Complication rates for preexisting baclofen pumps and ventricular shunts following scoliosis correction: a preliminary study

OBJECTIVE Many patients undergoing spinal fusion for neuromuscular scoliosis have preexisting neurosurgical implants, including ventricular shunts (VSs) for hydrocephalus and baclofen pumps (BPs) for spastic cerebral palsy. Recent studies have discussed a possible increase in implant complication rates following spinal fusion, but published data are inconclusive. The authors therefore, sought to investigate: 1) the rate of implant complications following fusion, 2) possible causes of these complications, and 3) factors that place patients at higher risk for implant-related complications. METHODS Cases involving pediatric patients with a preexisting VS or BP who underwent spinal fusion for scoliosis correction between 2005 and 2016 at a single tertiary children's hospital were retrospectively analyzed. Patient demographics, implant characteristics, spinal fusion details, neurosurgical follow-up, and implant complications in the 180 days following fusion were recorded and analyzed. RESULTS Overall, 75 patients who underwent scoliosis correction had preexisting implants: 39 had BPs, 31 VSs, and 5 both. The patients' mean age at fusion was 13.49 ± 2.78 years (range 3.62-18.81 years), and the mean time from the most recent previous implant surgery to fusion was 5.70 ± 4.65 years (range 0.10-17.3 years). The mean preoperative and postoperative Cobb angles were 62.4° ± 18.9° degrees (range 20.9°-109.0°) and 23.5° ± 13.3° degrees (range 2.00°-67.3°), respectively. No VS complications were identified. Two patients with BPs were found to have complications (unintentional cutting of their BP catheter during posterior spinal fusion) within 180 days postfusion. There were no recorded neurosurgical implant infections, failures, fractures, or dislodgements. Although 10 patients required at least 1 surgical procedure for irrigation and debridement of the spine wound following fusion, there were no abdominal or cranial implant wound infections requiring revision, and no implants required removal. CONCLUSIONS The results of this study suggest that spinal fusion for scoliosis correction does not increase the rates of complications involving previously placed neurosurgical implants. A large-scale, prospective, multicenter study is needed to fully explore and confirm this finding.