We read with great interest the article by Rodriguez et al.1 in which the authors present a study on mediastinal sampling using endobronchial ultrasound guided fine-needle capillary sampling (EBUS-FNC) instead of transbronchial needle aspiration (EBUS-TBNA), in which a negative pressure is applied via syringe. The authors report the interesting benefits of not applying suction to EBUS-guided puncture: the technique is simpler, procedure time is reduced, and adequate and representative material comparable to that of EBUS-TBNA may be obtained with less contamination from bleeding. However, a very important issue not highlighted in the article is the efficiency of the technique in obtaining cell blocks (CB). Madan and Guleria2 recently expressed concern regarding the possible differential yield when the histological core is obtained without the application of suction.

The preparation of CB from EBUS-TBNA samples is a simple way to provide additional information in the diagnosis of lung cancer. In our experience3 in lung cancer, CB obtained by EBUS-TBNA generally provided clinically relevant information in 83 of 270 patients (30.7%). Puncture was repeated up to three times and preservation of material for CB was possible in almost half of the aspirates. Of a total of 697 transbronchial aspirates, CB were available in 334 (47.9%). In 50 cases (7.2%), conventional extensions were not diagnostic while the CB were, and the malignancy diagnosed from extensions could be subtyped in four patients. Genetic analyses could also be performed from CB in 60% of patients with adenocarcinoma. CB samples were processed by air drying and coagulation on filter paper. For this, a blood sample is needed, and that is more likely to be obtained when negative pressure is applied with a syringe. Since the absence of suction produces less trauma, this technique is less likely to ensure a clot formation.

Several oncology societies4 have published guidelines recommending the necessary identification of molecular abnormalities for the initiation of specific treatment. Even today and certainly in the near future, EBUS will face the challenge of demonstrating its usefulness in the comprehensive analysis of all the newly available molecular studies5 in advanced neoplastic diseases, for which histopathological samples obtained by biopsy or surgical techniques have been routinely used.

It must be shown that useful CB can be obtained (even if they are prepared in liquid medium) by FNC instead of TBNA in an adequate proportion of cases, without increasing the number of punctures. A direct comparison is needed of the equivalence of the yield of the CB material obtained by each technique in terms of amounts (differences in cellularity) and quality for immunohistochemistry and molecular analyses, and not only the purely diagnostic yield. Until such evidence is available, we believe that FNC can be considered a useful alternative in cases where contamination by bleeding for immediate examination may render the sample inadequate.