MYD88 in the driver’s seat of B-cell lymphomagenesis: from molecular mechanisms to clinical implications

Abstract

More than 50 subtypes of B-cell non-Hodgkin lymphoma (B-NHL) are recognized in the most recent World Health Organization classification of 2016. The current treatment paradigm, however, is largely based on ‘one-size-fits-all’ immune-chemotherapy. Unfortunately, this therapeutic strategy is inadequate for a significant number of patients. As such, there is an indisputable need for novel, preferably targeted, therapies based on a biologically driven classification and risk stratification. Sequencing studies identified mutations in the MYD88 gene as an important oncogenic driver in B-cell lymphomas. MYD88 mutations constitutively activate NF-κB and its associated signaling pathways, thereby promoting B-cell proliferation and survival. High frequencies of the hotspot MYD88(L265P) mutation are observed in extranodal diffuse large B-cell lymphoma and Waldenström macroglobulinemia, thereby demonstrating this mutation’s potential as a disease marker. In addition, the presence of mutant MYD88 predicts survival outcome in B-NHL subtypes and it provides a therapeutic target. Early clinical trials targeting MYD88 have shown encouraging results in relapsed/refractory B-NHL. Patients with these disorders can benefit from analysis for the MYD88 hotspot mutation in liquid biopsies, as a minimally invasive method to demonstrate treatment response or resistance. Given these clear clinical implications and the crucial role of MYD88 in lymphomagenesis, we expect that analysis of this gene will increasingly be used in routine clinical practice, not only as a diagnostic classifier, but also as a prognostic and therapeutic biomarker directing precision medicine. This review focuses on the pivotal mechanistic role of mutated MYD88 and its clinical implications in B-NHL.

Introduction

With the introduction of high-throughput, next-generation sequencing, many studies have aimed to explain the diverse biology, clinical course, prognosis, and therapeutic response of B-cell non-Hodgkin lymphoma (B-NHL). This has increased our knowledge of lymphomagenesis by identifying many novel somatic alterations that affect signaling pathways involved in several B-NHL subtypes. In this rapidly evolving molecular landscape, it is important to translate newly obtained genetic knowledge directly into clinical benefit for patients.1

Ngo et al. were the first to identify an oncogenic, non-synonymous, gain-of-function mutation in myeloid differentiation primary response 88 (MYD88), leading to an amino-acid change of leucine to proline at position 265 (NM_002468.5, also referred to as position 273 in NM_001172567) of MYD88 [MYD88(L265P)].2 Other recurrent mutations in MYD88 were likewise identified; however, the impact of these mutations has been difficult to establish due to their low prevalence.3 This review, therefore, focuses on the present understanding of the role of MYD88(L265P) in NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) activation and its association with the B-cell receptor (BCR) cascade. In addition, we address the clinical importance of MYD88(L265P), including its prevalence across B-NHL subtypes, its predictive significance in patients’ outcome, and its potential as a therapeutic target.

Oncogenic mechanisms of MYD88(L265P)

Canonical NF-κB signaling

In normal physiology, MYD88 acts as a signaling adaptor in the canonical NF-κB pathway (Figure 1). This pathway is activated upon recognition of pathogen-associated molecular patterns (PAMP) by receptors containing a toll/interleukin-1 receptor (TIR) domain, such as toll-like receptors (TLR) and the interleukin receptors 1 (IL-1R) and 18 (IL-18R). After ligand binding, the TIR domain of these receptors interacts with the TIR domain of MYD884 and this process initiates the formation of the so-called ‘myddosome complex’. For this complex, activated MYD88 recruits IL-1R associated kinase 4 (IRAK4), a serine-threonine kinase, and together they phosphorylate IRAK1 or IRAK2. Phosphorylated IRAK1 and IRAK2 interact with tumor necrosis factor receptor-associated factor 6 (TRAF6), resulting in activation of transforming growth factor beta-activated kinase 1 (TAK1).5 Activated TAK1 continues signaling through the mitogen-activated protein kinase (MAPK) signaling cascade and cooperates with TAK1-binding protein (TAB) to activate the inhibitor of the NF-κB kinase (IKK) complex.

