Does the Addition of Intrathecal Baclofen Along with or After Soft-Tissue Hip Adductor Surgery Decrease the Need for Hip Reconstruction Compared with Soft-Tissue Surgery Alone for Children with Nonambulatory Cerebral Palsy?

Abstract

Purpose

Intrathecal baclofen (ITB) is a well-known treatment option for cerebral palsy (CP) spasticity. The combination of soft-tissue release and ITB for spasticity is common. This study compared patients who had soft-tissue release before ITB (PRE-ITB), soft-tissue release at the same time as ITB (ST-ITB), and no ITB (NON-ITB) but had soft-tissue release at a similar age as PRE-ITB.

Methods

Inclusion criteria were a spastic or mixed nonambulatory CP diagnosis, prior hip adductor surgery, no prior reconstructive surgery, and at least a five-year post-operative follow-up. Thirty hips were identified as PRE-ITB, 20 hips as ST-ITB, and 43 hips as NON-ITB. The primary outcome variables were the subsequent hip surgery during the study period and/or a migration percentage ≥ 50% at the final follow-up defined as “unsuccessful hip.”

Results

The mean follow-up duration was 9.0 years (SD 2.4) for PRE-ITB, 9.4 (SD 3.6) for ST-ITB, and 9.3 (SD 3) for NON-ITB. The odds of unsuccessful outcomes were not different between NON-ITB and PRE-ITB but were lower for the ST-ITB group. The need for subsequent osteotomies or revision adductor surgery was significantly higher in ST-ITB compared with PRE-ITB (p = 0.02) or NON-ITB (p = 0.015). The incidence of surgical site infection over the whole follow-up period was higher in ST-ITB (40%) compared with PRE-ITB (13.3%, p = 0.035) and NON-ITB (0, p < 0.001).

Conclusion

The addition of tone management with ITB did not reduce the need for later hip surgery but did increase the risk for surgical site infection.

Introduction

Hip displacement (subluxation or dislocation) is common for children with nonambulatory cerebral palsy (CP), with a reported incidence ranging from 64 to 90% [1–3]. The etiology of hip displacement has been ascribed to spastic contractures of the hip adductor and flexor muscles [4], but a lack of functional weight bearing and persistent proximal femoral lateral physeal tilt in CP has also been emphasized [5, 6]. This claim is supported by reports finding no difference in the rate of hip displacement between CP patients with “tone management” surgeries when compared with control groups [7, 8]. That said, soft-tissue hip adductor surgery has been suggested to have at least a temporizing effect [6], while others report more favorable outcomes [9, 10].

Although early reports favored soft-tissue releases alone as being a successful strategy to treat hip displacement [11–13], studies with longer post-operative follow-up revealed high recurrence rates and need for additional bony procedures [6, 14, 15]. Shore et al. [6] reported success rates of 27 and 14% for adductor surgery alone in Gross Motor Function Classification System (GMFCS) levels IV and V, respectively. Terjesen [15]; however, reported higher success rates of 50 and 66% for GMFCS levels IV and V, respectively. In addition, given the high risk of migration percentage (MP) progression in patients with quadriplegic CP younger than five years (13% per year) [16], early intervention with soft-tissue surgery alone as a means to delay the need for hip osteotomies may still be a plausible approach.

Despite the causative role traditionally attributed to spastic muscle contractures in the development of hip displacement in CP, no studies to date have investigated the role of systemic tone management as an adjunct to soft-tissue surgery. Intrathecal baclofen (ITB) therapy is one treatment commonly utilized for the surgical management of symptomatic spasticity and its effectiveness has been supported in the literature [17–20]. Baclofen is an agonist for GABA-B and inhibits the presynaptic neurotransmitter release and decreases the neuronal activity at the synapse to reduce muscle spasticity [21]. The goal of this study was to determine the role of adjunct ITB therapy in preventing the progression of hip displacement following adductor surgery without osteotomy. We hypothesized that the combination of early contracture release followed by spasticity management with ITB will reduce the recurrence of hip displacement and the need for subsequent hip osteotomies.

Materials and methods

The study design was a retrospective comparative study, performed at an academic tertiary-level children’s hospital. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008. Informed consent was not required due to the retrospective nature of this study and lack of any identifying information as determined by our institutional review board.

Inclusion criteria were patients with spastic or mixed spastic/dystonic CP diagnosis, GMFCS IV/V, prior hip displacement who underwent soft-tissue hip adductor surgery, no prior femoral or pelvic osteotomy, availability of postoperative pelvic radiographs, and at least 5 years of post-operative follow-up. Exclusion criteria were lack of available hip radiographs or having prior selective dorsal rhizotomy (Fig. 1). Hip displacement was defined as an MP [22] greater than 30% in one or both hips [1].

