Welcome international healthcare professionals

This site is no longer supported and will not be updated with new content. You are welcome to browse and download all content already included in the site. Please note you will have to register your email address to access the site.

You are here

CD79B and MYD88 mutations in diffuse large B-cell lymphoma

Human Pathology, 3, 45, pages 556 - 564


Mutations in 2 upstream components of the nuclear factor κB (NF-κB) pathway, CD79B and MYD88, are important information for new target therapy in malignant lymphoma. We examined the prevalence and clinicopathologic characteristics of CD79B and MYD88 mutation in a cohort of Asian diffuse large B cell lymphoma (DLBCL) patients. CD79B and MYD88 mutations were analyzed by Sanger sequencing in 187 DLBCL tissue samples. CD79B immunoreceptor tyrosine-based activation motif spanning exon 5 and 6 and MYD88 TIR domain spanning exons 3, 4 and 5 were amplified and sequenced. The cell-of-origin was determined based on immunohistochemical stains for CD10, BCL-6 and MUM-1 by Hans' algorithm. CD79B was mutated in 16 cases (8.5%), mostly involving the first tyrosine (Y196) of immunoreceptor tyrosine-based activation motif. For MYD88, L265P mutation was found in 31 cases (out of 161, 19.3%). In 11 of these, a CD79B mutation coexisted, which constituted 69% of CD79B mutants and 36% of MYD88 L265P cases. Clinicopathologic comparison between the mutant and the wild-type group showed that the mean age was older for both CD79B (66 versus 58 years) and MYD88 L265P mutant groups (64 versus 58 years). Survival analyses showed that neither CD79B mutation nor MYD88 L265P was a significant prognostic indicator. In conclusion, CD79B and MYD88 mutations are associated with an older age at onset in DLBCL with a significant overlap, which did not affect the outcome of the disease.

Keywords: CD79B, MYD88, Mutation, Diffuse large B cell lymphoma.

1. Introduction

Aberrant activation of nuclear factor κB (NF-κB) is known largely to promote oncogenic characteristics such as anti-apoptosis, proliferation, and changes in cell adhesion while tumor-suppressing effects have also been demonstrated in certain neoplasms [1] . Oncogenic NF-κB activities have been shown in various solid and hematologic malignancies, the latter including diffuse large B cell lymphoma (DLBCL), mucosa-associated lymphoid tissue (MALT) lymphoma, multiple myeloma, Hodgkin lymphoma and some leukemias [2] .

DLBCL can be distinguished according to the gene expression profiling into two major subtypes, the germinal center B-cell–like (GCB) type and the activated B-cell–like (ABC) type, and the minor unclassifiable group [3] . Constitutive activation of NF-κB pathway, which is normally transiently activated by antigen-dependent stimulation in B cells, is characteristic of ABC DLBCL [2] . The survival of ABC DLBCL is known to be inferior to that of GCB DLBCL, and the poor response to chemotherapy has been attributed to the anti-apoptotic effect of NF-κB [4] and [5].

In DLBCL, the mechanism of constitutive NF-κB activity has been traced to somatic genetic alterations in the upstream pathway components such as CD79A, CD79B, CARD11, A20 and MYD88 [2] . It has been reported that immunoreceptor tyrosine-based activation motif (ITAM) mutations of CD79B and, less frequently, of CD79A were present in DLBCL cell lines and biopsy samples, which were largely of the ABC subtype (21% of ABC DLBCL and 3% of GCB DLBCL, for CD79B) [6] . CD79 mutations were shown to increase surface B-cell receptor (BCR) expression and nullify the negative regulation of BCR, which were suggested to support the “chronic active” BCR signaling leading to constitutive NF-κB activation in ABC DLBCL [6] .

MYD88 is an adaptor protein of the toll-like receptors and interleukin-1 receptors; through association with interleukin-1 receptor-associated kinases, MYD88 mediates downstream activation of NF-κB and mitogen-activated protein kinases, the outcome of which includes secretion of interleukin-6 (IL-6) [7] . Mutations of MYD88 that are oncogenic, NF-κB–activating, have been identified in DLBCL by Ngo et al [8] . They showed that the most frequent and most oncogenic form was the L265P mutation of Toll/IL-1 receptor (TIR) domain, which was detected in 29% of ABC DLBCLs as well as in 9% of MALT lymphomas while rare in GCB DLBCLs. Recently, the MYD88 L265P somatic mutation was demonstrated to be highly recurrent (about 90%) in IgM-secreting lymphoplasmacytic lymphoma (Waldenström's macroglobulinemia; WM) [9] . The use of L265P detection in discriminating WM from a morphologically overlapping B-cell neoplasm, such as marginal zone lymphoma, and in monitoring progression to WM from IgM monoclonal gammopathy of undetermined significance was proposed in subsequent studies [10] and [11].

