Computed tomography for preoperative evaluation of need for sternotomy in surgery for retrosternal goitre

The purposes of this study are to evaluate the usefulness of available CT classifications of retrosternal goitre (RSG) to identify patients needing sternotomy and to examine the effect of neck extension on goitre position.

From the Scandinavian Quality Register for Thyroid and Parathyroid Surgery, all patients treated for RSG at Sahlgrenska (January 2005 through August 2012) were identified. Medical records and preoperative CT scans were retrospectively reviewed. Paired CT (normal position/neck extension) was done in three patients.

Of 1698 patients undergoing thyroid surgery, 158 (9.3 %) were registered as having RSG, of these 38 were excluded (no preoperative CT n = 27, no RSG at preoperative CT n = 11). Of 120 included patients (71 % females, median age 67 years, rate of malignancy 14 %), 104 were managed with a cervical approach only, 16 (13.3 %) needed sternotomy, of these 13/16 had growth below the aortic arch concavity. Predictors for sternotomy were goitre extension below the aortic arch concavity (positive/negative predictive value (PPV/NPV) 54/97 %, sensitivity/specificity 81/89 %, odds ratio (OR) 36.6, p < 0.001); main mass of RSG to the right of the midline (PPV/NPV 21/95 %, sensitivity/specificity 81/53 %, OR 4.9, p < 0.008); and main mass of RSG retrotracheal (PPV/NPV 31/92 %, sensitivity/specificity 50/83 %, OR 4.8, p < 0.005). The goitre was displaced cranially a mean 11 mm with neck extension, but the relationship to the aortic arch was unchanged.

RSG extension below the aortic arch concavity was confirmed as a significant risk factor for sternotomy, with a NPV for sternotomy of 97 % for less extensive goitres. CT in neck extension provided no additional clinically relevant information.