Preoperative magnetic resonance imaging screening for a surgical decision regarding the approach for anterior spine fusion at the cervicothoracic junction.

Abstract

In a study investigating the correlation between a set of designed criteria and judgments of surgical experience, 100 cervical magnetic resonance images from different patients were used. To demonstrate reliable and reproducible anatomic measurements that can aid spine surgeons in selecting surgical approaches for anterior spine fusion in the cervicothoracic region. Surgical approaches to the cervicothoracic junction vary among surgeons. Whereas sternotomy provides maximum exposure, less extensive approaches are preferred to minimize surgical trauma, provided surgical goals are not compromised. No quantitative criteria currently exist to determine before surgery the least invasive surgical approach for sufficiently exposing pertinent anatomy. Thirteen geometric variables designed to be clinically practical and to expose important anatomic relations were used to evaluate 100 sagittal scout cervical magnetic resonance image sequences. An experienced spine surgeon independently rated each image for the most appropriate surgical approach to the C5-T2 region. The ratings were tested for interrater reliability using a second spine surgeon. After testing for interrater and intrarater reliability, the geometric measurements were correlated with the surgeon's selected surgical approaches for each intervertebral segment (P < 0.05). Instrument manubrial thoracic distances, reflecting standardized heights of intervertebral discs above or below the superior tip of the manubrium, were the most reliable, reproducible, and correlative with the choice of surgical approach. All the measurements but one, the instrument manubrial thoracic distance for T1/T2, demonstrated interrater and intrarater reliability, with an interclass correlation of at least 0.70. The primary surgeon-investigator indicated the anterior approach with sternotomy (n = 3) or the transverse cervical approach (n = 97) for the C7/T1 exposure, and the anterior approach with sternotomy (n = 43) or the transverse cervical approach (n = 57) for the T1/T2 exposure. The interrater questionnaire reliability results indicated statistical agreement between the primary surgeon-investigator and the second cervical spine surgeon at all vertebral segments evaluated. Instrument manubrial thoracic distances showed the strongest significant correlation with the surgical approach, demonstrating a statistical power of 1. For the C7/T1 exposure, the instrument manubrial thoracic distance for C7/T1 was 1.9 +/- 2 cm (95% confidence interval [CI] = 1.41 to 2.22) for the transverse cervical approach, and -3.3 +/- 1.3 cm (95% CI = -4.79 to -1.75)] for the anterior approach with sternotomy. The instrument manubrial thoracic distance measurements for C5/C6, C6/C7, and T1/T2 also showed nonoverlapping 95% confidence intervals for the transverse cervical versus the anterior approach with sternotomy for the C7/T1 exposure. For the T1/T2 exposure, all four instrument manubrial thoracic distance measurements again showed statistically significant differences between approaches, with nonoverlapping 95% confidence intervals and a statistical power of 1. In addition, the measurements elaborating the anterior-to-posterior distance of the thoracic outlet and the measurements of the angle between the planes of the intervertebral disc and the sternum also showed statistically significant differences between approaches for the T1/T2 segment, with a statistical power of at least 0.9. Strong correlations exist between objective measurements and the choice of surgical approach for anterior spine fusion. Among investigated anatomic relations, the instrument manubrial thoracic distance correlated most reliably with the surgeons' choice of the anterior approach. Such objective measurements represent tools that cervical spine surgeons can use to determine the surgical approach.