INTRODUCTION  

Acute  lymphoblastic  leukaemia  (ALL)  is one  of  the  common  malignancies  of  the pediatric age group. It results from clonal proliferation  of  lymphoid  precursors  with arrested  maturation.  The  disease  can originate  from  lymphoid  cells  of  B-  or  T- lineage,  although  mixed  lineage  leukae- mias  do  rarely  occur.  Its  classification  is based  on  morphological  and  cytochemi- cal  characteristics  established  by  the French-American-British  (FAB)  group (Bennett et al. 1981). Today, the diagno-sis  of  ALL  is  further  facilitated  by  flow cytometry.  Flow  cytometric  analysis  al- lows  for  sub-classification  of  ALL  by  uti- lizing  flourochrome-labelled  monoclonal antibodies.  About  60%  of  ALL  patients have early pre-B immunophenotype. The blast  cells  usually  express  CD  19, cytoplasmic  CD  22,  CD  79α,  CD  10, terminal  deoxynucleotidyl  transferase (TdT)  and  75%  express  CD  34.  T-cell ALL  accounts  for  12-15%  of  ALL  cases. The  blast  cells  express  surface  CD  7, cytoplasmic CD 3, CD 2, CD 5 and TdT.  The  blast  cells  may  also  express aberrant antigens which do not belong to the  main  lineage.    Immunological characterization  of  acute  leukaemia  is therefore essential in defining cell lineage to  aid  treatment  options  and determination of prognosis. The  lineage  of  the leukaemic  blast usually  remains  similar  throughout  the disease  course,  although  the pattern  of antigen expressions may change slightly.  However,  a switch  of  lineage  of  blast cells  does  rarely  occur.  Pui et al. described  cases  which  showed  switch from  lymphoid  to myeloid  phenotype  or the  reverse  (Pui  et  al.  1986).  From literature  review,  phenotypic  lineage switch rarely occurs from T-ALL to B-ALL or  vice  versa  and  to  the  best  of  our knowledge, only one case has described this  phenomenon  (Besalduch  et  al. 1990).  Our  case  illustrates  an unusual phenotypic switch in leukaemic cells from a  patient  who first  presented  with  T-cell acute  lymphoblastic  leukaemia  who  had gone  into  remission  and  relapsed  seven months later with common pre-B ALL. 

CASE REPORT

A  10-year-old  boy  presented  with  two weeks history of low grade fever, lethargy and  poor  appetite.  The  patient  was  the youngest  of  four  siblings.  His  father,  a 36-year-old  man  worked  as  a printing officer  at a private  company  and his mother,  a 36-year-old  woman,  was  a housewife.

At  presentation,  the  patient  was  pale and  in  respiratory  distress.  His  blood pressure  was  105/40  mmHg,  pulse  rate of  168 bpm  and oxygen  saturation  was 100%  under  nasal  prong  oxygen.  Physi- cal examination revealed pleural effusion over  the  right  lung  and cardiomegaly.  There  were  hepatosplenomegaly  mea- suring 5 cm and 4 cm below the subcos- tal margins respectively.

Laboratory investigation results showed haemoglobin  (Hb)  4.9  g/dL,  total  white blood  cells  (WBC)  41.5x109/l,  platelets 37x109/l,  neutrophils  0.8x109/l,  lym- phocytes  38.6x109/l.  The  peripheral blood smear showed presence of numer- ous  blast  cells.  The  blast  cells  exhibited high  nuclear  to  cytoplasmic  ratio  with  in- conspicuous nucleoli (Figure 1).

