Is B-Lineage Acute Lymphoblastic Leukemia With a Mature Phenotype and L1 Morphology a Precursor B-Lymphoblastic Leukemia/Lymphoma or Burkitt Leukemia/Lymphoma? Open Access

In the French-American-British (FAB) classification scheme, acute lymphoblastic leukemia (ALL) is divided into 3 morphologic subtypes, L1, L2, and L3, according to cell size, amount of cytoplasm, and prominence of nucleoli of the leukemic blasts in Romanowsky-stained bone marrow aspirate smears.1 With the advent of comprehensive immunophenotyping with lineage-specific or lineage-associated antibodies, ALL is further divided into either T- or B-lineage ALL. B-lineage ALL with L1 and L2 morphology typically has an immature phenotype, whereas that with L3 morphology usually has a mature phenotype.

In the most recent World Health Organization classification of hematolymphoid neoplasms, B-lineage ALL with an immature phenotype is classified as precursor B-lymphoblastic leukemia/lymphoma, whereas that with a mature phenotype is classified as Burkitt leukemia/lymphoma, a subtype of peripheral B-cell lymphoma.2 Precursor B-lymphoblastic leukemia/lymphoma is characteristically negative for surface immunoglobulin light chain expression but positive for nuclear terminal deoxynucleotidyl transferase (Tdt) activity. In contrast, Burkitt leukemia/lymphoma typically shows surface immunoglobulin light chain restriction and nonrandom chromosomal translocations (8;14)(q24;q32), (2;8)(p12;q24), or (8;22)(q24;q11) involving the c-myc gene. This distinction is not only biologically relevant but also clinically significant, since the 2 entities respond to different treatment regimens.

The distinction between precursor B-lymphoblastic leukemia/lymphoma and Burkitt leukemia/lymphoma is usually straightforward by flow cytometric immunophenotyping. Sometimes, however, it may be difficult to distinguish these 2 entities based on immunophenotypic findings alone. For example, an immature phenotype has been reported in rare cases of Burkitt leukemia/lymphoma.3–5 Similarly, surface immunoglobulin light chain expression has been seen in rare cases of typical precursor B-lymphoblastic leukemia/lymphomas.6–12 Herein, we report 2 unusual cases of pediatric B-lineage ALL to further illustrate the difficulties in their diagnosis, classification, and management.

The patient was a 13-month-old boy who was transferred from an outside institution with complaints of irritability, right leg pain for 2 days, and low-grade fever for 1 to 2 weeks. He did not have any history of easy bruising, bleeding, shortness of breath, cough, vomiting, or diarrhea. On physical examination, he was found to have significantly pale conjunctivae and oral mucosa. There was also moderate cervical and inguinal lymphadenopathy and mild splenomegaly. Chest x-ray examination showed no pulmonary infiltrates or mediastinal enlargement. Laboratory studies were significant for profound anemia, thrombocytopenia, and leukocytosis (hemoglobin, 8.3 g/dL; hematocrit, 24.7%; platelets, 63 × 10³/μL; white blood cell count, 60.1 × 10³/μL) and a mildly elevated lactate dehydrogenase level (561 U/L). A bone marrow aspiration biopsy specimen revealed numerous lymphoblasts. The patient was treated with prednisone, vincristine, daunorubicin, and l-asparaginase as defined by the Children's Cancer Group (CCG) leukemia protocol 1961. Additional bone marrow aspiration biopsy specimens taken after therapy showed a “rapid early response” at day 7 with less than 5% residual lymphoblasts and complete remission by day 28. At our last follow-up 9 months after initial diagnosis, the patient remains in complete clinical, morphologic, and immunophenotypic remission.

The patient was a 12-month-old girl who presented initially with a nonprogressive, purplish skin lesion on the mons pubis of 2 months' duration without other systemic symptoms. A biopsy specimen of the lesion revealed a leukemic infiltrative process. Subsequent evaluation after referral to our institution revealed mild cervical lymphadenopathy and moderate hepatosplenomegaly. Laboratory studies revealed leukocytosis with circulating blasts, with otherwise normal hematologic indices (white blood cell count, 31.5 × 10³/μL; hemoglobin, 13.2 g/dL; platelets, 261 × 10³/μL) and a slightly elevated lactate dehydrogenase level (622 U/L). Chest x-ray examination did not reveal mediastinal or hilar adenopathy. Bone marrow aspiration showed replacement with lymphoblasts. She began chemotherapy with dexamethasone, vincristine, and asparaginase per CCG ALL protocol 1991 and achieved a “rapid early response” with clearance of bone marrow lymphoblasts by day 14 of therapy. However, daunorubicin was added after “unfavorable” cytogenetic abnormalities were found. She achieved complete morphologic and immunophenotypic remission by day 35 and continues to do well with chemotherapy 4 months from diagnosis.

