Outcome in pediatric acute lymphoblastic leukemia in children and adolescents using contemporary protocols of chemotherapy – experience of a single Pediatric Hematology-Oncology Center

The authors studied the outcome of a cohort of pediatric acute lymphoblastic leukemia (ALL) patients (33 children and adolescence) diagnosed and treated conforming to modern Chemotherapy Protocols (ALL ICBFM 2002, Interfant 06) in a single Center – Pediatric Clinic, Fundeni Clinical Institute, Bucharest, Romania. They analyzed the factors which determine the prognosis and the outcome of these patients in the course of multi-agent systemic chemotherapy to stand at the base of these Protocols: initial age, initial leukocyte count, blasts immunophenotype, cytogenetic and molecular abnormalities, initial response to cortisone, risk groups, time to obtain the complet remission, etc. Among the factors they discussed, a great value was proven to have the minimal residual disease (MRD) determination in certain check points of Protocols and revaluation of patients risk conforming to MRD values. Using the modern Protocols and continuously watching the evolution on therapy enable the authors to obtain results close to those of European and North American Pediatric Hematology Oncology Centers: OS 90.9% by 40 Mo and EFS 72.7% at the end of the same period of time.


INTRODUCTION
It is now widely accepted that pediatric ALL is a biologically heterogeneous disease.(1) This heterogeneity manifests also in what concerns the results of therapy and the outcome.
Modern pediatric ALL treatment Protocols employ risk-based therapy to reduce the toxicity in patients with low risk ALL (i.e.low risk of treatment failure), and to use aggressive therapy for those with a high risk of relapse and therefore an unfavorable outcome.
Newly diagnosed children are subjected to a Treatment Plan including chemotherapy via systemic and intrathecal administration and supportive care.Radiotherapy is now limited for emergencies like mediastinal mass producing respiratory obstruction, or for high risk patients with central nervous system (CNS) disease.(2,3) The contemporary multidrug chemotherapy regimens are working in three different phases: Induction (initial 4-6 weeks of therapy), which now achieves remission in over 95% children with ALL.(1-3) The principles of multiagent CxT are: combination of drugs, low therapeutic index and dose intensity (4).In this period of time, as secondary effects of therapy some patients can develop life-threatening toxicities.
Post-induction therapy -lasts 2-3 years and is used to remove any residual leukemia cells (1).It includes early intensification, re-induction, late intensification and maintenance.
A special attention is given to therapy of sanctuary sites for leukemia cells, primarily CNS involvement.Preventative CNS therapy was a major acquisition in the treatment of childhood ALL.The vast majority of contemporary Protocols have replaced cranial RxT with intrathecal therapy (TIT) and systemic intensification of specific agents (i.e.methotrexate); the risk of relapses at the level of sanctuary organs is now <5% in standard-risk (SR) ALL.(2,5,6) Cranial radiation is now reserved to treat patients with CNS disease at the time of diagnosis, and high risk of relapse.
Up to 20% of affected children treated with modern protocols will however relapse and their prognosis remains unfavorable.(7) A therapeutic breakthrough appears to be hematological stem cell transplantation (HSCT).In case of relapse, HSCT is indicated for high-risk patients (e.g., those who relapse on therapy, or with high MRD after reinduction therapy who have a great risk for systemic relapse).(1,3) HSCT can achieve a 5-year event-free survival (EFS) in nearly 60% of all risk patients.(8) Risk adapted therapy in childhood ALL is relying on risk stratification; we discussed this stratification in a previous paper.(9) Here, we intend to summarize risk factors in pediatric ALL: initial age and leukocyte count, CNS status at diagnosis, immunological subtypes of blast, cytogenetic and molecular abnormalities, response to Prednisone in the first 8 days of therapy induction failure, and -in particular -minimal residual disease (MRD) at the end of induction.
Recent studies shows that an important indicator of prognosis is the response of patient to treatment.It is possible now to quantify MRD using polymerase chain reaction (PCR) (detecting specific gene fusion products or specific arrangements of immunoglobulin/T-cell receptor genes), or flow cytometry, which use the aberrant leukemia cells phenotypes to highlight the presence of residual blast cells at the end of induction, or at other critical points in the course of therapy.
These techniques are 10-100 times more sensitive than morphologic evidence of residual blasts, and can detect as fewas 1 leukemia cell in 10,000 to 100,000 normal cells.(2,10,11) MRD negativity at the end of induction (<1/10,000 leukemia cells) is considered a marker of good prognosis; MRD positivity at end of induction and persistent or rising MRD levels at later check points in therapy signity an increased risk of relapse.(12,13)

Background
The aim of this study was to asses the results of therapy in a cohort of new diagnosed pediatric ALL using the contemporary Protocols -ALL IC-BFM 2002 and Interfant 06.(14,15) We focused on influence of variables like gender, initial leukocyte count, blast's phenotype, associated cytogenetic and molecular anomalies, initial response to cortisone, induction outcome, etc on overall and event-free survival (EFS).Also, we analyzed primary (due to disease) and secondary (due to therapy) complications and their impact on survival.We have to mention that using Protocols, all the patients were subjected to an uniform modality of treatment.