With the advent of various targeted therapeutic agents acting on NF-κB–related pathways [12] , knowledge on the frequency of individual NF-κB–affecting mutations and the clinicopathologic impact of such mutations is appreciated. We sought to examine the prevalence, clinicopathologic characteristics and possible overlap of the CD79B and MYD88 mutation in a cohort of Asian DLBCL patients.

2. Materials and methods

2.1. Selection of DLBCL cases and categorization

A total of 187 de novo DLBCL cases diagnosed from 1994 to 2005 at Samsung medical center, Seoul, South Korea were chosen based on the availability of the clinical follow-up data and tumor DNA. The diagnosis of DLBCL was made according to the 2008 World Health Organization classification [13] , and DLBCL associated with a low grade lymphoma such as MALT lymphoma or an immune-compromised setting was excluded from the selection. All patients underwent chemotherapy with or without other treatment modalities such as surgery and radiation therapy. The first-line chemotherapy regimen was mostly CHOP (cyclophosphamide, doxorubicin, vincristine, prednisolone; n = 89) or R-CHOP (Rituximab plus CHOP; n = 86) with the methotrexate-based regimen administered to the 12 primary central nervous system (CNS) DLBCL cases. The “non-GCB” of Hans' algorithm  [14] was designated as “ABC” in the present study. Immunohistochemical staining was performed on a 4-μm thick section of formalin-fixed paraffin-embedded tissue processed in an automated system (BOND-MAX, Leica, Wetzlar, Germany) using monoclonal antibodies against CD10 (Leica), BCL-6 (Leica), and MUM-1 (DAKO, Carpinteria, CA, USA). In 72 of 187 cases, the immunohistochemical staining was done in tissue microarray preparations consisting of 700-μm cores of tumor area. Epstein-Barr virus (EBV)–encoded RNA in situ hybridization was carried out using EBV in situ hybridization kit (Leica). Strong reactivity in the majority (>50%) of tumor cells was the criterion for EBV positivity, which resulted in a lower percentage of positive cases (3%) compared with that of our previous report [15] .

2.2. DNA isolation and sequencing

Genomic DNA was isolated from a 5-μm thick section of formalin-fixed paraffin-embedded tumor tissue using the QIAamp FFPE DNA Tissue Kit (Qiagen, Germantown, MD, USA). For sequencing CD79B immunoreceptor tyrosine-based activation motif (ITAM), polymerase chain reaction was performed to amplify the area spanning exon 5 and 6 using two sets of primers: CD79B-5F (5′-GGGCTGGGGGACACTAACACTC-3′), CD79B-5R (5′-TGGGTGCTCACCTACAGACCAC-3′), CD79B-6F (5′-CGGGGTCAGTGGCCACTATCTG-3′) and CD79B-6R (5′-AGCAGTCACTGAGGCCAGGGAG-3′). Sanger sequencing was done bidirectionally. For amplification of MYD88 TIR domain spanning exon 3, 4 and 5, primer pairs used were as follows: MYD88-3F (5′-AAGCCTTCCCATGGAGCTCTGACCAC-3′), MYD88-3R (5′-GCTAGGAGGAGATGCCCAGTATCTG-3′), MYD88-4F (5′-ACTAAGTTGCCACAGGACCTGCAGC-3), MYD88-4R (5′–ATCCAGAGGCCCCACCTACACATTC-3′), MYD88-5F (5′-GTTGTTAACCCTGGGGTTGAAG-3′), MYD88-5R (5′-GCAGAAGTACATGGACAGGCAGACAGATAC-3′). The polymerase chain reaction for MYD88 failed in 26 of the 187 cases, leaving 167 cases for analysis based on MYD88 mutation status. The study protocol was approved by the institutional review board of Samsung Medical Center.

2.3. Statistical analysis

SPSS software (version 19.0, IBM, Armonk, NY, USA) was used for statistical analyses. The level of significance for all tests was set at 0.05, two-tailed. The age was expressed as mean ± standard deviation or median (range). Association between categorical variables was assessed with Pearson's χ2 test or Fisher's exact test. Overall survival (OS) was defined as the period between the pathological diagnosis and disease-related death. Progression-free survival (PFS) was determined from the date of diagnosis to the date of first relapse or disease-related death. The univariate comparison of OS and PFS was made using the Kaplan-Meier method with log-rank test or Breslow test, the latter in case the survival curves crossed. Multivariate analysis was done with the Cox proportional hazards model.

