Table 1.

Applications of CRC patient-derived models in translational research

CRC Patient-derived model	Applications	Outcome	Advantages	Challeges	References
PDX
Primary or metastatic tumor implanted s.c.	Investigation of primary and acquired mechanisms of resistance	Amplification of ERBB2 gene as a driver of cetuximab resistance in KRAS/NRAS/BRAF/PI3KCA WT-metastatic CRC [45]; MET proto-Oncogene and FGFR1 amplification, ERBB2 and MAP2K1 activating mutation, IGF2 overexpression and EML4-ALK fusion protein production have been identified as cetuximab-resistance biomarkers [94–98]	Easy tumor monitoring and resecting	Slow expansion	[45, 46, 88, 94–98, 103, 104, 112]
	Discovery of prognostic and predictive biomarkers		Preserved intra-tumor and inter-patient heterogeneity	Difficult genetic manipulation
	Identification of new actionable targets	Pharmacologic GSK3α inhibition is sufficient to sensitize APC or β-catenin-mutant CRC to the anti-leukemic enzyme asparaginase displaying major tumor response. GSK3α inhibiting WNT-activating mutations, such as the RSPO3 fusion may predict asparaginase sensitivity [98]	Maintainance of original tumor architecture	Large collections and HTS difficult to realize
	Understanding of adaptive non-genetic processes sustaining residual disease	A reduced expression of EGFR ligands and high HER2/HER3 signaling have been displayed in mCRC cells surviving EGFR-targeted therapy. Pan-HER antibodies reduce residual desease and delay tumor relapse [103]		Lack of several immune system components
	Characterization of tumor heterogeneity and clonal evolution	Chemotherapy treatment eradicate actively proliferating cells while the resistant cells become dominant leading to a chemotherapeutic tolerance in CRC [104]		Progressive loss of human stroma cells and their replacement by murine counterparts
	Study tumor-stroma interactions
Primary or metastatic tumor implanted orthotopically	Investigation of mechanisms of invasion and metastasis	The tumorigenic potential of CRC stem cells (CCSCs) isolated from fresh human CRC have been evaluated through the CCSCs othotopically implantation into the wall submucosa of the ascending colon. The formation of spontaneous metastatic lesions was observed at local and distant sites (liver and mesenteric lymph nodes). Circulating CCSCs derived from CCSCs implanted in the colon can infiltrate and sustain distant metastasis [74]	Local invasive growth and tumor-host interactions in proper anatomical contex	Microsurgical skills	[74–77, 79]
	Study site-specific dependence on therapy		Spontaneous patient-like metastases development	Small animal imaging equipment required for tumor monitoring
Co-clinical trials and avatars	Real-time adaptive therapeutic decisions during clinical trials	Avatars with BRAF mutation show drug responses that mirror those in the corresponding patients, allowing to investigate the acquisition of resistance mechanisms [117, 118]	Best-matched PDX models for individual patients	Time-consuming	[117, 118]
				Not all tumor stages engraft
Humanized PDX models	Study how immune cells influence tumor progression	Anti-PD-1 therapy inhibits the tumor growth in MSI-H CRC, correlated with the increase of CD8 T cells and INF-γ-producing CD8+ tumor-infiltrating leukocytes, while fails in MSS-CRC, reflecting the patient's clinical outcome [123]	Mimicred human immune system in mice	Early onset of graft-versus-host disease	[123]
	Investigation of immunotherapies			Difficult and risky procedures during human stem cells collection in patients
PDO
Normal and tumor organoid cultures	Drug development	Identification of CRC patients not responding to irinotecan-based chemotherapy [133]; ERBB2-amplified, but not EGFR-amplified, PDO respond to lapatinib [63]; Combined inhibition of EGFR and KRASG12C is effective against colorectal with KRASG12C mutations [134]; WNT inhibition can improve the outcome of the 5-FU-based therapy [135]	Ease of genetic manipulation, in vitro functional studies and HTS	Lack of blood vessels, stroma and immune cells	[63, 132–152]
	Personalized medicine	Identification of specific drug sensitivities or resistances for each patient. CRC with KRAS and TP53 mutations is sentitive to trametinib alone or in combination with several targeted agents (celecoxib). Afatinib and the other EGFR inhibitors are effective against CRC with mutations in APC mutations and more effective in combination with HDAC inhibitors [85]	Fast expansion
	Modelling of cancer initiation and progression	Human colon organoids edited through CRISPR/Cas9 to induced mutation in tumor suppressor genes and oncogenes are tumorigenic in vivo but spontaneously develop metastasis only when implanted orthotopically into the naive microenvironment [148–151]	Healthy control organoids availability
	Study of single-cell tumor heterogeneity and clonal dynamics	In CRC clonal organoids have been identified distinct clonal organoids derived from the same tumor region with different drug responses [147]	Retained intra-tumor heterogeneity
			Possible transplantation in mice to substitute PDXs for in vivo studies
Air-liquid interface and co-culture approaches	Interrogation of tumor cells interactions with stroma and immune system	Air-liquid interface (ALI) recapitulates the interaction between tumor cells (PDOs) and tumor microenvironment components (stromal and native immune cells) and functionally models the immune checkpoint blockade with anti-PD-1 and/or PD-L1 that activates the tumor cytotoxicity mediated by CD8+TIL [153]	Preservation of endogenous immune stroma	Lack of blood vessels	[153–157]
	Study of tumorigenesis	Exploiting the co-culture of mouse intestinal organoids and fibroblasts, a rare fibroblast subpopulation that regulates tumor-initiating stem cells have been observed [101]
	Investigation of immunotherapies	Exploiting the co-culture of PDO derived from chemotherapy resistant mCRC and CD8 T cells, the efficacy of CEA-TCB immunotherapy have been evaluated. Low expression of CEA correlate with resistance to immunotherapy [137]