Establishment of a Patient-Derived Tumor Xenograft Model and Application for Precision Cancer Medicine

Abstract

Patient-derived xenograft (PDX) models can be created with the transplantation of cancerous cells or tissues from patients’ primary tumors into immunodeficient mice. PDXs are now in the spotlight as more accurate human cancer models compared with mouse tumor and human cancer cell lines transplanted into mice. PDX technology leads to breakthroughs with the introduction of novel, highly immunodeficient mice such as NOG (NOD/Scid/IL2Rγ^null), NSG (NOD/Scid/IL2Rγ^null), and NOJ (NOD/Scid/Jak3^null) mice. Xenograft efficiency differs by type of tumor, site of implantation, and tumor aggressiveness. Subcutaneous implantation is a standard method for PDX, and renal capsule or orthotropic implantation improves the efficiency. Despite positive test results in animal cancer models, significant numbers of novel drug candidates fail in clinical trials because conventional animal models such as murine tumor and human cancer cell line transplantation models do not always reflect the nature of human cancers. Since PDXs conserve the original tumor characteristics such as heterogeneous histology, clinical biomolecular signatures, malignant phenotypes and genotypes, tumor architecture, and tumor vasculature, they are currently believed to offer relevant predictive insights into clinical outcomes when evaluating the efficacy of novel cancer therapies. PDX banks with integrated genomic signatures are now established in many organizations including pharmaceutical companies. These PDX databases are becoming powerful tools for advancing precision cancer medicine.

1. Introduction

Laboratory mice have been used as key surrogate model systems to provide insights into the pathogenesis of human diseases and contributed to establishing novel drug candidates. Several different murine cancer models were established and novel anticancer drug candidates evaluated using these murine models (Table 1). The most commonly used models are transplantable murine tumors grown in syngeneic hosts and xenografts of human tumor cell lines grown in immunodeficient mice. However, even though murine models may suggest that novel drug candidates would be effective, a significant number of these drugs fails in clinical trials¹⁾ (Fig. 1) because of the considerable differences between mice and humans.²⁾ Recently, patient-derived tumor xenografts (PDXs) have been established and are becoming standard models for cancer research.^3,4) They are currently expected to become accurate models of human cancer and powerful tools for the evaluation of anticancer drugs in precision cancer medicine.

Table 1. Murine Model for Studying Human Cancers

Type	Subtype	Advantages	Disadvantages
Mouse tumor	Syngenic implant	Tumor exists in the presence of competent immune system	The character of murine tumor is not always mimic human tumors
	Induced	Defined mutations can mimic those identified in human tumors
	Genetically engineered	Can follow tumor development from early time point
Human tumor cell line	Native	Well established and easily evaluate the effects of anti-cancer drugs	Mice are immunocompromised, provide a less realistic tumor microenvironment
	Engineered	Genetic modification is posible	The character of tumor cell lines are not always mimic human tumors
	Subcutaneous
	Orthotopic
Patient-derived xenograft (PDX)	Subcutaneous	Provide realistic heterogeneity of tumor cells	Mice are immunocompromised, provide a less realistic tumor microenvironment
	Subrenal capsule	Keep the genetic characteristics of primary tumors	Technically complicated
	Orthotopic

