Integrated PET/CT in non-small cell lung cancer staging----Clinical and pathological agreement

Abstract Introduction Integrated PET/CT has become a fundamental tool in the preoperative assessment of non-small cell lung cancer (NSCLC) providing useful anatomical and metabolic information to characterize tumoral lesions and to detect unsuspected metastatic disease. Aim To compare the agreement between clinical and pathological staging before and after the use of PET/CT. Material and methods Retrospective study of patients with NSCLC who underwent potentially curative surgery throughout 10.5 years. Cohen's kappa coefficient was used to evaluate staging agreement. Results One hundred and fifty patients were evaluated, 78% males, with a mean age of 65 (±9.6) years. Thirteen percent were submitted to neoadjuvant chemotherapy. PET/CT was performed in 41%. Global agreement between clinical and pathological staging was 51% (kappa = 0.3639). There was a statistically significant difference between the staging results in patients who underwent PET/CT, when compared to the subgroup who did not ( p  = 0.003). For those with PET/CT false negatives occurred in less 39%, false positives in more 12% and clinical and pathological staging coincided in more 27%. The overall results reflected an improvement in the agreement between clinical and pathological staging in the PET/CT subgroup (67%, kappa = 0.5737 vs 40%, kappa = 0.2292). PET/CT accuracy was enhanced when patients re-staged after neoadjuvant therapy were excluded and a substantial staging agreement was obtained for those who had the exam only for staging purposes (73%, kappa = 0.6323). Conclusion Inclusion of PET/CT in NSCLC preoperative assessment improved the accuracy of the clinical staging, with a good level of agreement with pathological staging.