|Home | About | Journals | Submit | Contact Us | Français|
B-cell lymphomas with concurrent IGH-BCL2 and MYC rearrangements, also known as “double-hit” lymphomas (DHL), are rare neoplasms characterized by highly aggressive clinical behavior, complex karyotypes, and a spectrum of pathological features overlapping with Burkitt lymphoma (BL), diffuse large B-cell lymphoma (DLBCL) and B-lymphoblastic lymphoma/leukemia (B-LBL). The clinical and pathological spectrum of this rare entity, including comparison to other high-grade B-cell neoplasms, has not been well defined. We conducted a retrospective analysis of clinical and pathologic features of 20 cases of DHL seen at our institution during a 5-year period. In addition, we performed case-control comparisons of DHL with BL and International Prognostic Index (IPI)-matched DLBCL. The 11 men and 9 women had a median age of 63.5 years (range 32-91). Six patients had a history of grade 1-2 follicular lymphoma (FL); review of the prior biopsy specimens in 2 of 5 cases revealed blastoid morphology. Eighteen patients had Ann Arbor stage 3 or 4 disease and all had elevated serum lactate dehydrogenase (LDH) levels at presentation. Extranodal disease was present in 17/20 (85%), bone marrow involvement in 10/17 (59%) and central nervous system (CNS) disease in 5/11 (45%). Nineteen patients were treated with combination chemotherapy, of whom 18 received rituximab and 14 received CNS-directed therapy. Fourteen patients (70%) died within 8 months of diagnosis. Median overall survival in the DHL group (4.5 months) was inferior to both BL (p=0.002) and IPI-matched DLBCL (p=0.04) control patients. Twelve DHL cases (60%) were classified as B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and BL, 7 cases (35%) as DLBCL, not otherwise specified, and 1 case as B-LBL. Distinguishing features from BL included expression of Bcl2 (p<0.0001), Mum1/IRF4 (p=0.006), Ki-67 <95% (p<0.0001), and absence of EBV-EBER (p=0.006). DHL commonly contained the t(8;22) rather than the t(8;14) seen in most BL controls (p=0.001), and exhibited a higher number of chromosomal aberrations (p=0.0009). DHL is a high-grade B-cell neoplasm with a poor prognosis, resistance to multi-agent chemotherapy, and clinical and pathological features distinct from other high-grade B-cell neoplasms. Familiarity with the morphologic and immunophenotypic spectrum of DHL is important in directing testing to detect concurrent IGH-BCL2 and MYC rearrangements when a karyotype is unavailable. The aggressive clinical behavior and combination of genetic abnormalities seen in these cases may warrant categorization as a separate entity in future classifications and call for novel therapeutic approaches.
B-cell lymphomas commonly exhibit chromosomal translocations involving immunoglobulin genes (IG). The t(14;18)(q32;q21) results in the juxtaposition of BCL2 with enhancer elements of the IG heavy chain (IGH). This translocation is a hallmark of follicular lymphoma (FL) and is also found in 20-30% of de novo diffuse large B-cell lymphoma (DLBCL) (16, 56). Translocations involving MYC at 8q24 and IG partners, including IGH and kappa (IGK) and lambda (IGL) light chains, are characteristic of Burkitt lymphoma (BL), and are also found in 5-15% of unselected cases of DLBCL (2, 17, 24). B-cell lymphomas with concurrent IGH-BCL2 and MYC rearrangements are rare, and are characterized by highly aggressive clinical behavior, complex karyotypes, and a broad morphologic and immunophenotypic spectrum that overlaps with BL and DLBCL, and occasionally B-lymphoblastic lymphoma/leukemia (B-LBL) (30). Because these “double-hit” lymphomas (DHL) share pathologic features with other high-grade B-cell neoplasms, their diagnosis and classification may be difficult, and their rarity and poor response to therapy make selection of appropriate treatment challenging.
