A, A case harboring MYD88 mutation. The blue line indicates a heterozygous wild-type (WT) allele with lower fluorescence than in the negative witness (purple, green, and pink lines). B, A case harboring the MYD88 L265P mutation (red line). The green line represents the negative witness. C, Serial dilutions of a heterozygous case exhibiting the MYD88 mutation.
The MYD88 mutation was not present in 24 patients and was present in 34 patients. The global difference between the curves was significant (P = .03). WT indicates wild-type.
The MYD88 mutation was not present in 24 patients and was present in 34 patients. The global difference between the curves was significant (P = .002). WT indicates wild-type.
Pham-Ledard A, Beylot-Barry M, Barbe C, Leduc M, Petrella T, Vergier B, Martinez F, Cappellen D, Merlio J, Grange F. High Frequency and Clinical Prognostic Value of MYD88 L265P Mutation in Primary Cutaneous Diffuse Large B-Cell Lymphoma, Leg-Type. JAMA Dermatol. 2014;150(11):1173–1179. doi:10.1001/jamadermatol.2014.821
The activating mutation of MYD88 L265P is a frequent feature of primary cutaneous diffuse large B-cell lymphoma, leg-type (PCLBCL-LT), reported in up to 69% of the cases. Whether patients with MYD88 mutation display specific clinical and evolutive features has not been evaluated.
To identify the clinical characteristics associated with MYD88 mutation, confirm its high prevalence, and evaluate its effect on prognosis in patients with PCLBCL-LT.
Design, Setting, and Participants
A retrospective multicenter study was conducted using the medical records of patients from dermatology departments belonging to the French Study Group for Cutaneous Lymphomas. Sixty-one patients with a diagnosis of PCLBCL-LT made between 1988 and 2010 who were available for molecular study were included. Of these, 58 patients displaying interpretable results constituted the study group. Median follow-up was 33 months, and 39 patients (67%) were monitored until death.
Main Outcomes and Measures
Clinical features (age, sex, number of skin lesions, tumor stage, and location as leg vs elsewhere), MYD88 mutation (allele-specific TaqMan polymerase chain reaction assay), treatment regimen, and outcome were recorded. Baseline characteristics and outcome were compared according to the status of MYD88.
The median age of the patients was 79 years, and 59% were female. Skin lesions were located on the leg in 76% of the cases. Thirty-four of 58 patients (59%) harbored the MYD88 L265P mutation. Patients had similar clinical characteristics at presentation regardless of their MYD88 status, except that those harboring the MYD88 mutation were older (P = .006) and had more frequent involvement of the leg (P = .008). Patients harboring the MYD88 mutation had 3- and 5-year–specific survival rates of 65.7% and 60.2% vs 85.4% and 71.7% in patients with the wild-type allele. The MYD88 mutation was significantly associated with shorter disease-specific survival in univariate (P = .03) and multivariate (odds ratio, 3.01; 95% CI, 1.03-8.78; P = .04) analysis. There was no significant difference between the groups in their treatment regimens. Considering overall survival, in univariate (P = .002) and multivariate (odds ratio, 2.94; 95% CI, 1.18-7.30; P = .02) analysis, MYD88 L265P mutation was an independent adverse prognostic factor.
Conclusions and Relevance
This study confirms the high prevalence of MYD88 L265P mutation in PCLBCL-LT and shows its association with shorter survival. The clinical effect of MYD88 mutation activating the nuclear factor-κB pathway supports the use of targeted therapies at the time of relapse after conventional therapies.
