INTRODUCTION

Monitoring of residual disease (MRD) in leukaemic patients offers a means of evaluating the efficacy of the treatment given. In childhood acute lymphoblastic leukaemia (ALL), relapses are reported in at least 20% of children who are treated with contemporary programs of chemotherapy. However, monitoring MRD in childhood ALL may create the possibility of tailoring treatment according to patient’s response, where a more in­tensive therapy could be given to pa­tients with high levels of MRD and a less intense one to those without MRD. Vari­ous methods of detection of MRD in childhood precursor B-ALL, such as by morphology, karyotyping, in situ hybridi­zation, immunofluorescence, cell culture, Southern blotting and polymerase chain reaction (PCR) analysis have been de­scribed. However, the evaluation of resi­dual disease in acute leukaemia is most often carried out by morphological analy­sis of bone marrow aspirates with the criterion of free from residual disease (complete remission) denotes fewer  than 5 % of blasts in the bone marrow cell population. Unfortunately, this method does not allow identification of low levels of the disease and patients in morpho­logical remission may still have up to 1010 leukaemia cells (Campana & Pui 1995; Campana 2003). 

Sensitive techniques to detect MRD may allow better estimation of the leu­kaemia burden, provides information on residual disease and aids in therapeutic strategies. Flow cytometric detection of MRD is based on the identification of immunophenotypic combination markers expressed on leukaemic cells but not on normal haemopoietic cells. It is based on the identification of uncommon pheno­types or aberrant markers that are found at diagnosis. At certain points of time during and after chemotherapy, the same phenotypes or aberrant markers are analyzed and quantified. If expression of these markers is less than 1%, the pa­tient is considered free from MRD. How­ever, for the detection of a small number of leukaemic cells, a large number (105 - 106) of cells is needed for screening. Thus flow cytometry can be used as a tool to predict relapse of leukaemia pa­tients particularly when subsequent anal­ysis of marrow MRD levels shows an in­creasing trend.

The close relationship between the size of tumour burden and the curability of acute leukaemia is well established. In ALL, a high white blood cell count (WBC) at diagnosis is indicative of a large tu­mour mass and it is a poor prognostic indicator (Campana 2003, Pui & Crist 1994). Other presenting features such as age, sex, lymphadenopathy, organome­galy and the immunophenotype of blast cells are also known to have prognostic values in determination of re­mission in­duction rate, remission dura­tion and long-term survival of patients. Other pa­rameters such as haemoglobin and platelet count may play some role; how­ever, myeloid associated markers have been proven to have no adverse prog­nostic significance in children (Pui et al. 2008). These parameters have been used to classify patients into standard or high-risk groups before appropriate che­motherapy is started. From previous stu­dies, patients in the high-risk group that received standard risk of chemotherapy regime have a higher chance of relapse.

The aim of this study was to establish the detection of residual disease by im­munophenotyping using flow cytometry in childhood ALL during their induction and continuing chemotherapy.

MATERIALS AND METHODS

This descriptive cross sectional study was conducted in children with precursor B-ALL aged between one and 12 years, diagnosed in UKM Medical Centre (UKMMC) from June 2001 to October 2003. It was carried out under a protocol approved by the Institutional Review and Ethics Committee. We obtained bone marrow samples from patients with in­formed consent at diagnosis, and at vari­ous time points that have been defined for standardized analysis of chemothe­rapy treatment based on the UKALL 97 protocol (Working Party On Leukaemia In Children, Revised November 1999). The time points of analysis were: following induction chemotherapy (day-28) and during the continuation therapy (week-12 and week-20) for MRD as­sessment. The specimens of the patients were processed in the Haematology Unit, De­partment of Pathology UKMMC. All pa­tients were followed-up for five years to ascertain their survival status.

Bone marrow smears were made from the marrow aspirations of the patients for morphological analysis and the remain­ing of the marrow specimens were col­lected in ste­rile ethylenediamine tetra-acetic acid (EDTA) tubes for immuno­phenotyping analysis. Differential counts of the pa­tients from peripheral blood were also obtained at the same time. The marrow smears were stained for May-Grunwald-Giemsa’s (MGG) and perox­idase cyto­chemistry. The morphological assess­ments of bone marrow smears were ob­served by two Pathologists and the per­centage of lymphoblasts were recorded. Mononuclear cells from bone marrow aspirates collected in EDTA, were sepa­rated on a density gradient using lym­phoprep (HISTOPAQUE-1077), immu­nolabelled with a panel of mono

clonal antibodies conjugated with FITC and/or PE and analyzed by FACScan flowcyto­meter (Becton Dickinson) for immunophe­notyping. Analyses of data were done using Cell Quest software (Becton Dickison).