Fig. 1. Application of Highly Immunodeficient Mice for Biomedical Research

2. Establishment of Highly Immunodeficient Mice

The first known immunodeficient mice, nude mice, were discovered in 1962 by Dr. N.R. Grist.⁵⁾ They were named for their characteristic appearance of a lack of body hair. Because nude mice are also athymic, they cannot generate mature T lymphocytes, indicating that they cannot mount adaptive immune responses. Since the early 1960 s, nude mice have been used as recipients for the transplantation of human tumor cells, especially solid tumors, although there are limitations on transplantable tumor cells due to the intact natural immune response. In 1983, severe combined immunodeficient (SCID) mice lacking both functional B and T lymphocytes were first described Bosma et al.⁶⁾ SCID mice were shown to be better recipients of xenografts than nude mice with successful implantation of human hematopoietic stem cells and mature blood cells. Nevertheless, nude mice have been used as recipients of human solid tumor xenografts because of the ease of measurement of tumor cells in the hairless phenotype. Non-obese diabetic (NOD) mice was established as a model of non-obese diabetes mellitus by Kikutani and Makino, who found that these mice presented complex immunodeficiency traits such as dysfunction of natural killer (NK) cells, macrophages, and dendritic cells.⁷⁾ NOD mice were subsequently crossed with SCID mice to establish the NOD/SCID mice. NOD/SCID mice have the advantages of accepting normal and malignant human hematopoietic cells with high efficiency as well as human solid tumors.⁸⁾ Recently, NOG (NOD/Scid/IL2Rγ^null),⁹⁾ NSG (NOD/Scid/IL2Rγ^null),¹⁰⁾ and NOJ (NOD/Scid/Jak3^null)¹¹⁾ mice with complete loss of NK cells have been established based on NOD/SCID mice (Figs. 2, 3). Human cell and tissue implantation-based studies have improved rapidly with the introduction of these highly immunodeficient mice.¹²⁾

Fig. 2. Process of Drug Development

New drug development requires a long term (9–17 years) and substantial financial investment. Even if they pass preclinical studies using mice models, significant numbers of candidate drugs fail in clinical studies. More accurate in vivo evaluation systems are needed to avoid failure at the clinical stage.

Fig. 3. NOD/Scid/Jak3^null (NOJ) Mice

(A) Appearance of NOJ mice. (B) NOJ mice lack mature T lymphocytes (CD3⁺), B lymphocytes (CD19⁺), and NK cells (CD122⁺DX5⁺).

Signal regulatory protein alpha (Sirpα)-CD47 signaling plays an important role in graft rejection by macrophages, and Sirpα polymorphism in the NOD mouse strain contributes to the efficiency of human cell engraftment with murine strain specificity.¹³⁾ BALB/c strain immunodeficient mice are also useful for human cell and tissue transplantation with Sirpα polymorphism,^14,15) and BALB/c Rag-2^null/IL2Rγ^null and Rag-2^null/Jak3^null mice were established and shown to be useful recipients of human cell transplantation.^14,16) Nude Rag-2^null/Jak3^null mice strains were also developed as tools for cancer research as subcutaneous tumors can easily be detected in these strains, which is advantageous for in vivo imaging.^17–19)

3. Establishment of PDX Models

PDXs are established by the implantation of primary patient tumors into immunodeficient mice. PDX models serve as a platform for clinical trials by enabling the integration of clinical data, genomic profiles, and drug responsiveness data for precision cancer medicine (Fig. 4).

Fig. 4. Generation of PDX Models

Surgical specimens form cancer patients are divided into small sections and transplanted into immunodeficient mice. When tumors are grown in mice, xenografts are used for genomic analysis and maintained in cryobanks. Tumor sections are further implanted into the next set of immunodeficient mice for expansion and maintained in PDX tissue banks. in vivo drug responsiveness is screened in these models.

Several different types of immunodeficient mice are used to establish PDX models (Table 2). Nude mice have been used as recipients in the PDX model, which have the advantages of the hairless phenotype for easy detection of subcutaneous tumors. Although xenograft efficiency is low, especially for hematological malignancies, the nude mouse model remains useful for some solid tumors such as colorectal, pancreatic, and head and neck cancers.⁴⁾ NOD/SCID mice are more immunoficient, although they are not suitable recipients due to their relatively short life span, high incidence of thymoma, and sensitivity to genome-toxic drugs due to the scid mutation. BALB/c Rag-2/γc knockout (KO) and Rag-2/Jak3 KO mice are relatively resistant to stress and genome-toxic drugs, have long life spans, and are easily bred compared with NOD/SCID-based immunodeficient mice and are suitable recipients of solid tumor xenografts. NOG, NSG, and NOJ mice are commonly used recipients of xenografts and show high xenograft efficiency, including those of hematological malignancies. Jackson Laboratory, for example, routinely uses NSG mice to establish PDXs.²⁰⁾

