Should we bridge the cervicothoracic junction in long cervical fusions? A meta-analysis and systematic review of the literature

Abstract

OBJECTIVE Long posterior cervical decompression and fusion (PCF) is commonly performed to surgically treat patients with multilevel cervical pathology. In cases in which constructs may necessitate crossing the cervicothoracic junction (CTJ), recommendations for appropriate caudal fusion level vary in the literature. The aim of this study was to report the clinical and radiological outcomes of multilevel PCFs ending at C7 versus those crossing the CTJ. METHODS A systematic search of PubMed, CINAHL Plus, and Scopus was conducted to identify articles that evaluated clinical and radiological outcomes of long PCFs that ended at C7 (cervical group) or crossed the CTJ (thoracic group). Based on heterogeneity, random-effects models of a meta-analysis were used to estimate the pooled estimates and the 95% confidence intervals. RESULTS PCF outcome data of 1120 patients from 10 published studies were included. Compared with the cervical group, the thoracic group experienced greater mean blood loss (453.0 ml [95% CI 333.6–572.5 ml] vs 303.5 ml [95% CI 203.4–403.6 ml]), longer operative times (235.5 minutes [95% CI 187.7–283.3 minutes] vs 198.5 minutes [95% CI 157.9–239.0 minutes]), and a longer length of stay (6.7 days [95% CI 3.3–10.2 days] vs 6.2 days [95% CI 3.8–8.7 days]); however, these differences were not statistically significant. None of the included studies specifically investigated factors that led to the decision of whether to cross the CTJ. The cervical group had a mean fusion rate of 86% (95% CI 71%–94%) compared with the thoracic group with a rate of 90% (95% CI 81%–95%). Of patients in the cervical group, 17% (95% CI 10%–28%) required revision surgery compared with 7% (95% CI 4%–13%) of those in the thoracic group, but this difference was not statistically significant. The proportion of patients who experienced complications in the cervical group was found to be 28% (95% CI 12%–52%) versus 14% (95% CI 7%–26%) in the thoracic group; however, this difference was not statistically significant. There was no significant difference (no overlap of 95% CIs) in the incidence of adjacent-segment disease, pseudarthrosis, or wound-related complications between groups. CONCLUSIONS This meta-analysis suggests similar clinical and radiographic outcomes in multilevel PCF, regardless of inclusion of the CTJ. The lowest instrumented level did not significantly affect revision rates or complications. The ideal stopping point must be tailored to each patient on an individualized basis.