Complications of intrathecal baclofen therapy in children and young adults.

Objective: The primary objective of this study was to determine the frequency and nature of complications that occur during intrathecal baclofen (ITB) therapy, including nonsurgical complications as well as complications associated with both implantation surgeries and subsequent revisions.

Methods: The authors retrospectively reviewed the medical records of all patients who had baclofen pumps implanted at a tertiary children's hospital from 2006 through June 2019. The study employed appropriate descriptive statistics and statistical models to analyze patient demographics, improvements in tone, and clinical complications. The authors evaluated the associations of complications as well as changes in modified Ashworth Scale (MAS) scores with various preexisting conditions (e.g., presence of gastrostomy tubes [G-tubes] and ventriculoperitoneal [VP] shunts) and surgical variations (e.g., use of newer [Ascenda] catheters and subfascial versus subcutaneous catheter tunneling).

Results: One hundred forty-two pumps were implanted. Complications occurred in 111 patients (78.2%). The most frequent complications were catheter complications and pseudomeningoceles, each of which occurred in 63 (44%) patients. On multivariable analysis, pseudomeningoceles and percutaneous CSF leaks were significantly less likely when intrathecal catheters were placed via laminotomy versus spinal needle (OR 4.6, p = 0.044), and when catheters were passed from the posterior incision to the abdominal pump pocket deep to the thoracolumbar fascia rather than superficial to it (OR 2.7, p = 0.008). Preexisting G-tubes and VP shunts at the time of pump implantation were not associated with a significantly increased likelihood of pump malfunction or infection. Ascenda catheters were significantly less likely to have disconnections (p < 0.001) and obstructions (p = 0.016), and overall were less likely to have any catheter-related complications (p = 0.01). Patients with preexisting VP shunts at the time of implantation had a significantly greater mean decrease in MAS scores in both their lower extremities (1.8, p = 0.003) and upper extremities (1.3, p = 0.002) than did patients without shunts.

Conclusions: Various complications are associated with ITB therapy, most commonly catheter complications and pseudomeningoceles. There have been fewer catheter complications associated with the newest catheter model, while pseudomeningoceles have become less frequent since the surgical technique was modified to prevent them. VP shunts and G-tubes are not associated with significantly higher complication rates and shunts seem to be associated with greater efficacy of ITB therapy, as represented by a greater mean improvement in MAS scores.