Share this article on:

Precursor B-cell Lymphoblastic Lymphoma: A Predominantly Extranodal Tumor With Low Propensity for Leukemic Involvement

Lin, Pei M.D.; Jones, Dan M.D., Ph.D.; Dorfman, David M. M.D., Ph.D.; Medeiros, L. Jeffrey M.D.

American Journal of Surgical Pathology: November 2000 - Volume 24 - Issue 11 - pp 1480-1490
Original Articles

Precursor B-cell lymphoblastic lymphoma (B-LBL) is uncommon and accounts for less than 10[percnt] of cases of lymphoblastic lymphoma. We collected 25 cases of B-LBL, occurring in children and adults, and report the clinical and histologic features. Patients with concurrent precursor B-cell acute lymphoblastic leukemia (B-ALL) or a history thereof were excluded. There was no evidence of bone marrow disease at the time of diagnosis in 23 patients; two patients had focal ([lt]5[percnt]) involvement. Immunophenotypic analysis was performed in all cases using flow cytometry or immunohistochemical methods. The treatment and survival data available for a subset of patients with B-LBL were compared with those from a series of patients with B-ALL at our institution. The median age was 20 years (range, 5[ndash]68 yrs); 22 (88[percnt]) patients were younger than 35 years of age. There were 17 males and 8 females. The primary sites of disease were skin (nine cases), bones (five cases), soft tissue (four cases), lymph nodes, (three cases), breast (two cases), stomach and colon (one case), and mediastinum (one case). Clinical stage was stage I in 13 cases, stage II in seven cases, stage III in three cases, and stage IV in two cases. Histologically, each neoplasm was diffuse and composed of small to medium-sized lymphoid cells with blastic nuclear chromatin and a high mitotic rate. All cases were positive for B-cell antigens and terminal deoxynucleotidyl transferase. Thirteen (76.4[percnt]) of 17 cases analyzed were positive for CD10 and 13 (54.1[percnt]) of 24 cases assessed were positive for CD20. Of 14 patients with available survival data, all achieved complete clinical response after combination chemotherapy (13 patients) or surgical excision followed by local irradiation (one patient). Five (35.7[percnt]) patients subsequently relapsed, including the patient who had received only irradiation, and four of these patients died after a median survival time of 60 months. None of the patients had leukemia, although one patient developed extensive bone marrow involvement. Nine patients remained in complete remission and were alive at the last follow up (range, 6[ndash]144 months). Unlike precursor T-cell lymphoblastic lymphoma, which commonly involves lymph nodes and the mediastinum, B-LBL usually involves extranodal sites, most often the skin, and rarely presents as a mediastinal mass. With aggressive chemotherapy, patients with precursor B-LBL rarely develop leukemia and appear to have a better prognosis than do patients with B-ALL.

From the Division of Pathology and Laboratory Medicine, The University of Texas M.D. Anderson Cancer Center, Houston, Texas, U.S.A. (P.L., D.J., L.J.M.); and the Department of Pathology, Brigham and Women's Hospital, Boston, Massachusetts, U.S.A. (D.M.D.).

Address correspondence and reprint requests to L. Jeffrey Medeiros, MD, Division of Pathology, Box 85, The University of Texas M.D. Anderson Cancer Center, Houston, TX, 77030, U.S.A.; e-mail[colon] jmedeiro[commat]

Lymphoblastic lymphoma (LBL) and acute lymphoblastic leukemia (ALL) are high-grade neoplasms arising from precursor lymphocytes of B-cell or T-cell lineage. 6 Most cases of ALL are of immature B-cell lineage. In contrast, approximately 90[percnt] of LBL cases are of immature T-cell lineage. 5,9 Precursor B-cell LBL (B-LBL), without evidence of acute leukemia (B-ALL), is uncommon. 1,7,8,13,20,22,23

The neoplastic cells of B-LBL are morphologically indistinguishable from those of B-ALL and are typically of small to medium size, with fine chromatin, inconspicuous nucleoli, and a high mitotic rate. 9 Immunophenotypically, the neoplastic cells express terminal deoxynucleotidyl transferase (TdT) and B-cell antigens such as CD10, CD19, CD22, and CD79a. 5,25,28 Because B-LBL and B-ALL are histologically and immunophenotypically similar, these tumors are considered by many researchers to represent different clinical presentations of the same neoplasm and are grouped in the category of precursor B-cell lymphoblastic leukemia/lymphoma in the newly proposed World Health Organization classification of lymphoid neoplasms. 6

Patients with precursor T-cell LBL are commonly adolescents or young adults with a large mediastinal mass and supradiaphragmatic lymphadenopathy. 9 In contrast, the clinicopathologic features of B-LBL are less well defined. Previous reports of LBL have included T-cell and B-cell neoplasms, with the latter representing less than 10[percnt] of cases reported. 3,22,25,28,34 The criteria used for distinguishing B-LBL from B-ALL are also arbitrary and are applied inconsistently. In some studies, patients with B-LBL and acute leukemia are included, 27,34 but in other studies, patients with leukemic involvement are excluded.

