Staging of lung cancer. do we need a diagnostic ct of the brain after an integrated pet/ct for the detection of brain metastases?

Brain CT has been recommended in staging of patients with lung cancer because of its usefulness in the detection of metastases. Purpose of this study is to examine if a diagnostic brain CT (CT,) can be obviated when an integrated PET/CT (PET/CT) is available. 87 consecutive patients underwent a diagnostic brain CT and a whole-body PET/CT within a period of 3 weeks to stage a known primary tumour. CT examinations were evaluated by two experienced neuroradiologists on the detection of brain lesions (benign and malignant). The results of PET/CT and CT reading were compared and both readings were compared with the clinical results. Statistical analysis was done by measuring sensitivity, specificity, PPV, NPV and accuracy. The relative accuracies were compared by a McNemar (exact) test for correlated proportions. Considering the CT, as standard of reference, sensitivity, specificity, PPV, NPV and accuracy for the brain CT of PET/CT (CT2) and PET/CT were respectively 83%, 96%, 77%, 97%, 94% and 69%, 98%, 90%, 95%, 94%. Considering the clinical diagnosis as standard of reference these figures were for CT1, CT2 and PET/CT respectively 80%, 100%, 100%, 96%, 96% and 66%, 95%, 77%, 93%, 90% and 66%, 97%, 83%, 93%, 91%. There was no statistical difference between CT1 and CT2. The comparison of the additional CT in PET/CT with a diagnostic CT of the brain did not yield a statistical difference in the detection of brain lesions despite the inferior quality of the CT component of PET/CT. A diagnostic brain CT can be obviated when a PET/CT is available.