3. Results

3.1. Identification of CD79B ITAM mutations and clinicopathologic analysis

Sequencing for CD79B exon 5 and 6 of 187 DLBCL cases revealed 15 missense mutations and one nonsense mutation (8.5% in total), which were all in the ITAM domain ( Table 1 ). The prevalence of the first ITAM tyrosine (Y196) missense mutation (14 of 16) was in keeping with the previous report by Davis et al [6] . The CD79B ITAM mutations insignificantly prevailed in ABC DLBCL (10.8% in ABC versus 5.1% in GCB type, P = .4, Fisher's exact test), and the overall frequency of mutation was lower in our cohort (cf. about 16% in Davis et al [6] ).

Table 1 Details of CD79B ITAM mutations categorized according to the cell of origin

ABC type (13/128, 10.2%) GCB type (3/59, 5.1%)
Y196N (4×): TAC → AAC Y196N: TAC → AAC
Y196C (3×): TAC → TGC Y196S: TAC → TCC
Y196D (2×): TAC → GAC E197stop: GAG → TAG
Y196S: TAC → TCC  
Y196H: TAC → CAC  
L199P: CTG → CCG  

The CD79B mutant group's age at diagnosis was significantly older than the wild-type group (average 65.6 ± 12 versus 57.6 ± 13 years, P = .038, t test). In other clinicopathologic parameters, such as the sex, international prognostic index (IPI) risk, and treatment response, no significant difference was found ( Table 2 ). Of interest, EBV positivity, which may up-regulate the NF-κB pathway through the latent membrane proteins [16] , was found to be mutually exclusive with the presence of a CD79B mutation.

Table 2 Clinicopathologic characteristics of DLBCL with CD79B mutation

Parameter All patients (n = 187) CD79B P
Mutant (n = 16) Wild type (n = 171)
Age (range) 58 (20-88) 64 (49-85) 58 (20-88) .038 a
 ≤60 100 (54%) 6 (38%) 94 (55%) .18
 >60 87 (47%) 10 (63%) 77 (45%)  
Cell-of-origin       .40
 ABC 128 (68%) 13 (81%) 115 (67%)  
 GCB 59 (32%) 3 (19%) 56 (33%)  
Primary site       .98
 Nodal 59 (32%) 5 (31%) 54 (32%)  
 Extranodal 128 (68%) 11 (69%) 117 (68%)  
GI origin       .15
 GI 57 (30%) 2 (13%) 55 (32%)  
 Non-GI 130 (70%) 14 (87%) 116 (68%)  
CNS origin       1.0
 CNS 12 (6%) 1 (6%) 11 (6%)  
 Non-CNS 175 (94%) 15 (94%) 160 (94%)  
Sex       .60
 Male 114 (61%) 11 (69%) 103 (60%)  
 Female 73 (39%) 5 (31%) 68 (40%)  
IPI risk group       1.0
 L/LI 150 (80%) 13 (81%) 137 (80%)  
 HI/H 37 (20%) 3 (19%) 34 (20%)  
EBV       1.0
 Positive 6 (3%) 0 (0%) 6 (4%)  
 Negative 181 (97%) 6 (100%) 165 (96%)  
Lactate dehydrogenase       .43
 Normal 122 (65%) 9 (56%) 113 (66%)  
 Elevated 65 (35%) 7 (44%) 58 (34%)  
Treatment response       .7
 CR/PR 163 (88%) 15 (94%) 148 (88%)  
 SD/PD 22 (12%) 1 (6%) 21 (12%)  
 Not assessed 2      
Performance status       .38
 ECOG 0-1 163 (88%) 15 (94%) 148 (87%)  
 ECOG 2-4 23 (12%) 1 (6%) 22 (13%)  
 Not assessed 1      
Ann Arbor stage       .98
 Limited, I-II 129 (69%) 11 (69%) 118 (69%)  
 Advanced, III-IV 58 (31%) 5 (31%) 53 (31%)  
Bone marrow involvement       .38
 Positive 19 (10%) 0 (0%) 19 (11%)  
 Negative 168 (90%) 16 (100%) 152 (89%)  
Primary treatment       .92
 R-CHOP 86 (46%) 8 (50%) 78 (46%)  
 CHOP 89 (48%) 7 (44%) 82 (48%)  
 Other b 12 (6%) 1 (6%) 11 (6%)  

a t Test.

b Methotrexate-based polychemotherapy.

Abbreviations: GI, gastrointestinal; CNS, central nervous system; L/LI, low or low intermediate risk; HI/H, high intermediate or high risk; CR/PR, complete remission or partial remission; SD/PD, stable disease or progressive disease; ECOG, Eastern Cooperative Oncology Group.