Peripheral blood and bone marrow aspirate smears were stained routinely with Wright and Giemsa stains and evaluated morphologically.

Flow cytometric immunophenotyping was performed on peripheral blood and/or bone marrow aspirate specimens using the standard acute leukemia protocol in our laboratory with a FACSort flow cytometer (Becton Dickinson, San Jose, Calif). Briefly, the specimens were treated with ammonium chloride to lyse red blood cells and stained with various combinations of different fluorochrome-conjugated antibodies against CD2, CD3, CD4, CD5, CD7, CD8, CD10, CD11b, CD11c, CD13, CD15, CD16, CD19, CD20, CD22, CD23, CD25, CD33, CD34, CD36, CD38, CD45, CD56, CD117, FMC-7, HLA-DR, and κ and λ immunoglobulin light chains. Intracellular staining for CD3, CD22, lactoferrin, myeloperoxidase, and Tdt was also performed after permeabilization. All antibodies were purchased from BD PharMingin (San Diego, Calif). Approximately 10 000 events were collected and analyzed using the Paint-A-Gate computer software program (Becton Dickinson).

Conventional cytogenetics was performed on the bone marrow aspirate specimens according to the standard methods used in the laboratory. Fluorescent in situ hybridization (FISH) was performed using the LSI MYC dual color, break apart rearrangement probe for t(2;8), t(8;14), and t(8;22) translocations according to the manufacturer's instructions (Vysis Inc, Downers Grove, Ill).

The peripheral blood and bone marrow aspirate smears from both patients demonstrated a predominant population of blast cells that were small to medium with coarse to slightly immature chromatin, several small inconspicuous nucleoli, and scant cytoplasm, typical of L1 type lymphoblasts according to the FAB criteria (Figure 1). No cytoplasmic vacuoles or granules were identified.

Immunophenotypic analysis using a 4-colored flow cytometer revealed that the blast cells were positive for CD19, CD20 (low density and heterogeneous), CD22, CD38, and CD45 with surface immunoglobulin λ light chain expression in both cases. Representative scatterplots from the first case are shown in Figure 2. The leukemic blast cells in both cases lacked expression of CD5, CD10, CD34, CD23, FMC7, and Tdt. Myeloid and T-lineage markers were also absent.

Conventional cytogenetic studies showed an unremarkable male and female karyotype for cases 1 and 2, respectively. Interphase FISH analysis also failed to demonstrate the presence of chromosomal translocations t(2;8), t(8;14), or t(8;22) involving the c-myc gene in both cases. In case 2, however, a cryptic translocation/deletion involving the mixed lineage leukemia (MLL) gene at 11q23 was demonstrated by FISH analysis. No chromosomal abnormality involving the MLL gene was observed in case 1 by FISH.