Table 2. Comparison of Immunodeficient Mouse Strains

		NOG,	NOD	BALB/c	NOD/Scid	BALB/c
		NSG,	Rag-2/γc KO	Rag-2/γc KO		Nude
		NOJ	Rag-2/Jak3 KO	Rag-2/Jak3 KO
Immune system	Mature T cells	Absent	Absent	Absent	Absent	Absent
	Mature B cells	Absent	Absent	Absent	Absent	Present
	NK cells	Absent	Absent	Absent	Defective	Present
	Dendritic cells	Defective	Defective	Present	Defective	Present
	Macrophages	Defective	Defective	Present	Defective	Present
	Compliment	Absent	Absent	Present	Absent	Present
Leakiness		Very low	Very low	Absent	Low	Low
Irradiation tolelance		Low	High	High	Low	High
Lymphoma incidence		Low	Low	Low	High	Low
Benefits		Supprot engraftment of normal and malignant hematopoietc cells, and solid tumors	Relatively resistant to stress	Relatively resistant to stress		Hairless phenotype enhances assessment of tumor growth
				Support engraftment of human solid tumors
			Resistant to irradiation	Resistant to irradiation
				Breeding is easy
Considerations		Sensitive to irradiation and genome toxic drugs			Development of thymic lymphoma by 9 months	Intact NK activity limits engraftment
					Relatively short life span (ave. 36 wks)
					Sensitive to irradiation and genome toxic drugs

PDX success rates differ by tumor type. Digestive system tumors such as colon, gastric, and esophageal cancers tend to have high rates of PDX success. However, the success rates of breast cancer are generally low, especially of estrogen receptor-positive breast cancers. However, human estrogen supplementation can enhance the xenograft efficiency of PDXs.²¹⁾ In general, clinically aggressive and metastatic cancers tend to have high PDX model engraftment rates compared with less aggressive and primary cancers. We tried to establish PDXs from cholangiocarcinoma tissues using BALB/c Rag-2/Jak3 KO mice, with a 75% success ratio (12/16 cases) (Vaeteewoottacharn et al., unpublished data).

The site of tumor implantation is an important factor for the generation of PDXs (Table 3). Subcutaneous transplantation is commonly used due to the simple procedures involved and easy measurement of tumor size. The renal capsule is an ideal site for implantation due to its hypervascularity and escape from immunocompetent cells, and successful generation of renal capsule PDXs of several cancers has been demonstrated. Orthotropic implantation results in behavior more similar to that of patient tumors, however.

Table 3. The Site of PDX Implantation

Implantation	Benefits	Disadvantages
Subcutaneous	Simple procedure	Lower engraft ratio
	Easy measurement of tumor
	Suitable for massive expansion
Renal capsule	Higher engraft ratio	Well trained surgical skill is needed
		Assessment is not easy
Orthotropic	Higher engraft ratio	Well trained surgical skill is needed

PDXs can be useful to investigate the pathogenesis of rare tumors and in the search for anticancer reagents for their treatment. Primary effusion lymphoma (PEL) is a rare human immunodeficiency virus (HIV)-associated non-Hodgkin’s lymphoma associated with human herpes virus-8 infection.²²⁾ The prognosis of PEL is poor, with median survival of 6 months even with standard chemotherapy, and therefore it is expected that more specific targeting therapy will be revealed with molecular analysis. We succeeded in establishing a PEL PDX using NOD Rag-2/Jak3 double-deficient mice from a Japanese HIV-infected patient and evaluated the efficacy of anti-CD47 antibody in the PDX-bearing mice.²³⁾ We also established the PEL cell line GTO from the same patient²⁴⁾ and found that 88 of 3519 reagents examined were effective against the PEL-PDX cells, whereas 135 reagents were effective against the GTO PEL cell line. In total, 64 reagents exhibited anti-PEL activity in both PEL-PDX and GTO. YM155, a possible survivin inhibitor, was selected and its anti-PEL effects were confirmed both in vitro and in PEL-PDX mice.²⁵⁾ Thus, PDXs are effective tools for the investigation of rare tumors.

