Abstract
Purpose
The aim of this retrospective long-term study was to review and present the effects of treatment for 11 children with arthrogryposis multiplex congenital, or amyoplasia, followed from birth until skeletal maturity.
Methods
We evaluated walking ability, age of beginning to walk, required ambulatory devices, age of independent walking and muscle strength.
Results
Our series showed babies with severe limb involvements without spine abnormalities. Despite the initial severity of involvement, nine patients finally became ambulators with flexion contracture of less than 20° on hips and 15° on knees, and six were independent walkers before the age of 2.5 years. The two non-ambulators presented severe scoliosis at skeletal maturity, which needed spinal fusion.
Conclusion
We conclude that long-term ambulatory status at skeletal maturity is not correlated with the severity of condition at birth. A prognosis for ambulation at skeletal maturity will be done before 2.5 years of age. We believe that early aggressive management of children with severe arthrogryposis is warranted and justified.
Introduction
Arthrogryposis multiplex congenita (AMC) is a non-progressive disorder, present at birth and characterised by multiple joint contractures and defective muscles, with normal sensations [1–4]. This condition was first described by Otto in 1841. In 1923, Stern first used the term AMC. In 1932, the specific name of amyoplasia was given by Sheldon [5]. The most common feature is the lack of foetal movements. This can result from a large number of disorders, including neuropathies, abnormalities of the muscles, connective tissue disorders, conditions that limit the internal space of the uterus or defects of the uterine environment [6]. Because there are 150 different disorders or syndromes that have joint contractures as part of their manifestations, AMC is a diagnosis of exclusion, and making a precise diagnosis is very important for the evaluation of results [7].
The term ‘amyoplasia’ refers to the most common arthrogrypotic syndrome, which includes multiple congenital contractures, typical and symmetrical positioning of the limbs, and the replacement of muscles by fibrous and fatty tissues. It is sporadic, with no known hereditary pattern [5, 8]. The goal of treatment is to obtain the maximum possible function for each child by an early multidisciplinary approach [8, 9].
To our knowledge, this is the first report in the literature defining the initial presentation of babies and reporting the long-term results of treatment. The aim of this retrospective long-term study was to review and present the effects of our treatment modalities on final walking capabilities, functional status and active range of motion for 11 children with amyoplasia, followed from birth until skeletal maturity.
Materials and methods
From 1975 to 1986, 36 children with AMC amyoplasia-type were treated at our institution. Eleven were followed from birth to skeletal maturity. There were three males and eight females. The first visit was before 2.5 months of age (mean 20 days). The average follow-up was 17.5 years (age 14–25 years). Nine children had quadrimelic contractures and two had bimelic contractures of the lower limbs (Table 1). The hip joints were involved in 20 out of 22 children, including nine dislocations (six on the right side, three on the left), one of which was a bilateral anterior dislocation. Eight dislocated hips had severe contracture in many different positions. The remaining one had normal motion and was associated with contralateral acetabular dysplasia. Five hips had contracture in flexion associated with abduction; none were dislocated. The knee joint was involved 20 times, including ten contractures in extension (five recurvatum). We did not notice any dislocated knee in our series. The feet were involved in all patients; 21 in clubfeet and one was a calcaneovalgus foot. Six shoulders, ten elbows, 14 wrists and ten hands were affected by AMC. No child had spine abnormality at birth (Figs. 1 and 2).
Type and severity of the most frequent contractures at birth
Severity was determined by a loss of more than 50% of the normal range of motion
Newborn with hips and knees flexion contractures and no active motion. He presented bilateral clubfeet, but no dislocated hips
Newborn with hips and wrists flexion contractures and knees and elbows extension contractions. He presented bilateral clubfeet and left dislocated hip
Multidisciplinary management included, initially, an aggressive, non-surgical treatment, and, secondly, surgical treatment. Intensive physical and occupational therapies, orthoses and night-time splints were started immediately at birth. Four children were hospitalised in the first few weeks of life for this non-surgical treatment; the mean duration was 2.5 weeks (1–4 weeks). Physiotherapists a our institution are accustomed to treating newborns because congenital clubfeet are treated functionally.
