Learn. Share. Grow.

Title: Complications and Radiographic Outcomes of Posterior Spinal Fusion and Observation in Patients Who Have Undergone Distraction-Based Treatment for Early Onset Scoliosis

Authors: Jeffrey R. Sawyer, MD, Rodrigo GóesMedéa de Mendonça, MD, Tara S. Flynn, BA, Amer F. Samdani, MD, Ron El-Hawary, MD, Alan J. Spurway, MASc, John T. Smith, MD, John B. Emans, MD, Tricia A. St. Hilaire, MPH, Stephen J. Soufleris, BS, Ryan P. Murphy, BS, Children’s Spine Study Group

Journal: Spine Deform. 4(1):407-412

Date: January 01, 2016

Excerpt: This study reports on patients who are treated using observation or “watchful waiting” as opposed to posterior spinal fusion at the end of growth friendly surgical treatment.

Keywords: Complications, Early Onset Scoliosis, Growth Friendly Surgery, VEPTR, Infection


What is the purpose of this study?

To compare surgical and radiographic outcomes of early-onset scoliosis (EOS) patients treated with growth-friendly surgical treatments who stopped lengthening for >2 years without additional surgery to those who had posterior spinal fusion (PSF) at the end of lengthening.



  • 37 patients (21 females, 16 males) were studied: 12 (32%) were in the observation (OBS) and25 (68%) in the PSF group.
  • The OBS group obtained 88% of the T1 - T12 height obtained by the PSF groupwithout complications.
  • The PSF group had a slightly greater coronal Cobb angle (59_ vs. 51_ ) and maximal kyphosis (52_ vs. 50_ ) at the end of distraction, which were not statistically significant. Although there was some correction of the coronal Cobb angle (10_ , 17%) and maximal kyphosis (8_ , 15%) following PSF, the differences between the two groups were not statistically significant at final follow-up

Twenty-six complications occurred in 15 of the 25 patients in the PSF group (mean 1.73 complications per patient); 9 patients had 1 complication, 3 had 2 complications, 1 had 3 complications, and 2 had 4 complications(Table 4 ). Six patients (24%) required additional procedures after PSF, including 4 implant removals and 2 repeat/revisions, one 3 years after fusion and the other 8 months after fusion. The only complication in the OBS group was pain at latest follow-up in 1 patient.

Although a PSF has been considered the ''final'' procedure at the end of distraction, it has been shown that it may not be the last surgical procedure. In our series, at a mean follow-up of 9 years, 15% of patients with PSF have already required an additional surgical procedure.


Important Discussion Points

PSF after distraction-based treatment in EOS patients has been shown to have a small amount of curve correction and an increased frequency of complications compared to PSF in AIS patients

The decision to perform a spinal fusion at the end of distraction-based treatment is complex and multifactorial based on patient, family, and surgeon-specific factors. Factors such as underlying etiology, curve magnitude, implant and/or anchor failure, previous infection, and patient comorbidities all influence surgical risks and outcomes and should be considered. It also has been shown that patients who have multiple surgical procedures and their families have increased psychosocial stress and decreased quality of life [5,6] . PSF after lengthening also may be a difficult procedure because of altered spinal landmarks and spine stiffness.

The risk of complications from long-term observation, such as implant pain or failure, curve progression, and late infection, remains unknown.

Autofusion has been described in up to 81% of patients, with 62% having a completely autofused spine following treatment with either rib- or spine-based distraction methods

The choice of PSF may be biased by patients who required PSF because of factors such as coronal and/or sagittal imbalance, implant problems, anchor failure, and/or pain. It may be that the observation group in this study represents a set of patients who have unique patient and curve characteristics, such as solid anchors, good coronal and sagittal balance, adequate bone mineral density, and autofusion that provide long-term stability. Although factors such as age, sex, C-EOS classification, and coronal/sagittal curve magnitude at the end of lengthening were similar in these groups, these findings highlight the need for further study to identify other factors that may predict success with long-term observation.

It also is impossible to know whether patients treated with observation will in the future need to have a PSF; however, we have shown our patients to be stable at a mean of 4 years from last distraction.