In this study, we have collected 25 cases of precursor B-LBL, without evidence of acute leukemia at the time of diagnosis, to further define the clinicopathologic features of this neoplasm.

Back to Top | Article Outline


Cases of B-LBL were collected from the pathology files of the M.D. Anderson Cancer Center and Brigham and Woman's Hospital. These cases were initially reviewed in consultation or as in-house biopsy specimens. Clinical data were obtained by reviewing the medical record or information provided by the referring physician. All patients underwent bone marrow aspiration and biopsy as part of staging. Cases with tumor representing more than 25[percnt] of the bone marrow cellularity were regarded as B-ALL and were excluded. Patients with a history of chronic myelogenous leukemia in extramedullary lymphoblastic crisis or patients with extramedullary relapses of previously diagnosed B-ALL were also excluded. The diagnosis of B-LBL required compatible morphologic features, the presence of TdT, and at least one B-cell marker. For the purpose of this analysis, patients older than 15 years of age were regarded as adults.

Surgical biopsy and bone marrow trephine biopsy specimens were fixed and processed using formalin, B5, or Zenker fixatives. Routine hematoxylin and eosin[ndash]stained slides were prepared. Fine-needle aspiration smears were stained with Wright or Papanicolaou. Bone marrow aspiration and peripheral blood smears were stained with Wright-Giemsa.

Immunophenotypic analysis was performed using immunohistochemical methods in 19 cases. Frozen or formalin-fixed, paraffin-embedded tissue sections were stained using an avidin-biotin-complex immunoperoxidase technique (Vector Laboratories, Burlingame, CA, USA). Antibodies (monoclonal unless specified) specific for the following antigens were used on frozen sections[colon] CD1 (neat), CD2 (1[colon]50), CD3 (1[colon]50), CD4 (1[colon]20), CD5 (1[colon]50), CD7 (1[colon]50), CD10 (1[colon]10), CD19 (1[colon]30), CD22 (1[colon]100) (Becton-Dickinson, San Jose, CA, USA), and CD8 (1[colon]25) (Ortho Diagnostic Systems, Raritan, NJ, USA). Antibodies specific for the following antigens were assessed using formalin-fixed, paraffin-embedded tissue sections[colon] CD3 (polyclonal, 1[colon]150), CD20 (1[colon]700), CD45 (1[colon]300), CD45RO (1[colon]100), CD79a (1[colon]50), CD99 (1[colon]50), myeloperoxidase (1[colon]500), lysozyme (1[colon]2000) (Dako Corp, Carpinteria, CA, USA), CD34 (1[colon]80), and CD43 (1[colon]120) (Becton-Dickinson).

Flow cytometry immunophenotypic analysis was performed on nine cases, using a variable panel of a large number of antibodies specific for the antigens assessed immunohistochemically as described above and for additional antigens including CD23, CD56, and immunoglobulin heavy and light chains (Becton-Dickinson and Ortho Diagnostic Systems). Seven cases were analyzed by flow cytometry and immunohistochemical methods with no discordant results.

Terminal deoxynucleotidyl transferase was assessed using immunohistochemical methods and formalin-fixed, paraffin-embedded (18 cases) or frozen (four cases) tissue sections or by using flow cytometry (nine cases) (Supertechs Inc., Rockville, MD, USA). Seven cases were assessed using immunohistochemical and flow cytometry methods.

In selected cases, gene rearrangement analysis was performed using DNA extracted from fresh tissue or cells and Southern blot (restriction fragment length) analysis or polymerase chain reaction methods. With the Southern blot methods, the immunoglobulin heavy chain and T-cell receptor [bgr]-chain genes were assessed in five and four cases, respectively. Using the polymerase chain reaction, the immunoglobulin heavy chain and T cell receptor [ggr]-chain genes were assessed in two cases and one case, respectively.

Back to Top | Article Outline


Clinical Features

The clinicopathologic features are summarized in Table 1. There were 17 male patients and 8 female patients, ranging in age from 5 to 68 years. The presenting sites of primary disease included skin in nine cases, bones in five cases, soft tissue in four cases, lymph nodes in three cases, breast in two cases, stomach and colon in one case, and mediastinum in one case. Clinical Ann Arbor stage was as follows[colon] stage I in 13 cases, stage II in seven cases, stage III in three cases, and stage IV in two cases.