Figure 1.The role of MYD88 signaling in normal physiology and lymphomagenesis. Recognition of pathogens by TLR, IL1R, and IL-18R induces an immune response through activation of MYD88 and generates the myddosome complex with IRAK4 and IRAK1 or IRAK2, which is stabilized by HSP110. IRAK1 and IRAK2 activate the MAPK and NF-κB pathways through TRAF6 and TAK1, causing proliferation and survival of B cells. MYD88(L265P) allows for increased formation of the myddosome complex, preferentially with IRAK1, and constitutively activates the NF-κB pathway. In addition, the formation of the My-T-BCR supercomplex leads to increased activation of mTOR and the CBM complex, promoting lymphomagenesis. Lastly, constitutively active NF-κB increases autocrine signaling of IL-6 and IL-10, which further promote B-cell proliferation and survival via the alternative JAK/STAT signaling cascade.

The IKK complex consists of the kinase subunits IKKa and IKKβ and the regulatory subunit NF-κB essential modulator. After activation, this complex phosphorylates the inhibitor of NF-κB (IκB) proteins that are bound to NF-κB, which prevent migration of NF-κB to the nucleus. Phosphorylation of these IκB proteins results in ubiquitylation and proteasomal degradation of IκB and release of the NF-κB subunits. Subsequently, the NF-κB subunits, including RELA (p65)-p50 in the classical pathway and RELB-p52 in the alternative pathway, migrate to the nucleus where they bind to specific DNA-binding sites and induce increased expression of genes involved in B-cell proliferation and survival. In addition, expression of these genes is increased through interactions between the NF-κB subunits and other transcription factors, such as E1A binding protein P300 (EP300) and CREB binding protein (CREBBP).6

In the case of MYD88(L265P), the TIR domain of MYD88, in which L265P resides, is more highly activated compared with wildtype MYD88 and this increases downstream signaling and formation of the myddosome complex.2 Henceforth, MYD88(L265P) preferentially and constitutively recruits IRAK1 for the myddosome and, together with IRAK4, was found to be essential for survival of activated B-cell (ABC) diffuse large B-cell lymphoma (DLBCL) cell lines with MYD88(L265P).872 In addition, IRAK1 was shown to be co-immunoprecipitated with MYD88 in chronic lymphocytic leukemia (CLL) cells with MYD88(L265P) and stimulation of IL-1R and TLR induced a 5-fold to 150-fold increase of cytokine secretion compared to that of CLL cells with wildtype MYD88.9 However, Ansell et al.7 identified that in Waldenström macroglobulinemia (WM) cell lines, the myddosome complex consisted of IRAK4, TRAF6, and MYD88, but not IRAK1. The authors hypothesized that this difference in complex formation was instigated by the heterozygous nature of MYD88(L265P) in WM and the homozygous nature in DLBCL, which was strengthened by the finding that downstream signaling of TAK1 phosphorylation was highest in the DLBCL cell line with homozygous MYD88(L265P).7 Furthermore, the stabilizing effect of heat shock protein 110 (HSP110) on the myddosome complex, due to interference with the proteasomal degradation of MYD88, is stronger in ABC-DLBCL cell lines with MYD88(L265P) than in those with wild-type MYD88.10 As MYD88(L265P) constitutively activates the NF-κB pathway, it is regarded as an important oncogenic driver in B-NHL.1272

