criteria for the necessary prescription of temporary partial weight bearing. Following the hospital stay, radiotherapy and/or chemotherapy (20 of 34 cases) was performed in consultation with our oncology and radiotherapy consultants.
If possible, follow-up was conducted personally. As many patients were not alive at follow-up (25 of 33 patients), relatives or the treating doctor were questioned about possible complications and survival time using a questionnaire.
The statistical evaluation of the failure rates after compound osteosynthesis was carried out using Fisher’s exact test, logistic regression analysis, and calculation of the Kaplan-Meier survival curves.
Average surgery time was 84 minutes. Patients were hospitalized for an average of 11.7 days (range, 5-20 days). There was no substantial difference in the intraoperative blood loss (average, 1114 mL) between single-plate compound osteosynthesis und double-plate compound osteosynthesis. Table 2 shows comparative results of the single- and double-plate compound osteosyntheses.
Intraoperative complications were, in 1 case each, a perforation of the cortical bone during resection of the metastasis, a major intraoperative blood loss (3000 mL), and a fracture of greater trochanter during impaction of the condylar plate. The postoperative complications were nonspecific of the procedure, such as transitory syndrome, thrombosis, hematoma, temporary kidney failure, and death due to a subsequent spinal surgery. Single-plate compound osteosynthesis had a lower complication rate than double-plate compound osteosynthesis (1 complication in 12 procedures vs 7 complications in 22 procedures, respectively). Surgical revision was not necessary in any case.
Follow-up revealed that 8 of 33 patients were alive, with an average survival time after compound osteosynthesis of 14.2 months (range, 0-72 months). Failure of the osteosynthesis was observed in 7 (4 single- and 3 double-plate compound osteosyntheses) of 33 cases after an average of 8.9 months (range, 0.3-18 months). The failure rate of double-plate compound osteosynthesis at 14.3% was lower than that of single-plate compound osteosynthesis at 33.3% (Table 2). Due to the small number of cases, this correlation was not significant in Fisher’s exact test (P=.21). However, our observation is supported by the decrease of the P value to .17 in the logistic regression model with single- or double-plate compound osteosynthesis and fracture or osteolysis as explanatory variables for the failure of osteosynthesis.
Analysis of the time to osteosynthesis failure also showed clear differences between treatment with double- and single-plate compound osteosynthesis in the long term (Figure 3). Similar failure-free rates (76.4% and 77.1%, respectively) existed for the first postoperative year, while in the subsequent observation period no double- and 2 single-plate compound osteosyntheses required revision, with survival probabilities of 76.4% and 38.6%, respectively, after 5 years (Figure 3).
Figure 3: Kaplan-Meier survival curve in reference to single- and double-plate compound osteosynthesis (P value [Log-Rank-Test]=.4349). Abbreviations: DPCO, double-plate compound osteosynthesis; SPCO, single-plate compound osteosynthesis.
To our knowledge, this is the first study to clinically compare single- and double-plate compound osteosynthesis, and it reports the highest number of compound osteosyntheses of the proximal femur thus far. Due to the underlying malignant process that led to surgery, most patients had died, so follow-up and opportunities for clinical assessment were limited.
Stability of Osteosynthesis
Within the scope of this study, it was demonstrated that a double-plate compound osteosynthesis, with a mechanical failure rate of 14.3% (3 of 21 cases), can provide reliable stabilization of pathological fractures and extensive osteolyses of the proximal femur in the long term. There was a noticeably higher failure rate for single-plate compound osteosynthesis (33.3%; 4 of 12 cases). It was not possible to statistically substantiate this tendency in an explicit way due to the limited number of cases. No significant clinical data are available for comparison in the literature concerning the stability of these 2 methods; however, Friedl et al14 managed to establish biomechanically the effectiveness of the medial load transfer using an additional intramedullary plate. Overall failure rate of compound osteosynthesis within the study group was 21.2% after a mean period of 8.9 months, which is comparable to the literature.7,14,15
Due to varying survival times and follow-up periods and incomplete data on mobility, a comparison of stability with alternative stabilization methods is possible only to a limited extent.