MATERIALS AND METHODS
Our study is descriptive and observational (non-interventional).We used the same cohort of patients for assessing the value of newer diagnostic methods for risk stratification and therapy in pediatric ALL.

STUDY OVERSIGHT
The study was conducted in accordance with the Principles of the Declaration of Helsinki (DoH) (16); it was approved by the local Fundeni Clinical Institute Committee for Bio-Ethics and all families and adolescents >16 years old have signed Informed Consent.The study had no sponsors, and the investigators personally collected data and analyzed them.The first author wrote the draft of the manuscript, and all the authors reviewed and made the revisions to submit it for publication.

PATIENTS
The  The morphologic diagnosis of ALL was established in presence of>25% blasts in bone marrow, using the bone marrow aspirate smears submitted to May-Grünwald-Giemsa (MGG), Periodic acid-Schiff (PAS), α-Naphthyl Acetate Esterase and Myeloperoxidase (MPO) colorations.
Immunophenotyping of blasts was conducted in peripheral blood or bone marrow collections.Peripheral blood and bone marrow aspirate were harvested on EDTA (Ethylene diamine tetraacetic acid), then we proceed to isolation of blasts by density gradient centrifugation using Ficoll-Hypaque.Blasts were subject to flow cytometry using a Becton Dickinson FACS scan (FACS -Fluorescence-activated cell sorting) and the panel of monoclonal antibodies shown in Table 2. Cellular lines were defined by EGIL Score (EGIL 2002 -European Group for the Immunological Characterization of Leukemias).( 17) Coexpression of myeloid antigens was defined by simultaneous expression of ≥1 myeloid markers (CD 13, CD 33, CD 65) on more than 20% of lymphoblasts.
Cytogenetic studies were conducted on chrom osomes prepared in 24 hours culture on usual media (RPMI 1640, MEM Eagle, TC-199, IC 65) supplemented with bovine fetal serum (FBS), L-glutamine, antibiotics and phytohemagglutinin (PHA).Karyotypes were realized by standard cytogenetic technique and G-band analysis; chromosomal anomalies were assessed conforming to ISHCN (International System for Human Cytogenetic Nomenclature).We used also conventional fluorescence in situ hybridization (FISH) techniques for identification of specific translocations in blast cells.
Molecular biology studies were done on BM aspirate, after DNA extraction by standard techniques and PCR amplification of DNA fragments of interest, in presence of commercial primers for common ALL translocations/products of gene fusion (BCR-ABL, TEL-AML1, MLL-AF4, SIL-TAL).
We repeated the complex of analyses previously presented in different check points during therapy: D15, D33, W12, at the beginning and the every phase of therapy and the end of therapy.

Classification
Our patients were classified in 3 risk groups (SR, IR, HR) conforming to ALL IC-BFM 2002 Protocol Criteria (Table 3) and for infants allocation to SR and HR group was made conforming to Interfant 2006 Protocol criteria (Table 4).(14,15)

Treatment
Multiagent systemic chemotherapy and intrathecal therapy (TIT) was deployed conforming to Protocols above mentioned and depicted in Figure 1

Statistical analysis
Data obtained from individual sheets of patients were subjected to methods of descriptive statistics.Using the IBM SPSS statistics 20, Epi Info and Microsoft Excel 2010, we computed relative frequencies and realized the graphic representations.In order to test statistical signification of differences we used the χ2, log-rank tests and Cox proportional regression model.(18) To estimate survival (OS, EFS) we performed survivor analyses and generated the Kaplan-Meier survival plots.(19) Personal study -Results

OS and EFS in our patients
Table 5 presents the status of ALL patients at the end of study (maximal length of observation -40 months).
Table 6 depicts the length of OS and EFS in our cohort of patients.

Gender
EFS by 40 Mo (months) was 76,5% in girls, but only 66,8% in male patients.Due to the small number of patients, we could not asses the statistical significance of this difference.

Age of patients at diagnosis
The differencesof EFS in various groups of patients depending on the age at diagnosis is presented in Table 7.As it results from this table, the best prognosis was in the 1-6 years of age, while the age group < 1 years of age had the worst prognosis.

Initial leukocyte count
The EFS in correlation with initial leukocyte count is depicted in Table 8; due to the small number of patients, we could not appreciate the statistical significance of the obvious differences between the 3 groups of patients.

Immunophenotype of the blast cells
Distribution of phenotype in relation to the patient's age is presented in Figure 3.
Table 9 shows the EFS in our cohort of patients depending on the immunophenotype of the blast cells.

