Values of different specimen preparation methods for the diagnosis of lung cancer by endobronchial ultrasound guided transbronchial needle aspiration

Abstract

Background Endobronchial ultrasound guided transbronchial needle aspiration (EBUS-TBNA) has been become an important procedure for the diagnosis and staging of lung cancer. Our research identified the effects of different pathological preparation on the diagnosis of lung cancer for specimens obtained by biopsy. Methods Patients were clinically considered if lung cancer was accompanied by mediastinal or hilar lymph node enlargement between March 2014 and November 2017. Specimens obtained by EBUS-TBNA were treated by three methods: traditional smear cytology, liquid-based cytology (LBC) and histopathology. Results Of a total of 154 puncture sites from 153 patients, the total positive rate of combination for the three pathological treatment types (histopathology, direct traditional smear, and LBC) was 77.3%. The diagnostic positive rate for histopathology was 68.6%, direct traditional smear was 65.6%, and LBC was 60.4%; there was no significant differences among the three single pathological treatment types (P = 0.29), but there was a statistically significant difference between the combination of three treatments and any single pathological treatment type (P = 0.01). The diagnostic sensitivities of histopathology combined with traditional smear and histopathology combined LBC were 94.4 and 92.8%, respectively, the specificities and PPVs were both 100%, and the diagnostic accuracies were 95.5 and 94.2%, respectively; the sensitivities, specificities and diagnostic accuracies above were all higher than those of single specimen treatment and lower than those of the three combined. Conclusion When EBUS-TBNA is used for the diagnosis and staging of lung cancer, the use of histopathological sections combined with direct cytological smear should be sufficient and is the most economical choice.