Organoids of epithelial ovarian cancer as an emerging preclinical in vitro tool: a review

Using organoids, it would be theoretically feasible to introduce a nearly infinite amount of mutations in the genome itself, allowing for new genetic discoveries and insights [13, 24, 25]. Using the CRISPR/Cas9 technique, the genome of organoids can be relatively easily manipulated with high precision by introducing a double stand break at any location in the genome and insert or delete nucleotides or entire gene sequences [24, 25]. By doing so, gene expression, epigenetic changes or RNA behavior can be altered [25]. This technique has already been used in organoids derived from adult stem cells and its iPSC counterparts.

Monogenetic diseases can be modified, with potential clinical use [25, 26]. Mutations can be introduced or altered, modeling the tumorigenesis and potentially identifying and modifying driver genes in cancer organoids [13, 25]. Reporter cell lines (such as fluorescent proteins) could be introduced, allowing for quick identification of optimal conditions of organoids or as markers for genetic alterations [24, 25]. Lastly, CRISPR/Cas9 could be used in genome-wide screening, allowing to replicate the in vivo genetic changes [25].

Organoids provide an insight into embryonal and stem cell development and thus enabling detailed monitoring of each developmental step [18]. A major advantage of these human organoids over animal models is the ability to faithfully mimic human development. A striking example are organoids mimicking human brains, providing a unique insight into human neural growth [18]. Organoids could also be a good model to study the development of cancer. In a bottom-up fashion, the organoids are artificially (by CRISPR/Cas9) modified by a gain-of-function or a loss-of-function mutation, and thus inducing malignancy [27]. The major advantage of this bottom-up approach, over the top-down approach of traditional cell lines, is reducing unnecessary genetic background information. Furthermore, multistep carcinogenesis can be precisely studied using organoids [27].

Organoids could be used in order to test or screen drugs and their toxicities in certain pathologies or tissues [24]. The ultimate goals of these organoids would be to provide a preclinical platform for personalized medicine in oncological patients by correlating drug response to genomic, transcriptomic and proteomic alterations. This would reduce the burden of therapy on the patient and could potentially reduce cancer treatment costs [22]. Organoids could also be used for drug testing, streamlining efforts of drug companies for finding new cancer therapies [13, 28]. Furthermore, large biobanks of organoids would allow already developed drugs to be tested ex vivo, providing insight into possible resistance mechanisms [13]. Developing biobanks of organoids could allow for high throughput drug, predicting patient response before administering the therapy [12]. This has already been observed in a proof-of-concept study [29]. By using these biobanks, the ideal patient population for a particular drug could also be selected prior to the clinical trial [22]. By combining drug screening, cancer gene sequencing and biobanking, organoids could prove to be the key to true personalized cancer medicine [12, 13]. However, not all human drugs can be tested on organoids, since these lack angiogenic factors, immune cells or complete niche paracrine signaling [18]. Although, these limitations are being actively researched, different tumor-derived organoids from various organs and for multiple applications are already been established, as summarized in Table 3.

CRISPR/Cas9 could be used to correct germline mutations in patients, such as in cystic fibrosis [18, 25]. The mutated CFTR gene, involved in cystic fibrosis, has already been successfully corrected in patient-derived intestinal organoids [24]. Although controversial, potentially, these organoids could be transplanted into the affected individuals in order to correct the mutation in vivo [13, 18, 24]. Organoids could also be used for transplantation and tissue regeneration purposes [18, 29]. Using iPSC, it could be possible to generate healthy HLA-matched tissues.

Matrigel (Corning & BD Biosciences) is a substitute for an extracellular matrix, derived from mouse sarcoma cells. This is a stiff extracellular matrix, impairing drug penetration to the target tissues and thus potentially reducing the response of the cells compared to an in vivo setting [18, 22]. The non-human nature makes it unsuitable for human subjects due to potential immune reactions or other potential side effects, causing a major limitation of implementing organoid transplantation into humans [18]. Already some new synthetic hydrogel matrices have been developed, however, these are not yet used for multiple tissue types [20, 29].

Rising numbers of tumors are being treated with neoadjuvant treatment, resulting in decreased tumor cells at the time of surgery. This reduced number of tumor cells could, as in other tumor models, impair the yield of tumor-derived organoids [12].

One of the major limitations of current organoid technology is the lack of microenvironment. In vivo microenvironment consists of fluctuating gradients of growth factors and extracellular signals, immune cells, blood vessels, inflammatory responses, stromal niches, etc.

In cancer research, the microenvironment plays an essential role. For example, anti-angiogenic drugs, immunotherapy or other stromal-affecting drugs, cannot be tested in tumor-derived organoids [12]. As such, discrepancies of drug sensitivity in organoids compared to in vivo environments exist [22]. Additionally, stromal and micro-environmental factors can also alter the drug response of tumor cells, further complicating the usage of organoids [22]. No robust or effective standard protocol for microenvironment has been established to date, limiting the replicability of organoid studies.

A possible way to address these issues, is to co-culture stromal and immune cells with the organoids, although this is not widely utilized [22]. Other components of the microenvironment, such as the correct in vivo concentrations of growth factors and signaling gradients, are very difficult to mimic in organoids, further complicating matters. Research for a so-called “next-generation organoid” is already been conducted, trying combinations of vascular networks, structured matrix, bioreactors, etc. [29].

Another possible solution to the lack of microenvironment is using a microfluidic platform [12]. This novel technique, also called organ-on-a-chip, is constructed in a top-down fashion in order to fully mimic the in vivo state. Biochemical signals, extracellular environment, mechanical forces, concentration gradients, etc., are all simulated and adjustable [18, 31]. Furthermore, co-culturing immune cells is possible, expanding the possibilities for drug testing. However, further research and optimization is needed [12, 31].

It remains unclear whether organoid culture conditions favor healthy organoids over diseased ones. While healthy organoids are readily grown under niche-like conditions, the efficiency of establishing cancer organoid culture from patients is lower in some studies [32]. The reasons for this remain elusive. In pancreatic cancer, the existence of normal-like (Wnt-dependent) and “true” cancer organoids (Wnt-independent) inside one tumor sample has been demonstrated [13, 32]. This raises a question of how many different cancer clones can organoid technology expand beside normal-like organoids: all or only a subset.

Another unresolved issue is whether organoids actually can capture completely the in vivo intratumoral heterogeneity, with only limited knowledge about this aspect of tumorigenesis [13, 33]. It is also possible that not all the clones are represented in the biopsy specimen, and therefore cannot be expanded like in the in vivo counterpart.