Cytogenetic Profile in 7209 Indian Patients with de novo Acute Leukemia : A Single Centre Study from India

Background: Cytogenetics is one of the most important diagnostic parameters in the classification of acute leukemia. Recurrent chromosomal aberrations in acute leukemia have provided insights into the molecular mechanism of leukemogenesis. The variable frequencies of recurrent cytogenetic markers due to ethical/racial differences have been reported from Western and some Asian countries. Objective: We report cytogenetic data of largest cohort of 7209 adult and pediatric patients with de novo acute leukemia (AL) to determine the prevalence of various cytogenetic sub groups and compare with the Western and Asian population. Material & Methods: The AL patients included 2609 AML (adult: 2042, pediatric: 567), 3708 B-cell-precursor (BCP)-ALL (adult: 1300, pediatric: 2408) and 892 cases of T-ALL (adult: 480, pediatric: 412). Cytogenetic studies included conventional karyotyping and FISH using panel of probes. Results: The incidence of t(8;21) was high, comparable to other Asian countries. In comparison to our series and Western population, t(15;17) was more prevalent in Chinese population. Cytogenetic profiling of BCP-ALL revealed low prevalence of ETV6/RUNX1 in ours as well as other Asian population. The MLL aberrations in BCP-ALL and TLX1 & TLX3 aberrations in T-ALL occurred less frequently in our series as compared with Western population. Conclusion: The present study with a large cohort showed the heterogeneity of AL that involved various factors, such as age, gender and prevalence of distinct cytogenetic subgroups. Our data in comparison with other population based studies revealed differential distribution of some cytogenetic sub-groups indicating geographic heterogeneity due to differential environmental exposure which probably influenced underlying genetic susceptibility. Corresponding author. P. S. K. Amare et al.

As in AML, recurrent chromosomal aberrations observed in acute lymphoblastic leukemia (ALL) have provided insights into the molecular mechanism of leukemogenesis [11] [12].Around 80% of childhood and adult B-cell precursor-ALL (BCP-ALL) reveal recurring numerical and structural alterations such as translocations, deletions and copy number changes [13]- [18].
Recurring chromosomal abnormalities with variable prognosis have also been identified in 70% of T-ALL which includes hypodiploidy, 9p deletion, translocations involving T-cell receptor genes TCR-α and -β, transcription factor genes viz., TAL1, LYL1, TLX1/HOX 11 and TLX3/HOX11L2 [15] [22]- [24].Fluorescence in situ hybridization (FISH) and conventional cytogenetics are the main strategies for cytogenetic workup of acute leukemia in disease management [13] [25]- [27].Independent FISH analysis is used in several laboratories due to high sensitivity for detection of various cryptic structural and submicroscopic abnormalities, and also copy number changes in AL within short turnaround time [25]- [30].
The variable frequencies of recurrent cytogenetic markers due to ethical/racial differences have been reported in some published studies from Western and some Asian countries [31]- [36].In the current study, we present cytogenetic data of a largest cohort of 7209 patients with de novo acute leukemia from a single centre.The aim of the study is to analyze the cytogenetic pattern with prevalence of various recurrent cytogenetic markers and to compare it with reported literature.

Material and Methods
The study was conducted at the Department of Medical Oncology, Tata Memorial Hospital, Mumbai, India.It is the largest cancer diagnostic, treatment and research centre in India, with 40,000 -50,000 new cancer cases being diagnosed every year.The present study includes 7209 untreated patients that were newly diagnosed with AL between January, 2008 and December, 2015.Of these, 4600 cases were of ALL and 2609 cases were of AML that included patients diagnosed with acute promyelocytic leukemia (APL).Among the 2609 AML cases, 567 were pediatric and 2042 were adult cases, whereas the 4600 cases of ALL consisted of 2408 cases of pediatric BCP-ALL, 1300 cases of adult BCP-ALL, 480 cases of adult T-ALL and 412 cases of pediatric T-ALL.
The initial diagnostic criteria included standard clinical and laboratory findings including immunophenotyping.On the other hand, cytogenetic workup by FISH and conventional cytogenetics is an integral part of routine classification, risk stratification with appropriate treatment decision and monitoring of the disease.Cytogenetic studies were performed using standard protocols preferably in bone marrow aspirate and in peripheral blood in around 10% of cases, and occasionally on bone marrow morphology smears.At least 15 -20 GTG-banded metaphase cells were karyotyped and analysed as per ISCN 2009 and 2013 [37] [38].In ALL, ploidy analysis was done by counting chromosome number in 30 -40 metaphase cells.Ploidy was classified into various subgroups: Low and high hyperdiploidy, low and high hypodiploidy, near haploidy, diploidy and near triploidy/tetraploidy.
Age-related prevalence and variable ploidy association of recurrent chromosomal alterations were analyzed by the Pearson's chi-square test for determining the statistical significance (SPSS version 20).

