ABSTRACT
INTRODUCTION: Historically, a non-small-cell lung carcinoma diagnosis, without pathologic subclassification, provided sufficient information to guide therapy. Evidence now demonstrates that pathologic subtype classification is central in selecting optimal treatment. This review aimed to identify factors associated with a specific pathologic diagnosis. METHODS: All nonoperative cases of non-small-cell lung carcinoma (NSCLC) referred to the medical oncology divisions of the Ottawa Hospital Cancer Centre (2008) and Princess Margaret Hospital, Toronto (2007-2010) were identified. The charts were reviewed for demographics, diagnostic methods, and final diagnosis. Logistic regression was performed to identify variables associated with a specific diagnosis. RESULTS: Of 739 patient records analyzed, 377 (51%) were men, 299 (40%) were aged over 70 years, and 510 (69%) had an Eastern Cooperative Oncology Group performance status of 0-2. Three hundred and eighty five (52%) of patients were diagnosed in a tertiary academic center. The lung primary was sampled in 503 (68%) of patients. Computed tomography-guided biopsy (n = 370, 50%) and bronchoscopy (n = 179, 24%) were the most common techniques. Four hundred and seventy seven (65%) of biopsies were cytologic specimens alone, and immunohistochemistry was performed in 337 (46%) of cases. The most common diagnoses were adenocarcinoma (n = 338, 46%), NSCLC not otherwise specified (n = 254, 34%), and squamous cell carcinoma (n = 115, 16%). Overall, 456 (62%) of patients received a specific pathologic diagnosis. Factors significantly associated with attaining a specific pathologic diagnosis were diagnosis outside an academic center (adjusted odds ratios [OR] 2.1 [95% CI, 1.41-3.14]; P = .0003), histologic laboratory samples (adjusted OR 1.58 [95% CI, 1.003-2.49]; P = .049), and immunohistochemical testing (adjusted OR 1.82 [95% CI, 1.25-2.70], P = .0021). CONCLUSIONS: A significant minority of patients with NSCLC do not receive a specific pathologic diagnosis. In an era of individualized medicine, this may potentially impact optimal clinical management.
