Valgus intertrochanteric osteotomy for coxa vara of Bucholz Ogden Types II and III in patients older than 30 years

Transcription

1 Arch Orthop Trauma Surg (2011) 131: DOI /s ORTHOPAEDIC SURGERY Valgus intertrochanteric osteotomy for coxa vara of Bucholz Ogden Types II and III in patients older than 30 years Jan Bartoníbek Jaroslav Vávra Received: 19 December 2010 / Published online: 18 February 2011 Springer-Verlag 2011 Abstract Background Ischemic necrosis of the proximal femur resulting in coxa vara is a severe iatrogenous complication of the treatment of developmental dysplasia of the hip (DDH). Severe relative overgrowth of the greater trochanter and reduction of the neck result in insuyciency of hip abductors. Unequal limb length causes obliquity of the pelvis, compensatory scoliosis of the lumbar spine and valgus deformity of the ipsilateral knee. The purpose of this study was to investigate the evect of valgus intertrochanteric osteotomy in patients with coxa vara older than 30 years. Methods Fifteen female patients of the average age 43 years (range 31 60) with postdysplastic varus deformity of the proximal femur and shortening of avected limb of 2 4 cm were treated with valgus intertrochanteric osteotomy. None of them was operated on before. In eight cases, the varus deformity of the proximal femur was evaluated as Bucholz-Ogden Type II, in seven cases as Bucholz Ogden Type III. The average follow-up was 10 years (range 5 20). Results By the time of the last functional follow-up, three patients had already underwent conversion of osteotomy to total hip arthroplasty (THA), namely, 7.5, 11, and 12 years after osteotomy. All the patients evaluated the evect of osteotomy positively, including those treated later with J. Bartoníbek (&) Department of Surgery of the 1st Faculty of Medicine, Charles University, Thomayer University Hospital, Vídekská 800, Prague-4, Czech Republic bartonicek.jan@seznam.cz J. Vávra Orthopaedic Department of 3rd Faculty of Medicine, Charles University, Prague 10, Czech Republic vavra@fnkv.cz THA. The average preoperative Harris Hip Score was 83, the postoperative one was 93. Conclusions Based on results, the valgus intertrochanteric osteotomy appears to be a reliable treatment for postdysplastic coxa vara in patients older than 30 years. Keywords Coxa vara Intertrochanteric osteotomy DDH Introduction Ischemic necrosis of the proximal femur resulting in coxa vara is a severe iatrogenous complication of the treatment of developmental dysplasia of the hip (DDH). Bucholz and Ogden [1] hypothesized that speciwc vascular occlusion may progress in diverent morphological patterns and identiwed four distinct roentgenographic types Type I, II, III and IV. Thomas et al. [2] subsequently subdivided Type II into Type IIA and Type IIB. These types could be of importance in predicting the deformity of proximal femur in adulthood. Clinically the most important are the Bucholz Ogden (B O) Types II and III (Fig. 1). Residual deformity commonly called postdysplastic coxa vara is in Type II characterized by a short neck, relative overgrowth of the greater trochanter, sometimes also by the coxa plana and acetabular dysplasia and limb shortening of 2 3 cm. In Type III, these changes are even more signiwcant, Wrst of all the overgrowth of the greater trochanter, reduction of the neck to a mere notch between the greater trochanter and the mushroom femoral head, and limb-length discrepancy up to 7 cm [2]. Severe relative overgrowth of the greater trochanter and reduction of the neck result in insuyciency of hip abductors, in some cases accentuated by acetabular dysplasia. Unequal limb length causes obliquity of the pelvis, compensatory

2 1212 Arch Orthop Trauma Surg (2011) 131: this study. The average follow-up was 10 years (range 5 20), eight patients were followed up for more than 10 years. Methods Fig. 1 Types II and III of B O classiwcations. a, b, c Types II; d, e, f Types III scoliosis of the lumbar spine and valgus deformity of the ipsilateral knee. All these deformities may in middle-aged patients lead to osteoarthritis of the hip joint and degenerative changes of the lumbar spine. The method of treatment of coxa vara is osteotomy of the proximal femur, which produces the most eycient results between 7 and 12 years of age [3]. Wagner [4], Morscher [5] and other authors [6 8] described diverent types of valgus osteotomy of the proximal femur. However, long-term results of valgus osteotomy in children and adolescents with B O Types II and III were published only by Vávra et al. [3]. We would like to present our experience with