4. Future Directions

Recently, the US National Cancer Institute has decided to retire NCl-60,²⁶⁾ a panel of 60 human cancer cell lines, from its drug-screening program and will instead use PDXs.²⁷⁾ However, current PDXs require time for establishment and cannot host all patient-derived tumors. In addition, PDXs have been largely been developed within individual institutions. There is a growing recognition of the need to develop large collaborative groups to create large stocks of PDXs or PDX banks.²⁸⁾ Therefore several institutions have begun to develop repositories (libraries) of PDXs. Sixteen European institutions established EurOPDX, a consortium to store PDXs and have already accumulated more than 1500 samples in a PDX bank.^29,30) Jackson Laboratory has more than 450 cases and provides samples to researchers.²⁰⁾ Pharmaceutical companies are also establishing their own PDX banks, and Novartis published data on drug screening using 1000 PDXs.³¹⁾ Some Japanese institutions have also started to accumulate PDXs, although each repository is still small and they are not integrated. Since the character of tumors differs by region, race, etc., it is necessary to establish Japanese and Asian PDX banks in the near future to establish precision cancer medicine for Japanese and other Asians.

The current PDX model has some limitations. Humanized mice cannot reconstitute the complete human immune system even in the bone marrow, liver, and thymus transplanted BLT humanized mice, indicating that tumor immune reactions cannot be simulated completely by current PDX models. Jackson Laboratory established the Onco-Hu mouse strain, which is dually engrafted with human CD34⁺ hematopoietic stem cells and the clinically relevant PDX. Onco-Hu mice are expected to function as a novel platform for testing immunotherapies against human tumors (https://www.jax.org/jax-mice-and-services/in-vivo-pharmacology/oncology-services/onco-hu). However, in their current form some difficulties in PDX utilization remain in terms of differences from the complete human immune system and immune reactions against tumor cells.

Several attempts have been made to establish a more humanized microenvironment in mouse strains.³²⁾ Transgenic expression of human leukocyte antigen (HLA) support the development of HLA-restricted T cells, and KO of murine MHC class I and class II molecules is expected to decrease the risk of graft-versus-host disease.³³⁾ The introduction of human transgenes encoding human cytokines such as stem cell factor, granulocyte macrophage colony-stimulating factor, and interleukin-3 into NSG mice improves myeloid differentiation from transplanted human hematopoietic stem cells and xenograft efficiency of acute myeloid leukemia.^34–36) The human bone marrow microenvironment was also established from mesenchymal stem cells etc.³²⁾

It is expected that fully humanized orthotropic xenograft models will be established with improved humanized mice techniques and provide the preclinical platform to investigate human cancers.

Acknowledgments

This work was supported in part by the Research Program on HIV/AIDS (No. 17fk0410208h0002) from the Japan Agency for Medical Research and Development, AMED, the Adaptable and Seamless Technology Transfer Program (A-STEP, No. AS2311293E) from the Japan Science and Technology Agency, and a Grant-in-Aid for Science Research (No. 16K08742) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Conflict of Interest

The authors declare no conflict of interest.