Some authors report small mechanical failure rates of 0% to 2% after intramedullary osteosynthesis.12,13,16 In these studies either postoperative mobilization with partial load was conducted13,16 or no information about postoperative weight bearing is available.12 Van der Hulst et al17 describe a refracture rate of 16.1% after intramedullary nailing, but postoperative full load bearing was avoided. Due to various considerations, we assess the stability after intramedullary nail osteosynthesis as critical. Differing from compound osteosynthesis, the axial load forces, without tumor resection and bone cement filling, put a stress on the locking screws that can break under cyclical axial load and absent bone healing.18 Furthermore, progression of a nonexcised osteolytic lesion in a palliative situation can lead to partial loosening of osteosynthesis, eg, locking screws. Moreover, with defects of the proximal femur, bone parts cannot interlock due to the gliding femoral neck screw under compression. This constitutes a major difference from nonpathological pertrochanteric fractures for which the nail is designed and which result in reduced stability. In our opinion, a postoperative immediate full load capacity with corresponding functional advantages and consecutively improved life quality is of vital importance, especially if life expectancy is reduced. Moreover, no bone healing and no increasing stability are to be expected for pathological fractures. Therefore, primary stability with full weight bearing of the osteosynthesis is crucial.
Implantation of a tumor prosthesis gives patients the benefit of good primary stability, but implant dislocation must be taken into account.7 Long-term stability with loosening rates of up to 50% is a major drawback.8,9 Because these data are reported after resection of primary tumors and over longer observation periods, a comparison with our sample is only partially acceptable.
In our study of 34 patients, we encountered 8 complications (23.5%), most nonspecific to the operative procedure. In relation to compound osteosynthesis, which is a technically demanding procedure, major bleeding from the tumor or through the bone lesion was observed. In association with the osteosynthesis, no patient died perioperatively.
Although a low invasivity is reported for intramedullary nailing,10-13 complication frequency appears to be from 7.2% to 28.6%. In particular, attention must be paid to higher perioperative mortality mostly reported after fat embolism with rates between 7.1% and 14.3%.10,12,13,16,19 The fact that we observed no embolism during compound osteosynthesis can be explained by lower intramedullary pressure. However, mortality rate may also be influenced by procedure-independent aspects, such as patient comorbidities.
Published data indicates an increase in perioperative complication rates after endoprosthetic replacement compared to compound osteosynthesis for similar bone defects.1,8,9,20 Especially implant dislocation represents a specific complication of the endoprosthesis with a frequency of up to 30% after tumor prosthesis.1,7,20 Rompe et al7 observed in a comparative study of compound osteosynthesis of the proximal femur and implantation of a tumor prosthesis higher perioperative complication rates during endoprosthetic replacement with 1 femoral fracture and 3 dislocations. Noticeable in this study was better hip joint function for patients after double-plate compound osteosynthesis with a negative Trendelenburg’s sign (20 of 25 patients) compared to patients after tumor endoprosthesis (3 of 25 patients). This is attributable to the preservation of muscle insertions and the hip joint in compound osteosynthesis.
In clinical decision making how to stabilize osteolyses or pathological fractures of the proximal femur other aspects might be relevant. Therefore further considerations are listed in Table 3, comparing the different operative options.
The requirements for a stabilization procedure for tumor-caused bone defects of the proximal femur were fulfilled using double-plate compound osteosynthesis. Primary stability and satisfactory long-term stability were achieved during an average survival time of 14.2 months. After compound osteosynthesis, perioperative mortality was lower and complication rates were comparable to alternative therapeutic procedures such as intramedullary stabilization or tumor prosthesis implantation.
Due to a limited number of cases, we could not statistically substantiate a better long-term stability with an additional intramedullary plate in an explicit way. Our observation that double-plate compound osteosynthesis is associated with a lower failure rate than single-plate compound osteosynthesis must be confirmed in prospective randomized clinical trials.
Based on our data, we do not recommend single-plate compound osteosynthesis and, given the appropriate indication, propose double-plate compound osteosynthesis for stabilization of the metastatically infiltrated proximal femur.
- Friedl W, Mieck U, Fritz T. Surgical therapy of bone metastases of the upper and lower extremity [in German]. Chirurg. 1992; 63(11):897-911.
- Isler B. Surgical measures in metastatic lesions of the extremities and pelvic bones [in German]. Unfallchirurg. 1990; 93(10):449-456.
- Broos PL, Rommens PM, Vanlangenaker MJ. Pathological fractures of the femur: improvement of quality of life after surgical treatment. Arch Orthop Trauma Surg. 1992; 111(2):73-77.
