Recurrent mutations and targeted therapy in FL

In this issue of Blood, Krysiak et al expand our understanding of the genetic events which drive follicular lymphoma (FL), and provide insights into potential applications of targeted therapy.¹ 

FL is a pathologically and clinically heterogeneous disease. For the majority of patients, FL presents at advanced stage with multifocal lymph node involvement, and follows an indolent natural history with a favorable prognosis. In the modern era, most patients diagnosed with FL will have a life expectancy similar to age-matched controls, whereas ∼20% will experience a more aggressive course and ultimately die of their disease.²  Unique FL variants have also been identified with distinct clinical presentations, pathologic findings, and natural histories. These include pediatric-type FL and primary intestinal FL, each of which has been found to be biologically distinct from typical nodal FL cases.^3,4  The t(14;18) is present in the majority of FLs and serves as a defining molecular feature. This translocation, which results in constitutive expression of the antiapoptotic BCL2 protein, appears to be an early genetic event that alone is insufficient to result in lymphomagenesis. The development of lymphoma, therefore, is driven by a diverse collection of additional genetic events, the types of which determine the clinical and biologic heterogeneity within this disease.

Krysiak et al performed whole-exome sequencing on a training set of 24 FL biopsy specimens and, based on mutations identified in this cohort, along with those in the existing literature, proceeded to sequence 1716 genes in 113 FL biopsies drawn from 105 individual patients, most of whom were treatment naive. They identified 39 recurrently mutated genes, some of which have been previously identified, whereas others represent novel findings.

Newly identified mutations which warrant further investigation for their role in follicular lymphomagenesis include EGR1/2, POU2AF1, ZNF608, and HVCN1. Among previously identified mutations in FL, the investigators confirmed that histone modifiers are among the most commonly mutated genes, including MLL2, CREBBP, EP300, EZH2, and MEF2B. Additionally, 44% of cases had at least 1 mutation in 1 of 25 histone genes. These speak to a likely critical role for epigenetic regulators in FL pathogenesis. Frequent mutations of histone-modifying genes also suggest promise for epigenetically targeted therapies in FL. To date, the histone deacetylase (HDAC) inhibitors vorinostat and abexinostat have both shown encouraging activity in relapsed/refractory FL.^5,6  Similarly, tazemetostat, an inhibitor of the recurrently mutated EZH2, has shown early evidence of activity in an ongoing phase 1 study.⁷ 

Another interesting finding is that ∼40% of patients had mutations affecting the B-cell receptor (BCR) signaling pathway. These included previously identified recurrent mutations affecting CD79B, CARD11, and CXCR4, as well as BTK mutations in the tyrosine kinase domain. Multiple other critical components of the BCR pathway were also found mutated at lower levels, including CD22, BLNK, PLCγ2, BCL10, and NFκB2. These findings support a role for constitutive BCR signaling as a component of FL pathogenesis in some cases, and also have implications for therapy. The BCR pathway and Bruton tyrosine kinase (BTK) specifically have emerged as critical therapeutic targets in B-cell lymphomas, with US Food and Drug Administration (FDA) approvals for the BTK inhibitor ibrutinib in chronic lymphocytic leukemia (CLL), mantle cell lymphoma (MCL), and Waldenstrom macroglobulinemia (WM). In these diseases, the majority of patients with relapsed or refractory disease respond to ibrutinib, whereas development of resistance has been tied to mutations of BTK and PLCγ2.⁸  Despite evidence of recurrent BCR pathway mutations in a large proportion of patients with FL, the activity of ibrutinib in this disease has been decidedly more modest than in CLL, MCL, or WM, with an overall response rate of only 30%.⁹  Krysiak et al provide potential explanations for this. First, fewer than half of FLs have activating mutations within the BCR pathway, and are therefore likely to be driven by other mutations which confer proliferative and survival advantage that would not be targeted with ibrutinib. Among FL cases that are driven by BCR signaling, multiple mutations have now been identified which could confer ibrutinib resistance, including BTK, PLCγ2, CARD11, and NFκB. A similar phenomenon has been observed in diffuse large B-cell lymphoma (DLBCL) where the activated B-cell (ABC) type of DLBCL relies on chronic active BCR signaling which may render sensitivity to ibrutinib. The response rate within ABC-DLBCL in a phase 2 study was 37%, with CARD11 mutations, among others, being identified as potential mechanisms of resistance.¹⁰ 

These findings highlight that our application of targeted therapies in FL is still in its infancy. Although ibrutinib clearly has meaningful clinical activity in a subset of FL cases, determination of sensitivity cannot be predicted by histology alone. Rather, to select appropriate candidates for this therapy, we will need to identify cases driven by BCR signaling and that lack resistance mutations in BTK, or activating mutations in downstream signaling molecules. Similarly, the abundance of mutations in histone-modifying genes, along with early clinical activity of HDAC and EZH2 inhibitors, warrants further investigation as to which genes are being selectively modified by these agents to confer sensitivity so that optimal candidates for such therapy can be identified a priori. We will also need to identify recurrent genetic events enriched in patients who are at risk of early progression and ultimately dying of their disease, so they can be selected for relevant targeted therapies. As additional novel agents emerge in FL which target epigenetic modifiers, oncogenic signaling pathways, and the immune microenvironment, it will be critically important to understand the genomic substructure of a given patient in order to select among the available options with distinct mechanisms of action. In so doing, we can look forward to entering the era of personalized medicine in FL.