Subtrochanteric shortening with overlapping femoral resection in total hip arthroplasty for Crowe type IV adult dislocation of the hip

Abstract

Objective:  To summarize the surgical technique and clinical outcomes of subtrochanteric shortening with overlapping femoral resection in primary total hip arthroplasty (THA) for Crowe type IV adult dislocation of the hip (ADH).

Methods:  From January 2000 to December 2005, 12 patients with ADH (15 hips) were treated with primary cementless hip arthroplasty using the method of subtrochanteric shortening with overlapping femoral resection. There were three male and nine female patients (nine unilateral and three bilateral hips) with an average age of 56 years (range, 41–75). Subtrochanteric shortening with overlapping femoral resection and ‘V’ shaped derotational osteotomy were performed in all cases without soft tissue cutting release. Proximal femoral shaft splitting was performed as an adjunct in 10 hips.

Results:  The mean follow‐up time was 6 years (range, 3–8). There were no infections, nonunion, malunion, dislocation or nerve injury of traction. Postoperative X‐ray films showed that the acetabular cups were placed in anatomical position with 95% coverage of the acetabulum. Furthermore, initial stability of the femoral stem fixation was satisfactory and all osteotomies healed in 10–15 weeks. The Harris hip score had improved from 25–32 to 88–98 at one year after surgery (P < 0.01). All acetabular and femoral components were judged to be osteointegrated and well‐fixed during follow‐up. No components have needed revision.

Conclusions:  Subtrochanteric shortening osteotomy is a safe and predictable method for restoring the anatomic hip center in Crowe type IV ADH. The clinical outcomes of treating Crowe type IV ADH with THA were satisfactory.

Introduction

Crowe type IV adult dislocation of the hip (ADH) is usually treated by total hip arthroplasty (THA). The ideal therapeutic method is to fix the cup into the true acetabulum, as restoring normal biomechanics is required ¹ . However, contraction of the soft tissue accompanied by long‐term hip dislocation often makes reduction of femoral components impossible. Although release of soft tissue may solve this problem, the common lengthening of the lower limb by more than 4 cm during reduction is usually complicated by stretch injury of the nerve ² . In recent decades, subtrochanteric shortening resection has been considered to be an effective method that can facilitate reduction and avoid nerve palsy ³ , ⁴ , ⁵ , ⁶ . However, problems still exist when calculating the length of femur to resect. Traditionally, preoperative X‐ray measurement often results in inaccurate calculation and complications because the degree of contraction of soft tissue differs between patients and the exact amount of distraction which causes sciatic nerve dysfunction is unknown ⁴ , ⁵ , ⁶ . From January 2000 to December 2005, we performed 15 cementless THA with this technique (1, 2, 3, 4) in 12 Crowe type IV ADH patients. Measurement of the length to resect from the femur is based only on the overlapping distance of the osteotomy ends (1, 2). A summary of our experience and the clinical outcomes will be reported in this article.

Figure 1.

Diagrammatic representation of femoral osteotomy below the lesser trochanter and trial femoral component placed into proximal part of the femur and hip relocated.

Figure 2.

Diagrammatic representation of method for subtrochanteric shortening with overlapping femoral resection.

Figure 3.

Diagrammatic representation of a ‘V’ shaped osteotomy at the osteotomy site proximally and distally.

Figure 4.

Diagrammatic representation of reduction of the osteotomy and implantation of the stem.

Materials and methods

From January 2000 to December 2005, 15 Crowe type IV ADH hips in 12 patients were treated with primary cementless THA along with subtrochanteric shortening osteotomy. There were nine unilateral and three bilateral arthroplasties in three men and nine women, with an average age of 56 years (range, 41–75). X‐ray and computed tomography (CT) examinations were routinely performed to assess the associated bony deformity preoperatively. These deformities were graded according to the Crowe classification ⁷ . Only Crowe type IV patients were included. The surgical indications included severe hip pain with difficulty in walking and performing daily activities.

The adopted acetabular cups included Duraloc in six (Depuy, Warsaw, IN, USA), Reflection in two (Smith and Nephew Richards, Memphis, TN, USA), and Pressfit SIIin seven (LINK, Hamburg, Germany). The outside diameter of the acetabular cups ranged from 42 to 48 mm, and the thickness of the polyethylene insert ranged from 6.8 to 9.8 mm. Three AML (Depuy), three Summit (Depuy), two Synergy (Smith & Nephew Richards), and seven ribbed femoral components (LINK) were used. Cobalt‐chromium or ceramic femoral heads were applied with a diameter of 22.225 mm or 28 mm.