Incidence of Recurrent Cytogenetic Abnormalities in BCP-ALL Patients
Among the 2408 patients with pediatric Pre-B ALL, 1592 were males and 816 were females (M/F ratio of 2) in the age range of 2 months -15 years (median age 5 years) and 137 cases were ≥1 years.Among the 1300 adult patients, 930 were males and 370 were females (M/F ratio of 2.5) in the age range of 16 -77 years (median age 28 years).

Incidence of Recurrent Cytogenetic Abnormalities and Age and Ploidy Distribution in T-ALL
Among the 480 adult patients with T-ALL, 391 were males and 89 were females (M/F ratio of 4.3) in the age range of 16 -78 years (median age 24 years).Among the 412 pediatric T-ALL cases, 329 were males and 83 were females (M/F ratio of 4) in the age range of 1 -15 years (median age 10 years).
There was no association of ploidy or variable age groups with cytogenetic subgroups, except for association of 9p deletion with young adults ≤40 years (p = 0.0005) and with diploidy (p = 0.03) (Table 5(

Discussion
The comprehensive analysis of heterogeneity of acute leukemia of the large cohort in the present study has demonstrated the involvement of various factors that involves age, incidence pattern, gender and genetic factors.Several studies from various locations have shown the geographic heterogeneity in acute leukemias [6] [20] [27] [30]- [32] [34] [36] [39] [40].The cytogenetic pattern of acute leukemia in Indian population however is not clear due to lack of large scale studies.Hence, the main approach of the present study was to retrospectively analyze a large cohort to focus upon the epidemiology of various cytogenetic subtypes which may reveal the involvement of various etiological factors, thereby improving our understanding of the biology of the disease and thus help in better management of the disease.
Our combined approach of conventional cytogenetics and FISH significantly improved the detection rate of submicroscopic cryptic rearrangements within a short turnaround time.We could detect PML/RARA with RARA variants in 98% of APL cases.Table 1 summarizes the prevalence of recurrent chromosomal abnormalities in pediatrics and adult AML in present study and compares the incidence with different populations from USA, European and Asian countries.
The higher incidence of t (15;17) in Chinese population could be due to high prevalence of APL in their population [3] (16) in adult and pediatric AML also indicated that incidence of favorable markers are most common in children and young adults compared to older adults [3] [6] [8] [39] [44] [46].Although overall MLL frequency in both adult and children were comparable with Western populations, our study showed lower (3%) incidence of t(9;11) in pediatric AML, similar to that in Chinese population [8] [31] [39] [44]- [48].The translocation (4;11) occurred less frequently in pediatric AML probably due to lower number of infantile cases below one year in our series.The aberrations of chromosome 5, 7 and 8 were comparable with other studies however, these aberrations occurred with higher frequency compared to that in both adult and pediatric Chinese population [6] [31] [39] [40] [47].The RUNX1 and MLL amplification frequencies (both 1%) of adult AML in our study have been reported as rare secondary events with dismal outcome [49] [50].
In BCP-ALL, the age and gender distribution was comparable to reports from other geographic areas [14] [15] [17]- [19] [41].BCP-ALL has various cytogenetic subtypes which have significant impact on risk stratification and hence remain strong independent indicators of disease outcome [13] [14] [16] [17] [19]- [21].The incidence of high hyperdiploidy in children (27%) and adult ALL (9%) are similar to that of Western population [12] [16] [51].The overall hyperdiploidy rate observed in our study was probably due to slightly higher incidence of low hyperdiploidy in our patients.As reported in literature, hyperdiploidy in childhood ALL was strongly associated with gain of chromosomes 4, 10, 17 and 21 (Table 3(b)) [12] [51].The hypodiploidy incidence was comparatively higher in both pediatric and adult populations due to elevated frequency of high hypodiploidy.Near triploidy-tetraploidy, which is rare in childhood ALL was observed both in childhood and adult ALL with incidence of 1% -2% [13] [52].Near haploidy with chromosome numbers 23 -29 is a rare hypodiploid group in pre-B-ALL occurring in 0.3% -0.5% in ALL group in our study falling within the universal frequency of <1% [13] [53].
The ETV6-RUNX1 is another common favorable subset of pediatric BCP-ALL that occurs with an average frequency of 25%, but showed lower incidence of 12% in our study.This finding is consistent with our previous study [27] and another cohort of 928 patients [Nahar A et al. in communication].Data from various studies from Asian countries also indicated low prevalence (13% -19%) of ETV6/RUNX1 [54]- [56].Studies have reported incidence rates of <10% from other parts of India, however, these were smaller studies with sample size not large enough to focus on true prevalence [57] [58].The peak incidence of ETV6/RUNX1 was in lower age group of 1 -10 years as reported in our previous studies as well as in studies from other Asian countries [27] [52] [55] [59].Trisomy 21 and ETV6 allelic loss were most frequent additional abnormalities in ETV6/RUNX1 positive group, however these abnormalities were common in overall BCP-ALL patients [27] [59].