References

• 1) Perrin S., Nature (London), 507, 423–425 (2014).
• 2) Mestas J., Hughes C. C., J. Immunol., 172, 2731–2738 (2004).
• 3) Sausville E. A., Burger A. M., Cancer Res., 66, 3351–3354, discussion, 3354 (2006).
• 4) Cho S. Y., Kang W., Han J. Y., Min S., Kang J., Lee A., Kwon J. Y., Lee C., Park H., Mol. Cells, 39, 77–86 (2016).
• 5) Flanagan S. P., Genet. Res., 8, 295–309 (1966).
• 6) Bosma G. C., Custer R. P., Bosma M. J., Nature (London), 301, 527–530 (1983).
• 7) Kikutani H., Makino S., Adv. Immunol., 51, 285–322 (1992).
• 8) Lapidot T., Fajerman Y., Kollet O., J. Mol. Med., 75, 664–673 (1997).
• 9) Ito M., Hiramatsu H., Kobayashi K., Suzue K., Kawahata M., Hioki K., Ueyama Y., Koyanagi Y., Sugamura K., Tsuji K., Heike T., Nakahata T., Blood, 100, 3175–3182 (2002).
• 10) Shultz L. D., Lyons B. L., Burzenski L. M., Gott B., Chen X., Chaleff S., Kotb M., Gillies S. D., King M., Mangada J., Greiner D. L., Handgretinger R., J. Immunol., 174, 6477–6489 (2005).
• 11) Okada S., Harada H., Ito T., Saito T., Suzu S., Int. J. Hematol., 88, 476–482 (2008).
• 12) Shultz L. D., Ishikawa F., Greiner D. L., Nat. Rev. Immunol., 7, 118–130 (2007).
• 13) Takenaka K., Prasolava T. K., Wang J. C., Mortin-Toth S. M., Khalouei S., Gan O. I., Dick J. E., Danska J. S., Nat. Immunol., 8, 1313–1323 (2007).
• 14) Ono A., Hattori S., Kariya R., Iwanaga S., Taura M., Harada H., Suzu S., Okada S., J. Biomed. Biotechnol., 2011, 539748 (2011).
• 15) Goto H., Kariya R., Matsuda K., Kudo E., Katano H., Okada S., Leuk. Res., 42, 37–42 (2016).
• 16) Traggiai E., Chicha L., Mazzucchelli L., Bronz L., Piffaretti J. C., Lanzavecchia A., Manz M. G., Science, 304, 104–107 (2004).
• 17) Kariya R., Matsuda K., Gotoh K., Vaeteewoottacharn K., Hattori S., Okada S., In Vivo, 28, 779–784 (2014).
• 18) Gotoh K., Kariya R., Matsuda K., Hattori S., Vaeteewoottacharn K., Okada S., Biosci. Trends, 8, 202–205 (2014).
• 19) Tanaka A., Takeda S., Kariya R., Matsuda K., Urano E., Okada S., Komano J., Leukemia, 27, 1621–1627 (2013).
• 20) Shultz L. D., Goodwin N., Ishikawa F., Hosur V., Lyons B. L., Greiner D. L., Cold Spring Harbor Protocols, 2014, 694–708 (2014).
• 21) Zhang X., Claerhout S., Prat A., Dobrolecki L. E., Petrovic I., Lai Q., Landis M. D., Wiechmann L., Schiff R., Giuliano M., Wong H., Fuqua S. W., Contreras A., Gutierrez C., Huang J., Mao S., Pavlick A. C., Froehlich A. M., Wu M. F., Tsimelzon A., Hilsenbeck S. G., Chen E. S., Zuloaga P., Shaw C. A., Rimawi M. F., Perou C. M., Mills G. B., Chang J. C., Lewis M. T., Cancer Res., 73, 4885–4897 (2013).
• 22) Okada S., Goto H., Yotsumoto M., Intract. Rare Dis. Res., 3, 65–74 (2014).
• 23) Goto H., Kojima Y., Matsuda K., Kariya R., Taura M., Kuwahara K., Nagai H., Katano H., Okada S., Eur. J. Cancer, 50, 1836–1846 (2014).
