How Can We Predict Dominant Compensatory Curves in Congenital Scoliosis?

Background: Patients with congenital scoliosis may develop a large, unbalancing “compensatory” curve as they grow, which can be severely deforming. It is important to identify characteristics of patients who develop a “dominant” compensatory curve (one that is larger than the congenital curve) to help identify patients who may benefit from prophylactic treatment. Methods: By searching the Pediatric Spine Study Group database, we identified 307 patients 18 years and above with congenital scoliosis who had preoperative radiographs taken during at least 2 years of natural growth (period of no bracing or surgery). Seventeen patients (6%) had a dominant compensatory curve, and 290 had no compensatory curve or one that was smaller than the congenital curve. Of those 290 patients, 100 were randomly selected to serve as a control group, which we refer to as the “nondominant curve group.” We analyzed the type of congenital anomaly and its vertebral level, as well as the major curve angles of the congenital and compensatory curves at initial and latest follow-up. We compared vertebral level and type of anomaly between groups using χ 2 tests. Alpha = 0.05. Results: The congenital anomaly was at L4 or more caudal in 18% of patients in the dominant curve group and no patients in the nondominant curve group ( P <0.001). Similarly, the congenital anomaly was at T6 or more cranial in 59% of patients in the dominant curve group and 28% of patients in the nondominant curve group ( P <0.001). At the latest follow-up, the dominant curve group had a mean (and SD) congenital curve of 55±19 degrees and a compensatory curve of 73±24 degrees. The frequencies of wedge, hemivertebrae, and bar vertebral anomalies did not differ between groups. Conclusions: In pediatric patients with congenital scoliosis, a dominant compensatory curve was associated with vertebral anomaly at L4 or caudal, or T6 or cranial. These findings can help clinicians prioritize prophylactic treatment for patients who may be at high risk for developing a dominant compensatory curve. Level of Evidence: Level III.