A BRIEF REVIEW OF DRUG DEVELOPMENT TARGETING THE LTβ/LTβR PATHWAY
Companies mentioned: Biogen, Avalo Therapeutics, Pfizer, Roche, Mestag Therapeutics, Shanghai Junshi Bioscience
Attempts to modulate the immune activity of the lymphotoxin beta pathway have been long running and thus far unsuccessful. This is perhaps not so surprising, as both the pathways and the biologies regulated by the interaction of LTbeta (LTβ) with the LTbeta Receptor (LTβR) are complex. Many reviews are available (eg. Browning 2008. https://pubmed.ncbi.nlm.nih.gov/18613838/) that highlight these key features:
Control of secondary lymphoid organogenesis during fetal and post-natal development (lymph nodes (LN) and Peyer’s patches)
Control of the development of specific lymphocytes (innate lymphoid cells, NK cells, NKT cells, B cells)
Control of secondary lymphoid organ cellular architecture in adults (spleen, LN, PP)
Induction of cytokines critical for the induction of tertiary lymphoid structures (TLS) at sites of chronic inflammation in adults
These 2 last biologies – the control of local immune organs – stimulated research directed to antagonism of LTβ/LTβR interaction in the hope this would reduce chronic inflammation in such diseases as immune thrombocytopenic purpura, Sjogren’s Syndrome, lupus, rheumatoid arthritis (RA) and inflammatory bowel disease (IBD), all of which can show TLS development in diseased organs. Indeed, LTβR antagonism was robustly efficacious in many mouse models of chronic inflammatory and autoimmune diseases and pathway antagonists for clinical use were developed by Biogen and others. The Biogen asset, baminercept, contained the extracellular domain of LTβR fused to an IgG1Fc domain and was investigated in human clinical trials of RA and SS, where it failed to provide clinical benefit. A study in secondary progressive multiple sclerosis was withdrawn.
This brings us to the all-too-common issue of mouse pathophysiology failing to translate to human disease. This was somewhat of a surprise as the closely related TNF/TNFR pathway readily translated from mouse models of disease into clinical use for the treatment of RA, IBD, psoriasis and other immune-related diseases by TNF & TNFR-targeting drugs such as Remicade, Enbrel, Humira and many others. Why were antagonists of the LTβ/LTβR pathway different?
There are several issues. One is the complexity of the pathway, which contains multiple ligands and receptors in addition to LTβ and LTβR. Once all the players were revealed we found a more complex picture than we had first contemplated:
Image from De Trez 2012. https://doi.org/10.1016/j.imbio.2012.06.010
One can immediately see a complication: numerous ligands and receptors are expressed and they interact. Further, these ligands and receptors are found on a wide variety of cell types. This is the type of system we often encounter in biology, and this nearly always indicates that there are complex feedback loops between the functional pairs. That is, LTβ and the LTβR do not exist in a biological bubble, there are additional ligands (LIGHT, BTLA, CD160) and receptors (HVEM, Dcr3). Downstream of the receptors is a jumble of signaling pathways, notably the classical and alternative NF-κB pathways that control a wide array of cellular responses, including cell proliferation and cell survival. The NF-κB signaling cascades are highly regulated, and when any one component is shut down (or overstimulated) these cascades are disrupted. To add to the confusion, LTβR and the TNFRs are activating receptors; HVEM is an inhibitory receptor. BTLA and CD160 are inhibitory ligands while LIGHT can be an activating or inhibitory ligand depending on which receptor it engages. To oversimplify: LIGHT delivers pro-inflammatory signals through LTβR, and LIGHT, BTLA, and CD160 deliver inhibitory signals through HVEM. Several authors have even proposed that LIGHT functions as an “immune checkpoint” by binding to either LTβR or HVEM (I think this hypothesis is a bit forced). Finally, LIGHT can be blocked by the soluble decoy receptor DcR3 – the importance of this interaction is not at all clear.
