Introduction

Baclofen was first introduced as an anticonvulsive medication in 1962. It was abandoned as such shortly after due to adverse effects. However, the effect of baclofen on hypertonic muscle tone was clearly evaluated, and baclofen has been used ever since for the oral treatment of spasticity. Because of the blood-brain barrier, high doses of oral baclofen are needed to induce hypotonic effects, and this often results in side effects like nausea, sedation, and weakness [1]. The effect of Baclofen on the neural cells is not completely understood. It is known that baclofen interacts with the GABA-B receptors in the spinal cord. Abe et al. recently reported some evidence for pre- and postsynaptic effects of baclofen [2].

In severe spasticity, the side effects of higher doses of oral baclofen sometimes cause somnolence and even apnea. Intrathecal baclofen (ITB) therapy was first administered in 1984, and it was clearly shown that the application of baclofen directly into the intrathecal space achieved higher efficacy and lesser side effects due to direct site administration [1]. Unfortunately, it seems that the distribution range of intrathecal baclofen from the catheter tip in the cerebrospinal fluid is limited at least in pigs; in humans, it is not exactly known [3].

The first report of ITB therapy usage in pediatric patients mainly dealing with cerebral palsy was published in 1991 [4]. Although many publications thereafter covered all sorts of aspects such as surgical complications, pump infection, over dosage, pump failure, long-term dosing, trouble shooting, and others, most of the publications reported results in adult patients only, while few reports are available for pediatric cases [5,6,7,8]. Reports about ITB therapy in infants and small children under 6 years are rare, with only a few cases ranging from 5 to 6 years out of their cohorts, respectively [9,10,11]. A recent review published in 2018 quotes the lack of evidence for ITB therapy in pediatric cases [12]. We are reporting our results in infants and small children below 6 years of age, because we are convinced of the necessity to treat severely spastic infants with ITB therapy early instead postponing the therapy and forfeit crucial years of effective treatment to reduce painful spasticity, orthopedic complications like muscle tone–related fractures, contractures, and so forth.

Methods

We performed a retrospective analysis of 17 patients in whom the pump was implanted before 6 years of age out of our total collective of 135 pediatric cases of primary pump implantations for spasticity treatment between 2007 and 2018. The follow-up ranged from 12 months to 11 years. Surgical revisions and other unexpected events were recorded next to general aspects like diagnose, weight, length, and age at implantation. Daily dose at the last refill and other details like catheter tip positioning were recorded.

Implants, surgical technique, and follow-up

All implanted pumps (SynchroMed II, 20 ml volume) were products by Medtronic (Medtronic GmbH Germany, Earl-Bakken-Platz, 40,670 Meerbusch, Germany). These pumps are manufactured with a thickness of 19, 5 mm, a diameter of 87, 5 mm, a volume of 20 ml, and a tissue displacement of 91 cm3. They require an area of 8, 5–9-cm-diameter soft tissue for implantation, and the minimum incision requires 8 cm. In 2007, the 20-ml pumps were implanted subcutaneously; since 2008, the subfascial technique was utilized [13]. The implanted catheter system used from 2008 to 2011 was “Indura 1P” (model 8709) by Medtronic, which was discontinued from the market in 2011. This catheter was a single piece tube with good visibility in X-ray studies and fluoroscopy due to integrated barium. From 2011 on, we used “Ascenda” (model 8780 or 8781) by Medtronic, which is a two-piece system with reinforced tube walls to diminish the occurrence of tears in the catheter. Visibility in X-ray studies is somewhat poorer due to the lack of barium in the tube material, therefore demanding diagnostic catheter studies with contrast-enhancing test fluids to find or rule out catheter transections or disconnections.

The following details regarding surgical technique and follow-up are the authors’ preferred regimen, as there is no technical defined standard, and implantation technique differs between institutions. The incision is placed preferably in the right flank which we call the “pump triangle,” because many patients have gastrostomies on the left side. Nevertheless, left side implantation remains sometimes necessary because of severe scoliosis or coexisting ventriculoperitoneal shunts. The boundaries of the “pump triangle” consist of the 10th rib lower margin cranially, the navel medially, and the anterior superior iliac spine caudally. The catheters were implanted subcutaneously including a fascial anchoring system at the level of the lumbar spine from 2007 to 2011 and afterwards with placement underneath the thoraco-lumbar fascia. Vancomycin was placed at the implantation site since 2008. We ordered 3 days post-operative non-elevated patient positioning to reduce the risk of cerebrospinal fluid fistulas. All patients remained between 7 and 10 days in the hospital to carefully titrate the baclofen after pump implantation until a satisfactory tone reduction was achieved. Catheter location was documented by X-ray after mobilization. Refills were conducted in an ambulatory setting in our hospital and performed by the authors. We did not perform regular follow-up X-ray studies regarding the catheter tip position; therefore, catheter slides were not detected in all patients.