The (iAMP21), a distinct high risk cytogenetic subtype was identified with 1% incidence in pediatric BCP-ALL patients predominantly in older children of 3 -14 years age group which is in accordance with other reported studies [21] [60].
In our study, the t(9;22) incidence in childhood ALL was 6%, whereas it showed variable frequencies of 2% -6% in Western and other Asian populations [14] [36] [59].The t(9;22), which is the most common chromosomal abnormality in adult ALL occurred in 27% of the adult patients, falling within the universal range of 25% -30%.It was more prevalent in older children as well as adult ALL under 40 years.Literature also indicate rising incidences of t(9;22) in older patients [61]- [63].Geographic heterogeneity of variable frequencies of BCR/ABL1 has been reported [14] [62] [63].
The translocation (1;19) is another cytogenetic subgroup of BCP-ALL that occurs more frequently in children than adults.The incidence in both pediatric and adult patients in our study was closer to other reports [61]- [64] except in Chinese population which showed comparatively lower frequency [36].
Translocations involving MLL (11q23) gene occur in up to 5% of childhood and adult BCP-ALL [19] [61]- [63].The lower incidence of 2% -3% observed in the present study as well as in studies from other Asian countries indicate that MLL rearrangement incidence vary in different populations [36].Data from previous studies from our lab supports the finding that MLL rearrangement incidence is low in Indian population [30].The t(4;11) that is known for its adverse outcome is the common MLL translocation in childhood ALL with higher frequencies (60% -80%) in infants [61]- [63].The lower incidence of 38% of t (4;11) in children in the present study is due to the low number of infants in our study.The incidence of t(4;11) was also found to be lower in childhood AML in the present study indicating under representation of infantile leukemia in India.This could be due to poor tendency of referring infantile leukemia cases to tertiary centres.Pattern of age distribution in MLL and t(1;19) positive groups as observed in other studies, suggested that these translocations tend to occur in lower age group [19] [30] [52] [61]- [64].The ploidy distribution revealed association of favorable t(12;21)/ ETV6-RUNX1, t(1;19) as well as unfavorable t(9;22), MLL translocation subtypes with diploidy.
T-ALL commonly affects young males than females.This was evident from the data obtained in the present study which revealed 3 -4 times higher prevalence of T-ALL in young males [13] [15] [65].Hypodiploidy is common in T-ALL as compared with Pre-B-ALL.It was comparatively higher in the present study in both pediatric and adult cases [13] [61].On the contrary, MLL translocation frequency was lower in adult patients in our study than that of western population [66].As reported previously, 9p deletion and translocation involving T-cell receptors TCR-α (14q11.2) and TCR-β (7q34), transcription factor genes TLX1 (HOX11) (10q24) and TLX3 (HOX11L2) (5q35) occurred in more than 50% of cases [22]- [24] [67] [68].The incidence of TCR-α and TCR-β translocations in pediatric and adult T-ALL was similar to previous reports [65] [69].Among the most common partner genes, the TAL1, LYL1, LMO1, LMO2, TLX1 and TLX3 of TCR-α and TCR-β translocations, TLX1 is associated with better outcome, whereas TLX3 has been found to have increased risk of treatment failure [22] [23].
In comparison to previous reports, the incidence of TLX1 and TLX3 translocations in the present study was lower in both pediatric and adult patients [22] [65] [69].The study also observed that 9p deletion was common in young age group and associated with diploidy as has been reported previously [15] [22]- [24] [46] [69] [70].

Conclusion
The present study with a large cohort showed the heterogeneity of AL that involved various factors, such as age, gender and prevalence of distinct cytogenetic subgroups.Cytogenetic data when compared with data from other geographic areas revealed geographic heterogeneity due to differential environmental exposure which probably influenced the underlying genetic susceptibility.

Table 3 .
(a) Cytogenetic markers and age distribution in Pediatric and adult BCP-ALL; (b) Cytogenetic markers and ploidy distribution in Pediatric and adult BCP-ALL.

Table 4 .
Incidence of recurrent cytogenetic aberrations in present series, adult T-ALL (n = 480), pediatric T-ALL (n = 405) and comparison with other populations.

Table 5 .
(a) Cytogenetic markers and age distribution in Pediatric and Adult T-ALL; (b) Cytogenetic markers and ploidy distribution in Pediatric and Adult T-ALL.