Approximately 200 cases of DHL have been reported in the literature, mostly as case reports and small series (3, 4, 6, 8-10, 12, 13, 15, 18-20, 22, 25-27, 29-33, 36, 39, 40, 43, 47, 49-54, 59). Most reported patients have de novo disease, while a minority have a history of grade 1-2 FL and develop DHL secondarily, presumably by acquisition of a MYC translocation (8, 12, 18, 19, 26, 27, 31, 49-51, 54, 59). Few studies have directly compared the pathologic features and clinical outcome of DHL with BL and DLBCL (37, 42), and only a single series has classified DHL using criteria of the 2008 World Health Organization (WHO) Classification of Tumours of Haematopoietic and Lymphoid Tissues (18, 48).
We conducted a retrospective analysis of the clinical, morphologic, immunophenotypic, cytogenetic, and molecular genetic features of 20 cases of DHL seen at our institution to define further the clinical and pathologic spectrum of this rare entity and to classify cases according to the 2008 WHO Classification. In addition, we performed a case-control comparison with BL cases seen over the same time period, in order to identify distinguishing clinicopathologic features that facilitate early and correct identification of DHL. We also performed a case-control comparison with a group of International Prognostic Index (IPI)-matched DLBCL cases treated at the same institution to elucidate clinical differences between DHL and DLBCL. Based on the distinctive clinicopathologic features of DHL that we identified, we outline circumstances in which additional testing by fluorescence in situ hybridization (FISH) or PCR on diagnostic tissue samples may be helpful to confirm or exclude a diagnosis of DHL if a conventional karyotype is unavailable.
The Partners HealthCare Institutional Review Board (IRB) granted approval for the study before its initiation. The files of the Massachusetts General Hospital (MGH) Pathology Department were searched for cases of B-cell lymphoma on which cytogenetic and/or FISH analysis had been performed and had revealed concurrent IGH-BCL2 and MYC rearrangements. In some cases, tissue had been sent for cytogenetic analysis at the time of frozen section evaluation at the discretion of the frozen section pathologist, while in other cases lacking cytogenetic analysis, FISH to detect IGH-BCL2 and MYC and gene rearrangements had been performed at the time of diagnosis because BL was a diagnostic consideration based on the morphology and/or immunophenotype. A total of 19 cases from 2004-2009 were identified. One additional case (case 10) identified from this time period contained a MYC rearrangement and 8 copies of an intact BCL2 confirmed by paraffin FISH, and was included in the study because of the functional equivalence of multiple copies BCL2 to t(14;18) (11, 35, 58). Pathology, flow cytometry and cytogenetics reports, and H&E- and immunohistochemical-stained slides of these cases were reviewed by 4 hematopathologists to confirm the diagnosis and to classify each case according to the 2008 WHO Classification. Medical records were reviewed to determine the IPI score (1), prior history of lymphoma, clinical presentation, therapeutic regimen, response to therapy, outcome and overall survival (OS).
BL control cases were identified by review of the MGH pathology files from 2000-2008; a total of 29 consecutive cases in which a diagnosis of “Burkitt lymphoma” or “atypical Burkitt lymphoma” had been rendered were identified. Pathology, flow cytometry and cytogenetics reports, and H&E- and immunohistochemical stained slides of these cases were reviewed by 4 hematopathologists for a consensus diagnosis of BL according to criteria of the 2008 WHO Classification (28). Four cases that did not meet morphologic, immunophenotypic, and/or cytogenetic criteria for BL were excluded, yielding a total of 25 BL control patients whose medical records were reviewed to determine site of involvement by lymphoma, serum lactate dehydrogenase (LDH) level at presentation, outcome and OS. This control group included 7 patients ≤20 years old, none of whom were known to be HIV-positive, 9 HIV-positive adults, and 9 HIV-negative adults.
DLBCL control cases were selected from an IRB-approved clinical database of patients diagnosed with DLBCL and treated at the MGH from 2000-2008. Two IPI-matched controls were randomly selected and matched to each DHL case, yielding a total of 40 DLBCL patients whose pathology reports were reviewed to confirm the diagnosis, and whose medical records were reviewed to determine site of involvement by lymphoma, LDH level at presentation, outcome and OS.