The estimated 5-year OS and PFS rates of the DLBCL cohort were 73.6 % and 65 %, respectively, with the median follow-up of 88 months for surviving patients. Among the parameters known to be associated with worse prognosis in DLBCL [17] , high IPI risk group, elevated lactate dehydrogenase level, and CNS origin were significantly associated with shorter OS and PFS (P < .01 for each, log-rank test) in our cohort. Additionally, gastrointestinal tract origin was a factor of better prognosis (P = .006 for OS, P < .001 for PFS, log-rank test) as it had been reported [18] . High IPI risk group was an independent prognostic factor in multivariate analysis ( Table 3 ). However, the survival of the ABC type, another known adverse prognostic factor, was not significantly shorter compared with that of the GCB type (P = .25 for OS and P = .11 for PFS, log-rank test). Similarly, older age (>60 years) failed to show a significant association with an inferior prognosis (P = .24 for OS and P = .28 PFS, log-rank test). The R-CHOP regimen appeared to confer an improvement in the PFS of the high IPI group as compared to CHOP regimen although it was not statistically significant (P = .064, Breslow test, Supplementary Figure 1 ). On univariate analyses of the PFS as well as OS, the presence of a CD79B mutation was not a significant prognostic factor ( Fig. A). Looking into CD79B mutant group's survival separately for the IPI risk group and chemotherapy regimen suggested that, although not statistically significant, R-CHOP-treated CD79B-mutated patients with a low IPI risk may survive longer compared with the wild type patients ( Fig. B).

Table 3 Multivariate Cox proportional hazards model analysis

  Overall survival Progression-free survival
CD79B mutation 1.0 (0.33-2.98) 1.0 0.82 (0.27-2.45) .72
MYD88 L265P 0.57 (0.23-1.41) .22 0.57 (0.24-1.33) .20
Age a 1.02 (0.99-1.05) .13 1.02 (0.99-1.04) .23
GI origin 0.64 (0.30-1.34) .23 0.45 (0.22-0.91) .027
R-CHOP b 0.60 (0.29-1.26) .18 0.62 (0.32-1.18) .14
High IPI 3.80 (1.85-7.82) <.001 2.55 (1.30-5.01) .007

a Continuous variable.

b R-CHOP versus CHOP regimen.

NOTE. n = 175, excluding 12 primary CNS DLBCL cases.

Abbreviations: HR, hazard ratio with 95% confidence interval in parenthesis; GI, gastrointestinal tract; CHOP, cyclophosphamide, doxorubicin, vincristine, prednisolone; R-CHOP, rituximab plus cyclophosphamide, doxorubicin, vincristine, prednisolone; IPI, international prognostic index.


Fig. Effect on progression-free survival of CD79B and MYD88 L265P mutation. A, CD79B mutant (black) versus CD79B wild type (WT, gray). B, CD79B mutant versus WT in a subset of patients who were of low international prognostic index(IPI) and treated with R-CHOP regimen. C, MYD88 L265P (black) versus MYD88 WT (gray). D, MYD88 L265P versus WT in a subset of patients who were of low IPI and treated with R-CHOP regimen. E, Patients harboring both CD79B mutation and MYD88 L265P (Double mutant, black; all were of low IPI) versus patients having neither mutation and of low IPI (gray). F, Double mutant versus patients having MYD88 L265P without a CD79B mutation and of low IPI.

3.2. Detection of MYD88 L265P mutation and clinicopathologic analysis

Since CD79B and MYD88 activate NF-κB through separate pathways, the coexistence of these mutations needs to be considered in applying agents that target upstream pathways of NF-κB [2] and [12]. We sequenced exons 3, 4 and 5 of MYD88 in 161 of the 187 DLBCL samples including all 16 of the CD79B mutant cases. The L265P mutation of exon 5, and no other mutation, was found in 31 cases (19.3 %). The MYD88 L265P mutation coexisted with a CD79B mutation in 11 cases, which corresponded to 36% of MYD88 L265P mutations and 69% of CD79B mutations (P < .001, χ2 test). These proportions of overlap were comparable to those of the previous report [8] .

The clinicopathologic parameters of the MYD88 L265P group were not significantly different from those of the MYD88 wild-type group except for the older age (average 64.3 ± 12 versus 57.9 ± 13 years, P = .010, t test; Table 4 ). The primary site of disease was extranodal in 22 L265P cases (71%), which included testis (4 out of total 8 in the cohort), tonsil (3 out of total 14), stomach (2 out of total 15), and ileo-cecum (5 out of total 36) among the frequent sites. As seen with CD79B mutations, the MYD88 mutations did not significantly prevail in ABC DLBCL ( Table 4 ) (21.6% in ABC versus 14% in GCB type, P = .29, χ2 test). EBV positivity was again not observed among MYD88 mutant cases.

