Unfortunately, there really aren't any truly long-term outcomes study yet. Although I'm sure that some adjacent level problems were seen in the early post-op period for Harrington rods, I don't know that any of the surgeons really understood the depth of the issue. Although it appears that surgeons believe there won't be any BIG issues like there were with Harrington rods, there's no way to know for certain. We do know that there are still adjacent level issues, but surgeons are getting better at knowing what causes those issues, and are starting to compensate for them.

Anyway, here's a 3-year followup with a large cohort:

Spine (Phila Pa 1976). 2008 Jun 15;33(14):1598-604.
Operative treatment of adolescent idiopathic scoliosis with posterior pedicle screw-only constructs: minimum three-year follow-up of one hundred fourteen cases.

Lehman RA Jr, Lenke LG, Keeler KA, Kim YJ, Buchowski JM, Cheh G, Kuhns CA, Bridwell KH.

Walter Reed Army Medical Center, USA.

STUDY DESIGN: Preoperative review of a prospective study, single institution, consecutive series. OBJECTIVE.: To analyze the intermediate-term follow-up of consecutive adolescent idiopathic scoliosis (AIS) patients treated with pedicle screw constructs. SUMMARY OF BACKGROUND DATA: There have been no reports of the intermediate-term findings in North America following posterior spinal fusion with the use of pedicle screw-only constructs. METHODS: One hundred and fourteen consecutive patients having a minimum 3-year follow-up (mean 4.8 +/- 1.1; range, 3.0-7.3 years) with AIS were evaluated. The average age at surgery was 14.9 +/- 2.2 years. Radiographic measurements included preoperative (Preop), postoperative (PO), 2-year (2 years), and final follow-up (FFU). A chart review evaluated PFTs, Scoliosis Research Society scores, presence of thoracoplasty, Risser sign, Lenke classification, and complications. RESULTS: The most frequent curve pattern was Lenke type 1 (45.6%), followed by type 3 (21.9%). The average main thoracic curve measured 59.2 degrees +/- 12.2 SD Preop, and corrected to 16.8 degrees +/- 9.9 PO (P < 0.0001). Sagittal thoracic alignment (T5-T12) decreased from 25.8 degrees to 15.5 degrees at FFU (P = 0.05). Nash-Moe grading for apical vertebral rotation (AVR) in the proximal thoracic curve decreased from 2.0 Preop to 1.1 at FFU (P < 0.0001), and AVR in the thoracolumbar/lumbar spine decreased from 1.6 Preop to 1.1 at FFU (P < 0.0001). Importantly, the horizontalization of the subjacent disc measured -8.3 degrees Preop which decreased to -0.9 degrees PO (P < 0.001). PFT follow-up averaged 2.4 years with a 7.1% improvement in FVC (P = 0.004) and 8.8% in FEV1 (P < 0.0001). SRS scores averaged 83.0% at latest follow-up. Age, gender, Risser sign, or complications did not have a significant effect on outcomes. There were 2 cases of adding-on, 3 late onset infections, 1 with a single pseudarthrosis, but no neurologic complications. CONCLUSION: This is the largest (N = 114), consecutive series of North American patients with AIS treated with pedicle screws having a minimum of 3-year follow-up. The average curve correction was 68% for the main thoracic, 50% for the proximal thoracic, and 66% for the thoracolumbar/lumbar curve at final follow-up.