Formalin- or B+-fixed, paraffin-embedded tissues from diagnostic biopsies of DHL cases and BL control cases were characterized by immunohistochemistry using previously described methods (44) with some or all of the following antibodies, if not already performed at the time of diagnosis: CD20 (L26 epitope, prediluted, Ventana Medical Systems, Tucson, AZ), CD79a (prediluted, Ventana), Pax5 (prediluted, Ventana), Bcl6 (1:10 dilution, Dako, Carpinteria, CA), CD10 (prediluted, Ventana), Bcl2 (clone 124, prediluted, Ventana), Mum1/IRF4 (1:20 dilution, Dako), Ki-67 (clone 30-9, prediluted, Ventana), terminal deoxynucleotidyl transferase (TdT, prediluted, Ventana). One case that lacked Bcl2 protein expression using clone 124 (targeting amino acids 41-54) was subsequently analyzed using clone C-2 (targeting amino acids 1-205) (Santa Cruz Biotechnology, Santa Cruz, CA). In situ hybridization for EBV-encoded RNA (INFORM EBER Probe, Ventana) was performed as previously described (44).
In cases with material available for flow cytometry at the time of diagnosis, 4-color flow cytometry was performed using previously described methods (44) for the following antibodies: CD45-peridinin chlorophyll protein (PerCP), CD19-PerCP (SJ25C1), CD19-phycoerythrin (PE; 4G7), CD20-PerCP (L27), CD10-fluorescein isothiocyanate (FITC; W8E7), kappa light chain-FITC (TB28-2), lambda light chain-phycoerythrin (PE; 1-155-2), TdT-FITC (HT1, HT4, HT8, and HT9) (BD Biosciences, San Jose, CA), and CD10-PE (SS2/36) (Dako).
GTG-banded metaphases were obtained from unstimulated overnight lymph node or bone marrow cultures according to standard cytogenetic protocols in 17 tumors (11 DHL, 6 BL) at the time of diagnosis. Chromosome analysis was performed at a level of 400 bands or greater and the karyotypes were described according to the International System for Human Cytogenetic Nomenclature (ISCN) 2005 (46).
In DHL and BL control cases lacking a karyotype, FISH analyses to identify MYC, BCL6 and either BCL2 or IGH-BCL2 rearrangements were attempted on interphase nuclei of paraffin-embedded tissue at the time of diagnosis. In case 11, MYC and IGH-BCL2 FISH was performed on a cerebrospinal fluid (CSF) air-dried cytospin and in cases 12 and 17, FISH was performed on abnormal metaphases from tumor cells. All FISH probes were purchased from Abbott Molecular (Des Plaines, IL) and included the Vysis LSI MYC Dual Color Break Apart Rearrangement Probe to identify any rearrangement at 8q24, the Vysis LSI BCL6 Dual Color Break Apart Rearrangement Probe to identify any rearrangement at 3q27, the Vysis LSI BCL2 Dual Color Break Apart Rearrangement Probe to identify any rearrangement at 18q21, the Vysis IGH/BCL2 Dual Color Dual Fusion Translocation Probe to identify t(14;18)(q32;q21), and the Vysis CEP 18 (D18Z1) SpectrumOrange Probe. For formalin-fixed, paraffin-embedded tissue samples, interphase FISH was performed using standard cytogenetic protocols either on whole nuclei extracted from 50μm tissue sections or on 4μm tissue sections cut and fixed onto slides (57). Fifty to 100 nuclei were scored for each sample, and a sample was considered positive for the rearrangement if >15% of nuclei exhibited a break-apart and/or fusion signal.
All statistical analyses were performed using SAS version 9. Fisher’s exact test was used for comparing proportions and Student’s t-test was used for comparing means between 2 groups. Matching between the DHL and DLBCL groups was taken into account by using conditional logistic regression. OS was calculated from date of diagnosis to date of death or last follow-up. Distributions of OS were estimated using the method of Kaplan and Meier and differences in OS were assessed by the stratified log-rank test. Cox regression was used to analyze the effect of continuous variables on OS (5). Significance was tested at the α=0.05 level.
