Table 1 Characteristics of different cancer models

PDX model in immunodeficient mice

(1) Highly recapitulate tumor microenvironment

(2) High fidelity

(3) High predictive value

(4) Can be used in metastasis models

(1) Unsatisfactory take rate

(2) High cost

(3) Technically challenging

(1) Developing new immunodeficient mice

(2) Determining the most appropriate conditions and methods to improve the take rate

PDX model in humanized immunodeficient mice

(1) Can be evaluated for immunotherapy

(2) Partly recapitulate tumor microenvironment

Therapeutic effect is affected by immunity

Developing more comprehensive and functional immune system humanized mice

PDX model in zebrafish

(1) Transparent embryos facilitate the visualization of tumor processes and the tracking of fluorescently labeled cells

(2) Easy to culture, low cost

(3) Allow large-scale and high-throughput screening

(1) Not a mammal

(2) The rate of proliferation or the manner in which the transplanted human cells formed tumor masses differed from those of immunodeficient mice or human patients

Standardization of zebrafish xenograft techniques and application modalities

Cell line-derived xenograft model

(1) Low cost, easy establishment and high take rate

(2) Slightly recapitulate tumor microenvironment

(1) Can’t reproduce heterogeneity

(2) Can’t maintain the original cell properties

(3) Low predictive value

Large numbers of mice were used at relatively early stages of drug development to reflect heterogeneity among tumors

Circulating tumor cell-derived xenograft model

(1) High heterogeneity in the metastatic environment

(2) High tumorigenicity

(3) Partly recapitulate tumor microenvironment

(1) Difficulty in the isolation and counting of CTCs

(2) Organ metastasis is affected by the injection site of CTCs

(1) Capturing the heterogeneity of CTCs by drug sensitivity assay

(2) In vitro culture of CTCs

Patient-derived organoid model

(1) No ethical issues

(2) Time saving, suitable for high-throughput drug screening

(3) Maintenance of the gene expression profile of the initial tumor over an extended period of time

(4) Available for low malignant tumors

(1) Lack of uniform standards

(2) Tumor free microenvironment

(3) Limited tissue availability

(4) Labour intensive

(1) Develop high-throughput systems

(2) Development of uniform modeling standards

(3) Evolution of co-culture models

(4) Increased access to biobanks

(5) Use of circulating tumor cells

Genetically engineered mouse model

(1) Recapitulate tumor initiation and early development process

(2) Specific genes can be studied in detail

(3) Can be increased easily after establishment

(1) Can’t reproduce heterogeneity of human tumor

(2) Take a long time to be established

(3) Low predictive value

Investigating how a specific gene of interest could contribute to tumor initiation and relapse