Objectives: Pretreatment staging of patients with non-small cell lung cancer is critically important in determining an appropriate treatment plan. As positron emission tomography (PET) and computed tomography (CT) are proven complementary modalities in clinical staging, recent advances in PET technology have brought forth integrated PET/CT as the new standard. We tested the hypothesis that improvements in PET technology have not increased the sensitivity or specificity of PET in the staging of non-small cell lung cancer to an extent that surgical staging is no longer required. Methods: This is a retrospective, single-institution review of 336 patients from 1995 to 2005 with biopsy-proven non-small cell lung cancer who underwent [18F] fluoro-2-deoxy-d-glucose-PET before mediastinal lymph node sampling by cervical mediastinoscopy or thoracotomy. Clinical records, histopathologic reports, and PET findings were reviewed. Data were analyzed by the Pearson χ2 test. Results: Within the study population, 210 patients had routine PET and 126 had integrated PET/CT. For detecting mediastinal metastases the sensitivities of PET versus integrated PET/CT were 61.1% versus 85.7% (P < .05), specificities were 94.3% versus 80.6% (P < .001), positive predictive values were 68.8% versus 55.8%, negative predictive values were 92.1% versus 95.2%, and overall accuracy was 88.6% versus 81.7%. Conclusions: Improvements in PET technology have increased integrated PET/CT sensitivity at the cost of significantly decreased specificity. Although it may appear that integrated PET/CT incurs fewer false negative results, the dramatic increase in false positive results reinforces the notion that integrated PET/CT should be used only as an adjunct to clinical staging and that surgical staging remains the gold standard in non-small cell lung cancer.
ASJC Scopus subject areas
- Cardiology and Cardiovascular Medicine