New international recommendations of response for treatment of AML include morphologic complete remission with incomplete blood count recovery (CRi). This response criteria was defined following evaluation of new drugs used for the treatment of AML in first relapse (

Sievers et al.,
). The objective of our study was to evaluate this criterion in elderly patients with AML who are in first line of treatment. Between 1995 and 2006, 454 patients aged 55 years or older with previously untreated acute non promyelocytic leukemia received a conventional anthracycline and cytarabine induction chemotherapy in our institution. Ages were between 55 and 85 years (median 65 years). Two hundred and fourty-eight patients achieved a complete response (CR) (55%), 37 patients achieved CRi (8%), 104 patients had persisting leukemia (23%), and 49 died during remission induction therapy (13%). Multilineage dysplasia, secondary AML and blasts expressing CD34 were significantly more frequent in patients achieving CRi than CR (58% versus 29%, p=0.001, 33% versus 15%, p=0.007, and 79% versus 54%, p=0.01, respectively). No favorable prognostic karyotype was found in the CRi group but cytogenetic distribution did not differ statistically between the two groups. All patients who reached CR or CRi were scheduled to receive consolidation. Only 24 patients (65%) in CRi actually received this consolidation chemotherapy course and 11 patients (30%) had intensification (intermediate-dose cytarabine and/or autologous stem cell transplantation) whereas for patients achieving CR, 233 (94%) and 214 patients (86%) received consolidation and intensification, respectively (p<0,001 for both). None of the patients in CRi received an allogenic stem cell transplantation whereas 18 (7%) of CR patients had one (p=0,2). Disease-free survival (DFS) and remission duration were significantly different between patients in CRi and CR, with a median of 4 and 12 months, and 5 and 9 months respectively (p<0,001 and 0,03). The median overall survival (OS) was also significantly lower for patients in CRi versus CR, respectively 8 and 23 months (p<0,001). By landmark analysis, there was no difference in OS between patients in CRi and a group of 98 patients with induction failure surviving at least 60 days (p=0,4). We also noted that OS was better, in the group of patients in CRi, for those who finally achieved CR criteria after 1 or more course of post-remission chemotherapy (median 16 months, against 7 months for patients still in CRi, p=0,03). Our results show that the CRi criterion is not equivalent to CR in elderly patients who received intensive chemotherapy as the first line treatment of AML. This should be kept in mind when the results of new agents used in this setting are compared to historical data.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author