The median age at time of presentation was 20 years. Localized or migratory pain was the most common initial symptom and was present in 24 (92.3[percnt]) patients. Low-grade fever, alopecia, and weight loss of 5 to 10 pounds during several months were reported in two patients (case nos. 7 and 24), accompanied by night sweats in one (case no. 7). The duration of symptoms ranged from 2.5 to 36 months before treatment. The size of the largest primary lesion, as determined by physical examination or radiographic imaging in the cases of primary bone B-LBL, ranged from 2 to 11 cm[colon] less than 5 cm in eight cases, 5 to 10 cm in five cases, and more than 10 cm in three cases. Regional lymphadenopathy was documented in seven patients with primarily extranodal disease. No patients had hepatosplenomegaly or central nervous system involvement.

Serum lactate dehydrogenase levels were available for 16 patients. They were elevated in four (case nos. 1, 10, 14, and 24) and ranged from 681 to 1016 IU/L (normal range, 313[ndash]618 IU/L). The complete blood cell count was normal in 23 patients. Two male patients (case nos. 19 and 24) had mild anemia; their hematocrit ranged from 25[percnt] to 30[percnt] (normal male range, 40[percnt][ndash]54[percnt]).

Back to Top | Article Outline

Skin Presentation

Nine (36[percnt]) patients had lesions on the skin, the most common site of involvement. There were five females and four males, from 6 to 62 years of age. Eight patients were younger than 35 years of age. Grossly, the lesions were well-defined, raised masses with an erythematous or violaceous surface. The head and neck region was involved in six (67[percnt]) patients. Of these, three patients had tumors involving the scalp (two accompanied by cervical lymphadenopathy) and one patient each had a tumor involving the forehead, preauricular region, and external auditory canal. The remaining three patients had tumors involving the trunk overlying the scapula region, lower back, and abdominal wall.

Back to Top | Article Outline

Bone Presentation

Five (20[percnt]) patients had bone lesions. All were male between the ages of 15 and 68 years; three were younger than 35 years of age. Three patients had pathologic or radiographic evidence of multifocal disease. One patient had tumors involving the radius and talus (case no. 14). A second patient (case no. 12) had radiologic evidence of lesions involving the proximal femur, lumbar vertebral bodies and pedicles (L2[ndash]L4), seventh and ninth ribs, and bilateral clavicles. The third patient (case no. 10) had increased gallium scan activity of the bilateral tibiae, right shoulder, bilateral axillae, and the para-aortic region and biopsy-proven disease involving the distal metaphysis of femur, which extended into the surrounding soft tissue. The patient also had multiple small kidney lesions and a right intercostal rib lesion determined radiologically. Four patients had fractures of the involved sites. Radiologic imaging studies were characterized by lytic or sclerotic changes that mimicked benign or malignant primary bone lesions, such as trauma, fibrous dysplasia, and Ewing sarcoma. (Fig. 1).

Back to Top | Article Outline

Soft Tissue Presentation

Four (16[percnt]) patients had soft tissue tumors. All were males between the ages of 5 and 35 years. Three tumors involved the infratemporal fossa, the ankle, and the thoracic intercostal muscle. The patient with an ankle lesion also had biopsy-proven disease of bone at the same location, raising the possibility of a primary bone lesion with extension to the soft tissue. The fourth patient (case no. 16) had two masses involving the parotid gland and conjunctiva.

Back to Top | Article Outline

Lymph Node Presentation

Three (12[percnt]) patients had primarily nodal disease. There were two males and one female, ranging in age from 6 to 35 years. Two patients (case nos. 19 and 20) had inguinal lymphadenopathy and edema of the lower extremity. Case no. 19 also had extensive retroperitoneal and pelvic lymphadenopathy with displacement of internal organs. One patient (case no. 21) had supraclavicular lymphadenopathy with edema of the upper extremity.

Back to Top | Article Outline

Other Presentations

Two females, aged 8 and 17 years, had breast masses and enlarged axillary lymph nodes. Fibroadenoma was clinically suspected in both patients before biopsy. A 25-year-old man had malaise, weight loss, alopecia, and hematochezia. Endoscopy showed friable gastric and polypoid colonic lesions, and biopsies showed tumor involving the stomach and colon. Only one patient (case no. 25), a 30-year-old man, had a mediastinal mass accompanied by superior vena cava occlusion.

Back to Top | Article Outline

Histopathologic Features

Surgical biopsy specimens were obtained from 24 patients. Five were core-needle biopsy specimens, and 20 were incisional or excisional biopsy specimens. One patient was diagnosed only by fine-needle aspiration (case no. 23).