Fig. 1.

Flowchart for ITB group identification. The group was then divided into PRE-ITB and ST-ITB. ITB intrathecal baclofen, PRE-ITB soft-tissue release before ITB, ST-ITB soft-tissue release at the same time as ITB

The primary outcome variables were the subsequent hip surgery during the study period and/or an MP ≥ 50% at the final follow-up (defined for the purposes of this study as an “unsuccessful hip”) [6]. Secondary outcome variables included age at hip adductor surgery, age at hip osteotomies, MP, ITB therapy prevalence and age at implantation, topographic pattern (diplegia, quadriplegia), GMFCS level, patient comorbidity score (defined by a number of total points, one point for each comorbidity present, including gastrostomy, tracheostomy, epilepsy, and clinically relevant scoliosis [curve > 20°]), and demographic data.

All patients had undergone prior adductor surgery for the treatment of hip displacement including open tenotomies of the adductor longus, gracilis, and iliopsoas and ± adductor brevis myotomy [11]. The indications for hip soft-tissue surgery were an MP > 30% on hip surveillance anteroposterior pelvis radiographs and hip abduction less than 30° with the hip and knee extended. If hip abduction was < 45° after the release of the adductor longus and gracilis, a partial myotomy of the adductor brevis was also added. If the popliteal angle was ≥ 45° after completion of the adductor tenotomies, proximal hamstring recessions were also performed. After the surgery, knee immobilizers were utilized for 8–12 h a day for one month. Physical therapy, including stretching of the hip adductors and flexors, was performed three times a week for 6 weeks postoperatively.

Patients who underwent implantation of a programmable ITB pump (SynchroMed, Medtronic Inc., Minneapolis, MN USA) were assessed in our institution’s multidisciplinary spasticity clinic by a rehabilitation physician, neurosurgeon, and orthopaedic surgeon. Patients were determined to be candidates for ITB by consensus, with relative indications being symptomatic spasticity ± dystonia impacting comfort, caregiving, and/or seating tolerance. Preimplantation intrathecal trial dosing was not used. All patients with ITB were on a continuous protocol. The surgical details of implantation are reviewed elsewhere [23].

Patients with ITB were matched with a group of patients who did not have ITB pump placement using five prognostic criteria: worst hip MP (within ± 10%), pre-ITB radiographic age ± one year, CP topographic pattern, GMFCS level, and comorbidity score (within ± one point) (Figs. 1 and 2). Patients were then divided into three treatment groups: adductor surgery prior to subsequent ITB pump placement (PRE-ITB), adductor surgery concurrent with ITB pump placement (ST-ITB), and adductor surgery alone (NON-ITB, control group). The control group of NON-ITB patients was selected so that they were comparable with the PRE-ITB group for patient age at adductor surgery (within ± one year) (Fig. 2).

Fig. 2.

Flowchart for control group identification, which included 22 patients with 43 hips (NON-ITB) and had similar age for soft-tissue release as the PRE-ITB group. ITB intrathecal baclofen, NON-ITB soft-tissue release without ITB, PRE-ITB soft-tissue release before ITB

Kolmogorov–Smirnov tests and Shapiro–Wilk tests were used to assess the normality of numeric variables of age and MP. For the normally distributed numeric variables, student’s t-test assuming equal variances was used. Descriptive statistics (mean, median, standard deviation [SD], min–max) were calculated. For the non-normally distributed numeric variables, a comparison between groups was made with the Mann–Whitney U test. Chi-square analysis was used for the comparison of categorical values and for odds of “unsuccessful hip” outcome. A p value below 0.05 was considered statistically significant. SPSS (v. 27, IBM, Armonk, NY) was used for all statistical analyses.

Results

For both ITB treatment groups, 26 patients were identified (15 patients/30 hips, PRE-ITB; 11 patients/20 hips, ST-ITB). For the NON-ITB group, 22 patients were identified (43 hips).

The duration of follow-up for the entire cohort was 9.19 (SD 2.94) years. The mean follow-up duration was 9.0 (SD 2.4) years for the PRE-ITB group, 9.4 (SD 3.6) years for the ST-ITB group, and 9.3 (SD 3) years for the NON-ITB group.

There were no significant differences regarding GMFCS level, topographic pattern, hip radiograph age prior to ITB (p = 0.15), hip MP prior to ITB, and comorbidity score between PRE-ITB vs ST-ITB (p = 0.52) (Table 1). The NON-ITB group comorbidity score was not different than ST-ITB (p = 0.27); however, it was lower than PRE-ITB (p = 0.017) (Table 1). The NON-ITB group MP prior to ITB (matching age radiograph) was not different than PRE-ITB (p = 0.15) or ST-ITB (p = 0.98). There were no differences between groups regarding their index adductor surgery type (adductor longus/gracilis only vs ± iliopsoas tenotomy vs ± adductor brevis myotomy).