Diagnosis of precursor B-ALL was made based on the presence of surface expression of CD19 and cytoplasmic ex­pression of CD 22 with negative cytop­lasmic myeloperoxidase (MPO). The acute lymphoblastic leukaemia cases were included in the study if the blast cells expressed at least one of the fol­lowing combined immunomarkers:

         i.        CD19/CD34

        ii.        CD33/CD19

       iii.        CD13/CD19

       iv.        CD33/CD10

        v.        TdT

For MRD detection, bone marrow sam­ples from all patients during follow-up visits to the hospital were analyzed with the same immunomarkers that were present at diagnosis. Figures 1 and 2 illustrate the examples of cases that were MRD positive and MRD negative respec­tively. Morphological assessments were also compared with the flow cytometry immunophenotyping results. Bone mar­row samples from pediatric cases of idi­opathic thrombocytopenic purpura (ITP) and non-haematological malignancy without bone marrow involvement were used as controls.

Demographic and clinical data such as age, sex, race, presence of lymphadeno­pathy, hepatomegaly or splenomegaly and blood profile of the patients upon diagnosis and at day-28 post induction chemotherapy, were also recorded.

RESULTS

Sample population

From June 2001 to October 2003, 49 children aged one to 12 years with pre­cursor B-ALL were diagnosed in UKMMC. However, minimal residual dis­ease (MRD) could only be studied in only 38 cases. Eleven children had to be ex­cluded from the study because their bone marrow samples were not available at post chemotherapy or there were no suitable MRD markers detected at diag­nosis. Table 1 shows the ranges of age, sex and race of patients in this study.

Clinical findings at diagnosis

Most of the children with precursor B-ALL presented with lymphadenopathy (76.3%), hepatomegaly (89.5%) and splenomegaly (76.3%). Some of these cases had very large liver and spleen (≥5.0 cm and below subcostal mar­gin) i.e. 11 (28%) and 8 cases (21%) re­spec­tively.

White blood cells at diagnosis

Table 2 illustrates the haematological characteristics of the patients at diagno­sis. Most of the cases presented with white blood cell counts (WBC) of less than 50.0x109/L. Only four cases (10.5%) had WBC more than 50.0x109/L with the

highest count being 322.0x109/L (Table 2) and the lowest WBC was 2.0x109/L. The mean WBC was 29.07x109/L.  Most of the patients showed significant reduc­tion in the WBC at day-28 post chemo­therapy.

Residual leukaemic cells by flow cytometry (Table 3)

Residual leukaemic cells (MRD) had been analyzed by flow cytometry at the specified points of time as mentioned earlier. However, only 30 cases at day-28 of chemotherapy were available for MRD analysis, where 11 cases were found to be positive for MRD and 19 cases had no residual disease.  At week-12 of therapy, two cases had to be ex­cluded due to inadequate marrow sam­ples for MRD analysis but the number of cases that became negative for MRD had increased to 25 patients (69.0 %). 

However, at week-20, fewer samples were available for the analysis. There were inadequate specimens in ten pa­tients, one patient had defaulted the fol­low-up treatment and sought traditional medicine while another had succumbed to sepsis.

Correlation between demographic, clinical and laboratory data of ALL cases at diagnosis and the residual disease (MRD) at day-28 of therapy

There were only 30 cases for MRD as­sessment at day-28 of chemotherapy (Table 4). There was no correlation be­tween sex, race or age of the patients at diagnosis with MRD findings at day-28 (p>0.05). There was no correlation be­tween lymphadenopathy and hepatos­plenomegaly at diagnosis with MRD at day-28 (p>0.05). Correlation between WBC, haemoglobin (Hb) and platelet counts at diagnosis with MRD at day-28 were also not significant (p>0.05). Cases with a haemoglobin count of less than 8g/dl and WBC of less than 50.0 x 109/L were negative for MRD. There was also no correlation between the immuno­phe­notype of the blast cells and pres­ence of myeloid aberrant markers  at di­agnosis with MRD at day-28 (p>0.05).

Correlation between morphological and flowcytometric assessments of residual disease (MRD) of ALL cases

In this study, 35 cases achieved com­plete marrow remission at day-28 of in­duction chemotherapy by morphological examination. One case (Table 6; case 6) had a borderline blast cells percentage of 5%. Analyses for MRD by flow cytome­try at day-28 could be performed in only 30 cases and 11 cases were found to be positive for MRD (Table 5) but the mor­phological examination of their marrow specimens showed blast counts of less than 5%.