Clinical characteristics of DHL patients are summarized in Tables Tables11 and and2.2. The 20 patients included 11 men and 9 women with a median age of 63.5 years (range: 32-91). None were known to be HIV-positive. Six patients (30%) had a reported history of FL, all histologic grade 1-2 of 3, diagnosed 6 months to 12 years before their DHL diagnosis. Patients were treated with a variety of multiagent chemotherapy regimens that could be broadly divided into 2 categories: moderate intensity regimens (CHOP, R-CHOP and R-ICE) in 13 patients (70%) and high intensity regimens (CODOX-M/R-IVAC and R-EPOCH) in 6 patients (30%) (Table 1). All were treated with curative intent, except for patient 20 who received a single cycle of R-CHOP chemotherapy for palliation. Rituximab was included in 18 cases (90%) and 14 patients (70%) received CNS-directed therapy with intrathecal or high-dose systemic methotrexate. Treatment was not known in case 16. Eight patients (42%) achieved a CR, of whom 3 eventually died: 1 relapsed and died of disease, 1 had persistent neutropenia and died of infection, and 1 died of an unknown cause. After a median follow-up of 7.3 months (range: 1.5-26.6), 6 patients (30%) were alive (4 with no evidence of disease, 1 alive with disease who relapsed after 1 year in CR, and 1 diagnosed recently, undergoing chemotherapy) and 14 patients (70%) had died. Four patients remain alive in CR (median follow-up: 18.2 months, range: 8.9-26.6), all of whom were treated with R-CHOP. However, treatment regimen was not significantly correlated with response or OS (data not shown). The sole clinical variable that was significantly correlated with outcome was ECOG performance status (PS) of ≥2 (p=0.04). In addition, age at diagnosis was marginally significant in the univariate Cox regression analysis (p=0.06). The hazard ratio associated with older age (>60 years) was 2.7 (95% CI: 0.84-8.8).
Clinical comparison between the DHL patients and the 2 groups of control patients is summarized in Table 2. Patients in the DHL group had a median serum LDH level at presentation of 727 U/L, nearly twice that of either control group. An LDH cutoff of >630 U/L (>3 times the upper limit of normal in our laboratory) was found to be relatively specific for distinguishing DHL patients from those with BL (88% specificity, p=0.006) and DLBCL (82.5% specificity, p=0.005). DHL patients had a higher prevalence of bone marrow and CNS disease than either BL or DLBCL patients and a higher prevalence of extranodal disease compared with BL patients. OS for DHL patients was markedly worse than for BL (p=0.002) and DLBCL (p=0.04) patients (Fig. 1). The median survival time in the DHL group was only 4.5 months (95% CI: 4.0-7.9), and all observed deaths occurred within 8 months of diagnosis. In contrast, the BL group did not reach its median survival (median follow-up: 31.7 months) and the median survival time of the DLBCL group was 38.9 months (95% CI: 13.2-83.7).
Morphologic and immunophenotypic features are summarized in Tables Tables33 and and4.4. Diagnostic tissue samples included lymph nodes (13 cases), extranodal sites (8 cases), bone marrow (5 cases), and peripheral blood (3 cases). Cases were classified according to the 2008 WHO Classification as follows: 12 cases of B-cell lymphoma, unclassifiable, with features intermediate between DLBCL and BL (BCLU) (60%) and 7 cases of DLBCL, not otherwise specified (DLBCL-NOS) (35%). Case 20 with a leukemic presentation and lymphoblastic morphology and immunophenotype was classified as B-LBL (Fig. 2).
All 12 cases classified as BCLU exhibited a diffuse growth pattern with abundant mitoses and frequent apoptosis (Table 3). Five of 12 (41.7%) contained a conspicuous starry-sky pattern and monomorphous cell population closely resembling BL (“Burkitt-like” morphology, Fig. 3A-B). In 4 of 12 cases (33.3%), there was some morphologic overlap with BL but cell size was variable and a proportion of the cells contained more irregular nuclei and more prominent nucleoli than classic BL; these cases had morphological features intermediate between BL and DLBCL, and therefore were classified as BCLU rather than DLBCL-NOS (“intermediate” morphology, Fig. 3C) (21, 28). Three of 12 BCLU cases (25%) contained small cells with finely dispersed chromatin resembling lymphoblasts (“blastoid” morphology, Fig. 3D); all 3 expressed strong CD20 and surface light chain and 2 cases tested for TdT were negative. Despite the high-grade morphology seen in cases classified as BCLU, the Ki-67 proliferation index (PI) ranged from 25-100% (median: 80%) (Fig. 4A).