As a consequence of this profusion of ligands and receptors we can now overlay a bewildering number of functions in addition to the developmental and immune homeostatic roles played by LTβR. One can also see cross-talk with the TNF receptors, although that indicates a different issue which is the incomplete match between the mouse and human pathways. In mouse, one of the lymphotoxin subunits (alpha) can trimerize and bind to TNFR1 and TNFR2. This is not true of human cells (T cells and B cells) that express only heterodimers of the lymphotoxin subunits alpha and beta. How these complexities relate to the issue of poor translation from mouse models to human disease is unclear.
I’ll note here that there are other examples of failure to translate across mammalian protein families, eg. the TIGIT cluster of ligands and receptors has not yielded robust clinical translation despite abundant mouse tumor model data, and the chemokines and chemokine receptor families whose ligands and receptors are shared promiscuously have likewise shown limited clinical translation for both immune and oncology indications.
Antibodies to LTβR, LIGHT, BTLA and HVEM have all been evaluated as therapeutics in immune diseases and in oncology (note that this table may be incomplete):
Pfizer’s program was PF-07329640, a tetravalent agonist anti-LTβR antibody being developed to trigger lymphoid cell organization (ie. TLS) in the setting of solid tumors. This program was highly touted by Pfizer during their Oncology Innovation Day in February 2024, but a clinical trial was abruptly stopped after only 4 patients were dosed. This was almost certainly due to unexpected toxicity. Receptors like LTβR can generate non-physiologic responses if they are clustered on the cell surface unnaturally, and I’m going to guess that is what happened when the tetravalent antibody was used. Typical toxicity seen using agonist therapeutics of this receptor family is liver toxicity, which has notable derailed programs targeting related receptors TRAILR2, 4-1BB and OX40.
The therapeutic hypothesis here is solid, given that there is abundant evidence showing that the presence of TLS in tumors is strongly predictive of improved patient outcomes and better responses to immune-directed therapies such as the anti-PD-1 and anti-PD-L1 classes on antibodies (eg. Opdivo, Keytruda, Tecentriq and many others). The therapeutic potential of inducing TLS in the solid tumor setting would apply to many different cancer indications. This underscores the potential importance of anti-LTβR agonist antibody programs for companies like Pfizer and Roche that have developed anti-PD-1/PD-L1 therapeutics.
We also see toxicity emerging with the use of anti-BTLA antibody tifcemalimab plus anti-PD-1 (the Junshi program). One might imagine that the combination released an aggressive anti-tumor immune response, although details are scarce. A study of tifcemalimab plus anti-PD-1 plus standard of care chemotherapy in extensive stage small cell lung cancer patients revealed a Grade 3+ toxicity rate of 59% but also an overall response rate of 86.5%. This is a program we’ll want to watch.
Anaptys Bio’s antibody ANB032 is designed to signal through BTLA to dampen immune responses, rather than blocking it’s activity as tifcemalimab does. The company has indicated Phase 2 data in atopic dermatitis will be available in December 2024.
One last program to highlight is still preclinical, but very interesting. MST-0300 is a FAP-LTβR bispecific antibody, designed to activate LTβR only when the bispecific antibody is bound to the protein FAP that is highly expressed by cancer associated fibroblasts. The idea is to limit LTβR signaling to the tumor microenvironment, thereby triggering TLS formation and inducing better anti-tumor responses. MST-0300 is being developed by Mestag Therapeutics.
Finally, to end where we began, a recent study examined three patients from two families with functional hyposplenism (greatly reduced spleen), low circulating IgG, absence of tonsils, and complete lymph node aplasia (Ransmayr et al. 2024. 10.1126/sciimmunol.adq8796) . The patients suffered from recurrent bacterial and viral infections. Phenotypic analyses showed that the lymphoid organs were grossly disorganized and genetic analyses revealed biallelic loss-of-function mutations in LTβR. These mutations and their consequences highlight the critical role that LTβR plays in coordinating our immune environments, as in spleen and lymph nodes, thereby supporting robust immune function. We’ll look forward to more clinical data emerging in this space.
Stay tuned.