ABSTRACT: Recently, oncology has moved to a new clinical practice, more personalized, called Predictive Oncology (PO).
PO comes from our knowledge about tumor heterogeneity that implies that each disease, thus each patient, is unique. PO’s goal is to identify and administrate the right treatment to the right patient.
For this, PO requires to go through 3 majors steps:
1. A good characterization of the tumor to identify candidates,
2. A well-established panel of drugs targeting the identified candidates,
3. A relevant model to functionally test these candidates.
The first point could easily be addressed with recent technologies that now allow the Next Generation Sequencing (NGS) and/or the simultaneous analysis of transcriptomic profiles from thousands of patients. The last two points have not been efficiently achieved so far, which prevents PO to be really efficient.
Indeed, even if NGS allows the identification of potential targets, the presence of a molecular candidate does not necessary means obligatory functional response.
The number of drugs approved by the Food and Drug Administration remains limited and most frequent targets in solid tumors (for ex. RAS, P53, MYC, RB1 ...) still do not have specific drugs approved in clinic.
Finally, available pre-clinical models still present many major inconvenient:
* Chimiogrammes on 2D cultures are not sufficiently relevant to be really predictive of the in vivo situation;
* Patient derived xenograft (PDX) are not adapted for clinical use because not all tumors graft and the time to develop a PDX is too long (several months), thus incompatible with the history of the disease (especially for most severe patients). Furthermore the host (NOD-SCID mouse) is immuno-depressed, preventing to objectively test antibodies-mediated drugs.
Recently, the 3D cell culture technology has proven its superiority to predict drug response over classical 2D chimiogrammes. It consists in growing "mini-tissues", or organoid-derived from tumor/healthy tissues, thanks to the amplification of stem cells contained within the sample. The generated organoids are personalized and biologically relevant (organoids are expend form the patient’s stem cells which self-organized according to the architecture of the tissue they are originating from), they are genetically stable, their growth is compatible with patient’s disease history (organoids grow in few weeks), easy and convenient to achieve, even from small biological material quantities (0.5< x < 1cm3), and they can be amplified, frozen and thawed on demand. Moreover, organoids can be made more complex with the addition of other cell types (fibroblasts, immune cells ...). None of the actual available pre-clinical model regroups all these characteristics.
The constitution of a "next generation" biobank of liver samples (Metastases to the liver and Hepato Cellular Adenocarcinoma) will be very useful in the context of predictive oncology.
For this, a biopsy needs to be dissociated and grown in Matrigel, in presence of a well-defined list of growth factors. Once the culture is established, organoids can be frozen then defrost on demand.
Our main objective is to evaluate the feasibility for building a biobank of liver-derived organoids, from liver metastases of colorectal cancers, hepatocellular adenoma and adenocarcinoma (waste tissues).
Applications related to organoids derived from tumors are quasi indefinite, from drug screening assays, tests for novel therapies or original drug combinations, to patients’ stratifications or fundamental research.
In our case, we are interested in building this a biobank in the prospect of using it to build the "next generation of model for predictive oncology" to study liver-related cancers and related drugs testing. Briefly, we want to implement these organoids with cells from the microenvironment in order to makes the global model more pertinent for drug testing.
If successful, the generation of such biobank, including both tumor-derived organoids and healthy counterpart, could be really helpful for the scientific and medical community.