B-cell receptor signaling

In addition to the canonical NF-κB pathway, the BCR pathway plays an important role in B-cell survival and proliferation and oncogenesis of B-NHL with MYD88 mutations (Figure 1). In normal physiology, stimulation of the BCR activates NF-κB, as well as the phosphoinositide 3-kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR), and nuclear factor of activated T cells (NFAT) pathways. After antigen recognition by the BCR, Lck/Yes-related novel protein tyrosine kinase (LYN) is released from its inactive state through dephosphorylation of the C-terminal regulatory tyrosine by cluster of differentiation 45 (CD45) or an exogenous ligand for the Src-homology 2 (SH2) and SH3 domains of LYN, such as CD19. Activated LYN consecutively phosphorylates the immunoreceptor tyrosine-based activation motif (ITAM) domains of the coupled CD79A and CD79B heterodimers. These double-phosphorylated ITAM domains provide a docking site for the SH2 domains of spleen tyrosine kinase (SYK), which is activated by autophosphorylation or through transphosphorylation by LYN. LYN and SYK then activate Bruton tyrosine kinase (BTK) by phosphorylation, which is recruited to the membrane through interaction between the pleckstrin homology (PH) domain of BTK and phosphatidylinositol-3, 4, 5-triphosphate (PIP3) of the PI3K pathway or through interaction between the SH2 domain of BTK with the B-cell linker protein (BLNK) adapter molecule that also recruits phospholipase Cg2 (PLCg2) to the membrane.13 BTK activates PLCg2, initiating activation of the NF-κB pathway through formation of CBM complex, consisting of caspase recruitment domain family member 11 (CARD11), BCL10, and mucosa-associated lymphoid tissue lymphoma translocation protein 1 (MALT1). In addition, BTK activates the MAPK and PI3K pathways14 and PLCg2 triggers the NFAT pathway through calcineurin. The CBM complex subsequently attracts TRAF6, TAK1, and TAB, and promotes the degradation of IκB, which leads to the release of NF-κB subunits.151454

BTK is an integral protein in the BCR signaling cascade and has been found to be preferentially complexed to MYD88 in WM cells with MYD88(L265P) and not in MYD88 wildtype cells. Inhibition of BTK resulted in a decrease of the formation of this MYD88-BTK complex, but lacked effect on IRAK4/IRAK1 activity and vice versa, indicating a potential necessity of dual inhibition of IRAK and BTK for WM with MYD88(L265P).1816 MYD88 is frequently mutated in patients who also harbor a mutation in the 196 tyrosine residue in the ITAM domain of CD79B (NM_000626) and these patients seem to benefit most from BTK-inhibition treatment.19 The exact consequence of these double mutations in B-NHL is unclear, but Phelan et al.8 recently provided new insight into the mechanism of combined MYD88 and BCR-pathway activation as they identified a MYD88-TLR9-BCR (My-T-BCR) supercomplex. This supercomplex is generated by constitutive trafficking of the BCR towards endolysosomes that contain TLR9 and interacts with mTOR and the CBM complex, thereby promoting lymphomagenesis by activation of the mTOR and NF-κB pathways. Its presence was demonstrated in cell lines and biopsies of ABC-DLBCL, primary DLBCL of the central nervous system, and lymphoplasmacytic lymphoma and correlated with responsiveness to BTK inhibition. On the other hand, the supercomplex was not identified in CLL or mantle cell lymphoma, suggesting a different mechanism of BCR signaling in these entities. Therefore, the My-T-BCR supercomplex could potentially be used as a biomarker for predicting the efficacy of BTK inhibitors, as a classifier of B-NHL subtypes, or as a novel therapeutic target via inhibition of TLR9.8

Autocrine signaling

As described, increased formation of the myddosome complex with IRAK1, as well as activation of the BCR pathway, caused by interactions of BTK with MYD88(L265P), CD79B mutations, and the My-T-BCR supercomplex, result in constitutive activation of the NF-κB pathway. NF-κB not only activates the transcription of genes involved in cell survival and proliferation, but also results in autocrine signaling with IL-6 and IL-10. One consequence of this autocrine signaling loop is the phosphorylation of Janus kinase 1 (JAK1) and, subsequently, signal transducer and activator of transcription 3 (STAT3) with the assembly of a STAT3/STAT3 complex. This complex increases transcription of genes involved in several signaling cascades, including the PI3K/AKT/mTOR, E2F/G2M cell-cycle checkpoint, JAK/STAT, and NF-κB pathways. In addition, STAT3 activity represses the proapoptotic type I interferon (IFN) signaling pathway by downregulating IFN-regulatory factor 7 (IRF7), IRF9, STAT1, and STAT2 expression.2032

Another consequence of IL-6 signaling is the aberrant expression of hematopoietic cell kinase (HCK), as identified in primary WM cells and B-NHL cell lines.21 Increased levels of HCK promote lymphomagenesis, as HCK knockdown in B-NHL cell lines reduces survival and lowers the activity of the BCR, PI3K/AKT, and MAPK/ERK (extracellular signal-regulated kinases) pathways. Furthermore, BTK- and HCK-inhibition treatment of ABC-DLBCL and WM cells with MYD88(L265P) decreased HCK expression, whereas mutant HCK(T333M) (NM_002110.4) attenuated this effect. These findings suggest that HCK is downstream of MYD88(L265P) and that HCK should be regarded as a potential therapeutic target in B-NHL with MYD88(L265P).

