A Sensitive Risk Score for Directing Treatment in Younger Patients with AML.

Abstract

Abstract Treatment strategies for younger patients (<60 years) with AML who enter complete remission (CR) are frequently decided on a risk assessment based on cytogenetic characterisation. Most collaborative groups have devised prognostic cytogenetic groupings of favourable, intermediate or poor risk which have relapse risks of 25–35%, 50–60%, and 80–85% respectively. The precise lesions in each group might vary by study group, but the overall discrimination is consistent. Favourable and possibly intermediate risk cases may not receive transplantation, whereas poor risk cases will. Outcomes for poor risk cases have failed to improve over the years, and if not transplanted these patients are candidates for experimental therapy. There is concern that cytogenetic segregation may not on its own be sufficiently sensitive on an individual patient level, and the inclusion of other factors may better customise risk. We performed a Cox regression analysis on 1937 non-APL cases with complete data who entered CR in the MRC AML10 and 12 trials. Increasing age, cytogenetic risk group higher WBC, male sex, secondary disease, and incomplete response to course 1 were all identified as significantly (p= 0.05) associated with poorer survival. These respectively contributed 0.013 per year, 0.651 per cytogenetic group, 0.002 per WBC unit (/nl), 0.170 for males, 0.221 for secondary disease, and 0.195 for PR/ 0.390 for resistant disease following course 1 to an individual’s total risk score. The groups were separated into scores of <2.00 (good risk), 2.00–2.667 (standard) and >2.667 (poor risk) which gave excellent discrimination for survival from CR at 5 years of 63%, 47% and 24% (p<0.00001). The score was prospectively validated in 897 patients who entered the MRC AML15 trial and was confirmed to be predictive with survivals from CR of 69%, 61% and 42% (p<0.00001). The value of the new score was to move patients out of the old cytogenetic category. The major effect was to move 274 patients from intermediate risk to poor risk and 42 poor risk patients to intermediate risk (table). This score has implications for treatment approach since, relatively, 60% more patients were identified as high risk and are therefore candidates for experimental therapy, or may, from Mantel-Byar analysis, benefit from transplantation. MRC Risk Group Total Good Standard Poor Crosstabulation of old and new risk group classification New Risk Classification Good 309 28 0 337 (14%) Standard 51 1288 42 1381 (59%) Poor 2 274 353 629 (27%) Total 362 (15%) 1590 (68%) 395 (17%) 2347