Seventy-seven surgical procedures were performed overall; the mean number per child was 7 (3–16). Bilateral surgery was considered as two procedures, even if they were done during the same anaesthesia.
Sixty-six surgical procedures were performed on the lower limbs (Table 2); the average number was 6 per child (3–12). Closed reductions were attempted for seven unilateral hip dislocations, but all of them failed. Therefore, surgical reductions were done on six hips (five on the right side, one on the left) using a Smith–Petersen approach. A dysplastic hip, which dislocated completely later, remained in this position. No surgery was done on a bilateral hip dislocation at birth.
Location and type of procedures for the lower limbs
Nine surgical procedures were performed on the upper limbs (Table 3); the mean number of operations was 0.8 per child (0–4).
Location and type of operations for the upper limbs
FCU flexor carpi ulnaris, ECR extensor carpi radialis
Spine scoliosis was diagnosed during growth in four patients who were treated using a brace, starting at a mean age of 8.6 years (5–12). However, two of them finally needed spinal fusions at ages 14 and 19 years, respectively.
The age when patients started walking, the type of walking aids used and the age when patients began walking independently were recorded for all patients. Physical limitations in dependant children were evaluated and the significance of physical limitations was also determined in the children who were able to walk independently.
The final functional outcome was assessed using Hoffer's classification of walking status [10]: independent ambulator walks without any aid, community ambulator is able to walk with aids in the community and does not need a wheelchair, household ambulator is able to walk with aids in the household and use a wheelchair in the community, non-functional ambulator uses a wheelchair and is capable of transfer, non-ambulator always uses a wheelchair and is not capable of transfer.
The active range of motion was evaluated during growth. Muscle strength was quoted from 0 to 5, where 0 means no muscle activity and 5 refers to near normal strength.
Results
The age at which the patients were able to walk for the first time and their needs for walking aids are summarised in Table 4.
Characteristics of walking for the first time
The three immediate independent walkers and the two walkers with limited or little support were independent walkers from the age of 2.5 years until skeletal maturity (Fig. 3). There was one exception, who became a community walker. The three who used bilateral long-leg orthoses became two independent walkers and one household walker. Concerning the three who used bilateral long-leg orthoses and a walker, two became non-walkers and one became a late independent walker at the age 5.5 years.
3.5-year-old child (same as in Fig. 2) without walking aids
At skeletal maturity, the walking status and the age of acquisition of this final status is described in Table 5.
Characteristics of walking at skeletal maturity
The seven independent ambulators (six girls and one boy) had a normal range of hip motion or a mild flexion contracture of less than 20° at follow-up (five hips). Three children presented avascular necrosis before the age of 2 years, following open reduction for unilateral dislocation. They had normal knee mobility or a mild flexion contracture of less than 15° (five knees). All of them had plantigrade feet, and with the exception of one patient, they all had active upper limbs movements. Two of them had one elbow in flexion and one elbow in extension with an acceptable level of function for daily activities. These patients did not have spinal involvement, except one with mild scoliosis treated successfully by a brace. In this group, we noticed that four patients had muscle activity apparition or improvement after treatment of contractures. The quadriceps were concerned in two cases, with improvements in strength from 0 to 4 in one case and from 3 to 4 in the other case. The biceps brachii was concerned in two cases, with improvement in strength from 0 to 4 for both cases.
The patient who was a community ambulator was able to walk with limited support at the age of 2.5 years and independently at age 9 years. However, at 13 years of age, he needed one crutch to ambulate in the community. He presented lower limb active mobility without stiffness, despite a bilateral hip dislocation (unilateral dislocation at birth and a contralateral dysplastic hip, the latter dislocated). He had plantigrade feet and no upper limb or spinal involvement.