Table 1.

Patient demographics, hip migration percentages, and adductor surgery age of three groups

Patient demographics	PRE-ITB	ST-ITB	NON-ITB
GMFCS level (%)
IV	2 (13)	2 (18)	5 (23)
V	13 (87)	9 (82)	17 (77)
Topography (%)
Diplegia	0 (0)	0 (0)	3 (14)
Quadriplegia	15 (100)	11 (100)	19 (86)
Comorbidity score, median (min–max)	3 (1–4)	3 (0–4)	2 (0–4)
Hip-based information
Migration percentage prior to ITB, median (min–max)	26.5 (2–49)	31.6 (0–100)	30 (4–95)
Adductor surgery age in years, median (min–max)	3.7 (2.2–6.8)	6.1 (4.2–9.4)	3.7 (1.9–6)

ITB intrathecal baclofen, NON-ITB soft-tissue release without ITB, PRE-ITB soft-tissue release before ITB, ST-ITB soft-tissue release at the same time as ITB

The prevalence of hip osteotomies for the NON-ITB group (24/43, 56%) was not significantly different than for the ST-ITB group (12/20, 60%) (p = 0.79) or PRE-ITB group (11/30, 37%) (p = 0.15) at final follow-up. The need for subsequent osteotomies or revision adductor surgery was significantly higher in the ST-ITB group compared with the PRE-ITB (p = 0.02) or NON-ITB (p = 0.015) groups (Fig. 3). For future bony surgery type, all of the patients except one in the PRE-ITB group (varus derotation osteotomy only) had varus derotation osteotomy and Dega acetabuloplasty.

Fig. 3.

Future hip surgery incidence. Future hip surgery incidence was higher in the ST-ITB group compared with PRE-ITB or NON-ITB. When comparing PRE-ITB with NON-ITB, no difference was seen. ITB, intrathecal baclofen, NON-ITB soft-tissue release without ITB, PRE-ITB soft-tissue release before ITB, ST-ITB, soft-tissue release at the same time as ITB, ST isolated soft tissue, BS bony surgery, NS no surgery

The mean age for having osteotomies in the PRE-ITB group (10.5, SD 3.2 years) was not statistically significantly different from the other treatment groups (ST-ITB: 8.6, SD 3.1 years, p = 0.16; NON-ITB: 9.2, SD 2.7 years, p = 0.22). The duration between the index adductor surgery and the age at osteotomies was significantly higher for the PRE-ITB (6.5 years) group compared with the ST-ITB group (3.1 years, p = 0.02) but not with the NON-ITB group (5.8 years, p = 0.5).

The odds ratio of ST-ITB to be considered an unsuccessful hip was significantly higher than NON-ITB (Table 2). However, no difference was found between NON-ITB and PRE-ITB (Table 2). In total, 65% of the cohort were failures, with the ST-ITB group having the highest failure rate at 90%, the PRE-ITB group with 50%, and the NON-ITB group with 62.7%.

Table 2.

Hip outcomes in the PRE-ITB, ST-ITB, and NON-ITB groups

Hip outcome	PRE-ITB N = 30 (%)	ST-ITB N = 20 (%)	NON-ITB N = 43 (%)	Total N	Total %
Successful hip outcome	15 (50)	2 (10)	16 (37)	33	35.5
Unsuccessful hip outcome	15 (50)	18 (90)	27 (63)	60	64.5
Odds ratio for unsuccessful hip outcome
NON-ITB vs PRE-ITB	OR: 0.59		Reference
	CI 0.23–1.53
NON-ITB vs ST-ITB		OR: 5.32	Reference
		CI 1.09–26.3

Unsuccessful hip outcome is defined as a hip having a second surgery or last follow-up migration percentage ≥ 50%. The OR of unsuccessful hips was higher in the ST-ITB group

CI confidence interval, OR odds ratio, ITB intrathecal baclofen, NON-ITB soft-tissue release without ITB, PRE-ITB soft-tissue release before ITB, ST-ITB soft-tissue release at the same time as ITB

The incidence of surgical site infection over the whole follow-up period was higher in the ST-ITB (40%) group compared with the PRE-ITB (13.3%, p = 0.035) and the NON-ITB groups (0, p < 0.001) and includes only the spinal region. Surgical site infection was also higher for the PRE-ITB group compared with the NON-ITB group due to spinal infection related to pump insertion or catheter presence during subsequent spinal fusion (p = 0.025).

