Second Relapse of Pediatric Patients with Acute Myeloid Leukemia: A Report on Current Treatment Strategies and Outcome of the AML-BFM Study Group

BACKGROUND:

Children with acute myeloid leukemia (AML) still experience high rates of relapse. Facing increasing survival after first relapse, it appears critically important to examine the consequences of a second relapse in more detail. However, there is no population-based data available in pediatric AML and no reliable statement about general survival, patients’ characteristics or treatment approaches can be made.

Herein, we report current survival results following second relapse from the BFM study group, which represents to our knowledge the largest available dataset for this subgroup of patients.

PATIENTS AND METHODS:

Between 2004 and 2017, 1222 pediatric patients (age less than 18 years at initial AML diagnosis) with AML (no secondary leukemia, no Down syndrome, no acute promyelocytic leukemia) were registered in the population-based AML-BFM registry and trials in Germany, Austria, Czech Republic and Switzerland providing a longitudinal data collection with treatment, response rates, survival and disease characteristics. Central review of source documentation confirmed accuracy and consistency of all reported data. Only patients with a documented date of first and second complete remission (CR1 and CR2) and a diagnosed second relapse until the age of 21 are included. Statistical analyses were performed with SAS version 9.4 (SAS Institute). All living patients were censored at the time of last follow-up, but no patient later than 03/27/2020. The median follow-up after diagnosis of second relapse was 6,5 years.

RESULTS:

In all registered patients, 7% (83 out of 1222) met the strict criteria for a second relapse. For further analyses patients with a date of second relapse diagnosis after 12/31/2017 (n=6), two patients with isolated CNS relapse, who did not receive systemic chemotherapy, one patient with an underlying syndrome and one patient with insufficient data have been excluded.

The median age at second relapse was 9,2 years. Sixty percent (n=44) of the patients, who experienced a second relapse, did so within a year after first relapse diagnosis. Eighty percent (n=58) and 7% (n=5) had one or two previous HSCTs, respectively. Eighty-nine percent (n=65) received an anthracycline-containing re-induction (DNX-FLA) followed by FLA or another intensive treatment regimen before at first relapse.

In contrast to the standardized treatment approaches in first relapse, patients with second relapse received a wide range of therapy. Forty-six patients (63%) have been treated with an intensive cytotoxic treatment (Table 1). Seventeen patients (23%) got palliation only. Among the 25 patients (35%) who proceeded to HSCT, 21 patients (88%) had a prior HSCT.

Survival after second relapse was very poor with a 5-year pOS of 15±4% (see Figure 1A) and a considerable cumulative incidence of early deaths (until day 56 after diagnosis: CI ED 19±5%). Prognosis did not improve over time with consistent overall survival rates until 2017 (see Figure 1B). Causes of death include disease progression (n= 51, 70% of all patients), a combined SCT-related and disease-related cause (n=3, 5%) and SCT-related complications (n=4, 4%) or treatment-associated toxicity (n=5, 7%).

The 5-year pOS was 2±2% for patients with an early second relapse vs. 33±9% for those experiencing a second relapse more than a year after the first. (p<0.0001; Figure 1C). The timing of a first relapse and age did not show any influence on overall survival. The best response achieved in the respective bone marrow sample after up to two cycles with cytotoxic treatment have been categorized. Out of 45 patients with conclusive data 31.1% (n=14) achieved a third CR with a pOS of 36±13%, while 62.2% of the patients showed a nonresponse to the treatment (n=28, pOS 7±5%) or no evidence of leukemia, but also no peripheral regeneration (6.7%, n=3, pOS 0±0%).

CONCLUSION:

These data provide new insights into treatment strategies, prognostic factors and outcome of children with second relapse in pediatric AML. As expected, survival is poor, but nonetheless possible in this increasingly relevant subgroup of patients. These data may serve as foundation for urgently needed international clinical trials for relapsed and refractory AML in children.