This study therefore demonstrates that while high quality and high tumor content samples should be obtained and tested where possible, it is feasible to use low tumor content or cytology samples if these are the only sample available from the initial diagnosis of advanced NSCLC. Additionally, feedback from pathologists and molecular biologists on sample quality would help to minimize the costs of repeat testing and optimize the process of obtaining a quality result that the physician can take into consideration when making a treatment decision. The importance of ensuring

that samples are of sufficient quality/quantity has been confirmed in this study. The EGFR mutation frequency observed in the cytology samples implies that the pre-specified tumor content of 100 cells is still relevant Trichostatin A price within the clinical setting in order to avoid the issue of false-negative results in this sample type. In contrast, these data suggest that for histology sample analysis, it may be possible to reduce the criteria. Several groups have released recommendations for EGFR mutation testing practices which include guidance on good quality/quantity samples, Compound C but little guidance on how laboratories should deal with low tumor content or cytology samples [17], [18], [19] and [20].

Any samples used for diagnosis of NSCLC (e.g. biopsy, resection, Roflumilast cytology) should be tested for EGFR mutation status provided the laboratory performing the analysis is confident in the result. This confidence will depend on the method used, laboratory expertise, and the quality/quantity of the samples, typically those that contain sufficient tumor material to obtain an accurate

result, regardless of sample source. Testing of samples judged to be of low quality or low tumor content should be carried out using sensitive testing methods with or without a technique such as Laser Capture Microdissection (LCM), to enrich for the tumor cells. This technique was not attempted in IPASS, because while the technology is available in some institutions, it is not widely available and therefore not possible for all routine EGFR testing labs to employ. The Molecular Assays in NSCLC Working Group highlighted that LCM may be used to facilitate accurate test results by increasing the ratio of tumor to normal tissue, which is particularly important for techniques such as direct sequencing, which requires samples with ≥50–70% tumor cells for analysis [17]. However, the Working Group also noted that LCM can be laborious, and is unlikely to be acceptable for routine clinical sample analysis.