Results

The details of patient data are shown in Table 1. All suffered from bilateral spastic cerebral palsy of various etiologies, and all were level 5 according to the Gross Motor Function Classification Scale (GMFCS).

Table 1 Patients overview. Age in months, bodyweight in kilogram, length in centimeters; FU, follow up in months; catheter tip position regarding spinal vertebra; C, cervical; T, thoracic, in brackets the tip position in later FU if X-ray was performed, dose in μg/d at latest refill
Full size table

Three patients had to undergo revision surgeries within the first month after implantation. One pump had to be explanted due to pump infection after subcutaneous implantation within the first year of our program. The other revision surgeries were due to catheter events, two cases of catheter dislocations and one case of catheter transection due to opisthotonus. Of the catheter events that lead to revision surgery, two occurred in the very same dystonic patient (case #13 on Table 1) consecutively. Two more complications recorded were treated medically, an overdose during titration phase in the first week after implantation and an overdose at the first refill (pump refill with 2000 μg/ml instead of 500 μg/ml baclofen solution). Overdoses were treated by prompt dose reduction programming in the first and emergency refill with lower baclofen concentration and removal of higher concentration baclofen out of the catheter system, followed by bolus in appropriate concentration, in the second case. In the entire follow-up, we saw neither secondary pump infection nor cerebrospinal fluid fistula in this group of infants and small children, unlike others [14].

The youngest and smallest patient (case #17 in Table 1, Fig. 1) with 6, 4 kg, and 63-cm length at implantation had no complication at all and underwent uneventful implantation at 11 months of age. Nevertheless, we indicated ITB therapy already at 6 months of age in this battered child but considered his “pump triangle” to be too small to implant the 20-ml pump with respect to the diameter needed. In the meantime, he suffered two muscle tone–related femoral fractures at 7 and 8 months of age, before he reached the appropriate body size for pump implantation. Since implantation, there were no more tone-related fractures in this child.

Fig. 1
figure 1

Smallest infant implanted at 11 months of age, 6, 4-kg body weight, 63-cm length, using radiopaque Indura 1P catheter by Medtronic. a At 6 months, although already indicated for ITB therapy, there is not enough soft tissue space for implantation in the “pump triangle”. b X-ray post-surgery at 11 months. C: 4 years post-surgery

Full size image

Three pumps have been already exchanged due to expected battery expiry. We did not see any pump failures due to motor corrosion or other technical problems. Five children died years after pump implantation, each suffering from severe forms of epilepsy like Lennox-Gastaut or West syndrome; therefore, the causes of deaths were not related to pump implantation, but pneumonia and fatal fit in this severely handicapped patient group. The daily dose of baclofen was administered in a simple continuous fashion in all reported patients and we did not record any sudden loss of efficacy as reported by Saulino et al. [15].

In most but not all patients who underwent X-ray controls for various reasons, we saw a slide of catheter tip over the years. Significant sliding of several spinal segments was recorded after spine surgery with the correction of deformity by dorsal fusion and due to surgical failure shortly after implantation (Fig. 2). Sliding due to growth or utilization of expandable devices for scoliosis treatment like Vertical Expandable Prosthetic Titanium Ribs (VEPTR) or Magnetically Controlled Growing Rods (MCGR) mostly involved only a few spinal segments over years (Fig. 3).

Fig. 2
figure 2

Catheter slide due to surgical failure using a radiolucent Ascenda catheter by Medtronic. a X-ray at 3rd-day post-surgery, notice the small radiopaque mark at the tip of the catheter. Although catheter tip was located at T1 during implantation, it shows already significant sliding to T5. B: Few days later, the patient developed baclofen withdrawal, and the catheter tip is no longer visible on X-ray. The entire catheter lay rolled up behind the pump at revision

Full size image
Fig. 3
figure 3

Catheter slide due to spine surgery and elongation of magnetically elongated growing rods in neuromuscular scoliosis. a Scoliosis and tip position at T3. b Tip position after implantation of growing rods. c Tip position after frequent elongation of the growing rods

