Eleven years’ experience with Intrathecal Baclofen – Complications, risk factors

Abstract Objective Treatment with intrathecal baclofen (ITB) is commonly used in patients with severe spasticity. However, complications may occur after implantation of the ITB‐device, albeit mainly procedure‐ and device‐related problems. The aim of the study was to assess surgical‐ as well as catheter‐ and pump‐related complications and define their risk factors. Methods We retrospectively evaluated all patients with an implanted ITB‐device who were treated at the Department of Neurology, Hochzirl Hospital, Zirl, Austria, between 2006 and 2016. Results Twenty‐nine of 116 (25%) patients experienced 32 complications: 5 procedure‐ and 27 device‐related (4 pump‐ and 23 catheter‐associated) problems occurred. Risk factors for sustaining any complication were a spinal localization of lesion (odds ratio [OR] OR 2.71, p = .021), other catheter types than an Ascenda® catheter (OR 3.87, p = .041), a lower modified Rankin Scale (median 4 vs. 5; OR 2.86, p = .015) and a higher Barthel Index (median 53 vs. 0; OR 2.84, p = .006). The median time from the last ITB‐related surgery to the first complication was 18 (IQR 1‐57) months. Overall, 47% complications occurred within the first year after any surgical procedure regarding the ITB‐device, thereof 25% within the first month. Conclusions Procedure‐ and device‐related complications are frequent after implantation of an ITB‐device with catheter‐associated complications as the most frequently encountered problems. Patients with a spinal origin of spasticity, a lower modified Rankin Scale and a higher Barthel Index have a higher risk to sustain a complication.


| MATERIAL S AND ME THODS
We retrospectively evaluated all patients with an intrathecal ITBdevice, who were treated at our Department of Neurology, between 01. 01.2006 and 31.12.2016. Patients of any age and receiving care bothin inpatient as well as in outpatient hospital setting were included.
ITB-pump-related surgical interventions were classified as "primary implantation procedure" (pump implantation for the first time) and "replacement procedure" (pump surgery: replacement due to end of battery life 72 months after implantation at the earliest or revision because of a complication; catheter-only-revision after a complication; exchange of the entire system) according to earlier published studies (Awaad et al., 2012;Borrini et al., 2014). The retrospective analysis was done by reviewing patient's medical history and the data provided in our ITB-dedicated register, which was implemented in 01. 01.2006. Exclusion criteria were insufficient patient data. The majority of patients were admitted for ITB-evaluation, presurgical and postsurgical care as well as follow-up after implantation of the ITB-pump. The remaining patients were referred because of an occurring complication for diagnostic work-up and treatment but followed up at their local district hospital. According to earlier published studies, complications were divided into procedure-related and device-related complications, the latter into catheter-and pumpassociated complications (Guillaume, Van Havenbergh, Vloeberghs, Vidal, & Roeste, 2005;Kolaski & Logan, 2007;Rawlins, 2004).
Procedure-associated problems were defined as related to the surgical intervention occurring within the first 2 months after a surgical intervention and thus most likely being associated with the surgical intervention, such as scar complications, subcutaneous seroma, infections, or cerebrospinal fluid (CSF) leakages. Device-associated problems include complications related to the catheter or the pump and were classified as "early" when they occurred within the first 2 months or "late" when occurring later than 2 months after the surgical intervention.
We collected clinical data in all included patients: demographic information, spinal versus supraspinal localization of lesion responsible for the spasticity and data of any previous ITB-device-related surgery. Moreover, the type of the implanted catheter (Ascenda ® catheter/Model 8781 vs. other catheter types) and the following parameters at steady-state of ITB were assessed: Ashworth Scale, modified Rankin Scale, Barthel Index, the median ITB-dosage, median ITB-concentration, flow rate of ITB, mode of application, flow rate of ITB, and last follow-up. In patients admitted more than once, the follow-up was defined as period in months between the first and last consultation at our department. In patients with complications, we analyzed the number and type of complications, as well as the type of intervention. Patients with a suspected complication underwent a comprehensive diagnostic work-up including the evaluation of the clinical response to ITB, laboratory test and imaging (depending on the clinical situation conventional x-ray, computer tomography without or with contrasting agent, magnet resonance imaging or fluoroscopy). If the aetiology of the complication still remained unclear, a surgical exploration was done. In patients with a complication, we calculated the duration of the hospital stay from admission to the day of a consecutive intervention (revision surgery, surgical exploration, or nonsurgical intervention). Moreover, the time from the initial ITB-device implantation and from the last ITB-related surgery respectively to any occurring complication during the observational period was calculated in all patients with complications. In addition, risk factors for sustaining a complication were analyzed.