Primary cutaneous diffuse large B-cell lymphoma, leg-type (PCLBCL-LT) has been individualized in the World Health Organization1 classification as an aggressive disease preferentially involving the legs, with rapidly growing tumors in the elderly. Understanding the oncogenesis of diffuse large B-cell lymphoma (DLBCL) allows distinction of subgroups reflecting the cell-of-origin derivation at discrete stages of differentiation, namely activated B-cell–like (ABC) and germinal-center B-cell–like. Next-generation sequencing studies2- 4 identified recurring and multiple genetic alterations implicated in the oncogenesis of nodal ABC-DLBCL, contributing to the nuclear factor-κB pathway constitutive activity. Notably, mutations affecting the B-cell receptor signaling pathway (CD79B [GenBank 974])5 and other proteins affecting the nuclear factor-κB signaling (most commonly, mutations of MYD88 [GenBank 4615], CARD11 [GenBank 84433], and TNFAIP3/A20 [GenBank 7128]) have been reported.6- 8 In PCLBCL-LT, gene expression profile analysis, cytogenetic and genetic mutation studies, and protein expression profile with high interferon regulatory factor 4 protein and melanoma-associated antigen expression bring PCLBCL-LT closer to the nodal ABC-DLBCL than to the germinal-center B-cell–like type.9- 11
A recent study reported12 that among cutaneous B-cell lymphomas, PCLBCL-LT specifically exhibits a single driver mutation of MYD88 (c.794T>C causing L265P-leucin to become prolin) at a very high rate (69%). The protein MYD88, a Toll-like receptor–associated adaptor protein, is implicated in the signaling pathway during normal immune innate response. Furthermore, in a preliminary series, 18 patients11 with PCLBCL-LT harboring the MYD88 L265P mutation had a tendency to shorter specific survival; however, the finding did not reach statistical significance (P = .10). Another group10 confirmed the presence of MYD88 mutations in a series of 10 cases of PCLBCL-LT, although at a lower prevalence (40%) than in the preliminary series.11 These data prompted us to identify the MYD88 mutation in a large multicenter series with available clinical and follow-up data, establish its real prevalence, and determine whether PCLBCL-LT harboring the MYD88 L265P mutation displays specific clinical and evolutive features.
According to the French Public Health and Bioethical Law, our study was considered by the research direction of Bordeaux University Hospital as noninterventional without need for ethics committee approval (Article L1121-1 and Article R1121-3). Moreover, both data and samples were collected and then used anonymously through the tumor bank of CHU de Bordeaux with declaration of the sample collection and associated clinical data to the French Research and Education Ministry and its bioethical unit (DC-2011-1293). Cases were retrieved retrospectively from records of the French Study Group for Cutaneous Lymphomas, from January 1, 1988, to December 31, 2010. All patients had been included in a French multicenter study.13 The study had been performed after patients had been given information and provided written informed consent, according to the Declaration of Helsinki protocols. Inclusion criteria were (1) a clinical and histologic diagnosis of PCLBCL-LT according to the World Health Organization–European Organization for Research and Treatment of Cancer classification14 assessed by an expert panel of physicians and pathologists, (2) available clinical and follow-up data, (3) absence of extracutaneous localization at diagnosis, and (4) available skin tumor samples for DNA analysis (formalin-fixed, paraffin-embedded skin biopsies).
Extraction of DNA from skin biopsies and Sanger sequencing of MYD88 gene was conducted. The process has been previously described.12
Two primers were used to amplify a 132-base pair fragment located in exon 5 of the MYD88 locus: forward: 5′-GTCCCACCATGGGGCAAGG-3′ (Sigma-Aldrich Co LLC) and reverse: 5′-GTGATGAACCTCAGGATGCTG-3′. Allele-specific polymerase chain reaction (PCR) was performed to detect either the wild-type (WT) allele or the MYD88 mutated allele using two different probes designed with a centrally single base substitution corresponding to WT-MYD88 and L265P-MYD88: MYD88WT: [FAM]CAGAAGCGACTGATCCCCA[TAM] and MYD88L265P: [6FAM]AGAAGCGACCGATCCCCAT[TAM] (Sigma-Aldrich Co LLC). Real-time PCR was performed (Roche Light Cycler 480; Roche Diagnostics). The cycles were conducted as follows: denaturing for 5 minutes at 95°C, followed by 45 cycles of 95°C for 10 seconds, then 64°C for 15 seconds, then 72°C for 25 seconds, and then cooling for 30 seconds at 40°C. A patient harboring the MYD88 L265P mutation as determined by Sanger sequencing was used as a positive control, and healthy donors were used as negative controls. The sensitivity of PCR was assessed by PCR analysis of samples obtained by serial dilutions of tumoral DNA carrying the MYD88 L265P mutation (Figure 1).