The patient who was a household ambulator was able to walk with bilateral long-leg orthoses at the age of 5.5 years. She used a wheelchair in the community at the age of 7.5 years, but was able to walk with a small amount of support in the household at the age of 17 years. She still had a 35° hip flexion contracture, despite a bilateral derotation and extension osteotomies. She also had bilateral tensor fasciae latae tenotomy for severe abduction contracture. She had a 5° knee flexion contracture, despite bilateral posterior release. She presented active mobility in her hips and knees. Her feet were in a 20° equinus deformity, despite a posteromedial release. This girl had one elbow in flexion and one elbow in extension, with active flexed wrists, after four surgeries. She had a scoliosis associated with a hyperlordosis that was treated successfully by a brace. This patient had a total of 16 surgeries.
The patient who was a non-functional ambulator was able to walk with bilateral long-leg orthoses and was a walker at age 2.5 years; however, she lost this capability at age 13 years, due to spine imbalance. At the last visit, she obliged to use a wheelchair, but was capable of transfer. The girl did not have hip dislocations at birth but presented bilateral subluxation at the age of 10 years. Dega osteotomies were then performed. The hips were mobile but muscle activity was weak. Recurrent knee flexion contractures led to iterative posterior release. She had no active motion on her knees. Recurrent clubfeet led to repeated soft-tissue releases. She had no marked upper limb involvement, and she presented a severe double-curve scoliosis, which was treated with a brace since the age of 5 years. She finally had a spinal fusion at the age of 19 years. Her left dorsal curve was 95° and her right lumbar curve was 77° at that time.
The patient who was a non-ambulator was barely able to walk with both a walker and a bilateral long-leg orthoses at age 4 years, but he lost this capability at 9 years of age. Since then, he has used an electric wheelchair and is not capable of transferring alone. At the last visit, he presented with a pelvic obliquity, a bilateral hip flexion contracture due to a birth bilateral dislocation (30° on the right side and a 20° on the left). He had a bilateral knee flexion contracture despite posterior releases (30° on the right side and 40° on the left) and he also had severe equinovarus of the feet, despite posteromedial releases. His upper extremity joints were in poor functional position (shoulders in internal rotation, both elbows in extension), even after a right elbow posterior release and an ipsilateral humeral derotation osteotomy. He had no active muscles in his lower limbs or upper limbs since birth and no recovery was seen during growth. He developed a scoliosis, which was treated at first by a brace, but he required a spinal fusion at the age of 15 years. His left lumbar curve was 60° at that time.
Discussion
Treatment
Topography and characteristic postures at birth for our patients were similar to literature reports describing AMC amyoplasia-type [3, 8, 9]. Several authors inspired our multidisciplinary approach [11, 12].
Hips. We observed a similar prevalence of involvement (90% of hip contractures and 41% of hip dislocations at birth) compared to Friedlander et al. (82% of hip contractures and 43% of hip dislocations in 45 patients with a mean age of 11 years), but a higher prevalence than St. Clair and Zimbler (65% of hip contractures and 19% of hip dislocations in 26 patients with a mean age of 10 years) [1, 13]. Our series showed that the only contractured position, which is not related to dislocation, is the flexion–abduction one. All of the other positions should be considered to be at risk. We agreed with the St. Clair and Huurman approach that aggressive traction and physical therapy programmes, associated with soft-tissue releases, tenotomies and osteotomies, are necessary for all stiff hips. We observed that community walker patients presented less than 20° of hip flexion contracture. Hoffer et al. noted less than 30° in his independent walker patients [10]. Since Lloyd-Roberts in 1970, several publications referred to the relatively unsuccessful result of closed reduction of dislocated hips in AMC [3, 14–16]. Huurman and Jacobsen emphasised that ‘teratologic’ dislocation in AMC is much more rigid than the simple ‘idiopathic’ congenital dislocation, because the intrauterine hip modelling processes have been disturbed at an earlier time in gestation [17]. In the literature, the consensus is to perform open reduction for unilateral dislocation to avoid pelvic obliquity, sitting imbalance and consecutive scoliosis [2, 11–13, 17, 18]. We decided to perform open reduction for all six unilateral dislocations, and all patients became independent ambulators without spine involvement. On the other hand, the management of bilateral dislocation remains controversial. The rate of complication led most authors to recommend that bilateral dislocation left unreduced [2, 10, 11, 18]. Nevertheless, several authors published good results for most of their bilateral open reductions with a follow-up before skeletal maturity (between 65 months and 11.8 years) [12–14, 19]. In 2002, Yau et al. showed in their study a correlation between the final radiographic status and the St. Clair functional hip score. In other words, if hips remained dislocated, the function was poor [13, 16]. In our series, we chose not to reduce bilateral dislocations. One patient became a community ambulator and presented a good active range of motion, but the other is a non-ambulator and presented stiff hips and severe scoliosis. This last observation leans toward the surgical option for bilateral dislocation.