For the week-12 analysis, bone marrow aspirations could be performed only on 36 of 38 cases and the correlation be­tween marrow findings and MRD at week-12 was found to be statistically sig­nificant (p<0.05). Flow cytometric analy­sis at week-12 showed that there were nine cases with positive MRD although their morphological marrow assessments showed that the percentages of the blast cells were less than 5% (Table 5). One of the cases with positive MRD (case 6) had also not achieved complete marrow remission at day-28 (Table 6). Two cases were found to have blast cells of more than 5% of total nucleated cells by mor­phological examination and they were also positive for MRD by flow cytometric analysis.

Sequential determination of residual disease and survival

There were only 3 cases (Case 1, 3 and 6) with persistent MRD at week-20 of chemotherapy. These cases however, had no MRD assessment at day-28 (Ta­ble 6). Case 1 and 3 had marrow remis­sion morphologically at day-28 but case 6 had borderline blast cell percentage (5%). Case 1 relapsed after one year while case 3 after six months of treat­ment. Both succumbed to treatment complications.  To date, case 6 is still alive, which is more than 5 years after diagnosis.

There were four cases with persistent MRD from day-28 until week-12 of ther­apy but at week-20, three cases achieved a negative MRD status. How­ever, MRD analysis could not be done on one case (case 38) due to insufficient sample. To date, only one patient (case 5) had passed away after 3 years of di­agnosis.

There were 18 cases that were always negative for MRD. However, eight of them died within 12 weeks to 4 years af­ter di­agnosis. Case 8 had relapsed after 3 years of treatment; case 39 had de­faulted and relapsed while others re­mained in remission but unfortunately had succumbed to sepsis. Twenty four patients were still alive after 5 years of follow up.

DISCUSSION

Acute lymphoblastic leukaemia (ALL) is the most common malignant disease-af­fecting children and in Malaysia approx­imately 75% of ALL cases are younger than 20 years of age (Lim et al. 2002). Most of the cases in this study were in the good prognostic age group i.e. two to ten years of age (73.7%) but ma­jority of the cases were males (81%). A longer survival of cases in age group between two and ten years have been reported, which was more than one-and-a-half times as compared to patients  with age of less than two years or more than 10 years, when  treated in an iden­tical man­ner (Pui et al. 2008).

Gender has little or no impact on the success of remission induction and the relapse frequency during the first month of treatment. However, after discontinua

tion of therapy, boys continue to expe­rience a higher incidence of relapse due to relapse in the testes as well as in the bone marrow (Chessells et al. 1995). In this study, 12 of 31 (38.7%) male pa­tients had relapsed and died within five years. The number of female patients involved in this study was small for a comparative study.

Ethnically, the majority (78.9%) of the cases were Malays, followed by Chinese (13.2%) and Indians (7.9%). This obser­vation is consistent with the data of the National Cancer Registry published in 2002, where Malays have the highest incidence of leukaemias (Lim et al. 2002).

The sizes of peripheral lymph nodes, liver and spleen provide an indirect mea­surement of leukaemic cell burden. Sev­eral studies have demonstrated that massive lymphadenopathy; hepatome­galy and splenomegaly adversely affect remission duration and survival (Ojala et al. 1995). However, mild to mod­erate in­creases in the size of lymph nodes and abdominal organs do not ap­pear to influ­ence prognosis of ALL cases. In this study, most of the cases pre­sented with lymphadenopathy (76.3%), hepatome­galy (89.5%) and splenome­galy (76.3%), and enlarged liver and spleen of 5.0 cm or more were seen in 11 cases (28%) and 8 cases (21%) re­spectively. 

The total white blood cell count (WBC) at the time of diagnosis is the single most powerful determinant of remission induc­tion, remission duration, and long-term survival for all age groups (Pui et al. 2008). Patients with higher WBC at diag­nosis have a higher risk for treatment failure than patients with lower WBC, and often have bulky extramedullary disease. WBC of more than 50.0 x 109/L is an in­dication of poor prognosis (Pui et al. 1998; Shu & Chen 2005).

In this study, the clinical and haemato­logical characteristics of children with ALL at diagnosis were compared with the residual disease status detected by flow cytometry, to assess for any correlation. MRD detection on completion of induc­tion therapy was not significantly related to age, gender, race, lymphadenopathy and hepatosplenomegaly, leukocyte count, haemoglobin level, platelet count, immunophenotype of lymphoblast, and presence or absence of myeloid aberrant markers. It appears that the above fac­tors could not be used to predict MRD status at day-28.

