Surgical management of the congenital dislocation of the knee and hip in children presented after six months of age

Forty-five knee joints (24 children) were treated by modified Curtis and Fischer quadricepsplasty [6, 7]. Three modifications to the original technique were lateral incision instead of the midline, transverse division of rectus femoris at the musculotendinous junction, and preservation of both side retinacula and collateral ligaments. After gaining 90° knee flexion, the knee was immobilized in 90° flexion for four weeks. Children with bilateral dislocation of the knee were operated simultaneously. Thereafter, supervised physiotherapy was done for six weeks.

Once the knee dislocation was reduced, associated foot deformity was corrected with soft tissue release (14 clubfeet, 10 CVT) after three months. Twelve clubfeet and 9 CVT were with stiffness, whereas only two children with ligamentous laxity had foot deformity (2 clubfeet, 1 CVT). The feet were immobilized for six weeks. After immobilization, the correction was maintained with ankle-foot orthosis.

A reduction of the dislocated hip (32 hips) was done after three months following foot surgery. Surgical intervention for the hip dislocation included closed reduction (8 hips), open reduction (17 hips), and open reduction with femur osteotomy (7 hips) depending on the age. All surgeries were done through the anterolateral Somerville approach. The reduced hip was maintained in a spica for three months. Fixed abduction-splint was used for an additional three months in children with laxity. The stiff hip dislocation was mobilized after three months. No intervention was done in eight hips (4 children; one with stiffness, three with laxity).

Three stiff hips (3 children) and five hips (4 children) with ligamentous laxity required spica after closed reduction. Open reduction and spica application were done for six stiff hips (5 children) and 11 lax hips (7 children). A femur shortening with the open reduction was required in nine stiff hips (5 children) and only one child with ligamentous laxity.

All these children were treated and followed up clinically and radiologically at regular intervals. The mean age at quadricepsplasty was 7.95 months (6–16 months). The mean duration of follow-up was 87 months (24–181 months).

All children were evaluated for activity level, pain, limp, weakness, and use of a brace. The presence of deformity and the limb-length discrepancy was noted. The passive range of movement of the knee joint at the final follow-up was compared with the pre-operative range. The presence of extensor lag, degree of hyperextension, and flexion deformity of the knee was noted. The power of quadriceps and hip abductors was graded on the Medical Research Council (MRC) scale. The stability of the knee joint was checked in sagittal and coronal planes. All the hip joints were rated using the McKay criteria [8].

The pre-operative radiographs were studied for the severity of knee dislocation, Tonnis grades [9], and International Hip Dislocation Institute (IHDI) grades [10]. All the follow-up radiographs were evaluated for the development of femoral head, femoral neck, acetabulum, and hip joint congruity. The children older than six years were categorized into different Severin classes [11].

The mean passive flexion of knee joint improved to 101° at the final follow-up. The mean passive flexion of the knee joint noted in the laxity group and without the laxity group was 113° and 88°, respectively (p < 0.001). The children without laxity had more severe flexion deformity when compared with the children with laxity (p < 0.05). Three children showed grade I anterior instability in the laxity group. Two children demonstrated medial instability. No child with stiffness had knee joint instability. The median quadriceps power was 4.

McKay rating showed excellent results in nine hip joints. The rating was better in children with ligamentous laxity. The radiological parameter measurements were reproducible (intra-observer and inter-observer). Severin classification [11] was good in 67% of children. The radiological outcome was similar in both groups (Figs. 2 and 3). Three hip joints in children without joint laxity had subluxation. Coxa magna was seen in five hips (3 with laxity, two without laxity).

All except one child were community walkers in children with ligamentous laxity group (Fig. 4). One child was a household walker. The same child used braces for walking. Among the child without laxity, ten were community walkers. Seven children required braces for their ambulation (Table 2).

Secondary corrective surgery was required in four children. These surgeries were varus derotation osteotomy (n = 3) and acetabuloplasty (n = 1). Two children without laxity showed persistent subluxation. Avascular necrosis was noted in four hips (type I in 2, type II in 2; 4 children) (Fig. 5). No severe deformity was noted in any child. The frequency of secondary procedures and avascular necrosis were the same in both groups. One child developed a proximal femur fracture two years following hip surgery. The fracture was treated with osteosynthesis. The hip was subluxated at the last follow-up.