Low-power examination in the 24 tissue biopsy specimens showed diffuse infiltration by neoplastic cells dissecting between collagen or skeletal muscle fibers or invading through lymph node capsule into the perinodal fat. Twenty-one (87.5[percnt]) neoplasms showed marked crush artifact, with Azzopardi effect in some cases. A focal or diffuse starry sky pattern was present in 13 (54.2[percnt]) cases. In 8 (50[percnt]) of 16 large specimens, the tumor was compartmentalized by fibrous bands creating areas of nodular accentuation (Fig. 2). Ten (41.7[percnt]) neoplasms showed foci of necrosis, some of which contained marked karyorrhectic debris. The average mitotic count per high-power field (400[times]) was determined in the 24 tumors with tissue biopsy specimens[colon] one to five in 17 (70[percnt]) neoplasms and five to 10 in seven (30[percnt]) neoplasms. Rare reactive follicles surrounded by tumor were present in three (75[percnt]) of four nodal and 6 (30[percnt]) of 20 extranodal specimens. In primary skin lesions, a narrow Grenz zone separated the tumor from the overlying epidermis (Fig. 3). In bone lesions, the tumor infiltrated the medullary space with periosteal fibrosis, new bone formation, and bone remodeling (Fig. 4). The neoplasm involving the stomach and colon was located predominantly in the lamina propria and lymphoepithelial lesions were minimal. Single nonparatrabecular aggregates of neoplastic cells, representing less than 5[percnt] of the cellularity, were found in the bone marrow of two patients (case nos. 12 and 17). The bone marrow was negative in the other 23 patients.

Cytologically, all 25 cases were composed of immature cells of small to medium size, with fine chromatin, inconspicuous or small nucleoli, and scant basophilic cytoplasm. The tumors showed a range of nuclear irregularity and a variable amount of cytoplasm. The nuclei were round to oval in five (20[percnt]) tumors, indented (Fig. 5A) in 16 (64[percnt]) cases, and markedly convoluted (Fig. 5B) in four (16[percnt]) tumors. Some cases had neoplastic cells with small but distinct nucleoli and relatively more cytoplasm, similar to that of the FAB L2 subtype of ALL.

Back to Top | Article Outline

Immunologic Features

The immunophenotypic findings are summarized in Table 2. By definition, all 25 cases were positive for TdT and one or more B-cell antigens and were negative for T-cell antigens. Nineteen cases were assessed immunohistochemically using various antibodies and fixed, paraffin-embedded tissue sections. All cases analyzed were positive for TdT (n [equals] 18) (Fig. 6A), CD45RB (n [equals] 16), CD43 (n [equals] 13), and CD79a (n [equals] 12) (Fig. 6B). Twelve (66.7[percnt]) of 18 cases were positive for CD20, although five were only weakly or focally positive (Fig. 6C). All cases assessed were negative for CD1, CD3, CD5, and CD45RO.

Nine cases were assessed by flow cytometry. All cases were positive for CD19, CD45 (dim), TdT, and HLA-DR. CD10 was positive in eight (88.9[percnt]) tumors and CD20 was positive in three (33.3[percnt]) neoplasms. One case (case no. 7) was positive for CD2 and CD56. All cases assessed were negative for CD1, CD3, CD4, CD5, CD7, and CD8.

Eight cases were assessed immunohistochemically using frozen sections. All were positive for CD19. CD20 and TdT were positive in all cases assessed, three and two, respectively. Five (83.3[percnt]) of six neoplasms were positive for CD10. All cases assessed were negative for CD1, CD2, CD3, CD4, CD5, CD7, and CD8.

Seven cases were assessed for myeloid antigens by flow cytometry (two cases), immunohistochemical staining for myeloperoxidase (three cases) or lysozyme (one case), or cytochemically using napthol AS-D choroacetate esterase (two cases). Three cases expressed myeloid-associated antigens. Case no. 8 was CD13-positive by flow cytometry and focally immunoreactive for myeloperoxidase. Case no. 12 was positive for CD13 and CD33 by flow cytometry. Case no. 16 was focally immunoreactive for myeloperoxidase. Because these cases also expressed TdT and B-cell antigens, they were considered B-LBL with evidence of myeloid differentiation. Case nos. 8 and 12 expressed the B-cell antigens CD10, CD19, CD20, and CD79a. Case no. 16 was positive for CD20, CD79a, and cytoplasmic IgM.

