ALK in NHL: To B (cell) or not to B (cell)? Characterization of the entity “ALK⁺ DLBCL”

Expression of anaplastic lymphoma kinase (ALK) fusions in non-Hodgkin lymphoma (NHL) has been considered by many investigators to occur only in T or null cell CD30⁺ anaplastic large cell lymphomas (CD30⁺ ALCLs).^1-3  Although a significant percentage of diffuse large B-cell lymphomas express CD30 antigen and may show anaplastic morphology, ALK fusions occur rarely, if ever, in these tumors.⁴  An exception to the absence of ALK expression in B-cell NHL was reported in 1997 by Delsol et al⁵  who described a rare form of large B-cell lymphoma composed of monomorphic immunoblast-like cells that superficially resembled ALCL but lacked CD30 expression. These lymphomas expressed epithelial membrane antigen (EMA) like typical CD30⁺ ALCL but contained cytoplasmic immunoglobulin A (IgA) of a single light chain type in many cases. Interestingly, these tumors appeared to express full-length ALK with no evidence of nucleophosmin-ALK (NPM-ALK) expression (analysis for other ALK fusions was not performed as they had not yet been identified). Despite expression of full-length ALK, it was unclear whether the receptor played a causal role in these lymphomas because kinase activation of the protein was not detected.FIG1 

Three reports in this issue of Blood clarify the role of ALK in B-cell NHL. Gascoyne and colleagues (page 2568) demonstrate that the rare subtype of diffuse large B-cell lymphoma reported originally by Delsol and coworkers⁵  in 1997 mainly in adults (median age in the Gascoyne et al report, 48.5 years) likely owes its genesis to expression of the clathrin-ALK (CLTCALK) fusion, previously shown to occur in occasional classical T- or null-cell CD30⁺ ALCLs and the mesenchymal neoplasm known as inflammatory myofibroblastic tumor (IMT). The ALK-positive diffuse large B-cell lymphomas (ALK⁺ DLBCLs) described in the Gascoyne et al article include some of the cases originally reported by Delsol et al⁵  to express full-length ALK and the current study again suggests this to be the case, although the pathogenic importance of the normal receptor still remains undetermined. The article from De Paepe and colleagues (page 2638) independently describes similar cases of CLTCALK-expressing DLBCL; interestingly, unlike the cases reported by Gascoyne et al, the lymphomas in the De Paepe et al report occurred in children, lacked full-length ALK expression, and 1 of 3 cases expressed CD30. Confirmation that ALK⁺ DLBCL is not limited to adults is provided by Onciu and colleagues (page 2642) who describe 2 children with this entity but not with expression of CLTC-ALK but rather NPMALK. The cases reported by Onciu and colleagues were nonetheless quite similar to those in the other 2 studies, showing plasmablastic morphology, IgA restriction, and negative CD30 staining. Both cases in the Onciu report lacked full-length ALK.

Taken together, these articles define ALK⁺ DLBCL as a rare (<1% of DLBCL) CD20^-, CD30^- (usually), CD138⁺, EMA⁺ (usually), and IgA-restricted (frequently) immunoblastic/plasmablastic NHL. This entity occurs in both children and adults, appears to have a male predominance, and has a poor prognosis in many cases. CLTC-ALK fusions seem to characterize the majority of ALK⁺ DLBCLs, but NPM-ALK occurs in some cases. Now defined, ALK⁺ DLBCL will be an important entity for pathologists and oncologists to diagnose and treat; while CD20- and CD30-directed therapies are unlikely to benefit patients with this NHL, anti-ALK targeted therapies currently in development should hold great promise.