outcomes of the valgus intertrochanteric osteotomy (VITO) in patients older than 30 years. Group of patients and methods Group of patients The group comprised 15 female patients. The average age of patients was 43 years (range 31 60). All of them were as newborns treated conservatively for unilateral DDH by abduction device. None of them was operated on before. All of them had only one hip avected, including 12 left hips, 3 right hips. In eight cases, the deformity of the proximal femur was evaluated as B O Type II, in seven cases as B O Type III. The average Sharp angle [9] was 43 (range 34 50), the average Wiberg CE angle [10] was 18 (range 12 25). All the patients were operated on by the Wrst author of Indication for VITO was limping, hip pain, shortening of leg of 2 cm and more and low back pain. Preoperative planning based on radiographic examination was crucial to the success of the surgery. Anteroposterior (AP) radiographic view of the pelvis, AP radiographs of the avected hip in 15 of internal rotation were obtained in each patient. The femoral neck-shaft angle was measured on both sides, and the articular-trochanteric distance (ATD), i.e., relationship of apex of the greater trochanter and the femoral head [2], was determined from the pelvic radiograph. The resulting valgization angle, wedge size, and amount of lateral displacement of the femoral shaft were always a compromise based on the given anatomic situation of the avected hip joint. An AP view was traced on paper. The next step was to determine the angle and point of insertion of the plate blade. For a 120 angled blade plate, the angle of the plate insertion to the proximal femur (shaft axis) was determined by the intended valgization angle of +60 (a complementary angle to 120 ). The osteotomy was planned at the level of the top of the lesser trochanter. Subsequently, the osteotomy line was drawn parallel to the plate blade. In the next step, the wedge to be resected was drawn on the distal fragment. Then, the two drawn fragments were cut out, matched, and the impact of the lateral displacement of the distal fragment on the limb lengthening was determined. At surgery, the lateral longitudinal approach in a supine position of patient was used. After stripping the proximal portion of the vastus lateralis muscle, an L-shaped anterior arthrotomy running in the direction of the lesser trochanter was done. This improved visualization for the introduction of the plate blade and made it possible to release the insertion of the medial part of the capsule in the area of the lesser trochanter, including the iliopsoas tendon. The seating chisel was inserted. The line of osteotomy was made parallel to the inserted chisel. The angle of the wedge was determined on the basis of preoperative measuring and ranged between 30 and 50. After removal of the seating chisel, the plate blade was inserted in the proximal fragment. We chose a longer length of the blade by cm than originally determined by the depth of the seating chisel. As a result, the length of the lateral part of the blade corresponded to the planned lateral displacement of the protruding part of the proximal femur. Then, a careful valgus reduction of fragments was done by abduction of the limb. The plate was Wxed to the femoral shaft by boneholding forceps. After completion of the plate Wxation with screws, the range of motion in the hip joint and limb length were examined, drains inserted, and the wound closed.

3 Arch Orthop Trauma Surg (2011) 131: Table 1 Overview of patients N Age (y) SA (de) WA (de) BO ATD1 ATD2 LS1 LS2 LE F-U (Mo) HHS 1 HHS 2 Note II THA 132 mo after VITO III III ReWxation 4 Mo after VITO THA 90 Mo after VITO II II Non-union, rewxation 4 Mo after VITO III Varus displacement after VITO, THA 168 Mo after VITO III II II III II III Partial necrosis of femoral head, BHSA 112 Mo after VITO II BHSA 14 Mo after VITO II II N number of patient, SA Sharp angle, WA Wiberg (CE) angle, B O Bucholz Ogden type, ATD articular-trochanteric distance, LS1 limb shortening before surgery, LS2 limb shortening after surgery, LE lengthening evect, F-U follow-up, HHS 1 Harris Hip Score before surgery, HHS 2 Harris Hip Score after surgery, y years, de degrees, Mo months, THA total hip arthroplasty, VITO valgus intertrochanteric osteotomy, BHSA Bosworth s hip shelf arthroplasty Postoperative AP and lateral radiographs were obtained. The patient was immobilized with crutches, non-weight bearing on the avected extremity. If clinical and radiographic follow-up at 6 weeks was satisfactory, the patient was begun on progressive weight bearing; full