Therapeutic targeting of refractory/relapsed diffuse large B-cell lymphoma through systems biology

More than 13,000 people are diagnosed with non-Hodgkin lymphoma, a cancer of immune cells, annually in the UK. The most common non-Hodgkin lymphoma is Diffuse Large B-cell lymphoma (DLBCL). A large proportion (~40%) of patients with DLBCL are not cured, resulting in relapsed/refractory DLBCL (RR-DLBCL), and for these patients the outlook is dismal. There has been remarkable recent progress revealing differences in genetics and molecular biology of healthy B-cells compared to DLBCL cancer cells. We are learning more and more about the way genetic changes in B-cells can lead to the processes controlling cell survival, cell division and cell differentiation, going wrong in DLBCL. Despite this progress, the standard treatment for DLBCL has remained unchanged for more than a decade. Standing in the way of progress towards new treatments is the remarkable differences between one DLBCL case and the next. Drugs that are targeted to molecular-scale interactions in RR-DLBCL often only work in subsets patients. The challenge faced, is understanding the molecular makeup of each person's RR-DLBCL, and using this to target a drug the "Achilles heel" of that patient's cancer.
Systems biology simulations are equations representing how the molecules in a cell change over time. It is possible to simulate how many individual B-cells respond to treatment by solving these equations. I have recently used this approach to discover new molecular interactions and show that cell fates are remarkably predictable. Excitingly, I found that I could use these simulations to predict promising and unexpected ways to control cells, that were confirmed when tested in the lab. I also found that when I simulate molecular-scale differences found in DLBCL, the simulations recreate the uncontrolled cell divisions and cell survival of cancer cells. My simulations are strikingly similar to two common subtypes of DLBCL seen in patients.
In this project, my group will use these simulations of B-cells, and add genetic mutations found in DLBCL patients, to create virtual laboratories of different types of DLBCL responding to therapy. In our virtual labs we will show how different mutations on the genetic scale can cause molecular-scale differences leading to cell-scale changes that eventually change how well treatments works.
The beauty of investigating RR-DLBCL with virtual labs is that we can do things that would be difficult or impossible in traditional laboratories. Here, we will simulate targeting drugs to each and every important molecular process in RR-DLBCL cells, to predict the most effective drug targets. We will use these virtual experiments to predict effective drug targets, and then validate the best approaches in the traditional lab. We will use simulations to identify "biomarkers", a property of cancer cells that we can measure, to predict the most effective drug for each patient. We will then test these biomarkers and treatments using traditional laboratory techniques such as measuring proteins and drugging cells that have different genetic mutations.
Another impossible experiment, that can be done with our systems biology simulations, is tracking concentrations of all important molecules inside cells simultaneously all the way from diagnosis to end of treatment. We will use this to "rewind the clock" in our simulations and predict which DLBCL cells a patient has when they're diagnosed become treatment resistant RR-DLBCL cells when the patient is treated. Then, by virtually drugging every target we will find the best way to either: kill these cells before standard treatment is given, or make these cells sensitive to the standard treatment.
This work enables us to develop new approaches to treating DLBCL. We will show that we can measure biomarkers in patients and put these measurements into simulations to create personalised "virtual patients", which we can then use to help decide the best treatment for each person.

Diffuse Large B-cell Lymphoma (DLBCL) is the most common non-Hodgkin lymphoma and current treatment protocols lead to treatment refractory/relapsed disease (RR-DLBCL) in around 40% of patients. For these patients the outlook is dismal with secondary therapies only achieving median overall survival of ~6 months. Here I describe a novel interdisciplinary approach to gaining a mechanistic understanding of RR-DLBCL, rational therapeutic design and development of a personalised medicine approach to the disease. The goal of this project is to advance the scientific field by improving mechanistic insight into DLBCL, improve patient outcomes through identification of promising therapeutic techniques and contribute to the development of computationally-driven personalised medicine for substantial economic, commercial and societal impact within the timeline of this project and beyond.
The most immediate impact will be in the field of lymphoma research where we will provide a greater understanding of the mechanisms by which RR-DLBCL develops and link molecular understanding with lymphoma treatment/progression. In the longer-term, the systems biology research community broadly share a common goal: to construct virtual cells, organs and organisms to enable virtual experiments with clear substantial impact. The work here modelling B-cell fates in disease contributes to this and is likely to be incorporated in comprehensive models of whole organs and organisms as progress is made. Additionally, the successful application of computationally-generated insights to unmet needs in human health in this project will represent an exemplar project, motivating researchers, funding agencies and policy makers to accelerating progress towards the goal of virtual cells, organs and organisms.
Pharmaceutical companies will benefit from utilising systems approaches to accelerate drug discovery, stratify patients to identify sub-population in whom their drugs are likely to be efficacious, and reducing the cost of drug development by prioritising targets computationally predicted to be safe and effective (see support letter from NuCana PLC).
This work enables a personalised medicine approach. I aim to change the standard of care using computational models, personalised to patients as tools to facilitate therapeutic decision making in clinical practice. This approach has the potential for substantial impact beyond DLBCL, as it can be applied to any disease for which there is sufficient mechanistic understanding. The impact of this approach would be significant for health and wellbeing of patients receiving improved therapy, the economic competitiveness of the UK as it will attract substantial investment, and the efficacy of the NHS as it targets drugs to patients who will benefit. I expect initial clinical trials of personalised, model-assisted therapeutic decision making to be initialised on completion of this fellowship. Public outreach will also help improve understanding of systems medicine in the wider public and healthcare communities.
The impact of this fellowship on my career trajectory towards becoming a leader will be substantial. I will gain substantial expertise and experience through the delivery of this project, particularly in the areas of leading an interdisciplinary scientific team. These skills will complement my upward career trajectory to enable me to maximise my impact as I establish myself as a research and innovation leader.
The two postdoctoral research staff working on this project will both receive substantial training in computational biology, cancer biology, iterative experimental and systems studies, molecular and cellular biology, along with more general skills such as scientific communication, presentation skills, interdisciplinary collaboration and project management. This will position them well for high impact employment in the academic or commercial sectors.