Subtrochanteric shortening with overlapping femoral resection and ‘V’ shaped derotational osteotomy (1, 2, 3) was performed. The sagittal longitudinal split arthroplasty of proximal femur or both proximal and distal femur (Fig. 4) was added in 10 hips. No cutting release of soft tissue was performed.

Surgical technique

The modified Hardinge approach was applied. The incision was made along the greater trochanter to expose the femoral head, neck and true acetabulum. The femoral neck was cut at the level determined by preoperative templating. The true acetabulum was reamed sequentially and the acetabular cup seated anatomically. Then femoral medullary cavity reaming and broaching were accomplished according to the desired size. A subtrochanteric transverse osteotomy was performed (Fig. 1), then the femoral stem and head trial components were introduced into the proximal femoral medullary cavity and the hip joint was reduced temporarily. Traction was then applied to the distal femur while the length of overlapping femoral shaft was measured and resected for shortening (Fig. 2). A ‘V’ shaped osteotomy at the distal end of the proximal femur was performed with the apex of osteotomy rotated laterally. The degree of rotation was determined by measurements on CT scan, and the osteotomy apex centered anteriorly for a ‘V’ shaped osteotomy at the proximal end of the distal femur (Fig. 3). Then, the osteotomy ends were reduced and a standard femoral stem trial with the desired size was implanted. A sagittal longitudinal femur‐splitting arthroplasty was performed when implantation of the smallest stem trial was difficult due to a narrow femoral medullary cavity, or when there was instability at the osteotomy ends after implantation of the femoral stem trial. Finally, the standard femoral stem was inserted according to the size of trial components, the split femur tied with cerclage wires at one or multiple levels and the gap at the longitudinal split and the osteotomy site were filled with morselized autogenous bone (Fig. 4). The range of motion of the reduced hip joint was checked and then the wound closed once there was no tendency to dislocation.

Postoperative care

Prophylactic antibiotic was applied for 36–48 h. Bed rest was recommended for 6 weeks while the operated leg was placed in an abducted and neutral position without external fixation. Active motion was restricted in the operated leg, and muscular contraction training was encouraged. The patients were allowed to walk with double crutches after 6 weeks and with a single crutch after 10 weeks. Appropriate exercises for the gluteus medius were encouraged until the crutch could be discarded.

Evaluating criteria and statistical analysis

Preoperative and postoperative hip function was evaluated by Harris hip score. The muscle strength of the gluteus medius and gait were evaluated in accordance with the standards of five ⁸ and four classes (no, mild, moderate and severe) ⁹ , respectively. Trendelenburg's sign was also evaluated. X‐ray assessment of the femoral components and union at the osteotomy ends was according to the Engh ¹⁰ and Masonis criteria ⁹ , respectively. The preoperative and postoperative discrepancy in the lower limbs was measured by the foot pad method. Harris hip scores before and one year after surgery were compared using the paired Student's t‐test. P < 0.05 was considered to be significant.

Results

There were no infections, nonunion, malunion and other complications in this group. There were no difficulties in reduction or sciatic nerve stretching injury caused by too short an osteotomy. At the same time, there were no lax dislocations caused by too long an osteotomy. Postoperative X‐ray films showed that the acetabular cups were placed in the true acetabulum with more than 95% of acetabular coverage. In addition, the initial stability of the femoral stem was satisfactory and all osteotomies had healed by 10–15 weeks. The average length of femur resected was 37 mm (range, 24–55 mm). The average length achieved through lengthening was 30 mm (range, 25–35 mm). Leg‐length inequalities in nine patients with unilateral involvement were restored to an average of 28 mm (range, 20–34 mm) discrepancy after surgery, from their baseline of 52 mm (range, 46–65 mm) before surgery. The affected side was not equal to the healthy side for unilateral patients, but all our patients felt that the affected side was markedly elongated and more comfortable during walking.

The mean follow‐up time was 6 years (range, 3–8). The Harris hip score improved from 25–32 points preoperatively to 88–98 points one year after surgery (P < 0.01). Strength assessment of the gluteus medius was all above grade 4. Eight patients (67%) presented a negative Trendelenburg's sign. Limping gait improved from 67% with severe claudication and 33% with moderate claudication preoperatively to 67% without claudication and 33% with mild claudication one year after operation, indicating a remarkable improvement. All acetabular and femoral components were judged to be osteointegrated and well‐fixed on early and recent radiographic evaluation. So far, no progressive radiolucent lines or significant osteolysis around the stems has been observed. Also, no components have needed revision (Fig. 5).

Figure 5.

Radiographs showing the procedure (a) Preoperative anteroposterior radiograph. The hip of a 57‐year‐old woman with unilateral Crowe type IV ADH with a 65 mm leg‐length discrepancy before surgery. (b) Immediate anteroposterior radiograph after THA with a subtrochanteric shortening osteotomy. (c) Anteroposterior radiograph after follow‐up of 4 years shows healing of the osteotomy with no evidence of loosening of either component.

