Shoulder balance in Lenke type 2 adolescent idiopathic scoliosis: Should we fuse to the second thoracic vertebra?

Objectives: There are many different systems recommending upper instrumented vertebra (UIV) for Lenke type 2 adolescent idiopathic scoliosis (AIS), several of which suggest that all Lenke type 2 AIS patients should be fused to the second thoracic vertebra (T2). However, all previously proposed UIV selecting systems do not accurately predict postoperative shoulder balance. We investigated whether fusing to T2 could prevent postoperative shoulder imbalance and identified circumstances under which to fuse up to T2.

Patients and methods: We retrospectively collected all patients with typical Lenke type 2 AIS who received surgery by one spine surgeon in our hospital from 2010 to 2014. Lateral shoulder balance was assessed utilizing radiographic shoulder height difference (RSH), coracoid height difference (CHD), clavicle-rib intersection difference (CRID), and clavicle angle (CA). Medial shoulder balance was assessed by T1 tilt angle and first rib angle (FRA). Lateral shoulders were considered to be level if the absolute value of RSH was less than 10 millimeters. All patients were divided into two groups as follows: 1) T2 group: UIV of T2 (n=49); and 2) below-T2 group: UIV of T3 (n=24) or T4 (n=6). Patients were assessed before surgery and at final follow-up with a minimum follow-up duration of 24 months.

Results: Seventy-nine typical Lenke type 2 AIS patients were identified. Preoperative CHD and CA were significantly associated with postoperative lateral shoulder imbalance (both p=0.045), whereas the UIV level was not significantly associated with it. Both fusing to T2 and to below T2 could improve RSH (p<0.001 and p=0.001, respectively). Fusing to T2 slightly worsened CHD, CRID, and CA at last follow-up (all p<0.001), while fusing to below T2 improved these lateral shoulder balance parameters (p=0.042, p<0.001, and p=0.007, respectively). For medial shoulder balance, fusing to below T2 worsened T1 tilt angle and FRA at last follow-up (p=0.025 and p<0.001, respectively), while fusing to T2 effectively kept these medial shoulder parameters in balance. In addition, for patients with an elevated left border of T1, the T2 group had worse preoperative T1 tilt angle but gained better postoperative T1 tilt angle than the below-T2 group (p<0.001 and p=0.040, respectively).

Conclusion: Preoperative lateral shoulder balance, more so than the UIV level, can strongly influence postoperative lateral shoulder balance. Fusing to T2 can only effectively improve medial shoulder balance, not lateral shoulder balance (CHD, CRID, and CA). Moreover, a positive T1 tilt angle is an indicator for fusing to T2 to improve medial shoulder balance.