Back to Top | Article Outline

Molecular Studies

In seven tumors (case nos. 1, 3, 7, 10, 13, 16, and 24) assessed, the immunoglobulin heavy chain gene was rearranged. In one case involving the conjunctiva and parotid gland (case no. 16), a rearranged band of identical size was detected by the polymerase chain reaction in both specimens. In five neoplasms (case nos. 1, 3, 7, 10, and 13), T-cell receptor genes were assessed. The TCR [ggr] chain gene was rearranged in two (case nos. 7 and 13) of four cases analyzed. In a fifth case (case no. 1), assessed by the polymerase chain reaction, the TCR [bgr] gene showed no evidence of monoclonal gene rearrangement. The presence of T-cell receptor gene rearrangements in two (40[percnt]) of five cases is similar to that previously reported in B-ALL. 11,20

Back to Top | Article Outline

Treatment and Outcome

Survival data for 14 of 25 patients were available and are summarized in Table 3. Thirteen were treated with combination chemotherapy, including six patients with H-CVAD (hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone), alternating with methotrexate and cytarabine (Ara-C). Two patients (case nos. 13 and 24) were initially treated with CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone). One patient (case no. 4) was treated with P8904 (high dose 6-mercaptopurine, methotrexate, hydrocortisone, Ara-C, vincristine, and L-asparaginase). One patient (case no. 7) was treated with CCG 502 (vincristine, prednisone, doxorubicin, methotrexate, thioguanine, and carmustine). One patient (case no. 14) was treated with CCG 5941 (high-dose methotrexate and L-asparaginase). One patient (case no. 17) was treated with CVAD (cyclophosphamide, doxorubicin, vincristine, and dexamethasone), L-asparaginase, and Ara-C. One patient (case no. 19) was treated with cyclophosphamide, vincristine, L-asparaginase, methylprednisone, Ara-C, and methotrexate. Other agents used included etoposide (VP-16), [agr]-interferon, G-CSF, and leucovorin. POMP (methotrexate, 6-mercaptopurine, vincristine, and prednisone) was used as maintenance therapy and methotrexate and Ara-C for intrathecal central nervous system prophylaxis. One patient (case no. 9) was treated with surgery and local irradiation only. Protocols for relapses included high-dose methotrexate and vincristine (case no. 7), CHOP-bleomycin (case no. 9), or H-CVAD (case nos. 13 and 24). Five patients received radiation therapy. Five patients underwent surgery, including tumor debulking (case nos. 4, 9, 14, and 17) or orchiectomy for testicular relapse (case nos. 9 and 13). Case no. 9 also underwent splenectomy and staging laparotomy with sampling of multiple lymph nodes.

Follow up ranged from 6 to 144 months; the median follow-up interval was 48 months. Nine patients were alive in complete remission at the last follow up. Four of five patients (case nos. 7, 9, 13, and 19) who had a relapse died of progressive disease. One patient (case no. 9) was diagnosed in 1971 with localized tumor involving the skin and soft tissue of the flank. First treatment with surgical debulking and local irradiation led to complete remission. Relapse occurred in the thigh 1 year later and was again treated with local irradiation only. The second remission lasted for 2 years and was followed by dissemination to the cervical lymph node, breast, and testis, which required multiagent chemotherapy with CHOP-bleomycin and orchiectomy. There was no evidence of central nervous system or bone marrow disease throughout the patient's entire disease. The remaining three patients (case nos. 7, 13, and 19) who died had single or multiple relapses during a period of 4 or 5 years. Case no. 7 was a 10-year-old girl with stage IA disease of the skin that was initially treated with CCG 502 and CCG 1941. Three separate relapses occurred over 5 years before death. The immunophenotype of this tumor was unusual because the neoplastic cells expressed CD2 and CD56 in addition to TdT and B-cell antigens. Case no. 13 was a 68-year-old man with stage IA disease of the radius initially treated with CHOP and local irradiation. Relapse occurred 1 year later in the testis and required orchiectomy and more aggressive chemotherapy with H-CVAD. Second remission was followed by two more relapses with dissemination to the tibia, distal radius, ulna, thigh, chest wall, and penis in 4 years. Case no. 19 was treated with a modified VAD regimen because of coexisting postvaccination encephalopathy and chronic dependence on a respirator and phenytoin. Chronic use of anticonvulsants, such as phenytoin, has been shown to reduce the efficacy of chemotherapy for childhood ALL. 32

One patient (case no. 24) had a relapse but was alive and undergoing additional chemotherapy at the last follow up. He had tumor involving the stomach and colon and was initially treated with six cycles of CHOP. Repeat biopsy at the end of the treatment found residual disease in the stomach, cecum, and transverse colon. He was treated with five additional cycles of H-CVAD. Although multiple biopsy specimens of the colon and stomach remained negative at the end of H-CVAD therapy, bone marrow aspiration performed approximately 1 year after the initial diagnosis showed numerous lymphoblasts.

