CD28 fusions: an opportunity for young ATL?

In this issue of Blood, Yoshida et al identify concurrent fusions of CD28 with CTLA4 and ICOS in younger Japanese patients with adult T-cell leukemia/lymphoma (ATL).¹ 

With chemotherapy-based treatment approaches, overall survival in ATL has not significantly improved in the nearly 40 years since human T-cell lymphotropic virus and ATL were first described.²  Therefore, the identification of a novel, measurable, and druggable target is exciting.

The median age at presentation with ATL is ∼70 years in Japanese patients, whereas ATL arising in patients living in the United States, Europe, the Caribbean, and South America occurs at age 45 to 55 years.³  It is not understood whether the differences in age at presentation reflect differences in disease biology, underlying host genetics, immune response to the tumor or virus, or environmental factors.

Yoshida et al hypothesize that the tumors of younger patients with ATL would contain distinct genetic alterations, similar to other cancers that present in younger individuals, such as ETV6-RUNX1 fusion seen in childhood acute lymphoblastic leukemia. They identify concurrent CTLA4-CD28 and ICOS-CD28 fusions in 37.5% of ATL cases (3 of 8 cases) in those age <50 years, the presence of which did not affect survival. This is in contrast to earlier reports of peripheral T-cell lymphoma (PTCL) and ATL, where the presence of both fusions was rare.^4,5 

CD28 and ICOS are costimulatory molecules that potentiate T-cell activation on binding their respective ligands CD80/CD86 and ICOSL. In contrast, ligation of the coinhibitory CD28 homolog CTLA4 inhibits T-cell activation through binding ligands of CD28 with higher affinity, sequestering CD80/CD86 and initiating an inhibitory signaling cascade. The CTLA4-CD28 fusion consists of the extracellular and transmembrane domains of CTLA4 and the intracellular signaling domain of CD28, whereas the ICOS-CD28 fusion combines only the signal peptide from ICOS with the extracellular and intracellular domains of CD28. In other PTCL tumors, this ICOS-CD28 fusion was associated with CD28 overexpression and ICOS haploinsufficiency.⁶ 

Yoshida et al demonstrate in vitro that the expression of CTLA4-CD28 and, to a lesser degree, CTLA4-ICOS fusions could induce cellular proliferation when cocultured with cells expressing CD80 and CD86. In ATL cases with both fusions present, immunohistochemistry demonstrated that ATL tumor cells express CD80 and macrophages in the tumor microenvironment express CD86, suggesting that both intra- and intercellular actions could drive cellular proliferation in vivo. The authors report that cases with fusions had gene expression signatures associated with AKT and RAF signaling and, strikingly, LAG3 downregulation, which is known to negatively downregulate T-cell proliferation. Finally, the in vitro CD80/CD86-driven proliferation of cells expressing CTLA4-CD28 fusion could be suppressed with a CTLA4-blocking antibody in a dose-dependent manner.

Chemotherapy alone rarely cures ATL, and there is a desperate need for new and better therapies and for an understanding of how to apply our current therapies more effectively. For example, mogamulizumab is associated with better responses in leukemic rather than nodal disease, even more so in the presence of CCR4-activating mutations.⁷  Similarly, responses to combination therapy with zidovudine and interferon-α are significantly better in leukemic rather than nodal ATL.^8,9  This work suggests that there may be a rationale for targeting cases with CD28 fusions with an anti-CTLA4 antibody and/or targeting its downstream effectors, such as the phosphatidylinositol 3-kinase pathway.

Of course, these observations were made in a small number of cases, and it remains unclear why these dual fusions were not observed in a larger Japanese cohort,^4,10  which also included a small number of cases who presented at age <50 years. It is logical that these fusions should be investigated systematically in cohorts arising in the United States, South America, and Europe, where so-called young ATL is frequently seen. Presumably because of the small number of cases here, the age cutoff at 50 years is arbitrary, but in a larger cohort, perhaps a true biological entity may be defined. Understanding these biological differences and how to best select treatments and apply new therapies will be crucial to improving survival outcomes.