Comparison of Botulinum Toxin Type A Injection and Soft-Tissue Surgery to Treat Hip Subluxation in Children With Cerebral Palsy

Methods 

Medical records of children with CP who were admitted to our hospital between February 2004 and March 2007 were retrospectively reviewed. Among those cases, children who met the following inclusion criteria were selected as subjects: (1) children with bilateral spastic CP, (2) children whose first hip radiographs were taken under 6 years of age, and (3) children in whom radiographs of the hips were taken at least 3 times in intervals of more than 6 months. Children who had both soft-tissue surgery and BTX-A injection during the follow-up period were excluded from the study.

As a result, 194 subjects with spastic CP were enrolled. The type of CP was quadriplegia in 78 patients and diplegia in 116 patients. The age at which the initial radiographs were taken ranged from 18 to 70 months (mean age, 39.3±12.9mo). The subjects were divided into 2 groups according to the functional level based on GMFCS. GMFCS levels 1 and 2 were classified as a high-functioning group, and GMFCS levels 3, 4, and 5 were classified as a low-functioning group.

Based on the therapeutic intervention they received, subjects were divided into 3 groups: the nonintervention group (138 hips of 69 children), soft-tissue surgery group (soft-tissue surgery of hip adductor muscles, 130 hips of 65 children), and BTX-A group (BTX-A injection into hip adductor muscle, 120 hips of 60 children). In the BTX-A group, 7 patients received an additional BTX-A injection into both hip adductor muscles during the follow-up period.

In the BTX-A group, the average dose per kilogram of body weight was 3U/kg of muscle, and it was standardized by body weight over this time period. All injections were performed by using a 1-mL syringe with a 27-G needle. We dissolved the contents of 1 vial of BTX-A in 2mL isotonic saline. We injected the toxin into the target muscle under the guidance of ultrasonography.

Radiographs were obtained in a standardized position as described in a previous article.24 MP is measured by calculating the percentage of the femoral head that lies outside of the lateral border of the acetabulum as defined by bony landmarks on an anteroposterior pelvis radiograph.10

The severity of hip displacement was classified into 3 groups based on initial MP before therapeutic intervention: the mild subluxated group (20%≤MP<40%), moderate subluxated group (40%≤MP<60%), and severe subluxated group (60%≤MP<90%). The annual change in MP was calculated based on the changes of MP and duration between the first and last radiographs.

Statistical differences between the groups were assessed by a t test and chi-square test when 2 groups were compared, and analysis of variance was used when more than 2 groups were compared. The significance level was set at .05

Results 

The functional level of subjects based on GMFCS25 is described in table 1. The mean age when the initial pelvic x-ray was taken did not differ significantly between the nonintervention, soft-tissue surgery, and BTX-A groups. The age at which the final follow-up pelvis x-ray was taken, and the period of follow-up did not differ significantly between groups (see table 1). The distribution of the high-functioning and low-functioning groups based on GMFCS levels and the distribution of mild, moderate, and severe subluxated group based on MP on initial pelvic x-rays did not differ significantly between the 3 groups (see table 1).

Table 1. General Characteristics of Subjects

	Variables	Total (N=194)	Nonintervention Group (n=69)	BTX-A Group (n=65)	Surgery Group (n=60)
	Sex (male/female)	123/71	38/31	34/31	51/9
	GMFCS
	Level 1	15	10	3	2
	Level 2	43	14	17	12
	Level 3	53	14	18	21
	Level 4	46	17	11	18
	Level 5	37	14	16	7
	Initial MP
	Mild MP (20% ≤ MP <40%)	120	46	38	36
	Moderate MP (40% ≤ MP <60%)	70	22	26	22
	Severe MP (60% ≤ MP <90%)	4	1	1	2
	Initial age (mo)	39.3±12.9 (18–70)	37.7±14.6 (18–70)	39.9±13.0 (19–68)	40.4±10.6 (26–68)
	Final age (mo)	62.0±17.7 (37–174)	61.6±21.4 (39–174)	59.9±13.7 (37–86)	64.8±16.5 (45–100)
	Period of follow up⁎ (mo)	22.9±11.8 (18–108)	23.5±15.2 (19–108)	22.6±7.9 (18–39)	22.5±10.9 (19–52)

NOTES. Values are mean ± SD (range) unless otherwise noted. Comparisons were made by chi-square test and 1-way analysis of variance with Bonferroni post hoc test.

