Prognosis, therapy realization and toxicity in nonresponders of study AML-BFM 87

Background: Nonresponders (NR) are patients (pts.) with no or insufficient response to initial treatment, which may be caused by either initial risk f actors or poor therapy realization. In study AML-BFM 87, 49 NR of 307 patie nts (16%) did not achieve remission until the end of intensive chemotherapy and were analysed to assess the specific contribution of prognostic factor s, therapy realization and complications of therapy. Therapy and Methods: Therapy started with an 8-day induction therapy follow ed by a 6-week consolidation and two 5-day intensification blocks with high -dose cytosine-arabinoside and VP-16. Maintenance therapy was given for a t otal duration of 1.5 years. To evaluate the impact of treatment intensity i n NR, we compared the dose compliance (DC = dose given/intended dose), the dose intensity (DI = dose per time given), the treatment results, and toxic ity of the individual therapy phases in responders (CR) and NR. Results: In 19 of 19 NR therapy was stopped before starting intensification blocks. Twenty-six NR received at least one block of intensification, and seven patients between three and six intensification blocks. Six children e ntered maintenance therapy. Twelve patients received a bone marrow transpla nt (9 allogeneic, 3 autologous). Six (5 after bone marrow transplantation) of 49 NR are still alive for 64 to 108 months. In nearly all patients induc tion therapy could be applied according to protocol (mean DC: 98%,range 85% -100%), whereas therapy realization was more difficult in the 2(nd) phase o f therapy (mean DC: 92%, range 12%-113%). Deviations from the protocol in t he treatment blocks (changes of dose and/or schedule) were mainly attributa ble to persistence of blasts (n = 33) and septic complications (n = 24). Th e mean relative DI of 1.01 was according to protocol. Bleeding and infectio us complications in the individual therapy phases varied from 7% to 61%. NR compared to CR patients suffered significantly more often from bleeding du ring the first and second part of consolidation and from infections during the second part of consolidation. Withdrawal from protocol in NR was mainly due to persistence of blasts (n = 16), followed by bone marrow transplanta tion or other therapies (n = 13), and sepsis (n = 11). Conclusions: It is difficult to discriminate between nonresponse associated with blast persistence followed by complications and subsequent discontinu ation of therapy and nonresponse due to insufficient therapy in patients wi th complications. Our analyses revealed that therapy with 2 intensification s according to protocol was feasible in 13 NR. Patients' condition permitti ng, therapy should not be stopped prematurely, in order to sustain the opti on of BMT after blast cell reduction.