weight bearing was usually allowed after 3 months. Hardware was removed in all patients but one, usually within 1 year after osteotomy. Additional procedure: Bosworth shelf arthroplasty [11] was performed according to the original descriptions in two patients with residual acetabular dysplasia 1 and 9 years after VITO. Evaluation Evaluation covered intraoperative and postoperative complications, time to healing of the osteotomy, the limb length prior to operation and after healing of the osteotomy, articular-trochanteric distance and the Trendelenburg sign prior to and after the operation, progress or development of osteoarthritic changes after osteotomy and Harris Hip Score (HHS) [12]. Results The most important data are included in Table 1. Surgical site healed in all the patients without any complications. Radiological and functional results Prior to operation, 11 hips showed no signs of osteoarthritis, while in 4 hips, osteoarthritic changes of Grade I according to the Tönnis classiwcation [13] were found, including patient N. 12 who had partial avascular necrosis of the femoral head before surgery. Of 11 hips initially not avected by osteoarthritis, 1 patient (N. 10) developed osteoarthritis of Grade I, however, without any impact on the functional result. One patient (N. 6) developed osteoarthritis during 14 years after osteotomy that progressed to Grade III and required THA. In two of four osteoarthritic hips, osteoarthritis progressed from Grade I to Grade III during 7.5 years (N. 3) and 11 years (N. 1) and required THA. Two patients with Grade I osteoarthritis (N 12, N. 14) showed a mild radiological regression (Fig. 2). In patient N. 12, there occurred also resorption of necrotic segment of the femoral head after the operation (Figs. 3, 4). By the time of the last functional follow-up, three patients already underwent conversion of osteotomy to THA, namely (N. 3), (N. 1) and (N. 6), 7.5, 11 and 14 years after osteotomy, respectively. All the patients evaluated the evect of osteotomy positively, including those treated later with THA. As the main benewt, they considered reduction or elimination of unequal limb length, i.e., elimination of limp. Three patients (N. 3, N. 5, N. 7) reported pain in lumbosacral (LS) spine at the Wnal follow-up. Radiological

4 1214 Arch Orthop Trauma Surg (2011) 131: Fig. 2 Improving of osteoarthritis. a Patient N. 14, pelvic radiograph before osteotomy, B O Type II and osteoarthritis of Grade I of Tönnis classiwcation, shortening of left limb of 2 cm. b Pelvic radiograph 5 years after VITO, equalization of limb-length discrepancy, slightly improving of osteoarthritis Fig. 4 Resorption of necrotic segement of the femoral head. a Patient N. 12, pelvic radiograph before osteotomy, b 8 years after VITO, c 13 years after VITO and 2 years after Bosworth shelf arthroplasty, equalization of limb-length discrepancy, resorption of necrotic segment of the femoral head Articular-trochanteric distance and the Trendelenburg sign Fig. 3 Resorption of necrotic segment of the femoral head. a Patient N. 12, B O Type III, osteoarthritis of Grade I of Tönnis classiwcation, partial avascular necrosis of the femoral head of the left hip, shortening of the left limb of 3 cm and negative ATD of 1.5 cm. b Immediately after VITO, c 8 years after VITO, resorption of necrotic segment of the femoral head and positive ATD of 1.5 cm examination proved degenerative scoliosis of the LS spine. HHS was evaluated prior to and after the operation in 12 patients where the osteotomy evect lasted. The mean preoperative HHS was 83 (range 61 84), the mean postoperative HSS was 93 (range 78 98). Limb shortening The average preoperative limb shortening was 2.4 cm (range 2 4 cm) and postoperative limb shortening was 0.4 cm (range cm), i.e., the average limb lengthening was 2 cm (range cm) (Fig. 5). The mean ATD value was 0.4 cm (range 2.5 to +1.5) prior to operation and +1.0 cm (range 0.5 to +2.5) after the osteotomy healed (Fig. 6). Prior to operation, the Trendelenburg sign was positive in 12 patients, i.e., in all of those where ATD was less than 1.0 cm. At the last followup, the Trendelenburg sign was evaluated only in 9 of these 12 patients; 3 patients had already undergone THA by that time. The sign remained positive in Wve patients where ATD was less than 1 cm. On the other hand, in three patients where the ATD value was more than 1.0 cm prior to operation, the Trendelenburg sign was negative both prior to and after the osteotomy. Complications We encountered one intraoperative and three postoperative complications. The only intraoperative complication consisted in insertion of the plate blade too high, partially ov