Discussion

There are no randomized clinical trials that show the effectiveness of soft-tissue surgery for hip displacement in children with CP [24]. Our starting hypothesis was that doing tone reduction with ITB would decrease the need for further hip surgery following soft-tissue release. The odds of successful outcome after soft-tissue surgery were not different between NON-ITB and PRE-ITB but were lower for the ST-ITB group. This shows that ITB had no effect on the failure rate after soft-tissue surgery. Our results are consistent with a meta-analysis finding little evidence for ITB preventing hip displacement [25]. Intrathecal baclofen demonstrated documented benefit for tone management in 2019 [26]. Gerszten et al. showed in a case series of 48 patients with ITB implantation that the planned incidence of orthopaedic procedures decreased from 28 to 10 in a 53-month follow-up [27]. Gooch et al. [28] reported that there was no significant change in orthopaedic surgery frequency following pump placement.

The ITB pump requires subsequent surgical interventions due to battery expiration, cerebrospinal fluid leak, infection, catheter obstruction, and kinking and many patients need spinal fusions. All interventions may lead to ITB-related infections. In our series, the ST-ITB group had a higher infection rate (40%) requiring surgical intervention compared with the PRE-ITB (13.3%) and NON-ITB groups (0). All infections were localized to the spine incision. In a recent study, the infection rate after ITB was shown as 14.6% per patient in a cohort of 316 patients over a 15-year follow-up [29].

The comparison of the PRE-ITB and NON-ITB groups revealed no difference; however, both had lower failure rates compared with the ST-ITB group. This suggests that the addition of ITB does not improve upon the high recurrence rates associated with soft-tissue surgery alone. This revelation also supports the early intervention of contracture release having a beneficial effect in the nonambulatory population. However, we still have a 50% (PRE-ITB) and 63% (NON-ITB) failure rate for early intervention groups and the relatively low failure rate in PRE-ITB might be caused by smaller MP prior to ITB. The mean age for adductor/psoas tenotomy was 4.8 years in a Swedish population-based study with 20 years of hip surveillance, where the mean age was close to our whole cohort mean (4.34 ± 1.68 years), and the repeated hip surgery incidence was 45% [30].

The milestone studies by Reimers [22], Kalen and Bleck [12], and Onimus et al. [10] also emphasized the importance of early soft-tissue intervention in a spastic hip population. Our current results of having preventative soft-tissue surgery before severe hip dysplasia were supported in a study finding successful results in 90% of patients under 4 years of age and having an MP less than 33% [10]. Moreau et al. reported soft-tissue release was effective in 39 of 44 hips with five-year follow-up [9]. Shore et al. did not report a statistical difference in outcomes following soft-tissue surgery; however, the adductor tenotomy age (3.7 years) was lower in the group with successful outcomes compared with unsuccessful outcomes (4.4 years) [6]. Although the results were not significant, Presedo et al. [31] also reported a nonstatistically significant younger age for successful outcomes. Vidal et al. [32] found that children under four years of age had twice as much improvement in their hips compared with children over seven years at the time of surgery.

Our current study has a mean follow-up around nine years with a minimum of 5-year follow-up. The minimum follow-up bar was established at 5 years, which is higher than previously reported by Shore et al. [6] (2 years), Terjesen et al. [15] (not specified). The only study with more than five years minimum follow-up was Presedo et al. [31], with a minimum of 8 years of follow-up. Besides this, we also have distinct matching criteria which come at the cost of small sample size and the possibility of being underpowered. Inherent of retrospective case–control studies, our study has potential selection bias. In addition, postoperative physical therapy compliance cannot be assessed. Another limitation of this study was we did not integrate quality-of-life questionnaires. Our results should not be interpreted as ITB having no place in the treatment of CP spasticity in the nonambulatory population as we did not include quality-of-life questionnaires, activity, and participation outcomes. Although all ITB patients had continued the ITB treatment protocol, the high pump infection rate with the ST-ITB group might interfere with the spasticity management and postoperative physical therapy that is causing higher failure rates with the ST-ITB group.

In conclusion, the addition of tone management with ITB did not reduce the need for later hip surgery but did increase the risk for surgical site infection. Early soft-tissue release might have some impact, but the failure rate is more than 50%.

Author contributions

All authors contributed to the study’s conception and design. Material preparation, data collection and analysis were performed by AA, ACU, JJ H, KJR, FM, and MWS. The first draft of the manuscript was written by AA, ACU, JJH, and KJR. Critical revision for important intellectual content was made by AA, ACU, JJH, KJR, FM, and MWS. All authors read and approved the final manuscript.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Declarations

Conflict of interest

Ali Asma, Armagan Can Ulusaloglu, Jason J. Howard, Kenneth J. Rogers, Freeman Miller, and Michael Wade Shrader declare that they have no conflict of interest.

Ethical approval

This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Nemours Office of Human Subjects Protection Institutional Review Board.