Table 4 Clinicopathologic characteristics of DLBCL with MYD88 L265P mutation

Parameter All patients (n = 161) MYD88 L265P P MYD88 + CD79B c (n = 11)
Mutant (n = 31) Wild type (n = 130)
Age (range) 60 (20-88) 63 (34-88) 58.5 (20-86) .010 a 65 (49-85)
 ≤60 82 (51%) 10 (32%) 72 (55%) .021 4 (36%)
 >60 79 (49%) 21 (68%) 58 (45%)   7 (64%)
Cell of origin       .29  
 ABC 111 (69%) 24 (77%) 87 (67%)   9 (82%)
 GCB 50 (31%) 7 (23%) 43 (33%)   2 (18%)
Primary site       .75  
 Nodal 43 (27%) 9 (29%) 34 (26%)   4 (36%)
 Extranodal 118 (73%) 22 (71%) 96 (74%)   7 (64%)
GI origin       .67  
 GI 52 (32%) 9 (29%) 43 (33%)   1 (9%)
 Non-GI 109 (68%) 22 (71%) 87 (67%)   10 (91%)
CNS origin       .46  
 CNS 12 (7%) 1 (3%) 11 (8%)   1 (9%)
 Non-CNS 149 (93%) 30 (97%) 119 (92%)   10 (91%)
Sex       .67  
 Male 100 (62%) 22 (71%) 78 (60%)   8 (73%)
 Female 61 (38%) 9 (29%) 52 (40%)   3 (27%)
IPI risk group       .80  
 L/LI 130 (81%) 26 (84%) 104 (80%)   11 (100%)
 HI/H 31 (19%) 5 (16%) 26 (20%)   0 (0%)
Lactate dehydrogenase       .80  
 Normal 106 (66%) 21 (68%) 85 (65%)   7 (64%)
 Elevated 55 (34%) 10 (32%) 45 (35%)   4 (36%)
EBV       .58  
 Positive 5 (3%) 0 (0%) 5 (4%)   0 (0%)
 Negative 156 (97%) 31 (100%) 125 (96%)   11 (100%)
Treatment response       .37  
 CR/PR 139 (87%) 29 (94%) 110 (86%)   11 (100%)
 SD/PD 20 (13%) 2 (6%) 18 (14%)   0 (0%)
 Not assessed 2        
Performance status       .77  
 ECOG 0-1 141 (88%) 27 (87%) 114 (88%)   11 (100%)
 ECOG 2-4 19 (12%) 4 (13%) 15 (12%)   0 (0%)
 Not assessed 1        
Ann Arbor stage       .77  
 Limited, I-II 116 (72%) 23 (74%) 93 (72%)   9 (82%)
 Advanced, III-IV 45 (28%) 8 (26%) 37 (28%)   2 (18%)
Bone marrow involvement       .72  
 Positive 13 (8%) 3 (10%) 10 (8%)   0 (0%)
 Negative 148 (92%) 28 (90%) 120 (92%)   11 (100%)
Primary treatment       .049  
 R-CHOP 67 (42%) 8 (26%) 59 (45%)   4 (36%)
 CHOP 82 (51%) 22 (71%) 60 (46%)   6 (55%)
 Other b 12 (7%) 1 (3%) 11 (9%)   1 (9%)

a t Test.

b Methotrexate-based polychemotherapy.

c Coexistence of MYD88 L265P and a CD79B mutation.

Abbreviations: GI, gastrointestinal; CNS, central nervous system; L/LI, low or low intermediate risk; HI/H, high intermediate or high risk; CR/PR, complete remission or partial remission; SD/PD, stable disease or progressive disease; ECOG, Eastern Cooperative Oncology Group.

The Kaplan-Meier survival curves showed an insignificant tendency of the L265P group for better prognosis ( Fig. C). An older age (>60 years) is a known adverse prognostic factor although not reaching significance in our cohort. As the percentage of patients >60 years was larger in the L265P group than in the wild type ( Table 4 ), the effect of L265P mutation on the survival was examined in the subset of patients >60 years (n = 79 comprising 21 L265Ps and 58 wild types). Again, an insignificant upward shift of the survival curves was observed in the mutant group (P = .48 for OS and P = 0.24 for PFS, log-rank test), which suggested against the possibility of a favorable prognostic impact of L265P masked by the adverse effect of an older age. A lower proportion of L265P patients had been treated with the R-CHOP regimen compared with the wild-type group ( Table 4 ). Examination of the survival according to the chemotherapy regimen and IPI risk group raised the possibility that L265P patients subjected to R-CHOP may survive better than the wild-type patients in the low–IPI risk group ( Fig. D).

