Distraction-Based Surgeries Increase Spine Length for Patients With Nonidiopathic Early-Onset Scoliosis-5-Year Follow-up.

STUDY DESIGN

Retrospective, comparative.

OBJECTIVES

To determine if distraction-based surgeries will increase spine length in patients with nonidiopathic EOS and whether etiology affects final spine length.

SUMMARY OF BACKGROUND DATA

As early-onset scoliosis (EOS) has many etiologies, it is unclear whether etiology affects the spine length achieved with distraction-based surgeries. Since distraction may produce kyphosis, sagittal spine length (SSL; curved arc length of the spine in the sagittal plane) should be utilized.

METHODS

Patients with nonidiopathic EOS treated with distraction-based systems (minimum 5-year follow-up, 5 lengthenings) were identified from two EOS registries. Radiographic analysis preoperation, postimplant (L1), and after each lengthening (L2-L5, L6-L10, L11-L15) was performed with primary outcome of T1-S1 SSL.

RESULTS

We identified 126 patients (67 congenital, 38 syndromic, 21 neuromuscular) with a mean preoperative age of 4.6 years, scoliosis 75°, kyphosis 48°, and a mean of 12 lengthenings. After initial correction (p < .05), scoliosis remained constant (58° at L11-L15) and kyphosis increased (38° at L1 to 60° at L11-L15) (p < .05). SSL increased for the entire group from 27.1 cm preoperation to 35.3 cm at L11-L15 (p < .05) and during the distraction phase (29.2 cm at L1 to 35.3 cm at L11-L15) (p < .05). Preoperative SSL was higher in neuromuscular compared with congenital patients and maintained that difference until the 10th lengthening. Preoperative SSL did not differ between syndromic and congenital patients (28.0 cm vs. 25.6 cm); however, syndromic patients had greater SSL after implantation (L1: 30.5 cm vs. 26.8 cm) (p < .05) and maintained that difference until the 15th lengthening (37.1 cm vs. 34.3 cm) (p < .05).

CONCLUSION

At minimum 5-year follow-up, distraction-based surgeries increased spine length for all patients with nonidiopathic EOS; however, neuromusculars had higher preoperative spine length compared with congenital patients and maintained that difference until the 10th lengthening. Although congenital and syndromic patients had similar preoperative spine length, syndromic patients had greater SSL after implantation (L1) and maintained that difference until the 15th lengthening.

LEVEL OF EVIDENCE

Level III.