Prevalence

The described oncogenic mechanisms largely depend on the prevalence of MYD88(L265P) in B-NHL. Several studies, using Sanger sequencing, allele-specific polymerase chain reaction (PCR) analysis, or (targeted) next-generation sequencing, have demonstrated that the occurrence of MYD88(L265P) varies highly among the different subtypes of B-NHL (Table 1).108221832 The highest prevalence of MYD88(L265P) is found in lymphoplasmacytic lymphoma/WM, with approximately 85% of the patients being affected.372218 In DLBCL, the prevalence of MYD88(L265P) is highest (range, 44% to 73%) in extran-odal DLBCL, in immune-privileged sites,96 such as primary DLBCL of the central nervous system968886232218 and primary testicular lymphoma,108962322 primary cutaneous DLBCL, leg type,91897122 orbital/vitreoretinal DLBCL,989722 intravascular large B-cell lymphoma,95 and primary breast DLBCL.9922 The high prevalence of MYD88(L265P) in extranodal site-specific lymphomas, lymphoplasmacytic lymphoma, and WM may provide an indication for the origin of these lymphomas. Interestingly, B-NHL entities with a high prevalence of MYD88(L265P) are characterized by a monoclonal immunoglobulin M. Furthermore, the high occurrence of MYD88(L265P) in extranodal DLBCL may imply that B cells need to gain this mutation for survival and manifestation in extranodal sites.

Table 1.(A, B) Overview of reported frequencies of MYD88(L265P) in B-cell neoplasms according to the 2016 World Health Organization classification of lymphoid neoplasms¹¹⁰ (A) and other mature B-cell neoplasms with specific disease locations (B).

In DLBCL in general, a recent meta-analysis by Lee et al.,22 comprising 18 studies with a total of 2002 DLBCL patients, documented that 255 of 1236 (21%) cases of ABC-DLBCL harbored MYD88(L265P), compared with 44 of 766 (6%) cases of germinal center B-cell-like (GCB) DLBCL. Large sequencing studies, such as those by Reddy et al.,80 Schmitz et al.,81 Chapuy et al.,77 and Intlekofer et al.,79 have compared with 44 of 766 (6%) cases of GCB DLBCL with archaic cell-of-origin classification, based on immunohistochemistry or gene expression profiling, and have shown that MYD88(L265P) and other mutations transcend these classifications and should be put into context with emerging genomic classification systems. These large sequencing studies underscore the need to evaluate the status of not only MYD88, but also other genes involved in B-cell lymphomagenesis for diagnosis and during treatment with targeted therapies, as proposed by Sujobert et al.109

Overall, these results identify MYD88(L265P) as a diagnostic classifier for specific B-NHL subtypes. This is supported by a recent study by our group that identified MYD88 mutations as an independent marker, in a cohort of 250 patients with DLBCL, in addition to the routinely used MYC and BCL2 and/or BCL6 rearrangements and Epstein-Barr virus status (according to the 2016 World Health Organization classification110).83 Furthermore, MYD88(L265P) is absent in primary mediastinal large B-cell lymphoma9432 and primary cutaneous follicle center lymphoma,7371 and rarely present in hairy cell leukemia (1.1%),59573022 plasma cell myeloma (1.5%),10710643232218 Burkitt lymphoma (1.5%),742 follicular lymphoma (1.9%),6867232218 and CLL (2.5%).523828242218

Prognostic impact

In addition to its role as a diagnostic classifier, the prognostic value of MYD88(L265P) has been a topic of many studies involving B-NHL patients. Lee et al. performed a meta-analysis of three studies with accurate multivariate hazard ratios to investigate the prognostic value of MYD88(L265P) in DLBCL.22 This analysis, involving a total of 275 DLBCL patients, showed that DLBCL patients with MYD88(L265P) had a statistically significant inferior overall survival compared with DLBCL patients with wildtype MYD88. In addition, MYD88(L265P) was significantly associated with older age, high International Prognostic Index (IPI)-score risk groups, and extranodal localization. We also demonstrated this association of MYD88(L265P) with an inferior survival in our recent study in which we evaluated MYD88 status, together with CD79B, MYC, BCL2, BCL6 and Epstein-Barr virus status and clinical characteristics in 250 DLBCL patients.83 Additionally, we showed that the performance of the IPI score is improved by adding MYD88(L265P) as a poor risk factor.