His  chest  radiograph  showed  cardi- omegaly  with  mediastinal  widening.  The bone  marrow  aspirate  showed  presence of more than 90% blast cells which were homogenous  in  size,  showing  high  nuc- lear  to  cytoplasmic  ratio  with  basophilic cytoplasm  and  inconspicuous  nucleoli. Other  cell  lines  were  depressed.  The blast  cells  were  negative  for  peroxidase and  periodic  acid  Schiff  stains.  These findings  were  consistent  with  acute  lym- phoblastic  leukaemia,  FAB-L1.  The  tre- phine  biopsy  showed  similar  findings.  Flow cytometric analysis of the peripheral blood and bone marrow aspirate showed that the population was positive for T-cell markers; CD 2, intracellular CD 3, CD 4, CD 5, CD 7, CD 8, HLA-DR but negative for CD 19, CD 79a, CD 56, CD 117, intra MPO, CD 13 and CD 33 (Figure 2). Marrow  cytogenetic  study  revealed  a 46 XY normal karyotype with negative BCR-ABL translocation.

The patient achieved remission with UK ALL 97 protocol regime B induction che- motherapy.  However,  at  week  20,  his bone marrow aspirate revealed presence of  excess  blast  cells  of  10%.  Flow  cyto- metry  analysis  was  inconclusive  due  to sample  inadequacy.  He  subsequently received  Doxorubicin,  Vincristine,  Dex- amethasone  and Asparaginase  as  per delayed  intensification  1 chemotherapy protocol which was complicated by bron- chopneumonia  secondary  to febrile  neu- tropaenia.  The  repeat  bone  marrow  ex- amination  at  week  25 showed  persis- tence  of  excess  of  blast  cells  (7%).  The bone  marrow  examination  was  repeated at  week  29 and  showed  presence  of more  than 26%  blast  cells  which  com- prised  of  two  populations  of  blast  cells (Figure  3).  One  population  was  larger with  abundant  basophilic  cytoplasm (morphologically  L2 blast cells)  while the other  population  was  smaller,  more  ho- mogenous  with  very  scanty  cytoplasm (L1  blast  cells).  The  two  populations  of blast  cells  showed  variable  antigen  ex- pressions. Both populations of blast cells expressed  CD  19,  anti-HLA-DR,  CD  10 and  cytoplasmic  CD  22. One  population of  blast  cells  was  CD  34  positive  with negative  CD  20 while  the  other  popula-tion  was  CD  34 negative  with  positive cytoplasmic  CD  20. The  blast  cells  were negative  for  intra  CD  3, CD  7, CD  117, intra-MPO, CD 13 and CD 33 (Figure 4). A  diagnosis  of  precursor  B-cell  acute lymphoblastic  leukaemia  with  CD  10 positivity  (CALLA)  was  then  made  and the  patient  was  started  on  the German Relapse  Protocol  (GRP).  Repeat  bone marrow examination after four months of relapse  showed  excess  of  blast  cells (10%).  The  blast  cells  were mostly  large in  size  with  relatively  abundant  cytop- lasm.  A second  course  of chemotherapy was  given  for  eight  days.

Review  of  bone  marrow  examination after  completing  two  courses  of  chemo- therapy showed persistence of blast cells (10%  blast  cells).    In  view  of  this  condi- tion,  bone  marrow  transplant  was  no longer  a  therapeutic  option.  The  man- agement  was  then  changed  to  palliative care which included low-dose chemothe- rapy and prophylactic treatment for infec- tions.

DISCUSSION 

Leukaemic stem cells have heterogenous differentiation potential.  It has been well described  in  various  articles  that  many different  patterns  of  inappropriate  ex- pression  of  lineage  markers  were  ob- served  in  acute  leukaemias.  The  spec- trum of mixed acute leukaemias includes biphenotypic  leukaemia  with  markers  of more than one lineage of antigen expres- sions  present  on  the  same  blast  cells, bilineal  leukaemia  with  two  distinct  pop- ulations, and biclonal leukaemia with two distinct  leukaemic  cells  of  independent origins  (Hoffbrand  et  al.  1988).  These terms  have  recently  however  been  re- vised  in  the  WHO  classification  2008.  The  term  mixed  phenotypic  acute  leu- kaemia  (MPAL)  is  now  applied  to  both biphenotypic  and  bilineal  leukaemia  in general,  with  more  specific  terms B/myeloid  (B/MY)  and  T/myeloid  (T/MY) leukaemia  which  refer  to  leukaemias containing  the  lineages  specified,  irres- pective of whether one or more than one population of blast cells is seen (Borowitz et  al.  2008).  There  are  cases  of  MPAL based  on  one  criterion  at  diagnosis  (e.g biphenotypic  leukaemia)  which  change over time at relapse to the other (bilineal leukaemia),  or  vice  versa.  Furthermore, following  therapy,  persistent  disease  or relapse may occur as either pure ALL or AML.  In  some  cases,  this  phenomenon has been termed ‘lineage switch’. 