Back to Top | Article Outline


Advances in therapy and molecular genetics have significantly expanded our knowledge of B-ALL in the past 2 decades, yet relatively little is known about B-LBL. This is a result of the overall low incidence of the disease and few previous attempts to segregate this subgroup from B-ALL cases. 2,30 Currently, the complete remission rates for B-ALL have reached 95[percnt] to 99[percnt] and 60[percnt] to 85[percnt] in pediatric and adult patients, respectively, with survival rates of 60[percnt] to 70[percnt] in children and 20[percnt] to 40[percnt] in adults. 2,10,12,30 The worse survival rate in adult patients is related to the clinical heterogeneity of adult B-ALL. Separating the low-or intermediate-risk from the high-risk subgroups and adapting therapy to risk are the current strategies for improving overall survival of adults with B-ALL. 30

Our current study of 25 patients with B-LBL defines the clinicopathologic features of this neoplasm, showing its frequent extranodal presentation. Complete survival data of 10 adult patients showed eight long-term survivors, with seven in complete remission, suggesting that patients with B-LBL have a more favorable complete remission rate and remission duration than that reported for adult patients with B-ALL treated with similar regimens. 10,12 Although prognostic factors are not well understood for patients with B-LBL, presumably those identified in patients with B-ALL apply. In B-ALL, advanced age, high leukocyte count, circulating blasts, elevated serum lactate dehydrogenase levels, organomegaly, poor performance status, and delayed response to chemotherapy all are adverse prognostic factors. 2,30,32 In our series, in which patients with overt leukemia were excluded, patients with B-LBL differed from those with B-ALL by being younger (56[percnt] were 15 to 35 years of age), having localized disease involving skin and soft tissue sites, and having absence of hepatosplenomegaly, blasts in the peripheral blood, or central nervous system disease. Most patients in this study had normal serum lactate dehydrogenase levels and no constitutional symptoms. These favorable clinicopathologic features were translated into a complete remission rate of 100[percnt] and a median survival time of 60 months. Such results are better than that reported for patients with B-ALL of the same age group and suggest that the prognosis of patients with B-LBL is comparable to that of the low-risk or the favorable subset of intermediate-risk subgroups of adults with B-ALL. 10,12,30

Many of the known prognostic factors for patients with B-ALL, such as high serum lactate dehydrogenase levels, organomegaly, and central nervous system disease, are indirect measures of tumor burden. Their absence in the cases of B-LBL in this study suggests that one possible explanation for the survival advantage of patients with B-LBL lies in their having localized disease and low tumor burden. However, size of the tumor is also a measure of tumor burden, and tumor size showed no correlation with survival in our series. Half the patients with known tumor size in this study had lesions larger than 5 cm at the time of diagnosis. One patient with a 12-cm mass in the breast completely responded to therapy and was free of disease for 8 years.

An alternative possibility for the survival advantage of B-LBL patients is that this neoplasm has a low propensity for peripheral blood involvement. In this study, none of the 14 patients with long-term follow-up had overt leukemia, despite multiple and multifocal relapses. Only one patient eventually experienced extensive bone marrow involvement. The observation that some tumors grew to considerable size locally, without clinical or pathologic evidence of leukemia, is clinical evidence in support of this possibility.

Multifocal bone involvement by B-LBL has been described by others previously. 7 The pattern and extent of multifocal involvement suggest a close relationship with leukemic disease, but none of the patients in this study had overt leukemia. Most notably, case no. 14 had been symptomatic for 2.5 years and was found to have stage III disease without evidence of bone marrow or central nervous system disease at the time of diagnosis. Case no. 13, a 60-year-old, had an 11-cm localized lesion of the radius that was incompletely treated but did not progress for more than 1 year before experiencing a lasting 2-year remission with H-CVAD. The entire disease course of this patient spanned 5 years, with multiple extramedullary relapses but no evidence of leukemia. Localized bone involvement by B-LBL suggests that some cases may have a deficient capacity to disseminate. Secondary genetic changes are then needed for dissemination.

Despite the absence of unfavorable prognostic factors in the patients with B-LBL in this study, patients still require aggressive treatment similar to that for B-ALL. CHOP was the first chemotherapy regimen used in the two adult patients who died and one patient who had a relapse in this study. H-CVAD was used as frontline therapy in six patients and as salvage therapy in two patients and was successful in obtaining durable remissions, even in patients who had a relapse. This is in keeping with the conclusions of others that CHOP is insufficient therapy for B-LBL 4 and that H-CVAD is a much more effective regimen. 10,12,32,33 The favorable outcome of patients with B-LBL in this study also suggests that B-LBL is more responsive than B-ALL to an appropriately designed regimen. The complete remission rate of relapsed B-ALL in patients older than 40 years of age treated with H-CVAD as salvage therapy in our institution was 27[percnt], and the median duration of remission was 29 weeks. 12 In addition, when assessing H-CVAD as frontline therapy, we did not observe refractory cases or induction mortality, in contrast to 3[percnt] and 6[percnt], respectively, among patients with B-ALL reported from our institution. 10 The projected 5-year survival and remission duration rates are also higher in this series than those reported for B-ALL (39[percnt] and 38[percnt], respectively). 10,32,33

Similar to B-ALL, B-LBL is an immunophenotypically heterogenous group with a continuum of immunophenotypical profiles consistent with derivation from immature B-cell precursors at different stages of maturation. 2 In one tumor (case no. 16), cytoplasmic IgM was detected consistent with a pre-B-cell immunophenotype. In this study, we did not find a difference in clinical behavior between the immunophenotypic subgroups of B-LBL.