⁎ Period of follow-up is the number of months between measurements of the initial MP and final MP.

The annual change in MP was 4.4% in the nonintervention group, −0.7% in the BTX-A group, and −1.6% in the surgery group. The positive value of annual change of MP in the nonintervention group meant progressive hip displacement, whereas the negative value of those in the therapeutic intervention group meant improvement of hip displacement. Significant differences were shown between the BTX-A group and nonintervention group and between the surgery group and nonintervention group (P<.05) (table 2); however, there were no significant differences in values between the BTX-A group and surgery group (see table 2).

Table 2. Comparison of MP Between 3 Groups According to Therapeutic Intervention

	MP	Nonintervention Group (n=69)	BTX-A Group (n=65)	Surgery Group (n=60)
	Initial (%)	33.2±10.0	35.9±13.9	35.37±10.57
	Final (%)	37.9±13.5⁎	33.6±13.5†	31.8±11.9‡
	Change (%)	4.7±10.3	−1.6±8.4†	−3.3±6.1‡
	Change per year (%/y)	4.4±11.3	−0.7±6.5†	−1.6±4.4‡

NOTE. Values are mean ± SD. Comparisons were made by paired t test and 1-way analysis of variance with Bonferroni post hoc test. Change per year was calculated as (final MP − initial MP/period of follow up [%/y]). A positive number means worsening and a negative number indicates improvement.

⁎ P<.05, initial versus final MP by paired t test. † P<.05, nonintervention group versus BTX-A group. ‡ P<.05, nonintervention group versus surgery group.

In the therapeutic intervention group, patients were grouped by age of less than 3 years and aged 3 years or older according to the age at which the first therapeutic intervention began. Improvements of annual change in MP of the younger intervention group were significantly greater than the older intervention group (P<.05) (table 3). With regard to functional level, the annual change in MP was significantly different between the high- and low-functioning groups (P<.05). Improvements of annual change in MP were significantly greater in the high-functioning group (GMFCS levels 1 and 2) than the low-functioning group (GMFCS levels 3, 4, and 5) (table 4). The severity of hip subluxation before the intervention also had an effect on the annual change in MP. They were significantly different between the mild, moderate, and severe subluxated groups (P<. 05) (table 5). The best responses were shown in moderate subluxated hips, whereas further subluxation was noted in severely subluxated hips despite therapeutic intervention.

Table 3. Change in MP Related to Patients' Age at First Intervention by Group

		No. of Legs	MP (%)		Change Per Year (%)
			Initial	Follow-Up
	BTX-A group
	Younger⁎	30	40.2±16.9	37.3±19.2	−1.5±4.7
	Older†	100	34.6±12.7‡	32.5±11.1	−0.5±6.9‡
	Surgery group
	Younger⁎	16	43.7±7.1	33.1±8.4	−4.1±4.6
	Older†	104	34.5±11.5‡	31.9±13.7	−1.2±4.3‡
	Intervention group
	Younger⁎	46	41.5±14.3	35.8±16.3	−2.4±4.8
	Older†	204	34.5±12.1‡	32.2±12.5	−0.8±5.7‡

NOTE. Values are mean ± SD unless otherwise noted. Comparisons were made by independent t test.

⁎ Younger group: patient age at first intervention was <3 years. † Older group: patient age at the first intervention ≥3 years. ‡ P<.05, younger age group versus older age group.