The characteristics of the 11 cases harboring both the MYD88 and CD79B mutations are also summarized in Table 4 . The primary site of tumor was extranodal in 7 of the 11 cases, which comprised testis (3 cases), small intestine, tonsil, breast and brain. The age of double mutant cases (median 65 years) appeared older than that of the 5 cases having a CD79B mutation only (median 61 years) or the 20 cases harboring MYD88 L265P only (median 62.5 years), which was not statistically significant (P > 0.5 for each, Mann-Whitney test). The survival curves of double mutant patients, who all belonged to the low–IPI risk group, showed a slight upward shift in PFS compared with that of low–IPI risk group patients having neither CD79B nor MYD88 mutation (P = .67 for OS and P = .34 for PFS, Breslow test) ( Fig. E). Comparison of survivals between the double mutant patients and the MYD88 L265P patients not having a CD79B mutation and with a low IPI risk (n = 15) resulted in no difference (P = .69 for OS and P = .88 for PFS, Breslow test) ( Fig. F).

4. Discussion

We observed CD79B ITAM mutations and MYD88 L265P mutations in 8.5% and about 19.3% of DLBCL, respectively, with a significant overlap. The individual frequencies of mutations were lower than those reported by Staudt and colleagues [6] and [8], but the percentage of coexistence was comparable: Ngo et al [8] noted that 44% of CD79B mutations and 33% of MYD88 L265P mutations overlapped with each other among ABC DLBCL cases. Exome studies of DLBCL [19], [20], and [21] also found CD79B and MYD88 among recurrently mutated genes: the frequency of CD79B mutations ranged from 8% to 13%, and that for MYD88, including non-L265P mutations, from 7% to 16% with occasional coexistence of both mutations ( Table 5 ). These mutations were predominantly or even exclusively detected in ABC DLBCL in the works by Staudt and colleagues and the above exome studies, which is in keeping with the CD79B and MYD88 mutations activating NF-κB. In our DLBCL cohort, although both CD79B mutations and MYD88 L265P mutations appeared to be more frequent in ABC DLBCL, the difference was not statistically significant. This slight discrepancy may be explained by the difference in the method of determining the cell-of-origin. While other studies used gene expression profiling, we used the immunohistochemical method, which is known to be not perfect in predicting the gene expression profiling–based classification [17] and [22].

Table 5 Summary of reports on CD79B and MYD88 mutations in DLBCL

Reference Sequencing method Ethnic group or region Tumor site Gene Mutation frequency Effect on survival
Davis et al [6] Sanger sequencing North America and Europe Unrestricted CD79B 16% (n = 225) NA
Ngo et al [8] Sanger sequencing North America and Europe Unrestricted MYD88 (L265P) 30% (n = 242) NA
Morin et al [21] RNA sequencing North America Unrestricted CD79B

MYD88 a
8.3% (n = 96)

12.5% (n = 96)
Pasqualucci et al [20] Exome sequencing and Sanger sequencing North America Unrestricted CD79B

MYD88 a
11.7% (n = 111)

8.1% (n = 111)
Lohr et al [19] Exome sequencing North America Unrestricted CD79B

MYD88 a
16.3% (n = 49)

12.2% (n = 49)
Pham-Ledard et al [24] Sanger sequencing France Skin MYD88 (L265P) 69% (n = 16) NA
Gonzalez-Aguilar et al [23] Exome sequencing France CNS MYD88 (L265P) 38% (n = 29) Insignificant
Choi et al [22] Sanger sequencing Korean Unrestricted MYD88 (L265P) 6.5% (n = 124) Insignificant
Current study Sanger sequencing Korean Unrestricted CD79B

MYD88 (L265P)
8.5% (n = 187)

19.3% (n = 161)


a Various mutations including L265P.

Abbreviations: NA, not assessed; CNS, central nervous system.

The clinicopathologic characteristics of CD79B mutation in DLBCL has not been reported previously as far as we are aware, but several accounts of MYD88 mutation in DLBCL have been published. The MYD88 L265P mutation was detected in 38% (11/29) of primary CNS DLBCL [23] and 69% (11/16) of primary cutaneous DLBCL, leg type [24] . Although there were 12 cases of primary CNS DLBCL in our cohort, only one case (8%) had the L265P mutation, coexistent with CD79B Y196D mutation. No case of cutaneous DLBCL, leg type was included in our cohort. A recent study of a Korean DLBCL cohort [25] showed that MYD88 protein overexpression was a poor prognostic factor and that the L265P mutation, detected at 6.5% (8/124), was not associated with protein expression or survival. It was noticeable that similar to our study, the median age of the mutant cases was older than that of the whole cohort (64 versus 56 years). The authors stated that no significant difference existed in clinical parameters between the mutant and wild-type cases, however.

