Factors Influencing the Diagnostic Accuracy of Identifying the Histologic Type of Non–Small-Lung Cancer With Small Samples

Abstract Background The pathologic diagnosis has become a greater consideration in decision-making regarding the treatment options for lung cancer. Therefore, the accurate diagnosis of the tumor histologic type is essential, even when only small biopsy or cytology samples are available. However, the risk of a misdiagnosis with smaller biopsy samples is greater. The factors underlying the increased risk of a misdiagnosis in small samples are unknown. The aim of the present study was to identify the clinical and pathologic factors (other than immunohistochemical staining) that influence the pathologic diagnostic accuracy in small biopsy and cytological lung samples obtained by bronchoscopy. Patients and Methods We performed transbronchial lung biopsy or brushing and lavage to determine the preoperative diagnosis of 126 of 299 surgically resected lung cancer specimens. We assessed the diagnostic accuracy of the preoperative transbronchoscopic examination findings against that of the surgically resected lung specimens. Results On univariate analysis, the mean pathologic tumor size in the noncorresponding cases was larger than that in corresponding cases. Vascular invasion was also more prevalent in the noncorresponding cases. The tumor differentiation grade in the noncorresponding cases was poorer than in the corresponding cases. The noncorresponding cases were at a more progressed stage. On multivariate analysis, the pathologic tumor size and tumor differentiation grade were associated with the noncorresponding cases. Conclusion We found a larger tumor size and poor differentiation grade were indicative of lung cancer tissue with a greater content of heterogeneous cells. Therefore, a possibility exists of a false diagnosis using only these factors. Thus, treatment decisions should be made considering the pathologic diagnosis and other relevant factors.