Johnston CE defined the indication of the soft tissue lengthening for the treatment of the congenital dislocation of the knee [1] Different approaches to correct the congenital dislocation of the knee include minimal invasive technique, percutaneous surgery, needle tenotomy, traditional open lengthening of the quadriceps muscle [6], or modification of the quadricepsplasty [7].

There is no consensus in the literature regarding addressing the hip dislocation in nonsyndromic, syndromic, and arthrogryposis. The child with bilateral dislocation was left alone in Larsen syndrome and arthrogryposis [3]. However, recent trends showed good results with the early intervention for the hip dislocation in the child with stiffness [13, 14].

The closed reduction was adequate in cases with ligamentous laxity. However, closed reduction is rarely successful in arthrogryposis children [13] and Tonnis grade IV [15, 16]. These hips invariably require open reduction or open reduction with femur osteotomy. Few authors described good results with a medial open reduction [17], while others described similar results with the anterolateral approach [18]. Restricted range of motion of the hip was noted with the treatment of the hip dislocation [13].

In this study, most hips were addressed with closed reduction, open reduction, and open reduction with femur osteotomy. The children with stiffness required open reduction or open reduction with femur osteotomy. Children with ligamentous laxity were treated with closed reduction. The spontaneous reduction of the hip was noted in five hip joints in three children with laxity.

The ambulatory capacity of the syndromic, nonsyndromic, or arthrogryposis children depends on the associated deformities. The children with knee flexion contracture and hip dislocation were associated with nonambulatory status in arthrogryposis [19]. The early intervention with aggressive physiotherapy showed good function in a child with arthrogryposis [14]. An orthosis was needed in a few cases with foot deformities. Most children in the present study had community ambulation. The functional status was better in children with laxity.

We have compared our results with similar studies in the literature (Table 3). To the best of our knowledge, spontaneous reduction of the dislocated hips after walking age has not been reported in the literature.

Recently, treatment of CDH by closed reduction and short leg cast showed excellent outcomes in Tonnis grade II and III [16]. Bilateral dislocation and Tonnis grade IV is associated with the failure of the reduction of the hip joint. Similarly, closed reduction and dynamic cast were also described to maintain reduction [25]. Szepesi et al. [26] mentioned good long-term results with open reduction and early mobilization without spica cast.

All three children with spontaneous reduction of the hip had gross ligamentous laxity. All spontaneously reduced hips were Tonnis grade II/III. None of Tonnis grade IV hips showed a spontaneous reduction. These hip and knee joints were not reducible under anaesthesia. The dislocation of the knee required quadricepsplasty for reduction. They did not take any treatment for hip dislocation. The spontaneous reduction of the hip might be attributed to the relaxed hamstring. Persistent quadriceps contracture with dislocation of the knee leads to severe hamstring stretching. Once the knee flexion is achieved, hamstrings are relaxed. Stretched hamstring might play an important role in the dislocation of the hip joint. It can contribute to secondary hip dislocation. Following knee surgery, all children were immobilized in 90° of knee flexion. With knee flexion, hip might be flexed or abducted. These might help to relax the adductors and flexors of the hip joint. Hence, it might be the reason to achieve a spontaneous reduction of the hip. However, we did not get a spontaneous reduction in all dislocated hips. Four children (8 hips) were not treated for hip joint dislocation. Three hips remain dislocated at the final follow-up. The surgery is being planned to reduce the hip joint dislocation.

This study is a retrospective study. All children have not attained skeletal maturity. We intend to follow up on these children to skeletal maturity. The study includes a diverse population ranging from arthrogryposis multiplex congenita, hyper-laxity syndrome, and nonsyndromic children. Hence, the functional scoring would be different in stiff and lax joints. The five hips were normal, associated with CDK. However, the unilateral CDH was included with bilateral dislocation of the knee joints. The number of children in each group is less, so it is difficult to draw a robust, meaningful conclusion of the approach and spontaneous reduction of the hip. A future multicentric prospective study of such a rare congenital disorder will answer the exact outcome in both groups.

The study includes the largest case series of combined knee dislocation and hip dislocation who presented after six months of the age. The spontaneous reduction of the hip could be achieved after the reduction of the knee joint even after nine months of age in children with hyper-laxity. The outcome of the knee and hip surgeries would be different in the child with stiffness and the child with laxity.