Early clearance of leukaemic cells is a favourable prognostic indicator in child­hood ALL. However, identification of re­sidual leukaemic cells by light micro­scopy is subjective and lacks sensitivity. Common therapeutic protocols have considered 5% of blast cells in the bone marrow as the morphologic limit for con­sidering complete remission of ALL pa­tients. Even though intensification of chemotherapy, risk stratification and supportive care have improved the out­come of childhood ALL, 20% of children still succumb to relapse. Thus, a patient declared to be in complete clinical remis­sion might, harbour as many as 1010 leu­kaemic cells. The study of MRD has im­proved the estimation of the number of blast cells present in the bone marrow during complete remission and during or after therapy. Thus it may be helpful to design a risk-adapted therapy in patients with acute leukaemia. The ability to detect residual leukaemic cells while they are still sensitive to treatment would al­low for potentially curative intervention. The persistence of small numbers of leu­kaemic cells post induction chemothe­rapy in childhood ALL, indicates the re­quirement of prolonged therapy (2.5 to 3 years) to reduce the relapse hazard (Pui & Crist 1994; Hoelzer 1994).

The sensitivity of MRD detection using flow cytometry depends on the numbers of cells analyzed and also on the MRD markers used. For optimal flow cytometry results it is imperative that the bone mar­row specimen contain a large number of viable with a good separation of mono­nuclear cells. The sample preparation and live gate acquisition should be per­formed within 24 hours. If the samples have remained unprocessed for a longer duration, there will be changes in the FSC/SSC pattern and the increasing population of dying or apoptotic cells will interfere with the analysis and interpreta­tion of the results. Post chemotherapy bone marrow samples also have a large number of new regenerating normal lym­phoid cells (haematogones) which may express CD34, CD19, CD10 and TdT and may affect the analysis of results (Coustan-Smith et al. 2006).

We had used a two-color flow-cytome­try for measurement of MRD. Patients with a high tumour burden at diagnosis and have MRD more than 1% post che­motherapy are more likely to relapse early (Coustan-Smith et al. 2000). In this study, there were three cases with per­sistent MRD positivity and the correlation study between marrow findings and MRD status at week-12 was found to be statis­tically significant (p<0.05). Two of these patients had an early relapse and suc­cumbed to the disease. Marrow remis­sion status post chemotherapy of pa­tients with MRD less than 1% were questionable (Coustan-Smith et al. 2000). There were 18 cases in this study that had not shown MRD positivity at any point of time but eight of them died within a period of 12 weeks to 4 years after di­agnosis.

Coustan-Smith et al. (2006) showed that by using multi-coloured flow-cyto­metry, MRD can be detected up to 0.01% and the cases that showed less than 0.01% MRD positivity at the end of re­mission induction are likely to have an excellent treatment outcome, compared to those with more than 0.01% (Coustan-Smith et al. 2006).

In the Paediatric Haemato-Oncology Unit UKMMC, stratification of intensive treat­ment usually depends on the pre­senting WBC and age of the patient. Following induction chemotherapy, pa­tients who had more than 5% of blast cells in the bone marrow morphologically and pa­tients with nervous system (CNS) or tes­ticular relapse will be given a more inten­sive chemotherapy regimen. Thus MRD analysis by flow cytometry can also be used to design a risk-adapted therapy in childhood ALL. The ability to detect resi­dual leukaemic cells while they are still sensitive to treatment would allow more time, and potentially curative inter­ven­tion. The persistence of small num­bers of leukaemic cells post induction chemothe­rapy in childhood ALL indicate the re­quirement of prolonged therapy (2.5 to 3 years) to reduce the relapse hazard (Campana 2003; Hoelzer 1994).

This study was unable to prove that the risk of relapse was high in patients with persistent MRD since a few cases which were MRD negative also relapsed within 4 years after diagnosis.

CONCLUSION

This study showed that there was no sig­nificant correlation between MRD positiv­ity by flow cytometry with the demo­graphic, clinical or haematological data at diagnosis. However, there was a signifi­cant correlation between MRD positivity by flow cytometry with morphological analysis of blast cells by light micro­scopy. For better evaluation of MRD, a three or four  colour flow cytometry anal­ysis should be used and a larger cohort of ALL cases are required, and would then aid in providing a basis for future clinical decision making in the manage­ment of ALL cases.

ACKNOWLEDGEMENT

The authors gratefully acknowledge and thank Mrs Sivagengei for her technical assistance and advice regarding flow cytometry.