With current, more effective chemotherapy regimens, B-ALL with myeloid markers is no longer considered to have a poor prognosis. 10,30 The three cases with myeloid-associated antigens in this series involved the vertebra, skin, and conjunctiva. The first two patients were in complete remission with follow up of 1 and 2 years. Follow up for the third patient was not available, but this patient had localized disease (stage I) for nearly 1 year before seeking medical attention.

We also report a case of B-LBL in the mediastinum (case no. 25). Because no normal lymph node or thymus structure could be discerned in the biopsy specimen, the exact site of origin of the tumor cannot be definitively established. Our case had an immunophenotypical profile similar to a case described previously, 23 positive for pan-B-markers and TdT, but negative for CD10. There have been fewer than five cases of mediastinal B-LBL reported in the literature, with adequate immunophenotypic data available for two cases. 23,25,27 Medullary B cells are part of the normal structure of the human thymus and progenitor B cells capable of maturation have been shown in animal studies. 11,14,17 Thus, it is conceivable that B-LBL in this location represents a primary B-LBL of the thymus arising from a population of progenitor B cells, either native to or selectively homing to the thymus.

In conclusion, localized B-LBL has a tendency to affect extramedullary sites, especially the skin, soft tissue, and bones. Patients are unlikely to have constitutional complaints, and pain, localized or migratory, is often the earliest symptom. Aggressive systemic chemotherapy analogous to that designed for B-ALL resulted in favorable survival for patients with B-LBL. Ineffective treatment appears to be an important factor leading to recurrence and poorer prognosis. In patients who have a relapse, recurrent tumor tends to involve other extranodal sites, and leukemic involvement is uncommon.