The age of patients with a CD79B or MYD88 mutation was older than that of the respective wild-type group. It has been reported that the incidence of various genetic complexities, including translocations and gains, increases with age in DLBCL [26] . The apparent association with an older age (>60 years) for the CD79B and MYD88 mutation is in keeping with such observation, and thus may be a reflection of the general trend in increased vulnerability to somatic mutation with aging. The presence of CD79B mutations or MYD88 L265P was not a significant prognostic factor, although the possibility for a better outcome was raised. None of the previous studies of DLBCL showed an association of these mutations with survival, either: the primary limitation seems to be the small sample size in all instances. Incidentally, recent studies of MYD88 L265P in Waldenström's macroglobulinemia mentioned that no difference in survival was found between the mutant and wild type group [27] and [28]. As 69% of CD79B mutations coexisted with MYD88 L265P in our cohort, the age and the survival characteristics of CD79B mutant cases could be attributed to that of MYD88 L265P. If any significant effect on survival of the MYD88 L265P or a CD79B were to be demonstrated in a larger cohort of DLBCL, the status of both mutations would need to be looked into.

It has been suggested that the contribution of CD79B and MYD88 mutations to NF-κB activation may be “non-redundant”, or additive [8] . The cytotoxic effect of short hairpin RNAs blocking a BCR pathway component or MYD88 in ABC DLBCL cell lines was shown to be enhanced when two species of RNAs were co-applied in an ABC DLBCL line [8] . ABC DLBCL lines harboring both MYD88 L265P and a CD79B/CD79A mutation, namely, HBL1, TMD8 and OCI-Ly10, were susceptible to block of the BCR signaling pathway by inhibitors of protein kinase C, spleen tyrosine kinase, or Bruton's tyrosine kinase [29] . Whether the group of DLBCL patients with coexisting CD79B and MYD88 mutations could benefit from a combined regimen of newer drugs acting on the BCR signaling pathway and MYD88 pathway remains to be seen. The mutual exclusivity of CD79B or MYD88 mutations with EBV infection, another potential NF-κB activating factor, is interesting in that the redundancy is avoided.

In conclusion, mutations in CD79B and MYD88, two of the NF-κB upstream components, were identified in DLBCL at 8.5% and 19.3%, respectively, with a significant overlap. The presence of these mutations did not affect the clinical outcome of the disease but may offer additional therapeutic options to the patients as targeted agents are under development.