| Statistical analysis
Data were summarized in cross tables, and medians and interquartile ranges were calculated for ordinal variables, such as modified Rankin Scale, Barthel Index, or Asworth Scale. Complication and incident rates for the occurrence of complications were calculated and odds ratios comparing different subgroups regarding complication frequency were given. Since there were three patients with two complications, statistical analysis was conducted with cases including the second complication. Odds ratios for ordinal risk factors were obtained by dichotomizing the respective risk factor using the overall median value as the cut-off threshold. The influence of risk factors on the occurrence of complications was tested by chi-squared test for nominal and by Mann-Whitney U test for ordinal risk factors. A significance level of α = .05 (2-tailed) was applied. Statistical analyses were performed using SPSS, version 22.0 (IBM Corp., Armonk, NY, USA).
According to our country's law on retrospective research, this study did not require the approval of the ethics committee, but was performed based on ethical standards of the "World Medical Association Helsinki Declaration" (https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/).

| RE SULTS
One hundred and thirty-six patients with an ITB-pump were treated within the given 11-year period at our neurorehabilitation department. One hundred and sixteen patients (77 men and 39 women, mean age at implantation 39 ± 16.7 years, range 2-77 years) en-  Table 1). The mode of application after reaching steady-state of ITB was continuous in 95% and flex mode in 5% patients. The median follow-up was 42 months (IQR 17-91, mean 56 months), comprising 506 pump years. Seven patients died during follow-up (one suicide, three severe pneumonia, three sepsis). There was no permanent morbidity or death related to the ITB-pump-complication.

| Risk factors for complications
Analysis revealed a significant association between occurrence of a

| D ISCUSS I ON
This study provides a systematic assessment of procedure-and  TA B L E 2 (Continued) complications in 12/52 (23%) patients after ITB-device implantation (Paskhin, Dekopov, Tomsky, Isagulyan, & Salova, 2017): 5/12 complications were caused by catheter migration and the authors identified patients with severe dystonia of the trunk muscles to have an increased risk of spinal catheter migrations. Accordingly, the degree of independence and mobility represented with the Barthel Index and the modified Rankin Scale might be correlated with the occurrence of catheter problems based on mechanical stress.
In our cohort, no significant association was found between occurrence of an adverse event and gender, which is in line with a previous report (Borrini et al., 2014). Also age was no risk factor for a complication in our study which is in contrast with a previously published paper (Motta et al., 2007): in the study of Motta and coworkers, complications were statistically more likely in patients with an age of 10 years and younger. This might be related to the by far lower mean age of 13.7 ± 5.7 years of the 200 children and adolescents in their study compared to ours (mean age 39 ± 16.7 years). An Ashworth Scale higher than 3 was a further risk factor in this study whereas we did not find a significant difference of the Ashworth Scale in patients with and without complications. The divergent results might be explained by the different patient population, as about 90% of included patients in the study of Motta and coworkers were affected by cerebral palsy with thus clinically dominating dystonia, whereas we included a more mixed patient population with a traumatic etiology in more than half of the patients.
The impact of dose and concentration of intrathecally administered drugs on catheter tip granuloma is still discussed controversially. A direct correlation with high concentrations or high dose opiate drugs was suggested Yaksh et al., 2002), but also an association between intrathecal granulomas and low doses and concentrations of hydromorphone (Veizi et al., 2016) reported. We analyzed a possible association between the flow rate of ITB and possibly mechanically driven complications and found no relation to occurring complications.
Motta and coworkers compared complications before and after the introduction of the Ascenda ® catheter for the first time in paediatric patients recently (Motta & Antonello, 2016): 120/416 (29%) children with silicone catheters sustained major complications (infections, leakages, catheter-related problems), but only 1/92 children with an Ascenda ® catheter (leakage). We could confirm these F I G U R E 1 Temporal distribution of occurring complications within the first 7 years after the last intrathecal baclofen (ITB)-related surgical procedure (first implantation surgery, replacement surgery due to end of battery life or due to a revision after a complication) F I G U R E 2 Temporal distribution of occurring complications within the first 12 months after the last surgical procedure (first implantation surgery, replacement surgery due to end of battery life or due to a revision after a complication) results in our study as an implanted Ascenda ® catheter was associated with an approximately fourfold decreased risk for catheterrelated complications, most likely the consequence of technological improvements of the Ascenda ® catheter.
In our cohort, half of the complications occurred within the first 12 months after any ITB-related surgery, after primary implantation procedures as well as replacement procedures due to end of battery life or revision likewise, within the first month a quarter of the complications. Also Motta and coworkers did not observe a different event rate after initial device-implantations compared to complications after replacement surgeries (Motta & Antonello, 2014). Thus, the critical period is the first year and especially the first month after any ITB-related surgery.

| Study limitations
Limitations of our study are: First of all, the retrospective assessment of the data has to be taken into account. Furthermore, the Further prospective multicenter studies are necessary to evaluate risks and safety issues of treatment with an ITB-device in order to improve the risk management of complications related to the ITB-device.

ACK N OWLED G M ENTS
There are no acknowledgments.

CO N FLI C T O F I NTE R E S T
None declared.