DEVELOPMENT OF A SENSITIVE, HIGHLY CONTROLLED ASSAY FOR MOLECULAR-DETECTION OF THE PHILADELPHIA-CHROMOSOME IN PATIENTS WITH CHRONIC MYELOGENOUS LEUKEMIA

The Philadelphia chromosome (Phl), present in greater than or equal to 95% of chronic myelogenous leukemia (CML) patients, is a well-charact erized translocation that results in a unique chimeric gene product (B CR/ABL) with transforming capability, Molecular methods utilizing the polymerase chain reaction (PCR) to detect BCR/ABL mRNA transcripts has been useful for detecting minimal residual disease (MRD) after treatm ent, as well as for establishing the diagnosis of CML. Amplification-b ased assays for the BCR/ABL transcript, however, have shown variable r eproducibility and sensitivity. This variability may be largely due to technical differences and insufficient controls. In this report, we d escribe the development of a highly controlled, reproducible, and sens itive PCR assay to detect Phl that is well suited to clinical and, res earch applications, A validation study of 82 samples was performed con sisting of 25 dilutions of K562 cells (Phl +) into normal cultured B c ells, 26 pre- and posttransplant peripheral blood samples from CML pat ients, 16 peripheral blood samples for diagnosis of CML, and 15 periph eral blood samples from healthy individuals. RNA isolated from 3 to 5 million leukocytes was reverse transcribed (RT) and amplified by neste d primer PCR. The products were characterized using agarose gel electr ophoresis. Approximately 1000 Phl-positive cells admired with 10(6) no rmal cells were detectable after one round of amplification. In 60% of assays where one Phl-positive cell was admired with 10(6) normal cell s, a BCR/ABL product was detectable after nested primer PCR. Specific measures to ensure accurate results in routine testing included (a) as sessing RNA integrity and adequate cDNA preparation by detection of th e constitutively expressed ABL mRNA, (b) monitoring sensitivity with t he addition and detection of K562 RNA mixed with RNA from unknown samp les (failure to detect the ''spiked'' K562 RNA indicates the presence of inhibitors or ribonucleases within the unknown RNA sample), (c) det ection of nucleic acid contaminants by using negative controls in ever y assay, and (d) duplicate analysis of all samples and controls. Inter nally, this assay was 100% reproducible. Our results verify that neste d primer RT-PCR is a fast, sensitive alternative to cytogenetic or Sou thern blot analysis for monitoring MRD after treatment and for diagnos is of CML. In addition, the highly controlled detection scheme present ed here can be used as a general model for the development of other am plification-based detection assays.