Cell blocks prepared from EBUS-TBNA material in our series contained diagnostic material in a third of the samplings and provided additional information to non-diagnostic smears, increasing the accuracy of EBUS-TBNA by a seven percent, to a diagnostic yield of 80%. Cell blocks obtained during EBUS-TBNA provided clinically significant information for one third of the patients participating in the study (30.7%), through accurate typing of the disease, identification of metastasis in the mediastinum, and, in patients with adenocarcinoma, EGFR genetic analysis in cell block samples.
With the development of novel treatments for NSCLC that have different degrees of efficacy and toxicity in NSCLC subtypes, an accurate pathologic classification has become essential. Most patients with NSCLC present with advanced non-operable disease and surgical biopsies allowing additional pathologic and genetic analyses are not available [13
]. The difficulties of pathologic diagnosis have increased with the emergence of minimally invasive procedures like EBUS-TBNA. This technique provides conventional smears for cytology that have a good correlation with histological diagnoses. Feller-Kopman and colleagues [14
] compared the cytological samples obtained by EBUS-TBNA with core biopsies or surgical excision samples in a series of 88 patients, finding that diagnoses were equivalent in most patients. Cell blocks can be obtained by means of EBUS-TBNA, and, compared with conventional smears, allow the performance of sections suitable for larger immunohistochemical staining batteries [15
]. When cell blocks prepared with EBUS-TBNA material are used for NSCLC subtyping, the adequacy of tumour tissue available for immunohistochemistry is a key issue [17
]. That topic can be easily managed when the recovered samples are subject to rapid on-site evaluation, as in our study; thus the immediate evaluation of the sample increases the diagnostic yield and decreases the need for unnecessary repeated diagnostic procedures [18
]. The on-site cytopathologist confirms the adequacy of the recovered material, minimizing the rate of unsatisfactory samples and requests for further sampling when additional material is needed for cell blocks. Following this approach 4 (6.3%) cases initially diagnosed as NSCLC-NOS on the conventional smear could be adequately subtyped in our study.
We found that over a 75% of the recovered cell blocks contained diagnostic cellular material, a percentage similar to those in other series where cell blocks from needle core biopsies have been processed [3
], but lower than the figure attained by conventional smears [18
]. Cell-block analysis achieved the diagnosis in 50 cases out of the 189 samples (26.4%) in which conventional smears were non-diagnostic in our study. Thus, with cell-block processing, the diagnostic yield of EBUS-TBNA rose from 72.9% to 80%. Twenty-one of these diagnostic cell blocks were from malignant nodes that would not have been diagnosed if the blocks had not been obtained and clinically implies that 7 patients were diagnosed of mediastinal metastases (N2/N3 disease) solely by the cell block analysis. We attribute this increase in the diagnostic yield mainly to the contribution of cell blocks to haematic non-diagnostic smears (Figure ). One of the obstacles that bronchoscopists and cytopathologists have to deal during an EBUS procedure is a vascularised node; these nodes are more likely to contaminate the samples with red blood cells. In this situation the on-site cytopathologist may not be able for a proper diagnosis of the slides. These aspirates, processed as cell blocks, can be examined later on the pathology laboratory and sometimes harbour clusters of lymphocytes or malignant cells. Other situation apart from blood contamination is nodes or masses containing necrotic material.
Figure 2 Fine needle aspiration from a subcarinal node in a case of metastatic adenocarcinoma. Non-diagnostic conventional smear (figure 2a and b) and cluster of adenocarcinoma cells in the the cell block (figure 2c and d). a: Unsatisfactory specimen. Blood cells (more ...)
Cell-block processing allowed for the performance of EGFR mutational analysis in 60% of our patients with a diagnosis of metastatic adenocarcinoma and in two of them confirmed the presence of an EGFR mutation, which confer sensitivity to the tyrosine kinase inhibitors gefitinib and erlotinib [20
]. These findings agree with the few smaller studies that have focussed on the ability of EBUS-TBNA to obtain samples for EGFR gene mutation screening [21
]. Nakajima and cols. [21
] used this approach in a series of 46 patients with adenocarcinoma, detecting 11 patients with EGFR mutations. García-Olivé and cols. [22
] found nodal metastasis by means of EBUS-TBNA in 36 patients from a series of 51 patients with this diagnosis; these authors recovered cell blocks that were adequate for EGFR analysis through EBUS-TBNA for most of their patients, and were able to identify mutations in two of them. Other cancer-related genetic mutations may also be predictive biomarkers, and their detection in TBNA samples might be useful for choosing a lung cancer therapy [1
]. In this new scenario our study confirms the value of cell-block processing of the material recovered from malignant nodes using EBUS-TBNA.
In summary, cell-block preparation is a simple method that provides important additional information after EBUS-TBNA in lung cancer. In our study, it was possible to preserve diagnostic material for cell blocks from more than a third of the performed aspirates. This material supplemented the information from conventional smears in a third of the cases and increased the diagnostic yield of the technique by a seven percent. Overall, cell-block processing provided clinically significant information for on third of the lung cancer patients, and allowed for the performance of genetic analyses of EGFR mutations in a half of the samples showing metastatic adenocarcinoma, confirming the advantages of this processing method for the diagnosis and staging of lung cancer.