• 24) Goto H., Kojima Y., Nagai H., Okada S., Int. J. Hematol., 97, 624–633 (2013).
• 25) Kojima Y., Hayakawa F., Morishita T., Sugimoto K., Minamikawa Y., Iwase M., Yamamoto H., Hirano D., Imoto N., Shimada K., Okada S., Kiyoi H., Pharmacol. Res., 120, 242–251 (2017).
• 26) Shoemaker R. H., Nat. Rev. Cancer, 6, 813–823 (2006).
• 27) Ledford H., Nature (London), 530, 391 (2016).
• 28) Aparicio S., Hidalgo M., Kung A. L., Nat. Rev. Cancer, 15, 311–316 (2015).
• 29) Byrne A. T., Alferez D. G., Amant F., Annibali D., Arribas J., Biankin A. V., Bruna A., Budinska E., Caldas C., Chang D. K., Clarke R. B., Clevers H., Coukos G., Dangles-Marie V., Eckhardt S. G., Gonzalez-Suarez E., Hermans E., Hidalgo M., Jarzabek M. A., de Jong S., Jonkers J., Kemper K., Lanfrancone L., Maelandsmo G. M., Marangoni E., Marine J. C., Medico E., Norum J. H., Palmer H. G., Peeper D. S., Pelicci P. G., Piris-Gimenez A., Roman-Roman S., Rueda O. M., Seoane J., Serra V., Soucek L., Vanhecke D., Villanueva A., Vinolo E., Bertotti A., Trusolino L., Nat. Rev. Cancer, 17, 254–268 (2017).
• 30) Hidalgo M., Amant F., Biankin A. V., Budinska E., Byrne A. T., Caldas C., Clarke R. B., de Jong S., Jonkers J., Maelandsmo G. M., Roman-Roman S., Seoane J., Trusolino L., Villanueva A., Cancer Discov., 4, 998–1013 (2014).
• 31) Gao H., Korn J. M., Ferretti S., Monahan J. E., Wang Y., Singh M., Zhang C., Schnell C., Yang G., Zhang Y., Balbin O. A., Barbe S., Cai H., Casey F., Chatterjee S., Chiang D. Y., Chuai S., Cogan S. M., Collins S. D., Dammassa E., Ebel N., Embry M., Green J., Kauffmann A., Kowal C., Leary R. J., Lehar J., Liang Y., Loo A., Lorenzana E., McDonald R. E. III, McLaughlin M. E., Merkin J., Meyer R., Naylor T. L., Patawaran M., Reddy A., Roelli C., Ruddy D. A., Salangsang F., Santacroce F., Singh A. P., Tang Y., Tinetto W., Tobler S., Velazquez R., Venkatesan K., Von Arx F., Wang H. Q., Wang Z., Wiesmann M., Wyss D., Xu F., Bitter H., Atadja P., Lees E., Hofmann F., Li E., Keen N., Cozens R., Jensen M. R., Pryer N. K., Williams J. A., Sellers W. R., Nat. Med., 21, 1318–1325 (2015).
• 32) Theocharides A. P., Rongvaux A., Fritsch K., Flavell R. A., Manz M. G., Haematologica, 101, 5–19 (2016).
• 33) Morton J. J., Bird G., Refaeli Y., Jimeno A., Cancer Res., 76, 6153–6158 (2016).
• 34) Wunderlich M., Chou F. S., Link K. A., Mizukawa B., Perry R. L., Carroll M., Mulloy J. C., Leukemia, 24, 1785–1788 (2010).
• 35) Holzapfel B. M., Wagner F., Thibaudeau L., Levesque J. P., Hutmacher D. W., Stem Cells, 33, 1696–1704 (2015).
• 36) Saito Y., Ellegast J. M., Rafiei A., Song Y., Kull D., Heikenwalder M., Rongvaux A., Halene S., Flavell R. A., Manz M. G., Blood, 128, 1829–1833 (2016).