The correlation of MYD88 mutations with an inferior overall survival is also observed in several subtypes of extranodal DLBCL, such as primary cutaneous DLBCL, leg type111 and immune-privileged DLBCL.1128322 On the other hand, in a study by Xu et al.,84 MYD88 mutations were significantly more frequent in DLBCL patients who were refractory to chemotherapy with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone) (28%) compared with DLBCL patients who were chemosensitive (15%), but no statistically significant correlation with overall survival was found. The actual prognostic value of MYD88 in DLBCL requires further investigation, as other studies identified no effect of MYD88(L265P) on the survival of DLBCL patients.11411222

In other subtypes of B-NHL, such as CLL, splenic marginal zone lymphoma, and WM, MYD88(L265P9 is associated with a superior survival compared with wildtype MYD88.11611545 In WM, approximately 30-40% of patients present with concomitantly mutated MYD88 and CXCR4, a gene involved in homing of B cells in the bone marrow, and these patients present with a greater disease burden and reduced progression-free and overall survival.118117 With regards to CLL, Improgo et al.39 stated that MYD88(L265P) occurs mainly in patients with mutated IGHV or chromosome 13q deletions and both alterations are associated with a superior survival. Furthermore, WM patients with wildtype MYD88 had an increased risk of disease transformation, ibrutinib resistance and shorter overall survival.1181179

Table 2.Overview of several (ongoing) clinical trials with novel therapies targeting BTK, PI3K, mTOR and XPO1 in B-cell non-Hodgkin lymphomas in which MYD88(L265P) is frequent.

Targeted therapies

The oncogenic activity of MYD88(L265P), as well as its high frequency in several B-NHL subtypes, ensure that MYD88 and its affiliated signaling pathways are very interesting for targeted therapeutic strategies. As reviewed by Yu et al.18 and Weber et al.,119 several targets are conceivable for direct or indirect inhibition, such as IRAK1 and IRAK4 in the myddosome-complex, TAK1 in downstream signaling, BTK in the BCR pathway, TLR9 in the My-T-BCR supercomplex, and components of the concurrently activated PI3K/AKT/mTOR and HCK pathways (Figure 2).

Figure 2.Signaling cascades in mutated MYD88 B-cell non-Hodgkin lymphoma can be inhibited by several targeted therapeutic strategies. A combination of several therapies might increase efficacy and reduce the risk of relapse, depending on the molecular profile of the B-cell non-Hodgkin lymphoma.

Of these targets, inhibition of BTK has been the most extensively studied, regardless of the fact that BTK is not a MYD88(L265P)-specific target and is not directly involved with the myddosome complex. The BTK inhibitor ibrutinib is approved as treatment for CLL, mantle cell lymphoma, relapsed/refractory marginal zone lymphoma, and WM by the United States Food and Drug Administration (FDA). Additionally, the FDA permitted the combined use of ibrutinib and rituximab as the first non-chemotherapeutic regimen for WM patients. In early clinical trials in patients with relapsed/refractory DLBCL and primary DLBCL of the central nervous system, ibrutinib elicited an overall response rate of 80-85% in those with MYD88(L265P) alone or in combination with mutated CD79B.12019 Furthermore, in a randomized phase III trial, ibrutinib plus R-CHOP improved the overall survival of DLBCL patients younger than 60 years regardless of the cell-of-origin.121 Nonetheless, ibrutinib tends to produce many off-target effects and acquisition of resistance to the drug is common. For instance, ibrutinib resistance can be caused by the C481S mutation in BTK (NM_000061), which hampers the interaction between ibrutinib and BTK,122 but also by mutations in PLCg2,123 CARD11,120 and CXCR4.124 Given these drawbacks of ibrutinib, next-generation BTK inhibitors, such as acalabrutinib and zanubrutinib, are being developed and used for research. Studies demonstrated that acalabrutinib achieved an overall response rate of 95% in relapsed CLL125 and 81% in relapsed mantle cell lymphoma,126 and this medicine is now approved as treatment for mantle cell lymphoma by the United States FDA. Zanubrutinib achieved an overall response rate of 90% in WM, and was also shown to be well tolerated and to overcome the ibrutinib resistance mechanism induced by CXCR4 mutations.127