Table 4. Change in MP Related to Functional Level in Intervention Group

		No. of Legs	MP (%)		Change Per Year (%/y)
			Initial	Follow Up
	BTX-A group
	High functioning⁎	40	35.2±13.8	31.4±11.6	−2.4±5.2
	Low functioning†	90	36.2±14.0	34.5±14.2	−0.0±6.9‡
	Surgery group
	High functioning⁎	28	30.8±11.2	24.3±8.5	−3.4±4.8
	Low functioning†	72	37.2±11.1‡	34.5±13.3‡	−1.0±4.1‡
	Intervention group
	High functioning⁎	68	33.4±12.9	28.5±10.9	−2.8±5.0
	Low functioning†	182	36.7±12.6	34.5±13.7‡	−0.5±5.6‡

NOTES. Values are mean ± SD unless otherwise noted. Comparisons were made by independent t test.

⁎ High functioning: GMFCS 1 and 2. † Low functioning: GMFCS 3, 4, and 5. ‡ P<.05, high-functioning group versus low-functioning group.

Table 5. Change in MP Related to Initial MP in Intervention Group

		No. of Legs	MP (%)		Change Per Year (%/y)
	Botox group
	Mild MP	76	26.8±9.8	26.8±11.2	1.2±6.8
	Moderate MP	52	47.8±5.7⁎	41.9±8.0⁎	−3.7±4.6⁎
	Severe MP	2	71.1±0.0‡	76.7±0.0 ‡	3.9±0.0†
	Surgery group
	Mild MP	72	28.1±6.0	24.9±7.5	−1.4±3.3
	Moderate MP	44	44.8±4.5⁎	39.2±5.7⁎	−2.8±4.7
	Severe MP	4	68.5±0.0‡	79.7±0.0‡	9.6±0.0‡
	Intervention group
	Mild MP	147	27.4±8.2	25.9±9.6	−0.1±5.6
	Moderate MP	97	46.4±5.4⁎	40.6±7.1⁎	−3.3±4.6⁎
	Severe MP	6	69.3±1.4‡	78.7±1.5‡	7.7±2.9‡

NOTE. Values are mean ± SD unless otherwise noted. Comparisons were made by 1-way analysis of variance with Bonferroni post hoc test. Abbreviations: Mild MP, initial MP <40%; moderate MP, 40% ≤ initial MP <60%; severe MP, initial 60% ≤ MP <90%.

⁎ P<.05, mild MP group versus moderate MP group. † P<.05, moderate MP group versus severe MP group. ‡ P<.05, mild MP group versus severe MP group.

Discussion 

Hip displacement is common in children with CP and may progress from silent subluxation to hip dislocation if left untreated.4, 26, 27 It may cause many serious problems such as pain; contractures; problems with sitting, standing, or walking; fractures; skin ulcerations; and difficulty with perineal care, pelvic obliquity, and scoliosis.27, 28, 29, 30, 31 Early intervention is assumed to reduce the incidence of hip dislocation.1, 3, 5, 24, 32, 33 Soft-tissue surgery to spastic hip muscles has been widely used to prevent the hips of children with CP from dislocation.18, 22, 34 Favorable outcomes of soft-tissue surgery against progressive hip displacement were commonly noted in many previously reported studies.15, 16, 17 The use of BTX-A injection in terms of a prophylactic procedure against further hip subluxation or dislocation is a relatively new approach; thus, a few reports20, 21 have been published about the effects of BTX-A injection into hip muscles on MP changes. In the present study, hip displacement improved in both the surgery and BTX-A injection groups, whereas it progressed in the nonintervention group. These findings are in line with previous reports15, 16, 17, 18, 19 advocating the early intervention of hip subluxation or dislocation at risk. Interestingly, BTX-A injection into hip muscles showed a comparable effect on hip displacement to soft-tissue surgery in our short-term follow-up study. As far as we know, this is the first report on the effects of BTX-A injection into the hip adductor muscles compared with the effect of soft-tissue surgery for preventing further hip subluxation or dislocation.

There have been several reports about the combined effects of BTX-A injection to the adductors and a variable hip orthosis (such as sitting, walking, and standing) on gross motor function, hip displacement, and progression to surgery in a randomized clinical trial.4, 35 The role of hip bracing against progressive hip displacement remains controversial.3, 35, 36 In this study, the subjects did not wear the hip orthosis in both the intervention and control group. Therefore, the improvement of hip displacement shown in the intervention groups seems to reflect solely the effect of intervention. Further study is needed to determine whether there are the additional effects of hip orthosis against hip displacement if combined with surgical intervention or BTX-A injection.

