ABSTRACT
BACKGROUND: Multiplex gene-panel tests have recently been developed, including the Oncomine Dx Target Test multi-CDx system (ODxTT), and are commonly used to determine the adaptation of molecular-targeting drugs in non-small cell lung cancer. However, in actual clinical settings, we obtain false results owing to the small biopsy samples. We aimed to optimize tissue preparation methods to improve the success rate. PATIENTS AND METHODS: We investigated 88 biopsy samples. The area and nucleated cell count in the first cut section were quantified using a morphometric software. Pathological parameters, including "total tissue area" and "total nucleated cell count," were calculated by multiplying the total number of slides submitted to ODxTT. Optimal cutoff values to obtain the best success rate were also determined. Additionally, we morphometrically measured actual tumor cell proportions and attempted to determine the lower limit possible to detect mutations. RESULTS: Optimal cutoff values for "total nucleated cell count" and "total tissue area" were 132,885 and 32.94 mm2, respectively. The actual tumor cell proportions ranged from 4.6 to 97.7%. Even in cases with actual tumor cell proportions of less than 20% (ranging from 4.6 to 19.7%), there was no false negative. CONCLUSION: Thus, we proposed the pathological criteria for accurate ODxTT. Our result suggested that tumor cell proportions of less than 20% (around 5%) could be applicable for ODxTT. We hope that our results will help pathologists to choose between the multi-plex test (ODxTT) or single-plex test in routine diagnostics.