Knees. Compared to the published series, we observed a similar prevalence of involvement (90% of knee contractures at birth) than Thomas et al. (85% of knee contractures in 104 children with mean a age of 3 years) but higher than Södergård and Ryöppy (58% of knee contractures in 50 patients with a mean age of 18 years) [20, 21]. According to most authors, flexion contractures are more frequent and more difficult to treat with conservative methods than extension contractures [2, 21, 22]. In the literature, the consensus is to perform early serial casting and physiotherapy, which is sufficient to sometimes correct contractures in flexion or extension (sufficient in 13 cases of 20 involved knees in our study). In the case of conservative management failure to treat flexion contracture, the recommended surgery is posterior capsulotomy in addition to soft-tissue release (including hamstring tenotomy). We followed Lloyd-Roberts’ advice by avoiding supracondylar femoral osteotomy before the end of growth, because of the high recurrence rate [3, 21, 22]. DelBello and Watts recommended performing a distal femoral extension osteotomy for flexion contracture as a safe procedure even for large contractures. In opposition to Thomas et al., they used femoral osteotomies in young patients because walking capability improvements are large and develop quickly (32 non-ambulators became community or household ambulators in his series). In case of recurrence, DelBello and Watts advocated iterative procedures [23]. Brunner et al. used the Ilizarov technique as an efficient tool for knee and foot deformity correction; however, he noted a high recurrence rate in children younger than 10 years of age [24]. We had no experience with this procedure in our patients and preferred quadricepsplasty, as did Södergård and Ryöppy [20]. Because of the high recurrence rate, splints and orthoses always completed surgery. Thomas et al. and Hoffer et al. noted the importance of having a flexion contracture below 20° in order to walk independently [10, 21]. Our experience supports this conclusion. All of our ambulatory patients had a contracture of 15° or less. Murray and Fixsen presented functional results of 22 patients with amyoplasia followed up over an average of 7 years [22]. Their patients were treated by initial physiotherapy and splintage, as well as repeated posterior release and bony procedures, if necessary. He reported 14 community walkers in 22 patients, without a lot of detail on other involvements beside the knees. We presented the same seven community walkers of 11 patients with a similar management.
Feet. We observed that this was the most commonly affected area (100% of feet deformities at birth in our series) and rigid equinovarus was the most frequent foot deformity (95% of feet deformities at birth in our series). Rigid equinovarus was, in fact, much more frequent than calcaneovalgus or vertical talus, confirming the classical descriptions on the subject [1–3]. The realistic goal in treating this type of feet should be to obtain pain-free, plantigrade and rigid feet [3, 25]. For Menelaus, two important limitations of soft-tissue release operations for severe equinovarus are, firstly, the inadequate initial correction due to the insufficiency of skin and soft-tissue coverage on the medial side, and secondly, the relapse after cast removal, due to the wide gap between tarsal bones [26]. Most authors reported that conservative treatments, soft-tissue procedures, supramalleolar and tarsal osteotomies were inadequate and that more radical approaches such as talectomies are required either as a primary procedure for most rigid equinovarus in young children or after the failure of less radical treatment [3, 25–29]. A few authors also recommend triple arthrodesis for older children, after 10–12 years of age [28]. In our experience, complete soft-tissue releases or primary talectomies gave lasting results only in 6 out of 19 feet. Recurrences led to repeated soft-tissue releases or triple arthrodesis in 11 cases. The two patients with residual deformities were one household ambulator and one non-ambulator. All of our independent or community walkers had plantigrade feet.