Surface immunoglobulin light chain expression is characteristic of but not specific for Burkitt lymphoma/leukemia. In 1986, Van Eys et al1 for the first time described 4 cases of surface immunoglobulin-positive B-cell ALL with L1 morphologic features. Expression of CD34 and/or Tdt on the leukemic cells was not described. Cytogenetic data were not available. These patients, along with other B-lineage ALL patients in the study, were treated according to the AlinC 12 treatment regimens. The outcome of these 4 patients appeared to be the same as other B-lineage ALLs with L1 morphology and null or immature phenotype but significantly better than B-cell ALLs with L3 morphology that were presumably Burkitt leukemia/lymphoma. In 1988, Finlay and Borcherding7 reported 2 additional cases of B-cell ALL with L1 morphology and expression of surface immunoglobulin light chain. Cytogenetics was normal in one of the patients. The first patient was initially treated with the regimen typical for precursor B-lymphoblastic leukemia/lymphoma without success. He was then switched to an aggressive treatment regimen used for Burkitt lymphoma/leukemia. Due to severe neutropenia, the patient died of infectious complications. In light of experience with the first case, the second patient was treated aggressively after the initial diagnosis and was in remission 41 months later. The expression of the immature markers CD34 and/or Tdt was not evaluated in these 2 cases. Subsequently, Michiels et al6 and Shende et al8 each reported a case of Tdt-positive B-cell ALL without Burkitt characteristics. Both patients were treated with daunorubicin, l-asparaginase, vincristine, and prednisone typical for precursor B-lymphoblastic leukemia/lymphoma, and both patients achieved complete remission. More recently, precursor B-cell ALL or acute lymphoblastic lymphoma with surface immunoglobulin light chain restriction has been reported in both pediatric and adult populations.10–12 In those reported cases, the leukemia/lymphoma cells were positive for one or both of the immature markers CD34 and Tdt. It seems that all these cases reported so far probably represent precursor B-lymphoblastic lymphoma/leukemia with aberrant expression of surface immunoglobulin light chain and should be treated as such.

Similarly, neither the absence of surface immunoglobulin light chain restriction nor expression of the immature markers CD34 and Tdt is diagnostic of precursor B-lymphoblastic leukemia/lymphoma. Rare cases of Burkitt lymphoma/leukemia with negative surface immunoglobulin light chain expression have been reported.1,3–5 All these reported cases demonstrated L3 morphology and the presence of chromosomal translocations involving the c-myc gene. In those with available data, CD34 and/or Tdt expression was detected. All patients received intensive chemotherapy with central nervous system prophylaxis and had reasonably good clinical outcome. It appears that morphology is more specific than immunophenotyping in the diagnosis of Burkitt lymphoma/leukemia, although demonstration of chromosomal translocation involving the c-myc gene is necessary for correct diagnosis and management in those situations.

Precursor B-lymphoblastic leukemia/lymphoma and Burkitt lymphoma/leukemia are potentially curable with different treatment regimens. It is therefore imperative to distinguish these 2 entities to avoid inadequate therapy or overtreatment. As described herein, some overlapping features exist between precursor B-lymphoblastic leukemia/lymphoma and Burkitt lymphoma/leukemia, and the distinction between these based on immunophenotyping data alone might be extremely difficult. This is best exemplified by the 2 cases described herein. The leukemic cells in both cases lacked expression of the immature markers CD34 and Tdt. They also expressed surface λ immunoglobulin light chain protein, characteristic of a mature B-cell malignancy. Although the heterogeneous low-density CD20 expression suggests immaturity, lack of CD20 expression is not specific for precursor B-lymphoblastic leukemia/lymphoma. It has been observed in a variety of peripheral B-cell lymphomas, including Burkitt lymphoma/leukemia.13 Therefore, the immunophenotypic profile of these 2 cases fits better for Burkitt lymphoma/leukemia than precursor B-lymphoblastic lymphoma/leukemia. Morphologically, however, both cases have the typical L1 lymphoblasts as defined by FAB criteria in the classification of ALL. Cytogenetically, one case had a normal karyotype and the other had an abnormal karyotype with chromosomal abnormality involving the MLL gene. No chromosomal translocations involving the c-myc gene were demonstrated by conventional cytogenetics and FISH analysis in either one of these cases. Clinically, both patients have findings typically seen in precursor B-lymphoblastic leukemia rather than Burkitt lymphoma/leukemia, such as lack of bulky extramedullary disease or central nervous system involvement. The morphologic, cytogenetic, and clinical features favor the diagnosis of precursor B-lymphoblastic leukemia/lymphoma over Burkitt lymphoma/leukemia. Both patients were treated according to CCG ALL protocols. Both patients achieved a “rapid early response” and currently remain in remission with ALL therapy.

The cases reported herein and those described in the literature demonstrate the importance of a careful and multidisciplinary approach in the evaluation of leukemic blast cells with morphology, immunophenotyping, and cytogenetics. The expression of surface immunoglobulin light chain is insufficient to rule out the diagnosis of precursor B-lymphoblastic leukemia/lymphoma. Morphology appears to be a better predictor of immaturity of the leukemic cells, but cytogenetic analysis to show chromosomal changes involving the c-myc gene is essential for the diagnosis of Burkitt lymphoma/leukemia.