The  phenotypes  of  blast  cells  from  pa- tients  with  ALL  upon  relapse  most  often adhere to the original lineage (Greaves et al. 1980). However, blast cells which un- dergo a ‘lineage switch’ during the course of  disease  especially  during  relapse  had been  estimated  in  some  series  to  be  as high  as  8%  (Stass  et  al.  1986).  In  this case, a phenotypic lineage switch is seen where  the  blast  cells  initially  expressed T-cell  markers  (cytoCD3+,  CD2+,  CD4+, CD8+,  CD5+,  CD7+)  at  diagnosis.  How- ever at relapse, the blast cells were posi- tive  for  B-cell  markers  (CD19+,  in- traCD22  and  CD20)  and  also  showed positivity  towards  CD10,  CD34,  HLA-DR but  negative  for  cytoplasmic  CD3,  CD7, CD117 and anti-MPO.

The  mechanism  of lineage  switch  re- mains unclear. Most cases demonstrated consistent  chromosomal  findings  at  the time  of  diagnosis  and switch  and were therefore  thought  to be  clonal  in origin (Stass  et al.  1984,  Mantadakis  et al. 2007).  It  was  hypothesized  that chemo- therapy  might  have  induced  suppression of a clone apparent at diagnosis, allowing the  expansion  of  another  clone  with  dif- ferent  markers,  which  was  resistant  to treatment  (Stass  et al.  1984).  Alterna- tively,  the  therapy  might  have  a leuke- mogenic  effect  on a normal  hemopoietic cell  and  the  lineage  switch  might  then represent  the  induction  of  a new  malig- nancy  rather  than  transformation  of  the initial  leukaemic  cells.  In support  of  this view,  there  were  cases  described  with  a complete  variation  of  cytogenetic  mark- ers at diagnosis and relapse (Park et al. 2011)  and  after  allogenic  bone marrow transplantation  (Brito-Babapulle  et  al. 1989).  However,  therapy-induced  acute leukaemia  had a long  latency  period  of about 69 months as described in a large multicentre  GIMEMA  trial  reported  by Pagano et al. in 1998.

As  in this  case,  there  was  a rapid  li- neage  switch  with  an interval  of  just seven  months  from  first  presentation  to relapse. As the time frame was too short, it  was  very  unlikely  for  secondary  malig- nancy  to have  set  in.  Thus,  the most likely pathogenesis for the lineage switch in  this  case  was  chemotherapy-induced suppression  of  the apparent  leukaemic clone at diagnosis, while allowing the ex- pansion  of  another  clone  with  different markers  which  was  resistant  to treat- ment.  A similar  conclusion  was  also reached  by  Shivarov  et al.  who  reported a very early onset lineage switch from B- ALL  to AML in an adult  patient  within  a period  of  only  3 months  after  the initial diagnosis (Shivarov et al. 2009).

A  variety  of  genetic  anomalies  has been  reported  in cases  of  ambiguous lineage  leukaemias  especially  mixed phenotype  acute  leukaemia.  The t(9;22)(q34;q11) BCR-ABL1 translocation and  translocations  associated  with  the MLL gene are the two most common ge- netic  anomalies  associated  with  MPAL.

