Factors affecting the success of next-generation sequencing in cytology specimens.

Abstract

The use of cytology specimens for next-generation sequencing (NGS) is particularly challenging because of the unconventional substrate of smears and the often limited sample volume. An analysis of factors affecting NGS testing in cytologic samples may help to increase the frequency of successful testing. This study reviewed variables associated with all in-house cytology cases (n = 207) that were analyzed by NGS with the Ion Torrent platform during a 10-month interval. A statistical analysis was performed to measure the effects of the DNA input threshold, specimen preparation, slide type, tumor fraction, DNA yield, and cytopathologist bias. One hundred sixty-four of 207 cases (79%) were successfully sequenced by NGS; 43 (21%) failed because of either a low DNA yield or a template/library preparation failure. The median estimated tumor fraction and DNA concentration for the successfully sequenced cases were 70% and 2.5 ng/μL, respectively, whereas they were 60% and 0.2 ng/μL, respectively, for NGS failures. Cell block sections were tested in 91 cases, and smears were used in 116 cases. NGS success positively correlated with the DNA yield but not the tumor fraction. Cell block preparations showed a higher success rate than smears. Frosted-tip slides yielded significantly more DNA than fully frosted slides. Lowering the input DNA concentration below the manufacturer's recommended threshold of 10 ng (>0.85 ng/μL) resulted in a marked increase in the NGS success rate from 58.6% to 89.8%. The failure of NGS with cytology samples is usually a result of suboptimal DNA due to multiple pre-analytical factors. Knowledge of these factors will allow better selection of cytology material for mutational analysis.