In addition to studies on BTK inhibition, several phase I/II clinical trials have investigated the response of novel therapeutic targets (in)directly involved with MYD88 in patients with B-NHL. In relapsed/refractory WM, mTOR inhibition with everolimus produced an overall response rate of 50%.128 In several subtypes of relapsed/refractory B-NHL, PI3K inhibition with parsaclisib produced overall response rates ranging between 20% and 78%, with a low risk of adverse events and improved long-term outcomes.129 In in vitro assays, enzastaurin, a protein kinase C inhibitor, reduced proliferation and viability of DLBCL cells by regulation of the PI3K, MAPK, and JAK/STAT pathways; however, it also increased phosphorylation of BTK, suggesting the need for simultaneous treatment of enzastaurin with BTK inhibition.130 Patients with DLBCL are currently being recruited into a phase III clinical trial in which enzastaurin is combined with R-CHOP (NCT03263026).

The clinical trials mentioned above focus on therapeutic targets that are directly or indirectly involved with MYD88 activity; however these targets are not specific for MYD88(L265P) and patients are selected irrespective of the mutational status of MYD88. The lack of biomarkers in these clinical trials is a potential weakness, especially regarding the evolving field of genetics and precision medicine. Novel drugs targeting the oncogenic mechanisms of MYD88(L265P), such as inhibition of the interaction between TLR9 and MYD88 in the My-T-BCR supercomplex8 and between MYD88 and IRAK4 in the myddo-some,131 or direct inhibition of IRAK43911 or TAK1,7 would be interesting for B-NHL patients with MYD88(L265P) and have shown promising results in in vitro and in vivo studies. In addition, the use of immunomodulatory oligonucleotides (IMO) such as IMO 8400, an antagonist of TLR7, TLR8, and TLR9, could be an interesting targeted treatment for MYD88(L265P) B-NHL and especially for ABC-DLBCL with the My-T-BCR supercomplex. IMO-8400 has mainly been investigated in immune-mediated inflammatory diseases and only two phase I/II clinical trials with MYD88(L265P)-positive DLBCL and WM have been performed, showing that IMO-8400 is well tolerated in these patients (NCT02252146, NCT02363439, https://www.ider-apharma.com/wp-content/uploads/2015/12/IMO-8400-WM-ASH-Poster.pdf). More research is required on the MYD88(L265P)-specificity of the above-mentioned targets in order to determine their role in the treatment of B-NHL patients with MYD88(L265P) and, thereby, improve personalized medicine.

An alternative therapeutic approach for these patients, as reviewed by Weber et al.,119 is the induction of a T-cell mediated immune response towards tumor-specific neoepitopes that are derived from MYD88(L265P). In in vitro experiments, such neoepitopes, presented by major histocompatibility class I molecules, prompted a cytotoxic CD8 T-cell response.132 These tumor-specific T cells can be harvested from healthy donors133 or patients with B-NHL and primed to elicit a tumor-specific cytotoxic effect or theoretically used as a model for chimeric antigen receptor (CAR) T-cell therapy. Furthermore, in vitro assays of DLBCL showed that MYD88(L265P) tumor cells develop resistance against T-cell mediated cytotoxicity via upregulation of IL-10 and STAT3 and that inhibition of either IL-10 or STAT3 significantly attenuates this gain of resistance.134 To our knowledge, currently no clinical trials are underway to investigate this intriguing treatment concept.