In  the case  of  lineage  switch,  the  data are scanty, but two cases have been re- ported  to  involve  t (9;22)(q34;q11)  BCR- ABL1  translocation  in two  cases  of Ph+ ALL to Ph+AML (Pane et al. 1996, Rear- don et al. 1994).

In  a  study  of  immunophenotypic  and cytogenetic changes at relapse by Hur et al.  in  2001,  of    99  patients,  51  patients (51.5%) had immunophenotypic changes at  relapse  with  expression  of  aberrant markers  to  be  more  frequent  at  relapse than  at  initial  diagnosis  especially  in  B lineage  ALL  (41.1%  versus  10.3%).  Cy- togenetic  changes  at  relapse  were  ob- served in 28 of 46 patients (60.8%). The initial  abnormal  karyotypes  were  more frequently  associated  with  clonal changes  at  relapse  compared  to  initially normal  karyotypes  (78%  versus  43%) (Hur et al. 2001). 

Loss of CD10 and other antigens at re- lapse has been reported in about 15% of ALL  patients,  however  a  complete change  of  phenotype  from  T-cell  ALL  to B-cell ALL is rare. From literature search, only  one  case  in  1990  had  the  same phenomenon.  Besalduch  et  al.  in  1990 described  a  case  of  a  17-year  old  boy who  first  presented  with  massive  hepa- tosplenomegaly  and  a  mediastinal  mass and  was  diagnosed  to  have  T-cell  acute lymphoblastic  leukaemia.  He  was  given standard  chemotherapy  and  achieved remission. However, he relapsed 4 years later as precursor B-ALL. 

In a study by Stass et al. in a series of 264  newly  diagnosed    acute  leukaemia, of    the  239  patients  who  attained  com- plete  remission,  89  relapsed  with  six cases  (6.7%)  demonstrating  lineage switch  from  acute  lymphoblastic  leukae- mia  to  acute  myeloid  leukaemia  or  vice versa.  Of  the  six  cases,  four  had  T-cell associated  characteristics  either  at  diag- nosis (three cases) or at lineage conver- sion  (one  case)  suggesting  that  T  lymphoblasts  may  have  the  potential  for myeloid  differentiation  or  that  there  may be  a  multipotential  progenitor  cell  capa- ble of both T lymphoblast and myeloblas- tic  differentiation.  Of  30  patients  with  T- ALL  at  diagnosis,  10  of  the  89  patients who relapsed had T-cell ALL. Thus, 30% of  T-ALL  patients  who  relapsed  demon- strated  lineage  switch.  The  issue  is  hig- hlighted  here  to  show  the  heterogeneity of  T-lymphoblast  differentiation  which,  in this  case  evolved  to  B-lymphoblast  at lineage  conversion.  This  study  also  indi- cated  that  lineage  switch  in  acute  leu- kaemia  was  more  common  than  might have been expected, possibly because of the wider use of intensive chemotherapy. More studies are needed to identify the target  cells  in  leukaemogenesis  to  eva- luate  their  spontaneous  or  drug  induced capacity  for  lineage  deviation  and  diffe- rentiation,  and  to  elucidate  the  relation- ship between the genetic and phenotypic characteristics of leukaemic cell lineages. Prompt  recognition  of  lineage  switch may  be  helpful  in  selecting  an  effective therapeutic  regimen.  Ultimately,  therapy appropriate  for  the  leukaemic  phenotype at the time of lineage switch may be ad- visable. 

CONCLUSION

Conversion  of  T-cell  lineage  ALL  to  B-cell lineage ALL is a rare occurrence. This case illustrated  a phenotypic  lineage switch from a characteristic T-ALL to common precursor B-ALL.  In  our  case,  lineage  conversion occurred  seven  months  after  the  initial diagnosis which was postulated to be likely due  to  chemotherapy  which  eradicated  the dominant  clone  apparent  at  diagnosis, permitting  expansion  of  a  secondary  clone with  a  different  phenotype.  Rapid recognition  of  this  condition  is  needed  to tailor  treatment  decision  and  strategy  for disease monitoring.