A 3,239 -Patient Combined Eastern Cooperative Oncology Group (ECOG), M.D. Anderson Cancer Center (MDA) Analysis of the Effect of CR vs. Responses < CR on Long-Term Survival in Newly-Diagnosed AML Treated with Ara-C-Containing Regimens: Implications for Targeted Drug Development.

New definitions of response for patients with AML have been suggested, e.g. CRp, marrow CR, hematologic improvement. Although the criteria for these are less demanding than those for CR, some have proposed that their attainment can convert AML to a “chronic disease” associated with long-term survival. To test this hypothesis, we used ECOG and MDA data to assess the effect of the achievement of CR rather than responses < CR following initial Rx on probabilities of survival 3 and 5 years after diagnosis. 2,061 pts treated on 6 ECOG trials 1976–1999 and 1,651 pts given various ara-C-containing regimens at M.D. Anderson 1980–1999 formed our database. We excluded from the survival analyses the 473 pts (11% ECOG,16% MDA)who died within 4 weeks of beginning treatment. CR rates were 62% and 60% for ECOG and MDA pts, respectively. 11% of ECOG and 54% of MDA pts were age ≥ 60. CR rates in these pts were 46% ECOG and 49% MDA. Survival probabilities according to site and response to initial Rx are as follows:

We are assessing whether pts who survived 3 or 5 years despite not achieving CR with initial Rx attained CR with salvage Rx (e.g. allogeneic transplant) or had CRp or marrow CR rather than complete resistance after initial Rx. Nonetheless, our data indicate that in older pts given ara-C-containing regimens, long-term survival is essentially impossible without achievement of CR. Similarly, very few younger pts not achieving CR survive at 3–5 years. Therefore, it appears likely that newer “targeted agents” will have to contribute more than producing responses < CR and must fundamentally change the biology of the disease to improve long-term survival in AML.