Supplementary data

The following is the Supplementary data to this article

Download file

supplemental figure 1.doc


  • [1] N.D. Perkins. The diverse and complex roles of NF-kappaB subunits in cancer. Nat Rev Cancer. 2012;12:121-132
  • [2] K.H. Lim, Y. Yang, L.M. Staudt. Pathogenetic importance and therapeutic implications of NF-kappaB in lymphoid malignancies. Immunol Rev. 2012;246:359-378 Crossref
  • [3] A.A. Alizadeh, M.B. Eisen, R.E. Davis, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature. 2000;403:503-511 Crossref
  • [4] A.S. Baldwin. Control of oncogenesis and cancer therapy resistance by the transcription factor NF-kappaB. J Clin Invest. 2001;107:241-246 Crossref
  • [5] G. Lenz, G. Wright, S.S. Dave, et al. Stromal gene signatures in large-B-cell lymphomas. N Engl J Med. 2008;359:2313-2323 Crossref
  • [6] R.E. Davis, V.N. Ngo, G. Lenz, et al. Chronic active B-cell-receptor signalling in diffuse large B-cell lymphoma. Nature. 2010;463:88-92 Crossref
  • [7] T. Kawai, S. Akira. The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nat Immunol. 2010;11:373-384 Crossref
  • [8] V.N. Ngo, R.M. Young, R. Schmitz, et al. Oncogenically active MYD88 mutations in human lymphoma. Nature. 2011;470:115-119 Crossref
  • [9] S.P. Treon, L. Xu, G. Yang, et al. MYD88 L265P somatic mutation in Waldenstrom's macroglobulinemia. N Engl J Med. 2012;367:826-833 Crossref
  • [10] M. Varettoni, L. Arcaini, S. Zibellini, et al. Prevalence and clinical significance of the MYD88 (L265P) somatic mutation in Waldenström's macroglobulinemia and related lymphoid neoplasms. Blood. 2013;121:2522-2528 Crossref
  • [11] L. Xu, Z.R. Hunter, G. Yang, et al. MYD88 L265P in Waldenström macroglobulinemia, immunoglobulin M monoclonal gammopathy, and other B-cell lymphoproliferative disorders using conventional and quantitative allele-specific polymerase chain reaction. Blood. 2013;121:2051-2058 Crossref
  • [12] J.J. Steinhardt, R.B. Gartenhaus. Promising personalized therapeutic options for diffuse large B-cell lymphoma subtypes with oncogene addictions. Clin Cancer Res. 2012;18:4538-4548 Crossref
  • [13] S.H. Swerdlow, E. Campo, N.L. Harris, et al. WHO classification of tumours of haematopoietic and lymphoid tissues. (International Agency for Research on Cancer, Lyon, 2008)
  • [14] C.P. Hans, D.D. Weisenburger, T.C. Greiner, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood. 2004;103:275-282 Crossref
  • [15] S. Park, J. Lee, Y.H. Ko, et al. The impact of Epstein-Barr virus status on clinical outcome in diffuse large B-cell lymphoma. Blood. 2007;110:972-978 Crossref
  • [16] L.S. Young, A.B. Rickinson. Epstein-Barr virus: 40 years on. Nat Rev Cancer. 2004;4:757-768 Crossref
  • [17] L.H. Sehn. Paramount prognostic factors that guide therapeutic strategies in diffuse large B-cell lymphoma. Hematol Am Soc Hematol Educ Program. 2012;2012:402-409
  • [18] A. Lopez-Guillermo, L. Colomo, M. Jimenez, et al. Diffuse large B-cell lymphoma: clinical and biological characterization and outcome according to the nodal or extranodal primary origin. J Clin Oncol. 2005;23:2797-2804
  • [19] J.G. Lohr, P. Stojanov, M.S. Lawrence, et al. Discovery and prioritization of somatic mutations in diffuse large B-cell lymphoma (DLBCL) by whole-exome sequencing. Proc Natl Acad Sci U S A. 2012;109:3879-3884 Crossref
  • [20] L. Pasqualucci, V. Trifonov, G. Fabbri, et al. Analysis of the coding genome of diffuse large B-cell lymphoma. Nat Genet. 2011;43:830-837 Crossref
  • [21] R.D. Morin, M. Mendez-Lago, A.J. Mungall, et al. Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma. Nature. 2011;476:298-303 Crossref
  • [22] W.W. Choi, D.D. Weisenburger, T.C. Greiner, et al. A new immunostain algorithm classifies diffuse large B-cell lymphoma into molecular subtypes with high accuracy. Clin Cancer Res. 2009;15:5494-5502 Crossref
  • [23] A. Gonzalez-Aguilar, A. Idbaih, B. Boisselier, et al. Recurrent mutations of MYD88 and TBL1XR1 in primary central nervous system lymphomas. Clin Cancer Res. 2012;18:5203-5211 Crossref
  • [24] A. Pham-Ledard, D. Cappellen, F. Martinez, B. Vergier, M. Beylot-Barry, J.P. Merlio. MYD88 somatic mutation is a genetic feature of primary cutaneous diffuse large B-cell lymphoma, leg type. J Invest Dermatol. 2012;132:2118-2120 Crossref
  • [25] J.W. Choi, Y. Kim, J.H. Lee, Y.S. Kim. MYD88 expression and L265P mutation in diffuse large B-cell lymphoma. Hum Pathol. 2013;44:1375-1381 Crossref
  • [26] W. Klapper, M. Kreuz, C.W. Kohler, et al. Patient age at diagnosis is associated with the molecular characteristics of diffuse large B-cell lymphoma. Blood. 2012;119:1882-1887 Crossref
  • [27] S. Poulain, C. Roumier, A. Decambron, et al. MYD88 L265P mutation in Waldenstrom macroglobulinemia. Blood. 2013;121:4504-4511 Crossref
  • [28] C. Jimenez, E. Sebastian, M.C. Chillon, et al. MYD88 L265P is a marker highly characteristic of, but not restricted to, Waldenstrom's macroglobulinemia. Leukemia. 2013;27:1722-1728 Crossref
  • [29] T.L. Naylor, H. Tang, B.A. Ratsch, et al. Protein kinase C inhibitor sotrastaurin selectively inhibits the growth of CD79 mutant diffuse large B-cell lymphomas. Cancer Res. 2011;71:2643-2653 Crossref


a Department of Pathology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea, 135-710

b Samsung Biomedical Research Institute, Sungkyunkwan University School of Medicine, Seoul, Korea, 135-710

c Department of Pathology, Hallym University Sacred Heart Hospital, Anyang, Korea, 431-796

d Samsung Advanced Institute for Health Sciences & Technology, Sungkyunkwan University School of Medicine, Seoul, Korea, 135-710

e Department of Medicine, Division of Hematology and Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea, 135-710

lowast Corresponding author. Department of Pathology, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, Korea.

This study was supported by the Samsung Biomedical Research Institute Grant [SP1B20412].

☆☆ Disclosure: The authors have declared no conflicts of interest.