Correlations between BCR-ABL Ratio and Cytogenetic Response at Months 12 of the Randomized French SPIRIT Study on Behalf the Fi-LMC Group.

In September 2003, the French CML study group (Fi-LMC) activated the four armed randomized SPIRIT trial comparing imatinib 400mg with imatinib 600mg, imatinib 400mg + ara-C and imatinib 400mg + pegylated interferon in newly diagnosed CML in chronic phase. The molecular monitoring was centralized and performed, according to the EAC protocol, in duplicate on the same BCR-ABL cDNA among two French laboratories (Bordeaux and Lille). Data were expressed as BCR-ABL/ABLx100 on the new internationally agreed scale (IS) using a conversion factor of 1.33 and 0.80 for Bordeaux and Lille respectively. A good correlation between the two laboratories was observed. During the first year of follow-up the quantification was performed each three months (M0 to M12). In May 2006, 297 of the 492 enrolled pts had a follow up of more than 12 months, and 263 pts were analysed at M12. Using the IS, the median values of the BCR-ABL/ABL normalized ratios were 90%, 6,4%, 0.68%, 0.32% and 0.16% at M0, M3, M6, M9, and M12 respectively. At M12, 15%, 29%, 27%, 17% and 12% of the pts presented a BCR-ABL/ABL ratio lower than 0.01%, 0.1% ,1%, 10%, and 100% respectively. Overall, 71% of the pts presented a ratio lower than 1%. Conventional cytogenetic data were available for 193 pts at M12: 156 pts (80%) were in complete cytogenetic response (CCR), 25 pts (13%) were in major cytogenetic response (MCR) and 12 pts (6.2%) were in minor or a lack of cytogenetic response. Among pts with a BCR-ABL transcript ratio lower than 1%, all except four were in CCR. The last 4 cases were in MCR associated with BCR-ABL ratios of: 0.96% for one pt, 0.08% and 0.12% for 2 pts (who further reached CCR at 6 and 18 months) and 0.128% for the last one. Percentages of Philadelphia positive metaphase (Ph+) were 4%, 6%, 5 % and 15% respectively. Among the 30 pts with a BCR-ABL ratio > 5%, none were in CCR and only 18 pts were in MCR. Among the 27 pts presenting a BCR-ABL ratio in the 1%–5% intervals, 20 pts were in CCR and 7 pts were in MCR (5 pts with Ph+ lower than 10% and two pts with 16% and 20% Ph+).

In conclusion, our findings suggest that a ROC statistical test will be able to determine critical BCR-ABL/ABL ratios associated to CCR in our study. Nevertheless, at M12 the percentage of available data was superior using molecular than cytogenetic monitoring, suggesting that molecular monitoring expressed with the new international scale is probably sufficient fore more than 85 % of the pts treated by IM solely or in association.