5 Arch Orthop Trauma Surg (2011) 131: Fig. 5 Osteotomy in B O Type III. a Patient N. 10 pelvic radiograph before osteotomy, B O Type III, shortening of left limb of 3 cm and negative ATD 1.0 cm. b Pelvic radiograph 8 years after VITO, shortening of 1 cm of the left limb, positive ATD 1.0 cm and osteoarthritis of Grade I of Tönnis classiwcation the femoral neck (N. 3). This complication was treated by valgus re-osteotomy equalizing the limb length. Postoperative complications included one non-union and one varus displacement of 10. While in 14 patients osteotomy healed within 3 months, 1 patient (N. 5) developed a non-union that was rewxed by a 95 angled blade (condylar) plate. The non-union healed in 4 months. One patient (N. 6) fell down 1 month after the operation, which resulted in varus displacement of the proximal fragment by 15 and loss of lengthening by 1 cm. In this position, the osteotomy healed. Discussion Even though our study is retrospective and the patients were operated on within the interval of 16 years, it has certain benewts. All the patients were operated on for dysplastic coxa vara by one orthopedic surgeon, using the same type of osteotomy Wxed by the same type of implant. VITO was the Wrst operation in all the avected hips and osteoarthritic changes were maximally of Grade I. Coxa vara was in a greater detail speciwed by means of the B O classiwcation and ATD. B O classiwcation and ATD are very useful for a more detailed speciwcation of postdysplastic varus deformity of the proximal femur. In contrast, exact measuring of the neck-shaft angle in B O Type III is problematic due Fig. 6 Improving of articular-trochanteric distance. a Patient N. 11, pelvic radiograph before osteotomy, B O Type II, ATD 0.5 cm and shortening of left limb of 3 cm. b Pelvic radiograph 13 years after VITO, equalization of limb-length discrepancy, ATD 2.5 cm to the deformity of the femoral head and the femoral neck. The same applies in Type III to measuring of the Wiberg angle. On the other hand, the Sharp angle can be in these cases measured quite exactly. The average age of our patients was the highest of all the above-mentioned studies. In spite of this, the osteotomy has met its purpose in all the patients including those with a subsequent THA. DiVerent types of osteotomy to treat postdysplastic coxa were described [3 8]. Generally known is mainly the Wagner three-fragment osteotomy [4]. A similar technique was described by Morscher [5] and Graf et al. [8]. Part of all the three techniques is osteotomy of the greater trochanter. Its subsequent rewxation by a plate or screws is technically demanding and the plate or screw may pull out or a non-union may develop there [7]. Original is the longitudinal osteotomy of the proximal femur published by Papavasiliou and Kirkos [6]. We preferred a simple two-fragment osteotomy with lateral displacement of the femoral shaft, which we found benewcial in diverent indications already before [3, 14 16]. In spite of this, we could not avoid certain complications. In patient N. 3, the image intensiwer was not used and the blade of plate was misinserted. Non-union in patient N. 5 developed most probably due to thermic necrosis to the bone caused by oscillating saw. Varus displacement after the patient s fall could be partially

6 1216 Arch Orthop Trauma Surg (2011) 131: caused also by the plate blade inserted not deep enough into the femoral head. The mentioned complications show that osteotomy in postdysplastic coxa vara, particularly in Type III of the B O classiwcation, is a highly demanding procedure. Based on the experience gained in this and previous studies [3, 14 16] we consider important to reiterate the following steps. AP radiograph of avected hip should be taken in internal rotation of 15. Such radiograph will show the exact shape of the femoral neck and is the most suitable for preoperative planning. Preoperative planning should be done preferably by the classic method, i.e., using transparent paper, a pencil and scissors, rather than using a computer. Drawing and cutting requires greater concentration on details and better simulates the actual operation. An important step is the L-shaped anterior arthrotomy connected with release of the medial capsule from the proximal fragment and tenotomy of the iliopsoas tendon at the level of the lesser trochanter. This will signiwcantly facilitate reduction of the distal fragment to the plate. The plate blade must be inserted as far as the center of the femoral head. This increases its Wxation