Embed Size (px)
Citation preview
REVIEW Open Access
Applications of patient-derived tumorxenograft models and tumor organoidsGo J. Yoshida1,2
Abstract
Patient-derived tumor xenografts (PDXs), in which tumor fragments surgically dissected from cancer patients aredirectly transplanted into immunodeficient mice, have emerged as a useful model for translational research aimedat facilitating precision medicine. PDX susceptibility to anti-cancer drugs is closely correlated with clinical data inpatients, from whom PDX models have been derived. Accumulating evidence suggests that PDX models are highlyeffective in predicting the efficacy of both conventional and novel anti-cancer therapeutics. This also allows “co-clinical trials,” in which pre-clinical investigations in vivo and clinical trials could be performed in parallel orsequentially to assess drug efficacy in patients and PDXs. However, tumor heterogeneity present in PDX modelsand in the original tumor samples constitutes an obstacle for application of PDX models. Moreover, human stromalcells originally present in tumors dissected from patients are gradually replaced by host stromal cells as thexenograft grows. This replacement by murine stroma could preclude analysis of human tumor-stroma interactions,as some mouse stromal cytokines might not affect human carcinoma cells in PDX models. The present reviewhighlights the biological and clinical significance of PDX models and three-dimensional patient-derived tumororganoid cultures of several kinds of solid tumors, such as those of the colon, pancreas, brain, breast, lung, skin, andovary.
Keywords: Acquired resistance, Avatar models, Carcinoma-associated fibroblasts, Co-clinical trials, Heterogeneity,Immunodeficient mice, Organoids, PDX models, Translational research, Tumor microenvironment
BackgroundSince the first investigations to develop drugs using ani-mal models of leukemia were reported in 1950 [1], manytypes of murine models transplanted with human tu-mors have been developed to predict responses tochemotherapy. Many human tumor models were gener-ated in immunodeficient mice by subcutaneous ororthotopic injection of tumor cell lines that had beenpropagated in culture for several months to several de-cades. Although this model served as a gold standard,primarily because of ease of manipulation, tumor celllines frequently acquired unanticipated phenotypes dur-ing adaptation to in vitro culture conditions that oftendiffered between laboratories, resulting in minimal re-semblance to the parental tumors. For example, immor-talized cells cultured long-term in vitro exhibit changes
in tumor hallmarks caused by genetic and epigenetic al-terations. Thus, substantial limitations have precludedthe application of conventional xenograft models fordrug screening and estimating the pre-clinical efficacy ofdrugs [2].To develop a model more likely to mimic human tu-
mors, patient-derived tumor xenografts (PDXs) havebeen established as a useful tool for translational re-search [3–6]. Implantation of small pieces of tumors sur-gically dissected from cancer patients into highlyimmunodeficient mice allows tumor growth and subse-quent transplantation into secondary recipient mice.PDXs often maintain the cellular and histopathologicalstructures of the original tumors (Table 1) [7–9]. Fur-thermore, cytogenetic analysis of tumor cells from PDXshas revealed significant preservation of the genomic andgene expression profiles between PDXs and parental pa-tient tumors [10–14]. Notably, sensitivity to standardchemotherapeutics in PDXs closely correlates with clin-ical data in patients from which the PDXs are derived
© The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Correspondence: [email protected] of Pathology and Oncology, Juntendo University School ofMedicine, 2-1-1, Hongo, Bunkyo-ku, Tokyo 113-8412, Japan2Department of Immunological Diagnosis, Juntendo University GraduateSchool of Medicine, 2-1-1, Hongo, Bunkyo-ku, Tokyo 113-8412, Japan
Yoshida Journal of Hematology & Oncology (2020) 13:4 https://doi.org/10.1186/s13045-019-0829-z
[12, 13, 15–18]. All of these characteristics demonstratethat PDXs are more useful as predictive experimentalmodels of therapeutic responses.Tumor organoids are generated by three-dimensional
culture of primary cancer cells and have a high successrate [19–25]. Organoid cultures closely recapitulate themorphological and genetic/epigenetic features of theparent tumors (Table 1) [20, 21, 25, 26]. Accumulatingevidence reveals that primary cancer cells subjected tothree-dimensional culture give rise to many differenttypes of primary organoids, in which the heterogeneouscomposition of the original tumors is largely conserved[22, 27, 28]. This culture method provides promising op-portunities to establish large biobanks with relevant clin-ical materials that can be used to perform drugscreening and facilitate chemical discovery [24, 29].Tumor organoids are superior to conventional models,as organoids constitute a minute incarnation of anin vivo organ, and thus can better recapitulate the char-acteristics of the parental tumor, even after many pas-sages. Organoids have enormous potential for theidentification of optimal treatment strategies in individ-ual patients [28].
