Behavior of the un-instrumented lumbar curve following exclusive tethering of the thoracic region

Anterior vertebral body tethering (AVBT) has recently been utilized as a surgical alternative to posterior spinal fusion for a subset of pediatric scoliosis patients. Indications for AVBT are evolving and, while early results have been promising, there is a paucity of literature examining the behavior of the lumbar curve after exclusive tether of the thoracic region. It was hypothesized that thoracic tether leads to a spontaneous decrease in the un-instrumented lumbar curve.

The study population consisted of 166 patients with idiopathic scoliosis enrolled in the Pediatric Spine Study Group registry undergoing thoracic tether with a minimum of two-year follow-up. Exclusion criteria included: patients with non-idiopathic scoliosis, patients with prior spine surgery, and patients instrumented below L1.

Overall curve correction was notable, with mean pre-op, immediate post-op, and two-year follow-up angles of 51.3°, 29.7°, and 30.3° respectively for the thoracic curve and 32.7°, 22.9°, and 24.1° respectively for the un-instrumented lumbar curve. Overall, 124 subjects (74.7%) had a decrease in lumbar curve > 5° immediately post-op. Over a two-year follow-up period, 32 subjects (19%) had a continued decrease in lumbar curve > 5°, 91 subjects (54%) had minimal change, and 43 subjects (26%) had an increase in lumbar curve > 5°. In a subgroup analysis of 36 subjects with a decrease in thoracic curve > 5° from post-op to two-year follow-up, 11 subjects (31%) had a concomitant decrease in lumbar curve with only 5 (14%) showing an increase in lumbar curve > 5°. The changes in lumbar curve from pre-op to post-op and from post-op to two-year follow-up were found to be associated with changes in the thoracic curve for the same periods (rho = 0.603, p < 0.001; rho = 0.413, p < 0.001 respectively). When considering Lenke lumbar modifiers, the un-instrumented lumbar curve corrected an average of 35%, 27%, and 20% following surgery and 27%, 30%, and 17% at two-year follow-ups for A, B, and C curves respectively (p < .001 for all data points compared to pre-op).

Lumbar curves tended to mirror the behavior of the maximal thoracic curve in terms of correction or decompensation both during surgery and during the two-year follow-up. Furthermore, un-instrumented lumbar curves with a Lenke C modifier tended to achieve a lower percentage correction. These data provide clearer insights into the response of the lumbar curve following thoracic tethering and the effect of growth modulation.