in the proximal fragment. The osteotomy line runs parallel to the plate blade, at the distance of minimally 2 cm from the blade. Otherwise, there is a danger of cut out of the blade from the proximal fragment. The results of intertrochanteric osteotomies in coxa vara were published by several authors [3, 6 8, 17 19]. Perlau et al. [17] followed up 18 hips treated by isolated proximal femoral osteotomy for residues of congenital dysplasia for 5 10 years postoperatively. These residues concerned both the acetabulum and the proximal femur. The average age of patients was 33 years. Varus osteotomy was used in 11 cases and valgus osteotomy only in 7 cases. Only eight patients were satiswed with the outcome of the operation. Haverkamp and Marti [18] published the results of intertrochanteric osteotomy combined with acetabular shelfplasty in 15 dysplastic hips in 13 patients with secondary osteoarthritis. The average age of these 13 patients was 35 years; the average follow-up was 18 years. In three patients, conversion to THA was performed 9, 17, and 22 years after osteotomy. The result in the rest of the patients remained good. Bues and Morscher [19] presented the results of the Morscher valgus osteotomy for coxa vara in 16 hips (15 patients). The average age of patients was 26 years, the average follow-up 5 years. The average limb shortening was 2.3 cm prior to osteotomy, and 1.0 cm after osteotomy. Good result was achieved in 14 hips. Graf et al. [8] published the results of a triple valgus osteotomy of the proximal femur for coxa vara in 40 patients, at the average age of 17 years, with average shortening of 3 cm and the follow-up of 6 years. The average lengthening was 2.5 cm. Very good result was achieved in 10 patients, good and satisfactory result in 15 patients. Vávra et al. [3] presented the results of a two-fragment valgus intertrochanteric osteotomy with a lateral displacement of the femoral shaft in 77 children and adolescents (69 hips of B O Type III and 8 hips of B O Type II) at average age of 8 years at the time of operation (range 3 15 years). The average follow-up was 15 years (range 7 23 years). Based on the results, the authors set the optimal age for osteotomy between 7 and 10 years. Papavasiliou and Kirkos [6] evaluated the results of longitudinal osteotomy in 16 patients with coxa vara at average age of 16 years and with average shortening of 3 cm. The average follow-up was only 4.3 years. Equal limb length was achieved in all the patients. The Trendelenburg sign was negative 1 year after operation in ten patients. Good clinical and radiological result was achieved in all the patients. Lengsfeld et al. [7] presented the results of the Wagner valgus double osteotomy in 15 patients with coxa vara at the average age of 25 years, with the average follow-up of 10 years. Nine patients developed deformity after DDH, six patients after the Perthes disease. The average shortening was 1.8 cm prior to operation and 1.2 cm postoperatively. Osteoarthritic changes were present prior to operation in 8 patients, at the last follow-up in 14 patients. The average ATD was 0.74 cm prior to operation and cm postoperatively. The Trendelenburg sign was positive prior to operation in eight patients and after operation in two patients. No clear dependence of the Trendelenburg sign on the ATD value was identiwed. A very good result based on the Merle d Aubigné s score was recorded in Wve cases, good result in one, satisfactory in six, and poor in two cases. The literature analysis shows that the groups of patients are not but for a few exceptions large and the follow-up is not very long. Another disadvantage of the above-mentioned studies is heterogeneity of their groups of patients. They include patients with a diverent disease (DDH, Perthes, coxitis etc.), of diverent average age at the time of operation, in a diverent stage of osteoarthritis, with diverent types of osteotomy (varus, valgus), patients operated on previously. Some of the studies do not include data on the limb length; ATD was mentioned only by Lengsfeld et al. [7]. In most cases, the shape of the proximal femur and its varus deformity was not speciwed. The percentage of good results in individual studies varies, although most authors evaluated the evect of osteotomy positively [3, 6 8, 17 19]. The era of intertrochanteric osteotomies culminated in the 1970s and 1980s of the last century [20 22]. Nevertheless, even today they still are a useful procedure [23 25]. Based on our results we believe that VITO is a reliable treatment for postdysplastic coxa vara in patients older than 30 years.