Full size image

Discussion

The patient group presented distributes typical aspects of patients suffering from bilateral spastic cerebral palsy due to various reasons with level 5 on GMFCS. Therefore, it constitutes a typical patient group in whom ITB therapy is often indicated. Many of the patients were diagnosed with a mixed tone disorder (spasticity and dystonia) but only one patient presented an almost pure secondary dystonia. In this patient, the dose titration phase was considerably longer and even with a transient dose of 2000 μg baclofen, tone reduction was less satisfactory. Young age and low body weight are certainly not a contraindication for ITB therapy. The smallest patient in our cohort being successfully implanted without any complications at 11 months of age with a bodyweight of 6, 4 kg, and a length of 63 cm (Fig. 1) responded very well to ITB therapy, although we had to wait for about 5 months after indication of ITB therapy for him to grow enough to reach the required 8-cm diameter in the “pump triangle” of the right flank. Despite oral baclofen therapy, we observed two muscle tone–related femoral fractures until we could implant the pump in this particular child. In general, age and bodyweight are not principal limitations for the feasibility of ITB therapy in infants, but the diameter of the abdomen. Once patients have reached a body size comparable with an 8-cm incision in the “pump triangle,” implantation is possible. Smaller pumps than 20-ml volume are available, but keep in mind that smaller volumes go hand in hand with a higher frequency of refills.

The single pump infection in our patient group occurred after subcutaneous implantation in 2007, leading to a change of surgical strategy at our institution. After this event, all pumps were implanted in subfascial technique and vancomycin powder was administered at the pump implantation site. Since changing the surgical technique to subfascial implantation and administering vancomycin, we did not see any other pump infection, neither in the group of infants and small children reported nor in the other 119 pediatric cases, unlike others who reported up to 9, 8% infections in pediatric cases [6, 16]. Nevertheless, the use of vancomycin at the implantation site is the authors’ preference, as reports are published which do not support the efficacy of onsite vancomycin, but again, we had no single pump infection since 2007 [17].

All the reported 17 infants and small children receive the baclofen in a simple continuous fashion, and they were titrated over the years according to their needs during the refill sessions, showing mostly an increasing dose over several months after implantation as reported by Gilmartin et al. [18]. All refill sessions were performed by authors; therefore, there was a close relationship between patients, parents, caregivers, and treating doctors.

Catheter events leading to revision surgery are still occurring but we see them more seldom, and we recommend to implanting them in subfascial technique as well because we believe that there is more resistance for the catheter to slide if implanted in subfascial technique. In subcutaneous implantation technique, we have seen dislocations despite the anchoring system. One case of transection occurred in a dystonic child in the severe opisthotonic state while still in early titration phase and the catheter was cut in between vertebral arches (Fig. 4). At revision, an interlaminar fenestration was performed at the catheter entry zone to avoid future catheter transections in this particular child. Overall, 23% of cases required revision surgery, an incidence of complications similar to those reported by others [10, 19].

Fig. 4
figure 4

Catheter transection due to dystonia (radiopaque catheter Indura 1P). a X-ray after implantation. b Baclofen withdrawal on day 6 of titration phase after dystonic opisthotonus, which led to transection of the catheter

Full size image

Catheter tip placement should be considered in every case according to the spasticity and expected growth. Due to the assumed rather limited range of baclofen distribution in the cerebrospinal fluid away from the catheter tip, there seems to be a significant loss of concentration just a few centimeters away from the catheter tip as reported by Flack and Bernards [3]. Therefore, especially in quadriplegic patients in whom reduction of spasticity also at the upper extremities is an important goal, we recommend placing the catheter tip in the cervical region, like others [8]. Due to expected growth, there might be a loss of effect regarding the upper extremities following catheter slide over the years and even the need to exchange the catheter in the same session as pump exchange surgery due to battery depletion, which is normally due between 6 and 7 years after implantation depending on the performance of the pump.

Conclusion

ITB in infants and small children is as safe and effective as reported for elder pediatric cases and adults in the literature. We have successfully implanted a pump in an 11 months old infant with 6, 4-kg bodyweight, and 63-cm length once this patient reached the appropriate size to place an 8-cm incision in the “pump triangle” of the right flank. Subfascial implantation technique of the entire system seems to reduce infection and dislocation. Dose titration should be slowly and carefully to avoid overdosing, which can lead to bradycardia and apnea. As this is a retrospective analysis with a limited number of patients, the level of evidence is 4.