Upper limbs. Upper limbs were involved in 9 out of 11 patients (82%). The severity of stiffness increased from proximal to distal, as described in the literature [30, 31]. All authors agreed that the most important aim of the management is the ability to carry out daily activities, such as self-feeding and self-toileting. The second goal is to aid ambulation if crutches or a wheelchair are needed. With the exception of the non-ambulator boy, all of the patients in our series achieved those goals.
Spine. In our series, the prevalence for scoliosis in AMC during childhood and adolescence (36%) was lower than Yingsakmongkol and Kumar (66% of 46 patients) but higher than Drummond and Mackenzie (28% of 50 patients) and Herron et al. (22.5% of 80 patients) [32–34]. We had the same indications as Yingsakmongkol and Kumar for the four children with scoliosis in our series. He recommended treatment by bracing if the curve is less than 30° at the beginning of treatment and if the patient is an ambulator. If the curve is greater than 30° and the patient is a non-ambulator, the authors consider a brace as insufficient and could be used only to delay spinal fusion. They reported few ambulators in their series of 13 scoliosis treated by spine fusion compared to our study cohort (one independent, one community, four household ambulators and seven non-ambulators) [32]. Herron et al. also noted the high incidence of pelvic obliquity and hyperlordosis associated to hip flexion contracture or unilateral hip dislocation [34]. Our experience confirmed this observation.
Prognosis
It remains extremely difficult in comparing the results of treatment modes presented in various publications because of the lack of description in severity involvement before treatment. There is confusion in the use of the AMC term because it may be related to amyoplasia, to distal arthrogryposis or to other types of arthrogryposis syndrome. Furthermore, most papers are focused on one joint; sometimes, walking status is reported, but using this data is at risk of extrapolation because other joint involvements will often influence the outcome. To our knowledge, there is no review of the literature showing long-term functional and ambulatory studies in children followed up from birth until skeletal maturity. Another difficulty remaining in identifying prognosis factors is, as with the case of any rare diseases, the small number of patients available to generate data.
A prognostic factor seems to be the progressive muscle activity improvement after the recovery of a passive range of motion in a useful arc (sometimes late), because 4 out of 7 independent ambulators showed an increase in strength during growth. The two non-walkers in our series had severe scoliosis treated by spine fusion. The non-ambulator had recurrent lower and upper limb contractures without active motion; however, the non-functional ambulator was not limited by contractures in the lower and upper limbs, had weak active hip movements, active upper limbs but no active knee movement. Therefore, negative prognostic factors could be in the spine involvement severity, as well as in lower or upper limb contractures or weakness.
In our series, positive prognostic factors for independent or community walkers were active hips and knees, hip flexion contracture less than 20° and knee flexion contracture less than 15° without severe scoliosis.
Conclusion
Our patients seemed to be more severely involved than in other case studies; however, no authors have described patients as near from birth as this report. Despite this, 7 out of 11 children became independent ambulators in our study, which is more than two other reported series [10, 11]. We conclude that long-term ambulatory status at skeletal maturity is not correlated with the severity of condition at birth. A good idea of ambulation at skeletal maturity will be obtained before the age of 2.5 years old (6 out of 7 independent ambulators were independent before this age). The positive prognostic factors for walking abilities were active hips and knees, hip flexion contracture less than 20° and knee flexion contracture less than 15° without severe scoliosis. Therefore, we believe that early aggressive management of children with severe arthrogryposis is warranted and justified. This includes physiotherapy, occupational therapy and multiple surgeries.
Footnotes
Acknowledgments
We wish to thank François Fassier MD, Reggie Hamdy MD, Mehdi Aarabi MD and Tasima Haque from the Shriners Hospital for Children, McGill University, Canada, for their help with the preparation of this article.