Evaluation of the performance of a p53 sequencing microarray chip using 140 previously sequenced bladder tumor samples

Background: Testing for mutations of the TP53 gene in tumors is a valuable predictor for disease outcome in certain cancers, but the time and cost of conventional sequencing limit its use. The present study compares tradition al; sequencing with the much faster microarray sequencing on a commercially available chip and describes a method to increase the Specificity of the c hip. Methods: DNA from 140 human bladder tumors was extracted and subjected to a multiplex-PCR before loading onto the p53 GeneChip from Affymetrix. The sa me samples were previously sequenced by manual dideoxy sequencing. In addit ion, two cell lines with two different homozygous mutations at the TP53 gen e locus were analyzed. Results: Of 1464 gene chip positions, each of which corresponded to an anal yzed nucleotide in the sequence, 251 had background signals that were not a ttributable to mutations, causing the specificity of mutation calling witho ut mathematical correction to be low. This problem was solved by regarding each chip position as separate entity with its own noise and threshold char acteristics. The use of background plus 2 SD as the cutoff improved the spe cificity from 0.34 to 0.86 at the cost of a reduced sensitivity, from 0.92 to 0.84, leading to a much better concordance (92%) with results obtained b y traditional sequencing. The chip method detected as little as 1% mutated DNA. Conclusions: Microarray-based sequencing is a novel option to assess TP53 m utations, representing a fast and inexpensive method compared with conventi onal sequencing. (C) 2000 American Association for Clinical Chemistry.