All 7 cases classified as DLBCL-NOS exhibited morphological features outside of the spectrum of BL with predominantly large cells with oval or irregular nuclei; most had a conspicuous population of immunoblasts with prominent central nucleoli, while some contained mainly centroblasts with vesicular chromatin or cells with finely dispersed chromatin (Table 3, Fig. 3E-F). Some cases had frequent mitoses and single-cell necrosis, but all lacked a starry-sky pattern or areas of geographic necrosis. All had a Ki-67 PI of ≥80% (median: 85%, range: 80-95%) (Fig. 4B). Patients with tumors classified as DLBCL-NOS showed a trend toward improved OS compared to other DHL patients (p=0.18) (Fig. 5).
Among the 6 patients with a prior history of grade 1-2 FL, tissue from the initial FL was available for review in 5 (Table 3). The prior diagnosis of grade 1-2 FL was confirmed by morphologic examination and immunohistochemistry in 3 cases (case 1 with Burkitt-like BCLU, case 8 with BCLU with intermediate morphology, and case 19 with DLBCL-NOS). None of them showed evidence of a MYC rearrangement and a BCL2 rearrangement was present in 2 of 3 by FISH analysis of the initial lymphoma samples. In 2 cases (cases 10 and 11 with blastoid BCLU), review of the prior biopsies revealed morphological findings consistent with the blastoid variant of FL (38, 41, 55). In case 10 (Fig. 6A-C), a diagnosis of grade 1-2 FL had been made on a supraclavicular lymph node; 2 years later, DHL, with a MYC rearrangement and 8 copies of BCL2, developed in the testis. Subsequent FISH analysis of the supraclavicular lymph node revealed the same genetic abnormalities. In case 11 (Fig. 6D-F), the diagnosis of grade 1-2 FL had been made on 2 small retroperitoneal core needle biopsies; 6 months after the diagnosis, DHL developed in the central nervous system (CNS). Tissue from the prior biopsies was not available for FISH analysis to determine initial MYC status.
All 18 DHL cases tested for Bcl2 by immunohistochemistry were positive (Fig. 4C). Case 3 was Bcl2-negative with the commonly used antibody (clone 124), and positive with a different Bcl2 antibody (clone C-4) recognizing a longer amino acid segment of the protein (34). The majority of cases (18/20, 90%) had a germinal center B-cell like (GCB) immunophenotype (either CD10+, or Bcl6+, Mum1− if results of CD10 staining were not available), while only 2 (cases 2 and 15) had a non-GCB immunophenotype (CD10−, Bcl6+, Mum1+) (Fig. 4D-F) (14). Significant differences distinguishing DHL cases from BL controls included Bcl2 (p<0.0001) and Mum1/IRF4 (p=0.006) protein expression, Ki-67 PI <95% (p<0.0001), and EBER-negativity (p=0.006) (Table 4, Fig. 4A-F). Within the DHL group, Mum1/IRF4 expression or Ki-67 PI had no significant correlation with response to therapy, outcome or OS (data not shown).
Cytogenetic and molecular genetic features of DHL are summarized in Tables Tables44 and and5.5. Cytogenetic analysis, available in 11 cases, showed a complex karyotype in all (≥3 numerical or structural aberrations), with a higher median number of aberrations than BL controls (p= 0.0009) (Fig. 4G). Among the DHL cases, the MYC partner was frequently IGL (82%) or unknown (18%), while in BL cases the MYC partner was usually IGH (83%) or rarely IGL (17%), a difference that was statistically significant (p=0.001). There was no significant correlation between number of karyotypic abnormalities and response to therapy, outcome or OS within the DHL group (data not shown). Besides the t(14;18) and t(8q24), there was no cytogenetic abnormality common to all DHL cases. However, certain numerical and structural aberrations were seen frequently, including trisomy 12 (6 cases) and trisomy 7 (5 cases). Two cases had i(17)(q10), resulting in loss of heterozygosity for TP53 at 17p13.
