Time Dependent Response of the Unfused Lumbar Curve After Selective Thoracic Fusion for AIS: 2 -16 Year Radiographic and Clinical Follow-Up

Purpose: Selective thoracic fusion for King II/Lenke modifier ‘C’ curves remains controversial for patient selection, approach (asf vs.psf) and % thoracic correction to achieve overall balance In addition, the time dependent response of the lumbar curve (LC) and clinical outcomes have not been reported. The purposes of this study are to characterize radiographic changes of the unfused lumbar curve (LC) over time and to define the SRS-24 outcome for such procedures at 2-16 yr follow-up.

Methods: 40 consecutive pts with major thoracic (MT), compensatory ‘C’ modifier lumbar AIS curves underwent selective thoracic segmental instrumentation and fusion at a single institution (PSF:24, ASF:13, A+P:3) The lowest instrumented vertebra (LIV) was either T11(13), T12(14), or L1(13) and none were braced postop. Radiographs for all pts were analyzed at preop, 1wk, 2yr and latest f/u (n=30, mean=6.5 yrs, range=3-16 yrs). SRS 24 data was obtained at latest f/u (2-16 yrs).

Results: Selective fusion resulted in similar postop MT (preop:62 postop:34o) and LC (4832) Cobb measures that did not change with time. The LC apical translation from midline (AVT) at 1 wk was similar to preop, but sig at 2y (p=0.03) and long term f/u, (pre:29, post:30, 2yr:26, latest:24mm). LIV coronal tilt was sig postop (p<0.001), (26,16,16,16o) and was unchanged thereafter. The coronal LIV disc angle sig postop and continued to (8,5,4,3o). A sig in lordosis from T10-L2 occurred by 1 wk (p=0.01) and was subsequently maintained (4,1,2,1o). Surprisingly the LIV sagittal plumb (-8,-15,-6,-5mm) sig at 1 wk postop (p<0.05), but then sig to better than preop at 2y and long term f/u (p<0.05). The coronal C7 plumb was similar at each time frame (-18,-14,-16,-16mm). At latest f/u the C7 coronal plumb was 0-2 cm (16 pts), and 2-4cm (14pts). There were no cases of decompensation > 4 cm and no re-operations. Level of skeletal maturity (R1-3 vs 4-5) had no effect on changes of the LC over time.(p=0.63) ASF resulted in a sig more horizontal LIV and rectangular LIV disk, and better-maintained T10-L2 lordosis then PSF. SRS 24 data revealed 91% of pts responding would probably or definitely undergo the procedure again. SRS questionnaire scores were similar for patients with min 2 yr vs. 3-16 yr f/u and those with ASF vs. PSF.

Conclusions: Significant spontaneous improvement in lumbar curve parameters resulted after selective thoracic fusion, however the time course of improvement was varied and for some parameters occasionally followed a period of early worsening. Patient satisfaction was good irrespective of the approach used or level of skeletal maturity at the time of surgery.