Researchers embraced the opportunity to discuss the challenges of finding and using preclinical models for immuno-oncology at the 2nd in a series of Cancer Models Community events organised by Repositive. Already a major topic at the Hanson Wade Tumor Models Conference in Boston, participants in the follow-up Model Selection for Immuno-oncology and Combination Therapies discussion on July 18th in Cambridge, MA were eager to delve deeper into the merits and difficulties of available preclinical models for immuno-oncology as well as to theorize on the future.
Limitations of Preclinical Mouse Models
Syngeneic mice are currently the primary workhorses for preclinical immuno-oncology and combination therapy studies because their mostly intact immune systems allow testing of checkpoint inhibitors. In addition, they are relatively less expensive than genetically engineered mouse (GEM) models or humanised mice (Sanmamed et al., 2016). But while the value of preclinical models for immuno-oncology research relies on taking a systems approach, studies involving a number of syngeneic models have resulted in unexplained response variability.
Frustrations were palpable as scientists recounted unexplained differences in results based on handlers or even location of the experiments. For example, strain-matched syngeneic models from different preclinical oncology CROs can show significantly different tumour growth rates. Interestingly, these phenotypic differences might be linked to variability in mice microbiota as confirmed by co-housing mice in the same environment (Sivan et al., 2016).
Researchers also expressed the need for different preclinical mouse model endpoints in immuno-oncology studies – possibly focusing on immune cell infiltration rather than tumour shrinkage. The presence of tumour infiltrating lymphocytes indicate better clinical outcomes for a variety of cancers and a case can be made for the value of this measure (Barnes & Amir, 2017).
Additional tools such as in-life imaging were suggested for monitoring differences between genetically identical mice experiencing inconsistent responses to the same treatment. In-life imaging allows real time visualization of the progression of the disease within the mouse model (Bouvet & Hoffman, 2015). Nevertheless, the practical reality of teasing out every possible influence within the syngeneic model remains time-consuming and expensive.
Promise of 3D Systems
Discussion then turned towards whether to phase out in vivo preclinical models altogether for immuno-oncology, given the promising potential of the new ex vivo models such as 3D spheroids. The increasing complexity of these structures allow manipulation of driver mutations and finite control of conditions. These 3D models also exhibit relevant cell-to-cell interactions and gene expression of human tumours (Nath & Devi, 2016)
However, spheroids can be fragile with short life spans and issues such as heterogeneity of cells within the tumour complicates their development. Consistency in morphology and size has been difficult but progress is being made in overcoming these challenges (Raghavan et al., 2016). Spheroids may also not capture the uniqueness of the tumour microenvironment. Going forward, emerging improvements in microfluidic technologies potentially offer ways to mimic the factors in the microenvironment that influence metastasis (Sleeboom et al., 2018). For instance, at the Hanson Wade conference, one of the 3D models presented simulated the vascularization required for metastasis building on previous research (Jeon et al., 2015).
Whether ex vivo would truly replace in vivo is still open for debate. Some researchers assert that the in vivo or “4-legged” model is an opportunity to understand a systemic approach to cancer rather than trying to recreate it in ex vivo 3D models. There was consensus, however, that all these preclinical models required better characterisation. Even within syngeneic mouse models, research indicates the need for better expression data on immune-related pathways and regarding mutations in immune-cell specific genes (Mosley et al., 2016).
For anyone embarking on designing an experiment, finding the right preclinical model often necessitates a wide-ranging search. For immuno-oncology studies, this inquiry is further complicated by rapidly evolving model systems. Understanding and enabling comparable data across these models moves us another step closer to more efficient and faster biomedical research.
If you are trying to identify a preclinical model for your next immuno-oncology study, explore our world-leading inventory (containing over 5000 models!) and let our Cancer Models Scouts connect you with the right model for your project.
You can also stay up-to-date on this conversation as it develops. Follow us on Twitter, LinkedIn, or sign up to our mailing list for latest the developments. We will be organising a variety of events (online and in-person) throughout the year.