Allocation of patients to the risk groups in accordance with ALL IC-BFM 2002 and Interfant-06 Protocols of chemotherapy
We present in Figure 4, the allocation of our patients to the risk groups according to the two Protocols used in this study.
The criteria for allocation are depicted in Table 3 and Table 4.
We have to mention that in the IR group we noted 5 patients with initial age 1-6 years, but initial leukocyte count was >20,000/μL; 7 patients have initial age >6 yoa, but with on initial leukocyte count <20,000/μL; the other 2 patients from this group fulfilled both criteria: initial age >6 yoa, and initial leukocyte count >20,000/μL.
Discussing about the allocation of HR patients, we have to emphasize the next remarks: 2 patients had t(4;11)(MLL-AF4) translocation, one of them being PGR (prednisone good responder) and obtaining CR (complete remission) in bone marrow D33, and the other one -PPR (prednisone poor responder) -did not obtain CR D33. 4 patients had t(9;22) (BCR-ABL) translocation, wo being PGR and obtaining CR D33; the others were poor prednisone responders, only one with CR D33 (for the second PPR it was an induction failure).
We have also to mention that OS and EFS are estimated at 40 months of observation and that -due to small number of patients -we could not calculate the statistical significance, although the differences are obvious.

Outcome in correlation with response to therapy
The main check points related to response to therapy in the Protocols we used in this study were:

Initial response to cortisone (D8 of induction)
In our cohort of patients, they were 7 prednisone poor responders (PPR); it means a number of blast cells in peripheral blood >1,000/μL.The evolution of these patients is presented in Fig- ure 6.
From a practical point of view, poor initial response to cortisone compels the inclusion of patients in the HR group and intensification of therapy.

BM status D33
BM status M0, M1 in the D33 of induction means an appropriate response to induction therapy; BM status M2, M3 is the sign of induc-tion failure.In this study, 26 patients (78.8%) were in complete remission by the end of induction.Induction failure is a Protocol indication to move the patient in the high risk group and to further intensify the therapy.
The evolution of the patients in continuation of Protocol in correlation with their allocation to one of the risk groups is presented in Figure 7.

Minimal residual disease
MRD (by immunophenotyping or by molecular biology techniques -FISH and RT-PCR) was assessed in 10/33 patients (33.33%),only these having anomalies which could be demonstrated by the techniques we used.MRD was searched in two time points: TP1 (BM, D33 of induction) and TP2 (BM, D78 of therapy, before the initiation of Protocol M or the first HR block of chemotherapy).
In the standard risk group, MRD was assessed in 3 patients with t(12;21)(TEL-AML1) translocation.We could demonstrate the clearance of blasts in BM at the end of induction in all 3 patients (in one after the intensification of therapy).All these 3 patients remained in CR at the end of study.
In the high risk group, all the patients had cytogenetic or molecular abnormalities who could enable us to discover the MRD.MRD detected by FISH in 2 cases with t(4;11)(MLL-AF4) translocations was persistent by the both check points; at the end of study the evolution was unfavorable in these patients.

FIGURE 6. Evolution of prednisone poor responders
Abbreviations: PPR: prednisone poor responders; CR: complete remission. of fusion and treated by chemotherapy and tyrosine kinase inhibitor (TKI) -Imatinib in association, MRD was negative in both check points and final evolution was favorable.In 2 cases with t(4;11) translocation, MRD was positive and final outcome unfavorable.

Primary and secondary complications related to patient's risk group
We considered primary complications the events due to evolution of the disease, and secondary those due to infections, toxic, metabolic causes, etc. Table 11 summarizes the main complications in our cohort of patients.