- Friedl W. Pathological fractures of the proximal femur end [in German]. Zentralbl Chir. 1995; 120(11):873-880.
- Ganz R, Isler B, Mast J. Internal fixation technique in pathological fractures of the extremities. Arch Orthop Trauma Surg. 1984; 103(2):73-80.
- Branson JJ, Goldstein WM. Primary total hip arthroplasty. AORN J. 2003; 78(6):947-953, 956-969.
- Rompe JD, Eysel P, Hopf C, Heine J. Metastatic instability at the proximal end of the femur. Comparison of endoprosthetic replacement and plate osteosynthesis. Arch Orthop Trauma Surg. 1994; 113(5):260-264.
- Shin DS, Weber KL, Chao EY, An KN, Sim FH. Reoperation for failed prosthetic replacement used for limb salvage. Clin Orthop Relat Res. 1999; (358):53-63.
- Zehr RJ, Enneking WF, Scarborough MT. allograft-prosthesis composite versus megaprosthesis in proximal femoral reconstruction. Clin Orthop Relat Res. 1996; (322):207-223.
- Eingartner C, Pütz M, Schwab E, Weise K. Unreamed intramedullary nailing as minimal invasive palliative intervention in osteolysis and pathologic fractures of long tubular bones [in German]. Unfallchirurg. 1997; 100(9):715-718.
- Piatek S, Westphal T, Bischoff J, Schubert S, Holmenschlager F, Winckler S. Intramedullary stabilisation of metastatic fractures of long bones [in German]. Zentralbl Chir. 2003; 128(2):131-138.
- van Doorn R, Stapert JW. Treatment of impending and actual pathological femoral fractures with the long Gamma nail in The Netherlands. Eur J Surg. 2000; 166(3):247-254.
- Weikert DR, Schwartz HS. Intramedullary nailing for impending pathological subtrochanteric fractures. J Bone Joint Surg Br. 1991; 73(4):668-670.
- Friedl W, Ruf W, Mischkowsky T. Compound double plate osteosynthesis in subtrochanteric pathologic fractures. A clinical and experimental study [in German]. Chirurg. 1986; 57(11):713-718.
- Yazawa Y, Frassica FJ, Chao EY, Pritchard DJ, Sim FH, Shives TC. Metastatic bone disease. A study of the surgical treatment of 166 pathologic humeral and femoral fractures. Clin Orthop Relat Res. 1990; (251):213-219.
- Assal M, Zanone X, Peter RE. Osteosynthesis of metastatic lesions of the proximal femur with a solid femoral nail and interlocking spiral blade inserted without reaming. J Orthop Trauma. 2000; 14(6):394-397.
- van der Hulst RR, van den Wildenberg FA, Vroemen JP, Greve JW. Intramedullary nailing of (impending) pathologic fractures. J Trauma. 1994; 36(2): 211-215.
- Ueng SW, Shih CH. Augmentative plate fixation for the management of femoral nonunion with broken interlocking nail. J Trauma. 1998; 45(4):747-752.
- Gibbons CE, Pope SJ, Murphy JP, Hall AJ. Femoral metastatic fractures treated with intramedullary nailing. Int Orthop. 2000; 24(2):101-103.
- Windhager R, Ritschl P, Rokus U, Kickinger W, Braun O, Kotz R. The incidence of recurrence of intra- and extra-lesional operated metastases of long tubular bones [in German]. Z Orthop Ihre Grenzgeb. 1989; 127(4):402-405.
Drs Kinkel and Gotterbarm are from Orthopaedic University Clinic Heidelberg, Dr Stecher is from the Department of Trauma and Reconstructive Surgery, St. Vincentius Clinic, Karlsruhe, Dr Bruckner is from the Department of Medical Biometrics, University of Heidelberg, and Prof Holz is the former medical director of the Clinic for Trauma and Reconstructive Surgery, Katharinenhospital, Stuttgart, Germany.
Drs Kinkel, Stecher, Gotterbarm, and Bruckner and Prof Holz have no relevant financial relationships to disclose.
Correspondence should be addressed to: Stefan Kinkel, MD, Orthopaedic University Clinic Heidelberg, Schlierbacher Landstr 200a, 69118 Heidelberg, Germany.