Among the 9 cases lacking karyotypes, 8 had MYC and IGH-BCL2 rearrangements detected by interphase FISH (Fig. 4H-I). In case 10, FISH identified a MYC rearrangement and 8 copies of an intact BCL2; further analysis with a chromosome 18 centromeric probe revealed multiple signals, confirming chromosome 18 polysomy. BCL6 FISH analysis identified a rearrangement in a single case (case 4), confirming a 3q27 rearrangement suspected by karyotype (i.e. “triple-hit”) (Fig. 4G).
The purpose of this retrospective analysis was to define further the clinicopathologic spectrum of DHL in a well-characterized group of patients, to classify them according to the 2008 WHO Classification, and to compare the clinical and pathologic characteristics of DHL with 2 other high-grade B-cell lymphomas with which DHL may show morphologic and immunophenotypic overlap, DLBCL and BL. Our findings confirm that DHL has an aggressive clinical presentation and a poor prognosis with currently available therapies (6, 18, 19, 26, 30, 31, 36, 42, 47, 49). Our findings also suggest that DHL has clinical features distinct from both BL and DLBCL. DHL patients had a higher incidence of marrow and CNS involvement than either BL or DLBCL, and DHL patients had higher LDH levels at presentation compared with BL and IPI-matched DLBCL controls, with a median 3-4 times the upper limit of normal (Table 2). Importantly, DHL patients had a median OS of only 4.5 months, vastly inferior to that of either BL or DLBCL, and 70% of patients died within 8 months of diagnosis (Fig. 1). No clinical parameter emerged as predictive of superior OS within the DHL group, although PS <2 and age ≤60 years may suggest a slightly better prognosis. Our DHL patients were treated with either moderate intensity regimens used in DLBCL or high intensity regimens used in BL (Table 1). Neither regimen was associated with a superior outcome or OS, suggesting that highly intensive chemotherapy successfully used in BL may not offer clear benefit in DHL. Despite the retrospective nature of our data, our findings are in line with those of other studies and suggest the need for innovative therapeutic approaches to treat this disease (30, 37). Due to the high risk of CNS involvement, inclusion of CNS-directed therapy should be considered. Treatment of the majority of patients in our series with rituximab may help explain the somewhat lower mortality rate seen in our study as compared to others with similar follow-up in which fewer patients received rituximab (18, 26, 31, 42).
Because of its morphologic and immunophenotypic spectrum, recognition and classification of DHL has been difficult (30). All of the BCLU cases in our series bore some morphologic resemblance to BL, but had an immunophenotype unusual for BL such as positivity for Bcl2 or Mum1/IRF4 or a relatively low PI. All DHL cases diagnosed as DLBCL-NOS had a Ki-67 PI ≥80% and were morphologically distinct from BL, with conspicuous immunoblasts in most cases. In our study, classification as DLBCL-NOS was associated with a marginally superior OS compared with non-DLBCL morphology (Fig. 5), in contrast to the findings of Johnson et al (18) who also applied the 2008 WHO Classification to their cases and found that DLBCL morphology conferred a statistically significant OS benefit. In addition, our data showing that DHL overall has a worse prognosis that either BL or IPI-matched DLBCL underscores the importance of recognizing the presence of both MYC and IGH-BCL2 rearrangements for appropriate prognostic stratification. In particular, the fact that most DHL patients had a GCB phenotype but did so poorly suggests that pathologic prognosticators commonly used in DLBCL do not apply in DHL (14, 30).
In our series as with other published series of DHL, most patients presented with de novo disease, while a subset had a prior history of grade 1-2 FL from which the DHL was presumed to have transformed (18, 26, 30, 31, 49). Interestingly, review of the prior tumors in 2 cases revealed morphologic findings unusual for low-grade FL and associated with the blastoid variant of FL (38, 41, 55). Detection of a MYC rearrangement in the initial tumor from case 10 and the short time course from diagnosis of grade 1-2 FL to the development of DHL in case 11 suggest that both original tumors may have contained a DHL component. This is further supported by chromosomal studies of FL showing that secondary 8q24 abnormalities were found exclusively in patients with the blastoid variant of FL (38).