7 Arch Orthop Trauma Surg (2011) 131: Acknowledgments This study was supported by a grant from the Ministry of Health of the Czech RepublicGrant IGA MZ: NS Importance of intertrochanteric osteotomy and acetabular coverage procedures at adolescent and adults for preservation of long-term hip function. References 1. Bucholz RW, Ogden JA (1978) Patterns of ischemic necrosis of the proximal femur in nonoperatively treated congenital dislocation of the hip. In: The Hip. Proceedings of the sixth open scientiwc meeting of the hip society. St. Louis, Mosby, pp Thomas CL, Gage JR, Ogden JA (1982) Treatment concepts for proximal femoral ischemic necrosis complicating congenital hip disease. J Bone Joint Surg Am 74: Vávra J, Bech O, Zídka M (2000) Valgus osteotomy in the treatment of the ischemic necrosis of femoral head in children. Part II Clinical evaluation of the group. Acta Chir Orthop Traumatol Bech 67: Wagner H (1977) Korrekturosteotomien am Bein. Orthopäde 6: Morscher E (1980) Intertrochanteric osteotomy in the treatment of osteoarthritis of the hip. In: The Hip. Proceedings of the eight open scientiwc meeting of the hip society. St. Louis, Mosby pp Papavasiliou V, Kirkos JM (1997) Reconstruction of residual deformities of the hip. Clin Orthop Relat Res 341: LengsXed M, Schuler P, Griss P (2001) The long term (8 12 years) results of valgus and lengthening osteotomy of the femoral neck. Arch Orthop Trauma Surg 121: Graf R, Tschauner C, Klapsch W (1992) Dreifachosteotomie des proximalen Femurendes bei Coxa vara mit Hochstand des Trochanter major und Beinverkürzung. Operativ Orthop Traumatol 4: Sharp IK (1961) Acetabular dysplasia: acetabular angle. J Bone Joint Surg Br 43-B: Wiberg G (1939) Studies on dysplastic acetabula and congenital subluxation of the hip joint. Acta Chir Scand 83(Suppl 57): Bosworth DM, Fielding JW, Ishizuka T, Ege R (1961) Hip-shelf operation in adults. J Bone Joint Surg Am 43-A: Harris WH (1969) Traumatic arthritis of the hip after dislocation and acetabular fractures: treatment by mold arthroplasty. An endresult study using a new method of result evaluation. J Bone Joint Surg Am 51-A: Tönnis D (1987) Congenital dysplasia and dislocation of the hip in children and adults. Springer, New York 14. Bartoníbek J, Skála-Rosenbaum J, Douna P (2003) Valgus intertrochanteric osteotomy for malunion and nonunion of trochanteric fractures. J Orthop Trauma 17: Bartoníbek J, Havránek P (2004) Gunshot injury of the proximal femoral physis. Arch Orthop Trauma Surg 124: Bartoníbek J, Vávra J, Bartonka R (2010) Operative treatment of avascular necrosis of the femoral head after slipped capital femoral epiphysis. Arch Orthop Trauma Surg. doi: /s Perlau R, Wilson MG, Poss R (1996) Isolated proximal femoral osteotomy for treatment of residua of congenital dysplasia or idiopathic osteotoarthrosis of the hip. J Bone Joint Surg Am 78- A: Haverkamp D, Marti RK (2005) Intertrochanteric osteotomy combined with acetabular shelfplasty in young patients with severe deformity of the femoral head and secondary osteoarthritis. J Bone Joint Surg Br 87: Buess P, Morscher E (1988) Die schenkalsverlängernde Osteotomie mit Distalisierung des Trochanter major bei Coxa vara nach Hüftluxation. Orthopäde 17: Müller ME (1971) Die hüftnähen Femurosteotomien. 2. AuXage, Stuttgart 21. Pauwels F (1976) Biomechanics of the normal and diseased hip. Berlin, Springer 22. Schatzker J (ed) (1984) The intertrochanteric osteotomy. Berlin, Springer 23. Santore RF, Kantor SR (2004) Intertrochanteric femoral osteotomies for developmental and posttraumatic conditions. J Bone Joint Surg Am 86: Haverkamp D, Eijer H, Besselaar PP, Marti RK (2008) Awareness and use of intertrochanteric osteotomies in current clinical practice. An international survey. Int Orthop 32: Zweifel J, Hönle W, Schuh A (2010) Long-term results of intertrochanteric varus osteotomy for dysplastic osteoarthritis of the hip. Int Orthop. doi: /s