Discussion

After the acetabular component has been positioned in the true acetabulum for Crowe type IV ADH, reduction of the femoral head into the true acetabulum is impossible, due to contracture of soft tissue secondary to long‐term high dislocation of the femoral head. In recent decades, the method of subtrochanteric shortening osteotomy has been extensively applied in clinical practice ³ , ⁴ , ⁵ , ⁶ . This technique was first used in a child's congenital hip surgery to correct the abnormal anteversion and avoid the femoral head necrosis which can be caused by reduction under high tension ³ . This technique was adopted in THA of ADH patients for a totally different purpose: to rebuild the anatomical hip center, avoid nerve injury and improve the curative effect ¹⁰ .

However, there is a problem in calculating the length of shortening osteotomy correctly. If the resection is too long, it aggravates the inequality of lower limbs and lack of soft tissue tensioning can lead to a loosening dislocation. If the resection is too short, it leads to difficulty in reduction and nerve distraction injury. The traditional method is to subtract the safety lengthening (i.e. <4 cm) from the distance between the centers of the femoral head and true acetabulum based on X‐ray measurement ³ , ⁴ , ⁵ . However, because individual differences in soft tissue contracture surrounding the hip joint exist, and the threshold of distraction which causes sciatic nerve dysfunction is unknown in clinical practice, we believe measurement of the osteotomy length according to X‐ray films is inaccurate and may lead to the above‐mentioned complications. In the present study, we judged the resection length based on the overlapping measurement instead of on preoperative X‐ray measurement. Neither loosening dislocation nor nerve distraction injury occurred, confirming the reliability of femoral shortening by this technique. In addition, our postoperative measurement showed an average of 3 cm lengthening, which means the affected side was not exactly equal to the healthy side. However, all patients were comfortable during walking. This indicates that after overlapping femoral resection the lower limb can achieve an ideal functional length.

For Crowe type IV ADH, there are three different viewpoints about soft tissue release after the acetabular cup is placed at the anatomical site. Firstly, Harley and Wilkinson believe that the short bellies of the iliacus, tensor fascia lata and glutei muscles should only be shortened enough for prosthetic joint reduction without undue tension to be achieved by releasing these muscles from the pelvis with excision of the upper third of the ilium ² . As a result, the long bellies of the psoas, adductors and hamstrings can easily be stretched. Secondly, Veal, Haddad, and Cabanela state that only the long bellies of the hamstrings, adductors, and quadriceps should be shortened, and that tension of the hip abductors would not be a problem ¹¹ , ¹² , ¹³ . This point had been previously noted by Harris, who stated that with the hip fully dislocated, the abductors were frequently stretched around the greater trochanter when the patient sat down and thus did not allow excessive shortening (in the study of WH Harris as cited in Veal et al. ¹¹ ). So Veal and Cabanela have pointed out that it is not necessary to release these abductors ¹¹ , ¹³ . Finally, some others thought that both the long and short bellies of the muscles should be shortened, so that the prosthetic joint could be reduced either by releasing these muscles with cutting or tenotomy at their insertion, or by proximal femoral shortening and reattachment of the osteotomized trochanter and abductors ⁵ , ¹⁴ . In the current series, we did not release any short belly muscles such as the gluteus and the iliacus. There was no difficulty in reduction of the hip joint, which confirmed that difficulty in reduction of the femoral head in Crowe type IV ADH is attributable only to contracture of the long belly muscles.

Because patients with Crowe type IV ADH often have with associated bony deformities such as abnormal anteversion and a narrow femoral medullary cavity, there are other surgical techniques that may assist in subtrochanteric shortening resection. For example, in order to correct the anteversion deformity and simultaneously obtain rotational stability of the osteotomy ends, one of the rotational osteotomy techniques such as ‘Z’ shape ¹⁴ , ‘V’ shape ¹⁵ , oblique ¹⁶ , and transverse ¹⁰ could be added. In the current study, we performed derotational ‘V’ shaped osteotomy to correct the anteversion deformity and improve the stability of the osteotomy ends. Our clinical results showed that dislocation secondary to abnormal anteversion and nonunion at the osteotomy site secondary to instability of the osteotomy ends could both be avoided with this technique. For some cases where the femoral medullary cavity was excessively narrow, even the smallest prosthesis had no chance for implantation. We performed a sagittal longitudinal split of the proximal femur (as well as distal splitting for some patients) to successfully solve this problem. Follow‐up of X‐ray films showed all the femoral components to be osteointegrated.