Long-term survival rates in pediatric AML have been improved by intensification of firstline treatment during the last two decades. However, relapse remains a major obstacle to further improvement of prognosis. The aim of this study was to analyze treatment results in the total group of relapse patients from 1987 to 2007 (n=378) according to the intensity of the first-line treatment in studies AML-BFM -87, -93 and -98 (1987–2003), duration of first remission, modalities of relapse treatment, and achievement of a second remission.

Treatment: Intensity of first-line treatment increased mainly from study AML-BFM 87 to the following studies, AML-BFM 93/98. Until 1997, relapse treatment was not yet standardized. From 1997 to september 2001, treatment according to AML-BFM REZ 97 or IDA-FLAG protocols was recommended, and as of october 2001, according to the international protocol Relapsed AML 2001/01.

Patients: From a total of 1,251 patients with de novo AML registered in studies AMLBFM -87, -93 and -98, 378 patients relapsed until October 2007. The cumulative incidence of relapse was 30.2% and did not differ significantly between the three studies (31.5% vs. 28.0% vs. 31.5%, respectively). Patient characteristics (age, sex, FAB type, cytogenetics and risk group classifications) during first-line treatment were equally distributed over the three study populations with 26% standard risk (SR) and 74% high-risk (HR) patients. Relapses occurred mainly in the bone marrow (92%), combined or isolated in the CNS (15%) and other extramedullary sites (6%).

Results: The 5-year overall survival (OS) for the total group of patients was 23.5% (SE 3%). OS after relapse increased with study periods: initial treatment according to AML-BFM-87 = 18% (SE 4%), AML-BFM-93 = 21% (SE 4%), AML-BFM 98 = 31% (SE 4%), plogrank 0.012. Simultaneously, the proportion of patients who were transplanted in 2nd complete remission rose significantly between the study periods (41% vs. 46% vs. 59%, plogrank =0.004). Duration of 1st remission was a major prognostic factor (³1 year vs. <1 year from diagnosis): 5-year OS 36% (SE 4%) vs.13 % (SE 2%), plogrank <0.0001. From a total of 376 patients with documented relapse treatment, 303 patients received intensive chemotherapy consisting of either elements from upfront protocols or specific relapse therapy elements, 20 had other treatments and 52 no or palliative treatment. No significant difference in survival was observed between the different intensive re-induction schedules. The proportion of patients without or with palliative treatment decreased significantly from study AML-BFM 87 (23%) to studies AML-BFM 93/97 (11%), plogrank =0.004, illustrating a trend to recognize a realistic chance of cure for relapsed AML in children. 199 out of 364 patients with remission data (54.6 %) achieved a second remission (CR2). CR2 was the most relevant prognostic factor for survival (5-year OS: 43%, SE 4% vs. 1%, SE 1% for non-responders). This was also seen for OS after stem cell transplantation (SCT): 5 year OS in CR2 = 45% (SE 4%) vs. 7% (SE 5%) in non-responders. 153 of 199 patients (77%) were transplanted in CR2 (127 allogeneic, 26 autologous). OS was similar regarding both transplant types for this cohort (allogeneic/ incl. haploidentical SCT = 45%, SE 5% vs. autologous SCT = 50%, SE 10% plogrank 0.44) with better 5-year survival rates after SCT in late relapse (OS = 51%, SE 6% vs. 58%, SE 11% plogrank 0.54) than in early relapse (OS = 33%, SE 7% vs. 29%, SE 17%; plogrank 0.85). The percentage of early and late relapse was similar in autologous (27% early relapses vs. 73% late relapses) and in allogeneic transplants (35% vs. 65%, p 0.2).

Conclusion: In childhood AML, the implementation of a consistent and standardized relapse treatment substantially improved OS after first relapse over the last two decades. Due to a stringent treatment strategy within relapse protocols including stem cell transplantation and improvement of supportive care, the number of patients receiving no or only palliative treatment has declined continuously. Achievement of a second remission remains the most important prognostic factor regarding survival after AML relapse. Once CR2 is achieved, SCT is recommended. Patients who do not respond to relapse treatment are candidates for new therapeutic options.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution