A 17-year experience with ALL-BFM protocol in acute lymphoblastic leukemia: Prognostic predictors and interruptions during protocol

Abstract

Acute lymphoblastic leukemia (ALL) is the most common childhood cancer and despite the intense combination chemotherapy, cure rates are less than 90%. Several prognostic parameters, including nonneoplastic hematologic cell counts during induction phase, are suggested to predict outcome in ALL. We analyzed 242 ALL patients treated in our center to investigate individual prognostic parameters and the impact of delays on disease outcome. Age at diagnosis, risk groups, extramedullary involvement, t(9;22), prednisone response, bone marrow response at days 15 and 33, day 15 platelet count, day 33 lymphocyte, monocyte, and platelet counts, treatment delay, sepsis, and omission of day 64 cyclophosphamide were valuable predictors of survival in univariate analysis. However only the age, CNS involvement, omission of cyclophosphamide, and total delay during treatment were associated with survival in multivariate analysis. Omission of second cyclophosphamide dose had no impact on survival of standard risk group patients, but adversely affected the long term survival of medium risk group (MRG) patients. The second dose might be given with the first dose on day 36 to MRG patients to prevent delays. Day 15 and 33 platelet counts are promising predictors of survival in low income countries where assessment of minimal residual disease is difficult, but this data needs further consolidation.

Introduction

Acute lymphoblastic leukemia (ALL) is the most common malignant tumor of childhood and it is the fourth most common cause of mortality in children between 1 and 15 years in Turkey [1]. Almost 4000 new ALL cases are diagnosed annually in USA and the cure rates are about 80% to 90% in developed countries [2]. Significant progress has been made in the management of ALL patients with combination of effective anti-leukemic drugs and risk stratification of patients at the onset of therapy enabled treatment intensification for high risk patients. Currently, efforts are made toward decreasing the dose of chemotherapy applied, in order to prevent long-term complications, while preserving and even further improving the disease-free survival.

There are considerable differences between Western countries and developing countries for ALL, including age at diagnosis, immunophenotypes, death and relapse rates, causes of mortality, complications, and compliance to treatment. Unfortunately, data regarding the long term follow up of large patient cohorts from developing countries are so scarce.

Several clinical and biologic prognostic parameters have been established to predict outcome in ALL patients. Among these, age, white blood cell (WBC) count at diagnosis, immunophenotype, cytogenetic features of blasts, extramedullary involvement, prednisone response, complete remission at the end of induction therapy, and minimal residual disease (MRD) have been well-characterized in clinical trials [3]. More recently, individual response of nonmalignant hematologic cells to chemotherapy and the ability of nonneoplastic hematopoietic precursors to restore hematopoiesis early in the treatment were suggested as novel prognostic predictors in ALL [4], [5], [6], [7]. The ease of assessment of their levels with simple serial complete blood counts make them valuable prognostic parameters for use in developing countries. However, their validity and reliability in predicting disease outcome are highly debated and these markers have never been formally assessed in Turkish ALL patients.

The aim of this study is to retrospectively analyze all ALL patients who were treated in a single tertiary center of Turkey to investigate prognostic parameters which can be used to predict disease outcome and present the results of our modified Berlin-Frankfurt-Münster (BFM) 95 treatment protocol.