<h3>BACKGROUND CONTEXT</h3> Add-on phenomenon following posterior surgery for Lenke 1A curves has been studied and lowest instrumented vertebra (LIV) selection, skeletal immaturity, and L4 right vertebral tilt have been identified as risk factors. To minimize the risk, LIV has been recommended to be set at the last substantially touched vertebra (LSTV); however, add-on phenomenon has been observed even when LIV was set at LSTV and immediate postop X-rays showed minimal residual curve. To study the possible causes in this scenario, we can keep LIV consistently at or immediately adjacent to LSTV and compare characteristics between curves with and curves without add-on phenomenon. <h3>PURPOSE</h3> To study risk factors for add-on phenomenon while keeping LIV consistently at or immediately adjacent to LSTV. <h3>STUDY DESIGN/SETTING</h3> Case control study, single institution. <h3>PATIENT SAMPLE</h3> We identified 21 consecutive patients who had Lenke 1A curves, preop upper thoracic bending X-rays, and minimum of 2 year follow-up from patients who underwent posterior scoliosis surgery at a single institution from 2006 to 2017. <h3>METHODS</h3> We used billing records to identify patients who underwent posterior scoliosis surgery from 2006 to 2017 at a single institution. From the 21 patients who met our criteria, we obtained age at presentation and at time of surgery from medical records. The main thoracic curve (MT), proximal thoracic curve (PT), and PT bend out Cobb angle, % PT correction on bend out = (PT - PT on bend out)/PT, Risser stage, and L4 tilt were determined from preop X-rays. Upper instrumented vertebra (UIV) was either T3 or T4. LIV was at or immediately adjacent LSTV such that if LSTV was: a) T12, LIV was primarily L1 (LSTV+1); b) L1, LIV was L1; c) L2, LIV was L1 if L1 was neutral on bending xrays (LSTV-1) and L2 if not (LSTV); d) L3, LIV was L2 (LSTV -1). Residual curve, lowest instrumented vertebra adjacent disc angle (LIVDA), and lumbar curve were measured in first (avg 2 months) and last follow-up (avg. 40 months). Add-on phenomenon was defined as (1) an increase in the residual Cobb angle of at least 10°, (2) an increase in the lumbar curve of 5 ° (3) increase in LIVDA of greater than 5 °. <h3>RESULTS</h3> The residual curve progressed more than 10 ° in 3 patients between two follow-ups, 1/3 of whom had progression of LIVDA of greater than 5 ° and 2/3 of whom had progression of the lumbar curve of more than 5 °. These 3 patients were the add-on group (AG). None of the other 18 patients met any of the criteria for add-on phenomenon and were the non-add-on group (nAG). There were no significant differences between the two groups in the age at the time of presentation or surgery, Risser stage, preoperative MT and PT. One out of 3 patients in AG and 10/18 patients in nAG had L4 tilt to the right (p=0.59). LIV=LSTV-1 did not appear to be a risk factor (p=0.38); all 8 of those patients were in the nAG while 3 patients in AG had LIV either at LSTV or LSTV+1. PT bend out Cobb angles were statistically larger (25.7o +/- 2.1o, range 24 to 28, versus 10.8o +/- 5.5 °, range 0 to 21, p=0.0002) and the % PT correction with bend out was statistically lower (5.0 +/- 2.7%, range 3 to 8, versus 53 +/- 19%, range 25 to 100, p=0.0004) in the AG compared to the nAG. <h3>CONCLUSIONS</h3> We demonstrated an association between add-on phenomenon and limited correction of PT on bend out in Lenke 1A curves where the LIV was primarily at or immediately adjacent to LSTV. If % PT correction with bend out is less than 10%, the surgeon may consider treating the curve as a Lenke 2A curve and set UIV at T1 or T2. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.
Read full abstract