Back to Top | Article Outline


1. Chimenti S, Fink-Puches R, Peris K, et al. Cutaneous involvement in lymphoblastic lymphoma. J Cutan Pathol 1999; 26[colon]379[ndash]85.
2. Cortes JE, Kantarjian H. Acute lymphoblastic leukemia[colon] a comprehensive review with emphasis on biology and therapy. Cancer 1995; 76[colon]2393[ndash]417.
3. Cossman J, Chused TM, Fisher RI, et al. Diversity of immunologic phenotypes of lymphoblastic lymphoma. Cancer Res 1983; 43[colon]4486[ndash]90.
4. Furman WL, Fitch SH, Hustu O, et al. Primary lymphoma of bone in children. J Clin Oncol 1989; 7[colon]1275[ndash]80.
5. Grogan T, Spier C, Wirt DP, et al. Immunologic complexity of lymphoblastic lymphoma. Diagn Immunol 1986; 4[colon]81[ndash]8.
6. Harris NL, Jaffe ES, Diebold J, et al. World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues[colon] report of the clinical advisory committee meeting[mdash]Airlie House, Virginia, November 1997. J Clin Oncol 1999; 17[colon]3835[ndash]49.
7. Iravani S, Singleton TP, Ross CW, et al. Precursor B lymphoblastic lymphoma presenting as lytic bone lesions. Am J Clin Pathol 1999; 112[colon]836[ndash]43.
8. Kamps W, Poppema S. Pre-B-cell non-Hodgkin's lymphoma in childhood. Report of a case and review of the literature. Am J Clin Pathol 1988; 90[colon]103[ndash]7.
9. Knowles DM. Lymphoblastic lymphoma. In Knowles DM, ed. Neoplastic Hematopathology, 1st ed. Baltimore, MD[colon] Williams [amp] Wilkins, 1992[colon]715[ndash]48, 1295[ndash]314.
10. Kantarjian HM, O'Brien S, Smith TL, et al. Results of treatment with Hyper-CVAD, a dose-intensive regimen, in adult acute lymphocytic leukemia. J Clin Oncol 2000; 18[colon]547[ndash]61.
11. Kawamoto H, Ohmura K, Katsura Y. Presence of progenitors restricted to T, B, or myeloid lineage, but absence of multipotent stem cells, in the murine fetal thymus. J Immunol 1998; 161[colon]3799[ndash]802.
12. Koller CA, Kantarjian HM, Thomas D, et al. The HCVAD regimen improves outcome in relapsed acute lymphoblastic leukemia. Leukemia 1997; 11[colon]2039[ndash]44.
13. Link MP, Roper M, Dorfman RF, et al. Cutaneous lymphoblastic lymphoma with pre-B markers. Blood 1983; 61[colon]838[ndash]41.
14. McKenna HJ, Morrissey PJ. Flt3 ligand plus IL-7 supports the expansion of murine thymic B cell progenitors that can mature intrathymically. J Immunol 1998; 160[colon]4801[ndash]9.
15. Medeiros LJ, Carr J. Overview of the role of molecular methods in the diagnosis of malignant lymphomas. Arch Pathol Lab Med 1999; 123[colon]1189[ndash]207.
16. Millot F, Robert A, Bertrand Y, et al. Cutaneous involvement in children with acute lymphoblastic leukemia or lymphoblastic lymphoma. Pediatrics 1997; 100[colon]60[ndash]4.
17. Montecino-Rodriguez E, Johnson A, Dorshkind K. Thymic stromal cells can support B cell differentiation from intrathymic precursors. J Immunol 1996; 156[colon]963[ndash]7.
18. Nanni G, Bertoncini M, Garbini A, et al. Primary lymphoblastic lymphoma of the ileocecal valve manifesting as an ileocecal intussusception. NY State J Med 1983; 83[colon]240[ndash]2.
19. Nathwani BN, Diamond LW, Winberg CD, et al. Lymphoblastic lymphoma[colon] a clinicopathologic study of 95 patients. Cancer 1981; 48[colon]2347[ndash]57.
20. Ozdemirli M, Fanburg-Smith JC, Hartmann DP. Precursor B-lymphoblastic lymphoma presenting as a solitary bone tumor and mimicking Ewing's sarcoma. A report of four cases and review of the literature. Am J Surg Pathol 1998; 22[colon]795[ndash]804.
21. Relling MV, Pui CH, Sandlund JT, et al. Chronic anticonvulsant therapy reduces the efficacy of chemotherapy for childhood acute lymphoblastic leukemia (ALL). Blood 1999; 94[colon]289a.
22. Sander CA, Medeiros LJ, Abruzzo LV, et al. Lymphoblastic lymphoma presenting in cutaneous sites. J Am Acad Dermatol 1991; 25[colon]1023[ndash]31.
23. Sander CA, Jaffe ES, Gebhardt FC, et al. Mediastinal lymphoblastic lymphoma with an immature B-cell immunophenotype. Am J Surg Pathol 1992; 16[colon]300[ndash]5.
24. Schmitt IM, Manente L, Di Matteo A, et al. Lymphoblastic lymphoma of the pre-B phenotype with cutaneous presentation. Dermatology 1997; 195[colon]289[ndash]92.
25. Sheibani K, Nathwani BN, Winberg CD, et al. Antigenically defined subgroups of lymphoblastic lymphoma[colon] relationship to clinical presentation and biologic behavior. Cancer 1987; 60[colon]183[ndash]90.
26. Smith RG. Parosteal lymphoblastic lymphoma. Cancer 1984; 54[colon]471[ndash]6.
27. Soslow RA, Baergen RN, Warnke RA. B-lineage lymphoblastic lymphomas is a clinicopathologic entity distinct form other histologically similar aggressive lymphoma with blastic morphology. Cancer 1999; 85[colon]2648[ndash]54.
28. Streuli RA, Kaneko Y, Variakojis D, et al. Lymphoblastic lymphoma in adults. Cancer 1981; 47[colon]2510[ndash]6.
29. Stroup R, Sheibani K, Misset JL, et al. Surface immnoglobulin-positive lymphoblastic lymphoma[colon] a report of three cases. Cancer 1990; 65[colon]2559[ndash]63.
30. Stock W. Treatment of adult acute lymphoblastic leukemia[colon] risk-adapted strategies. 1999 American Society of Hematology Education Program book, New Orleans, Louisiana.
31. Taubenberger JK, Jaffe ES, Medeiros LJ. Thymoma with abundant L26-positive `asteroid' cells. A case report with an analysis of normal thymus and thymoma specimens. Arch Pathol Lab Med 1991; 115[colon]1254[ndash]7.
32. Thomas D, Kantarjian H, Smith TL, et al. Primary refractory and relapsed adult acute lymphoblastic leukemia[colon] characteristics, treatment results, and prognosis with salvage therapy. Cancer 1999; 86[colon]1216[ndash]30.
33. Thomas D, Kantarjian H, O'Brien S, et al. Outcome with the Hyper-CVAD regimen in lymphoblastic lymphoma (LL). Proc Am Soc Clin Oncol 1999; 18[colon]11a.
34. Zinzani PL, Bendandi M, Visani G, et al. Adult lymphoblastic lymphoma[colon] clinical features and prognostic factors in 53 patients. Leuk Lymphoma 1996; 23[colon]577[ndash]82.

Lymphoblastic lymphoma; Acute lymphoblastic leukemia; Extranodal; Immunophenotype; B cell

[copy] 2000 Lippincott Williams [amp] Wilkins, Inc.