Clinical and follow-up data were collected from a French multicenter clinical study.13 Data are described using median and range for quantitative variables and number and percentage for qualitative variables. Comparison between subgroups of patients according to MYD88 status was performed using a χ2 test or Fisher exact test, as appropriate. Disease-specific survival was calculated from diagnosis to the date of disease-related death or censoring. Patients whose deaths were unrelated to lymphoma were considered censored. Observed or overall survival duration was calculated from diagnosis to the date of death. The survival curves were established using the Kaplan-Meier method. Prognostic factors were identified by univariate analysis using the log-rank test and by multivariate analysis using a Cox proportional hazards model. Factors significant at the P = .20 level in univariate analysis were included in a stepwise regression multivariate analysis with entry and removal limits set at P = .20. Multivariate analysis was systematically adjusted for age. Statistical analysis was performed using SAS, version 9.3 (SAS Institute Inc).
Sixty-one patients with PCLBCL-LT from 12 French centers met the inclusion criteria. Three cases were excluded because the DNA was not amplified for sequencing or by TaqMan PCR assay. Therefore, 58 patients constituted the study group. The main clinical characteristics, treatment regimens, and follow-up of these 58 patients according to their MYD88 status are reported in the Table. The female to male ratio in the study group was 1.42. Age ranged from 50 to 96 years (mean, 76; median, 79 years). Skin lesions were localized on the leg in 44 cases (76%). The distribution of the TNM disease stage at diagnosis was category T1, 34%; T2, 45%; and T3, 21%.
The status of MYD88 was first determined by Sanger sequencing using DNA extracted from paraffin-embedded material as in a previous study,12 but only 32 of 61 cases displayed interpretable results, which is most probably attributable to preanalytical factors (eg, use of various fixation procedures) and heterogeneity of storage duration and conditions between centers and samples (data not shown). Allele-specific TaqMan PCR permitted analysis of 58 of 61 available cases (95%). The MYD88 L265P mutation was found in 34 of 58 (59%) PCLBCL-LT samples. Comparison of serial dilutions of a WT and heterozygous mutated case with the standard curve showed that the cycle threshold of the mutated and WT alleles was detected at a similar level for mutated cases, supporting the hypothesis that the mutation was always heterozygous (Figure 1).
The clinical features of patients exhibiting the MYD88 mutation compared with WT cases are summarized in the Table. Patients with MYD88 L265P mutation were older (P = .006) and more frequently had skin tumors located on the leg (P = .008), whereas no significant difference was observed between the 2 groups for sex, number of skin lesions, TNM stage at diagnosis, and treatment regimens (rituximab plus polychemotherapy with or without anthracyclines vs other treatment regimens).
The evolution and final status of MYD88 in the entire series are presented in the Table. Complete response was obtained in 42 cases (72%), without a significant difference between patients harboring the MYD88 mutation and those without the mutation. Twenty-four of 42 patients (57%) had cutaneous relapses and 21 of 58 individuals (36%) experienced extracutaneous spreading. Median follow-up was 33 months, and 39 patients (67%) were monitored until death. Nineteen of 58 patients (33%) died of lymphoma, whereas 10 (17%) died from unrelated diseases. The 3- and 5-year overall survival rates of the entire study group were 66.0% and 51.8%, respectively.
Patients harboring the MYD88 L265P mutation had 3- and 5-year–specific survival rates of 65.7% and 60.2% vs 85.4% and 71.7% in patients with WT lymphoma, respectively (P = .03) (Table and Figure 2). In univariate analysis, disease-related death was significantly associated with MYD88 mutation (P = .03) (Figure 2), location on the leg (P = .02), and multiple skin lesions (corresponding to T2 to T3 categories vs a single lesion corresponding to T1 category) (P = .02), whereas age and sex did not reach significance. Multivariate analysis systematically adjusted for age identified MYD88 mutation (odds ratio [OR], 3.01; 95% CI, 1.03-8.78; P = .04) and multiple skin lesions (OR, 5.12; 95% CI, 1.12-23.26; P = .03) as adverse prognostic factors, whereas location on the leg (P = .25) and age (P = .13) did not reach significance.
