Ishida, Wataru, Ramhmdani, Seba, Perdomo-Pantoja, Alexander, Elder, Benjamin D., Holmes, Christina, Theodore, Nicholas, Gokaslan, Ziya L., Wolinsky, Jean-Paul, Sciubba, Daniel M., Bydon, Ali, Witham, Timothy F., and Lo, Sheng-fu L.
BACKGROUND CONTEXT Occipitocervical fusion is the gold standard stabilization method for several pathologies involving craniovertebral junctions, including basilar invagination, atlantoaxial trauma and neoplasm. However, its postoperative clinical course is oftentimes complicated by dysphagia and dyspnea, presumably due to loss of mobility across occiput-C1-C2 joints, which could lead to aspiration pneumonia and acute airway obstruction, respectively, and are life-threatening complications in this procedure. The exact mechanism and its predictors still remain widely unstudied, but a previous article involving healthy volunteers elaborated on the importance of the occipital and external acoustic meatus to axis angle (O-EAa, "the angle formed by the McGregor line and the line connecting the external acoustic meatus and the midpoint of the inferior end plate of C2"), when predicting the narrowest oropharyngeal airway space (nPAS, "the narrowest anterior–posterior distance from the posterior pharyngeal wall to the back of the tongue between the levels of the uvula tip and the epiglottis tip") on X-rays, which theoretically should directly correlate with postoperative dysphagia. Here, we aimed to evaluate whether pre- and postoperative radiographical measurements such as O-EAa and nPAS could predict postoperative dysphagia in adult patients who underwent occipitocervical fusion procedures. PURPOSE To identify predictors of postoperative dysphagia in patients with adult occiptocervical fusion surgery. STUDY DESIGN/SETTING Retrospective, single-center. PATIENT SAMPLE Patients who underwent occipitocervical fusion at a single institution. OUTCOME MEASURES Prolonged postoperative dysphagia (more than 30 days). METHODS Single-center, retrospective data review from 2010 to 2016 identified 57 patients who underwent spine surgery involving occipitocervical fusion procedures. Among those, 51 patients had more than one-year follow-up as well as an adequate postoperative assessment of swallowing functions. Clinical records of those 51 patients (average age: 53.3, female: 54%) were collected and statistically analyzed. A total of 16 patients (31.4%, group (A)) were diagnosed with postoperative dysphagia associated with occipitocervical fusion, who were compared with 35 patients without this complication (group (B)) in terms of baseline characteristics, operative data and radiographical findings including perioperative O-EAa and nPAS at patients' neutral positions. Intergroup comparison of binary variables was performed via Fisher's exact test. Intergroup comparison of continuous variables was achieved using unpaired t -tests. ROC curves were drawn to evaluate the diagnostic value of O-EAa and nPAS for postoperative dysphagia. Furthermore, the Pearson's correlation was assessed between O-EAa and nPAS. Finally, a multilinear regression model was created to identify factors influencing nPAS. All reported p values are two-sided and p values <.05 were regarded as statistically significant. RESULTS There were no statistically significant difference in terms of baseline characteristics including age, sex, BMI, smoking and comorbidities or operative data such as estimated blood loss, operative time and operated levels. While preoperative O-EAa and nPAS were similar between the two groups (P=.39 and.55, respectively), postoperative O-EAa ((A) 89.3° versus (B) 104.5°, p<.01) and nPAS ((A) 9.9 mm versus (B) 13.6 mm, P=.01) as well as perioperative changes in O-EAa ((dO-EAa), (A) −4.1° vs. (B) 0.6°, P=.01) and nPAS (dnPAS, (A) −2.0 mm vs. (B) 0.2 mm, P=.01) were significantly different. The cut-off value of dOEA-a <−2.4° to predict postoperative dysphasia yielded sensitivity of 81.8% and specificity of 100% (AUC=0.939), which was comparable to dnPAS >−1.1 (sensitivity 81.8%, specificity 100%). The Pearson's correlation coefficient between these two parameters were 0.747 (p<.001). In a linear regression analysis, nPAS (mm) was found to be affected by O-EAa (estimate=0.2, P=.04) and sex (estimate=−4.8 (female=1), P=.03), but not by BMI in this model (R square=0.42). CONCLUSIONS Perioperative changes in O-EAa and nPAS were associated with postoperative dysphagia. Intraoperatively, O-EAa measurement on fluoroscopic images could be utilized as a surrogate marker for nPAS, thereby potentially allowing us to decrease the risk of this crucial complication. These data may prompt multicenter, prospective studies to further validate them and translate them into better patient care in the future. [ABSTRACT FROM AUTHOR]