Liquid biopsy

Until now, comprehensive genomic analysis for accurate diagnosis and classification of B-NHL has been based on DNA isolated from lymphoma tissues. For most patients, the collection of this tissue is a highly invasive procedure with the risk of severe complications.135 An alternative and less invasive method of sampling is the so-called ‘liquid biopsy’, using blood plasma or cerebrospinal fluid, instead of lymphoma tissue. These fluids contain circulating tumor DNA (ctDNA) that is secreted or released during apoptosis or necrosis of the tumor cells, and may harbor somatic mutations, such as MYD88(L265P). Besides being a less invasive method of sampling, ctDNA allows detection of spatial differences between lymphoma cells spread throughout the body, which is not possible with tissue biopsies.

The high frequency of MYD88(L265P) in several B-NHL subtypes make this mutation perfectly appropriate for screening by ctDNA, as already demonstrated in DLBCL,136 primary DLBCL of the central nervous system,137 and intravascular large B-cell lymphoma.95 With the highly sensitive and specific method of digital droplet PCR (ddPCR), even low amounts of ctDNA can be detected, potentially providing information about minimal residual disease, clonal evolution over time, and spatial differences between the lymphoma cells. As demonstrated in patients with DLBCL and WM, ddPCR analysis of liquid biopsies can aid in monitoring the disease course, because of the highly sensitive identification and quantification of the variant allele frequency of MYD88(L265P).13831

An alternative technique for ctDNA analysis is targeted next-generation sequencing. The benefit of this technique over ddPCR is the possibility of identifying multiple variants at the same time, as was shown by Bohers et al.139 and Kurtz et al.140 in liquid biopsies from 30 and 217 DLBCL patients, respectively. The mutational burden of most of their patients, with a median of 117 variants per patient, was sufficient for disease monitoring. This novel way of disease monitoring could enhance evaluation of treatment responses (Figure 3). In their studies, the tumor burden, as measured by positron emission tomography-computed tomography scans, was significantly correlated with the variant allele frequency of ctDNA both during and after treatment.140139 Given this recent progress in ctDNA analysis, liquid biopsies are a minimally invasive method for evaluation of the molecular profile and can be used for analysis of tumor burden, disease progression, and treatment efficacy in patients with B-NHL.

Figure 3.Schematic representation of the potential use of liquid biopsies during disease progression in B-cell non-Hodgkin lymphoma. After diagnosis, a hypothetical patient was treated with immune-chemotherapy. During therapy, the lymphoma was significantly reduced, as evidenced by a complete metabolic remission on positron emission tomography-computed tomography (PET/CT) scans and minimal residual disease by the analysis of circulating tumor DNA (ctDNA) mutation frequency. Thereafter, the residual B-cell lymphoma developed again, gradually increased, and induced a significant clinical relapse. Following comprehensive (genetic) diagnostic procedures, including histological confirmation, liquid biopsies, and PET/CT scans, the patient was treated with BTK inhibition as a second-line therapy, consequently reducing the lymphoma and leading to a partial metabolic remission. Lastly, residual lymphoma cells harboring a BTK(C481S) mutation gained resistance to the BTK inhibition therapy; these cells expanded unimpeded and resulted in another clinical relapse. In this schematic representation, the mutation frequency throughout the course of the patient’s disease is plotted. The two detection limits indicate the sensitivity of PET/CT and the liquid biopsy (e.g., ctDNA with digital droplet polymerase chain reaction analysis).

Conclusions and future perspectives

Routine diagnostics in B-NHL are moving forward from classical morphology and immunohistochemistry towards the implementation of genetic analysis. In several subtypes of B-NHL subtype, MYD88(L265P) plays a crucial role as a driver of lymphomagenesis and can be used as a diagnostic classifier, as well as a prognostic factor and predictive biomarker. B-NHL with MYD88(L265P) can be (in)directly targeted by several novel therapeutic strategies and prospective clinical trials investigating these strategies are ongoing. We expect that that these theranostic strategies will be guided by analysis of MYD88(L265P) in liquid biopsies to monitor disease progression and determine response to therapy. Altogether, given the significant clinical relevance of MYD88(L265P), we advocate evaluation of MYD88 mutational status in routine diagnostics of B-NHL.

Footnotes

• Check the online version for the most updated information on this article, online supplements, and information on authorship & disclosures: www.haematologica.org/content/104/12/2337

• Received May 17, 2019.
• Accepted September 19, 2019.