Patients harboring the MYD88 L265P mutation had 3- and 5-year overall survival rates of 55.8% and 38.4%, vs 81.3% and 62.6% in patients with WT lymphoma, respectively (Table and Figure 3). In univariate analysis, death was significantly associated with MYD88 mutation (P = .002), location on the leg (P = .006), and multiple skin lesions (corresponding to T2-T3 categories vs a single lesion corresponding to T1 category) (P = .008), whereas age and sex did not reach significance. In multivariate analysis systematically adjusted for age, MYD88 mutation (OR, 2.94; 95% CI, 1.18-7.30; P = .02) and multiple skin lesions (OR, 2.94; 95% CI, 1.03-9.26; P = .04) were identified as independent prognostic factors, whereas location on the leg (P = .23) and age (P = .79) did not reach significance.
The present study shows that the MYD88 L265P oncogenic mutation indicates a poorer outcome in a specific group of patients with PCLBCL-LT and confirms its high prevalence (59%). To our knowledge, this is the largest study to date evaluating MYD88 status in PCLBCL-LT.
The rate of MYD88 mutation in PCLBCL-LT may have been decreased by DNA degradation in specimens from different sources with various fixation procedures and sample storage conditions. In view of their clinical effect, the results of the present study required the development of an allele-specific TaqMan PCR assay targeting the L265P mutation, because Sanger sequencing was contributive in only 52% of the specimens. Use of an allele-specific TaqMan PCR resulted in efficient analysis of 58 of 61 samples (95%); moreover, the mutation was detected by the TaqMan assay in 4 patients interpreted as harboring the WT allele in Sanger sequencing.
Compared with the levels of MYD88 mutation in PCLBCL-LT, lower rates have been reported8,15 in nodal ABC-DLBCL (17%-29%). In PCLBCL-LT, Koens et al10 reported a lower rate (40%) in a series of 10 patients using Sanger sequencing. Our data are also in line with those obtained in Waldenström macroglobulinemia showing higher rates of detection when using a real-time PCR-based technique such as allele-specific PCR or high-resolution melting analysis for detection of the MYD88 L265P mutation.16,17 The increased rate of the MYD88 mutation in PCLBCL-LT is similar to rates observed in immune-privileged site-associated DLBCL, such as in the primary central nervous system and testis, where MYD88 mutation has been reported15 with high frequencies of 75% and 71%, respectively. We can presume that PCLBCL-LT and central nervous system and testis DLBCL, ABC-type, may represent a distinct group of lymphomas driven by similar or equivalent oncogenic hits. Frequent involvement of the testis and central nervous system in PCLBCL-LT has been described.18,19
To our knowledge, the present study is the first to demonstrate an association between MYD88 mutation, older age, and location on the leg. Moreover, we were able to demonstrate a prognostic effect of MYD88 mutation with a 3-year survival rate of 65.7% in patients with MYD88 mutated tumors compared with 85.4% in those with WT tumors (P = .02) (Figure 3). Multivariate analysis considering either specific survival or overall survival showed that MYD88 L265P mutation was an independent prognostic factor (P = .04 and P = .02, respectively), with systematic adjustment for age.
Nevertheless, elderly patients harboring MYD88 mutation have a poorer outcome and may benefit from personalized medicine, especially targeted therapies, at the time of relapse after conventional therapies. Mutation of MYD88 predicts a poorer outcome considering MYD88 signaling and nuclear factor-κB pathway activation as actionable targets. Therapeutic strategies targeting either B-cell receptor signaling (as ibrutinib, the Bruton tyrosine kinase inhibitor, or fostamatinib targeting the spleen tyrosine kinase), or the interferon regulatory factor-4 and the transcription factor Spi-B (targeted by lenalidomide),20 or NF-κB proteasome inhibitor (bortezomib), or MYD88 inhibitors,21 seem promising in ABC-DLBCL and especially in PCLBCL-LT. Because PCLBCL-LT is a rare disease, clinical trials including nodal and extranodal ABC-DLBCL cases are needed to evaluate responses rates to such targeted therapies according to putative predictive mutations, especially those of MYD88 and other genes encoding for members of the B-cell receptor signaling pathway, such as CARD11 or CD79B.5 Indeed, such ABC lymphomas usually carry several genetic alterations shown at nodal2,8 as well as extranodal sites for PCLBCL-LT10,11 or for central nervous system lymphoma.15
A prognostic effect of the MYD88 L265P mutation identified in PCLBCL-LT on overall survival and a high rate (59%) of this mutation were observed in the present study. These findings support the hypothesis that this actionable oncogenetic event has a major role among the genetic alterations of PCLBCL-LT.