In our series, MYC was most frequently translocated with IGL. This is in contrast to some reports of DHL that show MYC to be more frequently translocated with IGH or a non-IG partner (18, 42) and to the majority of BL in which IGH is the most common MYC translocation partner (23). The pathogenetic significance of t(8;22) in DHL is not entirely known, but its presence implies that t(14;18) occurs primarily followed by t(8;22), since neither the expressed nor rearranged IGH is available to participate in a balanced translocation with MYC. The t(8;22) may play a diagnostic role by raising the possibility of DHL if early karyotypic analysis reveals a t(8;22) rather than a t(8;14). Careful review of the karyotype for the presence of t(14;18) or metaphase FISH to identify a cryptic IGH-BCL2 rearrangement may be warranted.
Despite the importance of conventional cytogenetic analysis in confirming a diagnosis of DHL, nearly half of the cases in our series lacked a karyotype and we relied on FISH to make the diagnosis. In many pathology laboratories, including our own, not all lymphoma cases are routinely sent for cytogenetic analysis, either due to insufficient tissue or lack of recognition of the importance of such analysis to a particular case at the time of surgery. In addition, some laboratories may not have access to routine cytogenetic analysis. In contrast, FISH is readily available at most large and reference laboratories and can confirm the diagnosis of DHL without the need for cell culture. Since DHL is relatively rare, with estimates ranging from 3-5% of high-grade B-cell lymphoma (18, 42), sending all cases of high-grade B-cell lymphoma for FISH analysis may be unnecessary. We suggest that when a karyotype is not available, FISH for both MYC and IGH-BCL2 should be limited to the following circumstances, in order to increase cost-effectiveness and minimize unnecessary testing:
Finally, our findings suggest that DHL warrants separate categorization in future classifications due to the clinical and molecular genetic features that distinguish it from other high-grade B-cell neoplasms, perhaps reflecting a different underlying pathogenesis (17, 30, 45). The basis for the extremely aggressive clinical behavior of DHL is likely related to MYC-induced growth promotion combined with the anti-apoptotic effect conferred by BCL2 overexpression (30). As in our series, the vast majority of reported cases exhibit a complex karyotype, pointing to a role for clonal evolution and further gene dysregulation. Most DHL cases studied using gene expression profiling have shown a molecular signature intermediate between DLBCL and BL (17), while fewer cases more closely resemble BL (7). Comparative genomic hybridization studies demonstrated that even such cases of gene expression-defined BL that do not fulfill WHO criteria for BL (so-called “discrepant BL” cases) show significant alterations from pediatric and adult BL and DLBCL (45). It is possible that DHL represents a neoplasm in which a primary genetic aberration and induction of a specific gene-expression profile was subsequently shifted in a different direction due to the acquisition of additional genetic changes (17). This may explain the morphologic and immunophenotypic heterogeneity, but shared molecular genetic and clinical features, seen in these cases.
In summary, we describe 20 patients with B-cell neoplasms with concurrent IGH-BCL2 and MYC rearrangements. Our findings confirm these neoplasms to be clinically aggressive with a poor prognosis. High intensity regimens used to treat BL have yet to demonstrate a clear benefit in DHL patients. DHL may show morphologic and immunophenotypic overlap with BL or DLBCL, or less frequently with LBL or FL with blastoid morphology. Awareness of this pathologic spectrum is important in directing ancillary testing to detect IGH-BCL2 and MYC rearrangements, particularly in laboratories where conventional cytogenetics is not routinely performed or available. Even with the use of the current WHO Classification containing intermediate categories, classification of DHL remains problematic. Despite its morphologic and immunophenotypic heterogeneity, the distinct clinical and molecular genetic features of DHL may warrant categorization as a separate entity in future classifications, as well as the development of novel treatment strategies.
The authors are grateful to Dr. A. John Iafrate of the MGH Diagnostic Molecular Pathology Laboratory for his assistance with interpretation of FISH results.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.