Personalized medicine aims at bringing the right drug to the right patient at the right time. Getting many things right at the same time is challenging. There are many potential treatments for cancer patients and many of these can be combined with each other resulting in a lot of potential combination treatments. This growing number of treatment options makes it increasingly difficult to determine the right one for an individual patient. Biomarkers that help to predict response to a treatment are highly needed to guide treatment selection to maximize efficiency for both patients and payers.
Two recent publications demonstrate that a novel approach to personalized cancer medicine holds a lot of promise: Personalized cancer vaccines (reference 1 & 2 below). A possibly expanding arsenal of treatment options will increase the need for novel diagnostics designed to provide information across different treatment strategies. This is evident when looking at exciting ways how vaccination could complement checkpoint inhibition.
Checkpoint inhibitors (for example anti-PD-1 and anti-PD-L1 therapy) are the frontrunners of immunotherapy and provided a lot of excitement with durable responses in multiple indications. Conceptually, they disrupt mechanisms that allow tumor cells to avoid destruction by activated immune cells. Despite their great potential, not all patients benefit from treatment with checkpoint inhibitors. Personalizing treatment with checkpoint inhibitors requires to identify those patients that are likely to benefit.
Two approaches appear particularly promising. As checkpoint inhibitors are designed to disrupt a mechanism that allows tumor cells to avoid destruction it seems plausible that it will work best in tumors that actually use this mechanism. Consequently, multiple assays to detect PD-L1 expression in tumor cells (and immune cells) have been developed to predict response to checkpoint inhibitors. The second approach is similarly straightforward. Tumor cells only need to avoid destruction by activated immune cells if there are activated immune cells in the first place. Patients with a high density of infiltrating cytotoxic T-cells and high expression of PD-L1 are thus most likely to benefit from treatment with checkpoint inhibitors.
Conversely, patients with a low density of infiltrating cytotoxic T-cells and a low expression of PD-L1 are unlikely to benefit from checkpoint inhibitors. They could benefit from other treatment approaches. Alternatively, other treatment approaches could trigger modifications in the patient’s immune response that improve the prospects of treatment with checkpoint inhibitors. Specifically, vaccination could help to increase the density of infiltrating cytotoxic T-cells.
Personalized cancer vaccines support the patient’s immune system in recognizing tumor cells by drawing attention to deviations from normal cells that result from the unique set of mutations the tumor accumulated. Some of these, the immune system can detect itself but patients can benefit if the immune system gets some extra help. Sequencing the tumor and analyzing the sequence with powerful algorithms can reveal opportunities for tumor cell detection that the immune system missed. As a result, immune cells are activated and recruited to the tumor. In the best of cases, the infiltrating cytotoxic T-cells are able to control tumor growth. The recent reports on personalized cancer vaccines provide examples for patients that experience tumor regression after vaccination. There is also first evidence that patients that relapse after vaccination may benefit from subsequent treatment with checkpoint inhibitors.
All of this is exciting news but many challenges remain. Some of these relate to the concept of personalized cancer vaccines. In addition to the vaccine, many things need to be in place for the immune system to clear the cancer. These include functional MHC restriction, sufficient numbers of macrophages and dendritic cells that successfully present the neo-antigen, and sustained survival of T-cell clones in the body despite potential peripheral selection. Some of the challenges relate to checkpoint inhibition. The idea is to enable activated cytotoxic T cells to kill tumor cells by blocking the interaction between PD1 and PDL1. However, we know that patients that appear to be prime candidates for treatment with checkpoint inhibitors fail to respond. Conversely, we know that some patients do respond despite low levels of PDL1 expression.
One of the most interesting remaining challenges it thus to better understand which patient will benefit best from which treatment approach. Or from which sequence of therapeutic interventions. Novel diagnostics will be crucial to provide that information. It seems likely that the local tumor microenvironment and the spatial distribution of infiltrating immune cells will be essential cues. Eventually, personalized cancer therapy will require both novel therapies and novel diagnostics. Promising results with personalized cancer vaccines raise hopes that we will see an expanding arsenal of therapeutic options. Developing novel diagnostics to select vaccination or checkpoint inhibition or a combination or something else as the most promising treatment strategy will be an important next step.
- Ott, P. A.et al. Nature (2017)
- Sahin, U.et al. Nature (2017)
Author: Dr. Dr. Florian Leiss, Senior Program Leader, Definiens