Immune-mediated Therapy in Oncology

 Immune-mediated Therapy in Oncology

Immune-mediated Therapy in Oncology

The Marcus Aurelius of Cancer Therapy

The famous words of wisdom by Marcus Aurelius, “The impediment to action advances action. What stands in the way becomes the way,” a great metaphor for how immune-checkpoint inhibitors are becoming the main approach to tackling the current challenges in cancer therapy. Evolving from being a major impediment to the main path forward, immune checkpoints are hopefully a means by which a curative treatment for cancer can be achieved.  At the same time, in this case, not all roads lead to Rome! So the main questions become which marker should be followed to get to your destination; how do you get there faster; how do you follow the path with the least resistance i.e. the lowest toxicity possible?


The immune system has a complex and very selective approach to distinguish between “self” and “foreign”, thereby attacking only the foreign invaders while leaving the normal cells alone. This unique ability to maintain a balancing act is brought about by molecules known as “checkpoints” or “impediments”, which immune cells need to overcome in order to initiate an immune response. Tumor cells exploit these checkpoints to block an attack by immune cells and we now realize that over the past few decades, this cellular masking has been the reason for the failure of immune-mediated therapy in oncology.


In the last few years, therapies have been developed that remove these blockades with initial efforts focused on two main targets CTLA4 and the receptor/ligand pair of PD1/PDL1. These efforts have scored impressive results recently, particularly in patients with non–small cell lung cancer (NSCLC); metastatic melanoma; Hodgkin lymphoma and bladder cancer, and are showing promise in clinical trials involving patients with many other types of cancer. Unfortunately, except in HL, characterised by high and widespread PD-L1 amplification and overexpression, where the response rate is as high as >60%, the ORR in other indications has been more modest around 15-to 30%. Therefore, major efforts have been focussed on selecting the responder patient population to optimise treatment regimens.


The examination of PD-L1 levels was an obvious prognostic biomarker to analyze in such selection processes. However, PD-L1 expression measured by immunohistochemistry in patients is not necessarily a strong prognostic marker in patients. In general, all efficacy parameters for both Nivulomab and Pembrozulimab are higher in PD-L1 positive patients, a substantial % of PDL1 negative patients (10% for Pembrozulimab in NSCLC) are still responders. This has led to the contradictory situation where Pembrozulimab is approved only in PD-L1+ patients while Nivulomab is approved in all patients. Moreover, the criteria used to define the PD-L1 status of a tumor differs vastly: for nivolumab, the cut off used is 1% to 5% of positive cells in a biopsy while pembrolizumab has a much wider range of 1% to 50%, while that for atezolizumab (anti-PD-L1) it is 1% to 10%.  It is not clear what is the correct cut-off and whether measurements in biopsies are truly optimal, as they might represent a snapshot in time, especially given the data that both efficacy can be related to positivity in both the tumor cells as in the immune infiltrate whose status and amount can change with the progression of the disease and the treatment previously received. Further complications are added by the evidence that these markers might have different levels of predictivity in different tumour types. It is understandable, therefore, that the FDA convened a public workshop to discuss these issues and has mandated a comparison of the different PD-L1 diagnostics.


In the near future, the availability of tumor tissue and blood from the many clinical trials that are testing the immune-checkpoint inhibitors coupled with the application of a variety of technologies, like multiplex cytokine profiling, whole exome sequencing for mutational/neoantigens burden determination, gene expression signatures determination, should permit to achieve a more complete understanding of the tumor immunophenotype and identify more comprehensive and selective biomarkers to achieve a better patient selection for single agent therapies.


At the same time, the scientific community has found another way to turn road blockers into therapies. An immune-checkpoint blockage is only one aspect of the interaction between the immune system and tumour cells, which is a multistep cyclic process that requires the coordination of numerous factors, both stimulatory and inhibitory in nature to amplify and broaden immune-cell responses. Inhibitory molecules like IDO, TIM3, VISTA, A2AR or stimulatory ones like OX40, GITR, ICOS, CD27 are expressed or secreted by various stromal cells (list MDSC, Macrophages and Fibroblast) within the immune privileged tumor microenvironment that can affect the activity of immune checkpoint inhibitors The current hypothesis is that combinations of various agents acting at different stages in this cycle will permit to optimize immune-oncology therapies and improve efficacy to a wider population and reduce  resistance, potentially making single agent and related biomarkers less relevant.


This hypothesis has received its first validation by the recent approval of the combination of the two checkpoint inhibitors ipilimumab and nivolumab, which increased the response rate seen with the respective monotherapy. Unfortunately, this combination has the major drawback of toxicity, as many patients experience unusual toxicities related to an excessive immune response leading to pneumonitis, hepatitis, colitis and other immune-related disorders. Many see another limitation that we can call “financial toxicity” given the substantial costs of these treatments, further compounded by the fact that even this combination does not achieve 100% ORR.


For these reasons efforts to identify the right segment of patients that benefit most from immune-therapy combinations have become even more relevant. At the same time complex as each combination target comes with its own set of potentially predictive markers and further markers could be modulated in unexpected ways by synergistic mechanistic interactions.


The predictive algorithms of BioXcel’s PharmGPS® Imuuno-Oncology Platform have been designed to build comprehensive interaction maps amongst two targets that take into consideration their signature in the tumour microenvironment. This interactive map can be used to generate testable hypotheses on which markers to prioritize for analysis and which biomarker analysis to conduct to capture the highest number of relevant parameters.


For instance, PharmGPS® shows that the combination of PD1 and TIM3 will have a strong overlapping effect on Tregs activation in NSCLC, suggesting that monitoring the presence of activated Treg could predict for responses. In the same context, the combination of anti-PD-1 and anti-CTLA -4 will modulate a larger number of pathways, affecting CD8 + proliferation,  CD4+Cd25+FoxP3+ Tregs activation, NK cell signalling and memory T cell expansion. Thus, a multivariate analysis of the immune signature associated with the CD8+ Ki67 index, the numbers of CD4+FoxP3+ T cell vs CD8+ effector memory T cells in the patient tissues samples would constitute a more robust and predictive approach.


Thus, biomarkers derived from analysis of the immune-activation signature stand out as prime parameters, to not only predict the therapy outcome in selected patient populations but also as a strong mechanistic approach to identify rationally designed combinations that can further expand the number of responding patients. BioXcel’s PharmGPS® Immuno-Oncology Platform virtual validation suite is ready to guide the traveller on this journey toward the next generation of immuno-oncology therapies.


Written by Luca Rastelli, Ph.D., VP, Oncology R&D, BioXcel