PDCellX Models – Beviglia Cancer Models.com

PDCellXs for better prediction of clinical efficacy:

Preclinical tumor modeling plays a key role in the discovery and development of therapeutics; one of the reasons for cancer drug failure in the clinic is testing efficacy in preclinical models that do not recapitulate the most important aspects of human cancer.

by Lucia Beviglia, Ph.D.

One of the limiting factors for lack of clinical cancer efficacy is evaluating anti-tumor activity of drug candidates in poorly predictive preclinical models. For decades drug discovery programs have relied on results from in vivo xenografts that do not recapitulate the heterogeneous complexity of human cancer. These xenografts are derived from cultures of human tumor cells that had undergone multiple in vitro passages and clonal selection, becoming highly homogeneous and poorly tumorigenic. Very often remarkable preclinical efficacy results have not been translated to phase II clinical trials, which led several investigators to turn to potentially more predictive tumor models, i.e. Patients-Derived Xenografts (PDXs). These models are minimally passaged in vivo only and have been largely used in most laboratories.

For decades, drug discovery programs have relied on results from in vivo xenografts that do not recapitulate the heterogenous complexity of human cancer. These xenografts are derived from cultures of human cancer cells that had undergone multiple in vitro passages becoming highly homogeneous and poorly tumorigenic.

However, PDXs initiated with tumor biopsy fragments have shown many limitations, e.g. poor engraftment, slow growth, and variability. These fragments are spatially segregated subclones varying in composition of tumor tissue, stroma, and vascular components. Therefore not all fragments are able to form tumors and the tumors that grow do not have a similar tissue composition causing intra-group variability in the response to treatment. Long latency time and slow proliferation of PDXs explain delayed study timelines and high costs. Moreover, these PDXs do not address the phenotypical diversity of human cancer cells present in the primary tumor.
In contrast the more innovative PDCellXs, initiated with single cells dissociated from a surgical biopsy instead of fragments, are able to address the intratumor cellular heterogeneity. The PDCellXs present an opportunity for analysis of the different tumor cell populations, i.e. those drug resistant, tumorigenic, and/or more invasive, as well as analysis of specific markers and targets enabling a correct assessment of the drug mechanism of action. In addition, dissociated tumor cells can recapitulate the metastatic human disease when implanted at the orthotopic sites and can be characterized ex-vivo in downstream functional analyses to follow-up in vivo efficacy results. PDCellXs enable optimized engraftment, better tumor take rate and growth, and consequently short study time-frame to minimize costs and meet project deadlines. In conclusion, several known limitations described for PDXs can be overcome with PDCellXs.initiated

With the emerging preclinical investigations of cancer immunotherapies there is a need to analyze the immune cell profile associated with tumors. Although most of the immuno-oncology studies are performed in syngeneic models for the presence of immune system, they do not capture the true human immune component complexity and, therefore, are not ideal. The humanized models created by introducing the human immune system in immunocompromised mice have proven to be potentially more clinically predictive than others available for testing new immunotherapies. Ideally the immune system engrafted in mice should derive from the same patient from which the tumor biopsy is excised. This approach presents some difficulties, however it is not impossible. While creation of those models is awaited, it is still imperative to select the most suitable models among those available to increase the probability of clinical success. While fully murine syngeneic models might be the best available approach for first screening of immunotherapies, e.g. checkpoint inhibitors, humanized systems created with both human tumor and immune components are necessary to validate the proof-of-concept data before landing to clinical testing. If we consider other immuno-oncology therapeutics being developed, e.g. the adoptive immunity with CAR-T cells, modified human T cells and human tumor cells are required to be implanted in the immunocompromised NSG mice. In this case, PDCellXs remain the best valuable models to use for validating efficacy. There is also a number of in vitro cell functional assays, including the 3D systems, that can incorporate the dissociated tumor cells derived from patient surgical biopsies and human immune cells. The ideal 3D system must recapitulate the multicellular microenvironment present in the in vivo models and study the interactions of tumor cells with stromal and immune cells. Importantly, the tumor cells derived from PDCellXs retain the original phenotype acquired in the patient or animal model enabling a better clinical prediction.