Introduction Differentiation and accurate classification of NSCLC (Squamous, Adenocarcinoma, Large cell carcinoma) is crucial in determining the prognosis and selecting targeted chemotherapy regimens. However, it is not always possible to subtype the tumours particularly if the biopsy samples are small and such undifferentiated tumour is referred as NSCLC not otherwise specified (NOS). It has been shown that 25% of bronchial biopsy specimens and 40% cytological specimens result in a diagnosis of NSCLC-NOS. However, the frequency of NSCLC-NOS with EBUS-TBNA samples is not known.

Methods We looked at the cytology reports of all patients with an EBUS-TBNA diagnosis of NSCLC over a period of 13 months. In patients with a diagnosis of NSCLC-NOS, we obtained further information on the details of the EBUS procedure and the cytological methods used.

Results Of the 243 patients who underwent EBUS-TBNA, 78 with a diagnosis of NSCLC were included. A confident initial cytological sub typing of NSCLC was possible in 68 (87%). Analysis of the remaining 10 patients with a diagnosis of NSCLC-NOS showed that biopsies taken from the lymph nodes were deemed adequate for cell block and immunohistochemistry (IHC) in all but one patient. Despite this, IHC was performed on 3 out of 9 samples. IHC was able to subtype the tumour in these cases. The Haematoxylin and Eosin (HE) and IHC profile of the 10 patients are shown in Abstract S39 Table 1.

Conclusion Thus we have shown that adequate tissue samples can be obtained at EBUS-TBNA and the frequency of NSCLC-NOS is less (7/78=9%) compared to the histological bronchial biopsy samples. In cases, where morphological sub typing of NSCLC on HE is not possible, immunohistochemistry should be performed.