DISCUSSION
ALL in children is considered now a highly curable cancer; OS is 90% in the most privileged countries.(2,20) This was achieved by escalating the intensity of treatments -chemotherapy and HSCT (hematopoietic stem cell transplantation) -but it is related to a much more toxicity of therapy.
The major events who are the backbone on actual therapy are risk adapted treatment (based on biological features of the disease), CNS prophylaxy and the improvement of supportive care.The modern ALL therapy is based on risk stratification, up front, at diagnose and later on and by response to therapy in the different phases of Protocols.
One of the main aim of our study, was to assess the relationship between principal biological characteristics of the disease and the therapeutical outcome.In this study, we were able to demonstrate the value for predicting the response to therapy of such biological factors as: Age at diagnosis -the best prognosis was in the 1-6 yoa group, while the ALL children aged less than 1 y at diagnose had the worst prognosis; Initial leukocyte count (<20,000/μL: 80% EFS at 40 Mo, between 20,000-100,000/μL: 62.5% and >100,000/μL only 55.6% EFS at 40 Mo); Blast's immunophenotype, with B-common ALL having the best outcome (81% EFS by 40 Mo), and pro-B the worst one (only 33.3% EFS by 40 Mo); Associated cytogenetic and molecular anomalies -The impact of these biological variables is visible especially for HR group patients.In our cohort, EFS was 33.33% in HR patients with favorable abnormalities (hyperdiploidy, t(12;21) (TEL-AML1), and t(1;14)(SIL-TAL1)), the last one MRD was assessed by multicolor flow cytometry, immunophenotyping or by molecular techniques (FISH and RT-PCR) on 2 time points (D33 and D78 of therapy).Demonstration of blast's clearance by these two time points is a factor of good prognosis, whereas MRD positivity predicts an unfavorable outcome.
We have to emphasize the fact that the results of our study are similar to the conclusions of other contemporary studies.(21)(22)(23)(24)(25)(26)) Also, we should to mention that the major limitations of personal study are the relative small number of patients (33 pts), and the limited period of observation (only 40 months); these would not enable us to assess the statistical significance, even the most differences are obvious.
From a more general print of view, the validity of our study was also demonstrated by different outcomes related to allocation of patients to risk groups, which are mostly based on variables above discussed: EFS by 40Mo being 88.89%, 85.7% and 60% for SR, IR and HR groups respectively.
An argument in favor of increased toxicity of contemporary Protocols of chemotherapy is the list of secondary complications in our cohort of patients (Table 11).
The major challenges for the development of newer, and better Protocols for treatment in pediatric ALL remain: the adverse prognosis of infant ALL, the role in leukemogenesis and the peculiarities of constitutional predispositions, some biological anomalies (like hypo-diploid ALL, ALL with iAMP21 (intrachromosomal amplification of chromosome 21), BCR-ABL like pe-diatric ALL, early T-cell precursor ALL), as well as induction failure, refinement of MRD diagnose, and the new targeted therapies.(27) For our country the main problems remain: the better endowment of laboratories and of clinical wards, as well as a better co-operation between national pediatric centers and a more active participation at international Protocol studies, in order to mitigate the differences between outcomes in comparison to Center from higher income countries.

CONCLUSION
The use of modern Protocols of chemotherapy in pediatric ALL lead to the assumption that this atrocious disease could be highly curable (more than 90% EFS was reported by some Centers).
Recent worldwide studies, as well as our study on a small cohort of 33 consecutive patients emphasize the value of such biological factors for the better use the current treatment modalities, as initial age of patients, leukocyte count at diagnosis, blast's immunophenotype, cyto genetics and molecular characteristics.Also, they demonstrate the impact of monitoring the evolution of patients on therapy -especially by MRD -for adapting the Protocols to patient's peculiarities.
We should to emphasize that remaina lot of major challenges for development of newer, more active, Protocols in pediatric ALL: the poor prognosis of infant as well as the adolescent ALL, some biological specificities of the disease like hypodiploid ALL, ALL with iAMP21, BCR-ABL like pediatric ALL, early T-cell precursor ALL, etc.
To solve all these challenges it is necessary the refinement of preclinical studies, as well as the amplification of cooperation between the Pediatric Hematology Oncology Centers.

FIGURE 5 .
FIGURE 5. OS and EFS in correlation with risk group and associated cytogenetic/molecular abnormalitiesAbbreviations: SR: standard risk; IR: intermediate risk; HR: High risk; OS: overall survival; EFS: event-free survival.
Patients were eligible if their age at inclusion was under 18 year and if they were diagnosed and treated in the Pediatric Clinic, Fundeni Clinical Institute, Bucharest,

TABLE 3 .
ALL

TABLE 7 .
EFS in different age groups in our patients *Number of months from diagnosis to relapse, to death due to the toxicity of therapy or to the abandonment of study.Abbreviations: yoa: years of age; EFS: event-free survival.

TABLE 9 .
EFS correlated with the immunophenotype of the blast cells

TABLE 5 .
Status of ALL patients at the end of study Abbreviations: SR: standard risk; IR: intermediate risk; HR: high risk; CR: complete remission.

TABLE 6 .
The leght of OS and EFS in our cohort of patients Abbreviations: SR: standard risk; IR: intermediate risk; HR: High risk; OS: overall survival; EFS: event-free survival.

TABLE 10 .
OS and EFS correlated with the risk group of patients*

Outcome at the end of study OS (%) EFS (%) SR IR
*OS by 40 months in the whole cohort of patients: 90.9%; EFS by 40 months in the whole cohort of patients: 72.7%.FIGURE 3. Correlation between blasts immunophenotype and the patient's age FIGURE 4. Allocation of patients to risk group

TABLE 11 .
Primary Induction failure is an indication of most recent Protocols (ALL IC-BFM 2002, Interfant 06) to assign the patient to HR group and to further intensify the therapy.