There is general agreement in the literature that a lower preoperative MP leads to a favorable final outcome after soft-tissue surgery,10, 15, 16, 17, 18, 19, 23 although the suggested numeric value of MP for good response varies from 60% to less than 30%.15, 23, 37 As for BTX-A injection, there has only been 1 report showing the relationship between preinjection MP and radiologic outcome within 7 months after injection.20 In their study, patients with an initial MP of 30% or more were more likely to show a decrease in MP after BTX-A injection compared with less than 30% of initial MP. In the present study, we also found that the annual change in MP was greater in the moderate subluxated group compared with the mild subluxated group. It is thought that hip displacement in the mild subluxated group was so mild before the intervention that it could not lead to much change. Further hip displacement was observed in the severe subluxated group despite therapeutic intervention. The poor outcomes in this group are in accordance with the findings of a previous study23 in which all hips with more than 60% of preoperative MP showed unsatisfactory results to soft-tissue surgery. They suggested that soft-tissue surgery was ineffective for all hips that are more than 60% subluxated. In the present study, the small number of cases in the severe subluxated group and unequal distributions of cases with mild, moderate, and severe subluxated groups are limitations in finding the relationship between the initial MP and outcome. However, progressive hip displacement in the severe subluxated group suggests that the outcome to therapeutic intervention is closely related to initial MP.

There is general agreement in the literature that functional level and walking ability have a strong impact on progressive hip displacement in children with CP5, 17, 38; however, there is no consensus on whether the severity of involvement or functional ability are related with the outcomes after soft-tissue surgery. Some researchers did not reveal any relationship between the severity of disability and radiologic outcomes of soft-tissue surgery,17, 39 whereas significant relationships between walking ability and favorable outcomes were found in other reports.10, 16, 22 This inconsistency is caused by differences among the series regarding the degree of neurologic involvement, age at surgery, and duration of follow-up. As for BTX-A injection, the association between the radiologic outcome of BTX-A injection and severity of involvement has not yet been published as far as we know. In the present study, the better response in the high-functioning group compared with the low-functioning group suggests that the functional level can be considered as a factor relating to the outcome.

We found that the age at intervention was also related to outcome. The younger intervention group showed better results than the older intervention group. For soft-tissue surgery, some previous reports10, 15 showed successful results in children who had surgery before 4 years of age. On the other hand, age at the time of surgery was not predictive of outcome in other cases.16, 18 Differences among the series regarding the definition of a good outcome, age at surgery, and follow-up period may have led to these inconsistent results. Regarding BTX-A injection, the influence of age on the improvement of hip MP after toxin injection was published in a previous report.20 In their report, children who were less than 24 months old were more likely to have a better outcome compared with patients who received the injection after 24 months. The results of the present study also support early intervention for better outcomes to prevent further hip subluxation. Further hip subluxation in the nonintervention group suggests that spontaneous improvement can hardly be expected once the hip is subluxated. These findings indicate that early therapeutic intervention should be performed to obtain the best results in children with bilateral spastic CP as soon as hip subluxation is noted.

Conclusions 

Hip displacement progressed in the nonintervention group, whereas it improved in both the soft-tissue surgery and BTX-A injection groups. The effects of both soft-tissue surgery and BTX-A injection on hip displacement were not significantly different in the present short-term follow-up study. These findings suggest that BTX-A injection may replace soft-tissue surgery as a prophylactic procedure against further hip subluxation or dislocation if timely reinjected. The improvement in hip displacement after therapeutic intervention was greater in young children and high-functioning groups compared with older children and low-functioning groups. Hip displacement was progressive in severe hip subluxated group despite therapeutic intervention. The better responses in the younger age group and progressive hip displacement in the severe hip subluxated group and nonintervention group suggest that early intervention should be performed to prevent further hip displacement or dislocation as soon as hip subluxation is noted.