This important paper reviews the results of growing-rod spinal instrumentation in 140 children treated by members of the Growing Spine Study Group over an eighteen-year time period. This series represents the largest cohort of patients with early-onset scoliosis treated with growing-rod instrumentation in the literature. The authors do not report on patients with a single diagnosis; instead, they include those with neuromuscular, idiopathic, and congenital scoliosis. The mean age at the time of implantation of the growing rods was six years (range, 1.7 to ten years), and the initial curve magnitude averaged 75° (range, 32° to 147°). An average of 6.4 procedures (range, two to fifteen) were performed per child. Surprisingly, only seventy-four surgical procedures were unplanned.
The authors state that thoracic growth occurs most rapidly between birth and five years of age, and that severe curves in these very young patients may result in trunk shortening and pulmonary failure. The literature supports this statement, as Goldberg et al.1, Vitale et al.2, and our study3 have shown restrictive lung disease in children whose spine was fused before the age of five years. Therefore, it is this population that is most in need of preservation of thoracic growth. The decision to embark on growing-rod treatment of children with scoliosis who are between the ages of eight and ten is questionable, as definitive spinal fusion in this age group is unlikely to produce the same negative effect on respiratory function as is seen in very young children.
In their historical papers, Moe et al.4 and Mineiro and Weinstein5 reported a 50% to 91% incidence of complications in patients treated with Harrington instrumentation without fusion who subsequently underwent further serial distraction of the instrumentation. Bess et al. reported that 58% of their patients had at least one complication. Despite more modern spinal instrumentation, the complication rate of this surgical program remains high. Bess et al. sought to identify which technical factors were most commonly linked to the occurrence of a complication. Patients treated with a single rod or a subcutaneous rod had a higher prevalence of complications than patients treated with submuscular dual growing-rod instrumentation. This is valuable information for the pediatric spine surgeon.
Kaplan-Meier analysis showed that the likelihood that a given patient would have a complication increased as the number of surgical procedures increased. However, the average number of procedures in the study group was only six, with a range of two to fifteen. The authors state, "With eleven procedures the wound complication-free rate decreased to 60%." The reader should recognize that the study is likely underpowered in terms of patients with high numbers of surgical procedures. Only fifty children had undergone definitive spinal fusion at the conclusion of the growing-rod program. The mean duration of follow-up was only five years, and ninety children were still undergoing spinal lengthening. If the likelihood of complications increases with the number of surgical procedures, it is sobering to realize that the complication rate reported in this paper is likely to increase as these children undergo future lengthening procedures.
The authors note that an older age at the time of the initial growing-rod implantation reduces the risk of complications. A child who is two years of age at the time of the initial growing-rod instrumentation and undergoes lengthening every six months through the age of ten years without a complication is likely to have seventeen procedures. A child who is five years old at the time of growing-rod insertion will have approximately six fewer planned surgical procedures. Therefore, is the influence of age on the prevalence of complications the result of fewer surgical procedures, the fact that older children are bigger so wound dehiscence from bulky implants is less likely, or the fact that the youngest children who require growing rods have more severe deformities and more stress on their implants as a result of larger residual curves?
Bess et al. comment on the complications of Mehta casting, specifically skin irritation, lumbar lordosis, and early satiety. The authors cite reports in the literature stating that casting and bracing may lead to stunting of chest wall development and poor pulmonary function. With a 58% rate of complications per child treated with growing-rod spinal instrumentation, continued attempts to control these curves, which are amenable to casting in the infantile and very young pediatric age groups, seem merited. Skin irritation and early satiety are easily addressed with cast trimming. Wound dehiscence and deep infection from multiple surgical lengthening procedures are not as easily treated. Early-onset-scoliosis clinics must critically assess thoracic development in individuals treated with casting, and future studies must also address this in children treated with growing rods and VEPTR instrumentation.
In conclusion, Bess et al. reported a substantial complication rate for young children with early-onset scoliosis treated with a growing-rod spinal instrumentation program. Orthopaedic surgeons treating this challenging patient population will note the decrease in implant-related complications, due to fewer broken rods, in the group treated with dual rods and the decreased wound complication rate in the children treated with submuscular rods. The results of this study emphasize the need for continued research into better surgical strategies and smaller-profile implants for young patients with this challenging condition. Figure E-3 beautifully illustrates both the challenges and the rewards of treating early-onset scoliosis, as the patient experienced seven years of growth of the spine and chest despite the typical complications that may occur. While it is acknowledged that complications may be expected, expandable instrumentation that is either spine or chest-based is the best answer today for promoting pulmonary development in very young children with severe progressive spinal deformity. We look forward to more reports from this study group in the future.