Developing an effective quality evaluation strategy of next-generation sequencing for accurate detecting non-small cell lung cancer samples with variable characteristics: a real-world clinical practice

Purpose Next-generation sequencing (NGS) has been widely used in determining molecular profiling of non-small cell lung cancer (NSCLC). However, low-quality sequencing data may be generated with formalin-fixed and paraffin-embedded (FFPE) samples that have passed pre-sequencing quality control (QC). Therefore, quality evaluation of sequencing data is also critical for accurate tissue genotyping. Herein, we aimed to developed a grading QC algorithm, and provide a recommendation to refine and optimize NGS-based molecular diagnostic strategies. Methods We interrogated 1260 NSCLC samples using hybrid capture-based targeted DNA NGS, and quantified the sequencing data as high, medium and low quality, according to a grading QC algorithm. Then, we explored the relationship between sequencing quality and sample characteristics, and compared the concordance rates of results between NGS and conventional molecular tests for FFPE samples with variable characteristics. Results We found that high-quality data were associated with samples with shorter storage time and lower DNA degradation in resection samples, and were associated with intra-hospital samples, adequate DNA quantity, and lower DNA degradation in biopsy samples. Moreover, accurate NGS results can be achieved in samples with high-quality data, but not samples with medium-quality data, especially for rearrangements detection. Conclusion Our study demonstrates that the real-world clinical adoption of an effective QC strategy for NGS is necessary to ensure accurate results from FFPE samples of NSCLC with variable characteristics. Validation of actionable alterations by additional methods is highly recommended in cases with low QC score, particularly for the detection of rearrangements.