Factors involved in the decision to perform a selective versus nonselective fusion of Lenke 1B and 1C (King-Moe II) curves in adolescent idiopathic scoliosis.

STUDY
DESIGN: A retrospective evaluation of 203 adolescent idiopathic scoliosis patients with Lenke 1B or 1C (King-Moe II) type curves.
OBJECTIVES: To evaluate the incidence of inclusion of the lumbar curve in the treatment of this type of deformity as well as radiographic factors associated with lumbar curve fusion. SUMMARY OF BACKGROUND DATA: In patients with structural thoracic curves and compensatory lumbar curves, many authors have recommended fusing only the thoracic curve (selective thoracic fusion). Studies have shown that correction of the thoracic curve results in spontaneous correction of the unfused lumbar curve; however, in some cases, truncal decompensation develops. Though there have been various attempts to define more accurately what type of curve pattern should undergo selective fusion, controversy continues in this area.
METHODS: Measurements were obtained from the preoperative standing posteroanterior and side-bending radiographs of 203 patients with Lenke Type 1B or 1C curves from five sites of the DePuy AcroMed Harms Study Group. Patients were divided into two groups depending on their most distal vertebra instrumented: the "selective thoracic fusion" group included patients who were fused to L1 or above and the "nonselective fusion" group included patients fused to L2 or below. A statistical comparison was conducted to identify variables associated with the choice for a nonselective fusion.
RESULTS: The incidence of fusion of the lumbar curve ranged from 6% to 33% at the different patient care sites. Factors associated with nonselective fusion included larger preoperative lumbar curve magnitude (42 +/- 10 degrees vs. 37 +/- 7 degrees, P < 0.01), greater displacement of the lumbar apical vertebra from the central sacral vertical line, (3.1 +/- 1.4 cm vs. 2.2 +/- 0.8 cm, P < 0.01), and a smaller thoracic to lumbar curve magnitude ratio (1.31 +/- 0.29 vs. 1.44 +/- 0.30, P = 0.01).
CONCLUSIONS: The characteristics of the compensatory "nonstructural" lumbar curve played a significant role in the surgical decision-making process and varied substantially among members of the study group. Side-bending correction of the lumbar curve to <25 degrees (defining these as Lenke 1, nonstructural lumbar curves) was not sufficientcriteria to perform a selective fusion in some of these cases. The substantial variation in the frequency of fusing the lumbar curve (6% to 33%) confirms that controversy remains about when surgeons feel the lumbar curve can be spared in Lenke 1B and 1C curves. Site-specific analysis revealed that the radiographic features significantly associated with a selective fusion varied according to the site at which the patient was treated. The rate of selective fusion was 92% for the 1B type curves compared to 68% for the 1C curves.