Accepted for Publication: April 8, 2014.
Corresponding Author: Jean-Philippe Merlio, MD, PhD, Equipe d’accueil, University Bordeaux, bâtiment 3B, 2ème étage, Zone Nord, 146 rue Léo Saignat, 33076 Bordeaux, France (email@example.com).
Published Online: July 23, 2014. doi:10.1001/jamadermatol.2014.821.
Author Contributions: Drs Pham-Ledard and Grange had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Pham-Ledard, Beylot-Barry, Merlio, Grange.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Pham-Ledard, Cappellen, Merlio, Grange.
Critical revision of the manuscript for important intellectual content: Pham-Ledard, Beylot-Barry, Barbe, Leduc, Petrella, Vergier, Martinez, Merlio, Grange.
Statistical analysis: Barbe, Leduc.
Obtained funding: Beylot-Barry.
Administrative, technical, or material support: Beylot-Barry, Martinez, Cappellen.
Study supervision: Beylot-Barry, Vergier, Merlio, Grange.
Conflict of Interest Disclosures: None reported.
Funding/Support: This work is part of the REV-LEG Programme Hospitalier de la Recherche Clinique grant to Centre Hospitalier Universitaire Bordeaux number 2011-28, and received funding from Celgene. This study was designed based on a clinical study conducted on this series13 funded by the Société Française de Dermatologie. This work was conducted as part of the SIRIC BRIO program (Site de Recherche Intégrée sur le Cancer-Bordeaux Recherche Intégrée Oncologie), grant INCa-DGOS-Inserm 6046.
Role of the Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: The following physicians and pathologists of the French Study Group for Cutaneous Lymphomas provided data or samples: Pascal Joly, MD, PhD, and Philippe Courville, MD, PhD (Departments of Dermatology and Pathology, Hôpital Charles Nicolle, Centre Hospitalier Universitaire, Rouen, France); Martine Bagot, MD, PhD, and Caroline Ram-Wolff, MD, PhD (Departments of Dermatology and Pathology, Hôpital Saint-Louis, Paris, France); Stéphane Dalle, MD, PhD, and Brigitte Balme, MD, PhD (Departments of Dermatology and Pathology, Hôpital de l’Hôtel-Dieu, Lyon, France); Eve Maubec, MD, PhD (Department of Dermatology, Hôptial Bichat, Paris); Michel D’Incan, MD, PhD, and Pierre Dechelotte, MD, PhD (Departments of Dermatology and Pathology, Hotel-Dieu, Clermont-Ferrand, France); Sophie Dalac, MD, PhD (Department of Dermatology, Hôptital du Bocage, Dijon, France); Gaëlle Quéreux, MD, PhD (Department of Dermatology, Hôpital de l’Hôtel-Dieu, Nantes, France); and Olivier Dereure, MD, PhD, and Eric Frouin, MD, PhD (Department of Dermatology, Hôptial Aviecenne, Bobigny, France). Martina Carlotti, PhD (Equipe d’accueil 2406, Histology and Molecular Pathology of Tumors, Universitaire Bordeaux, Bordeaux, France), Nathalie Carrere, PhD (Tumor Bank and Tumor Biology Laboratory, Centre Hospitalier Universitaire Bordeaux), and Christine Alfaro, MS (Department of Dermatology, Centre Hospitalier Universitaire Bordeaux), provided advice and help with acquiring and analyzing the data. No financial compensation was given. Most samples were collected through the Tumor Bank of Centre Hospitalier Universitaire de Bordeaux.