DETECTION OF BCR-ABL TRANSCRIPT IN CHRONIC MYELOGENOUS LEUKEMIA PATIENTS BY REVERSE-TRANSCRIPTION-POLYMERASE CHAIN-REACTION AND CAPILLARY-ELECTROPHORESIS

Background and Objective. Capillary electrophoresis (CE) has become an attractive alternative to SLAB gel analysis for direct and accurate d etection of amplified product,(1-5) and a few cycles of polymerase cha in reactions (PCRs) could be sufficient for both quantitative and qual itative analysis. We try to assess: 1) whether CE could be a practical , non-isotopic method for direct detection of the presence of amplifie d bcr-abl obtained by a reverse transcription (RT)-PCR (qualitative an alysis) and 2) whether it is possible to quantify PCR products using a competitive RT-PCR measuring peak areas of CE electropherograms (quan titative analysis). Design and Methods. The two types of bcr-abl chron ic myelogenous leukemia (CML) associated transcript products were gene rated by RT-PCR (qualitative analysis) from 1 mu g of total RNA extrac ted from bone marrow samples of 34 CML patients at diagnosis (median a ge 47.5; range 18-65; median Sokal's score 0.9; range 0.53-2.78). The PCR products were analyzed by SLAB-gel electrophoresis (SGE) on 2% aga rose gels and by CE (128 runs; median 3.3 times for each sample). Furt hermore, we assessed the amount of PCR product (quantitative analysis) by a competitive RT-PCR approach and by CE (bcr-abl transcripts were expressed as transcript per mu g of total RNA examined). Results. CE s eparation of PCR products obtained by qualitative RI-PCR showed baseli ne resolution for the two peaks corresponding to the two types of bcr- abl junctions: the b2-a2 type (343 base pairs, 10 patients) was reveal ed at 9.33 min [standard deviation (SD) = 0.1] and the b3-a2 type (418 base pair, 24 patients) at 10.03 min (SD = 0.25). By quantitative ana lysis we found that there is great interpatient variability in bcr-abl expression at diagnosis: the median value of the amount of bcr-abl tr anscript was 78,000 bcr-abl transcript/mu g total RNA ranging from 17, 300 to 750,000. The amount of bcr-abl transcript at diagnosis was rela ted to the number of blast cells (mean value 128,859 vs. 331,722 in pa tients with 0% blast coils and >1% blast cells, respectively; p = 0.00 4) and sokal's score (mean value 156,865 vs. 408,800 in patients with Sokal's score <0.8 and >1.2, respectively; p = 0.003). Interpretation and Conclusions. Our results confirm that CE analysis offers greater r esolution and enhanced sensitivity for detection and quantification of bcr-abl PCR product in the study of this leukemia. Qualitative analys is by CE of bcr-abl product provides a rapid technique (less than 20 m in) for the analysis of subnanogram amounts of DNA fragments. CE run t imes are short, the capillary can be reused and full automation may be feasible with data acquisition by a computer-controlled step. Competi tive/quantitative analysis of bcr-abl as analyzed by CE allowed fewer reactions and more precise quantification. (C)1998, Ferrata Storti Fou ndation.