PDX models maintain in vivo structureA major advantage of PDX models in cancer research isthe retention of the original tumor architecture. Al-though cancer cell lines are frequently transplanted intoimmunocompromised mice as a reproducible method toestablish tumor tissues in vivo, the resultant tumors donot fully exhibit the distinct histopathological character-istics observed in clinical settings. One important con-sideration in this regard is that cells within tissues aresurrounded by an extracellular matrix (ECM), a mesh-work of proteins and proteoglycans consisting of lam-inin, collagen, and fibronectin, which regulates theintegrity of epithelial structure formations. The ECMprovides structural support, stability, flexibility, andshape to the tissue, and also mediates cell polarity, intra-cellular signaling, and cell migration [30, 31]. Intracellu-larly, ECM-induced signaling pathways are transmittedmainly through integrin molecules, which are heterodi-meric transmembrane receptors that mediate cell adhe-sion to the ECM. Integrins act as bridges between the
ECM and the internal cytoskeleton by transducing keyintracellular signals through association with clusters ofkinases and adaptor proteins in focal adhesion com-plexes [32]. These interactions are distinctly mediated byspecific integrin heterodimer binding to individual ECMcomponents [33]. The signals are transmitted via thecytoskeleton to nuclear transcription factors and ultim-ately change the gene expression profile [34]. As such,highly specific and localized signaling cascades can beactivated by the ECM in association with various avail-able growth factors through sequences of reactions in-volving networks of proteases and sulfatases [35, 36].The tumor microenvironment comprises the ECM and
stromal cells, including endothelial cells, pericytes,carcinoma-associated fibroblasts (CAFs), immune cells,and proinflammatory cells [37, 38]. One of the mostprominent cell types in the tumor stroma is CAFs, whichare activated fibroblasts. CAFs often exhibit an α-smooth muscle actin (α-SMA)-positive myofibroblasticphenotype, which is typical of fibroses and wound heal-ing, and is a pro-inflammatory phenotype [36, 39].Tumor-promoting CAFs influence tumor hallmarks topromote cancer growth, progression, and metastaticspread as well as remodeling of the ECM [36, 39, 40].CAFs contribute to ECM remodeling by secreting TGF-β1, which is synthesized as an inactive multidomaincomplex, and is activated through multiple extracellularmechanisms that require integrins, proteases, andthrombospondin-1 [41, 42]. TGF-β1 has a potential rolein the regulation of ECM remodeling during cancer pro-gression. Exposing CAFs to TGF-β promotes anisotropicfiber organization and increased matrix stiffness throughincreased α-SMA expression and RhoA activation [43].These changes in cellular contractility result in the CAF-mediated changes in the ECM architecture.Tumors in the PDX model have similar pathohistologi-
cal and genetic characteristics to the parent tumor, andexhibit similar susceptibility to anti-cancer therapies [7,10, 11, 15, 16]. Therefore, PDX models based on graftingcancer tissue subcutaneously, orthotopically, or under kid-ney capsules in immunodeficient mice provide relevantpre-clinical cancer models. The mouse strains widely usedfor tumor formation and propagation are classified into (i)nude mice, which lack a thymus and are unable to pro-duce T cells; (ii) non-obese diabetic severe combined im-munodeficiency (NOD-SCID) and SCID-beige mice,which lack T and B cells; and (iii) NOD-SCID IL2R-γ null(NSG or NOG) mice, in which T, B, and NK cell activityis completely absent. Due to differential immunologicalimpairments in these models, it is expected that more per-missive mouse strains such as NOD-SCID, SCID, andNSG, will increase the efficiency of xenotransplantationover that of nude mice. In fact, a very low engraftmentsuccess rate (10–25%) was reported after transplanting
Table 1 Comparison between PDX models and organoids
PDX Organoids
Genetic/epigenetic alterations Similar Similar
Pathohistological characteristics Similar Similar
Response to anti-cancer drugs Similar Similar
Use of immunodeficient animals Yes No
Reliability as pre-clinical models Yes Yes
Quantity of cells for establishment Large Small
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 2 of 16
tumor fragments of different pathohistological types intonude mice, whereas transplantation into NOD-SCID miceresulted in a higher engraftment rate (25-40%) for breastcancer, non-small cell lung cancer and malignant melan-oma [44]. An extremely novel study evaluated the use ofzebrafish for PDX, and reported that transparent zebrafishlarvae allow for visualization of single tumor cells andtheir response to treatment, thereby providing a rapidscreening platform [17].However, it is important to note that human stromal
cells are gradually replaced by murine counterparts aftertransplantation into immunodeficient mice, which sug-gests that implanted human cancer cells retain the poten-tial to recruit murine stromal cells to their niche.Importantly, there are some differences between ligandssecreted by human and murine fibroblasts. Human-derived interleukin (IL)-2 stimulates efficiently the prolif-eration of murine T cells, whereas mouse IL-2 stimulateshuman T cells with significantly lower efficiency [45, 46].Furthermore, the T cell stimulating potential of IL-4 ap-pears to be species-specific [45]. On the other hand, IL-15binds to human or mouse IL-15 receptor α with equallyhigh affinity [47], suggesting that human-derived IL-15can function on murine cells. However, human NK cellsare weakly sensitive to murine IL-15 [48]. In order to re-solve the issues related to species specificity, co-implantation of human CAFs and tumor cell suspensionsextracted from PDXs into secondary recipient mice couldprovide an optimal setting for evaluating human tumorcell-stroma cell interactions.
Patient-derived tumor cell organoid culture mimicparental tumorsFrom a histopathological perspective, PDX modelslargely retain parent tumor architecture, as describedabove. Further, at the cellular level, both inter-tumoraland intra-tumoral heterogeneity, as well as the pheno-typic and molecular characteristics of the original cancerare also preserved in PDX models [12, 49, 50]. From thisperspective, human cancer tissues or tumor cells derivedfrom PDX models that mimic the biological and molecu-lar characteristics of the original cancer can be employedto provide clinically relevant donor cells for in vitromodeling. Tumor tissues from PDX models can be usedto generate three-dimensional tissue explants in vitro orprimary cell cultures on a petri dish. Use of PDXs hasthe advantage that the original tissue can be seriallypropagated in vivo, making additional materials availablefor repeated experiments and alleviating the limitationsinherent to collecting multiple patient samples. To morefully represent the range of biological and molecularcharacteristics of clinical samples, a panel of establishedxenograft lines covering multiple cancer tissues of originand subtypes is required for selection of the most
appropriate model to address particular experimentalquestions.Furthermore, mounting evidence suggests that orga-
noids are three-dimensional constructs comprised ofmultiple cell types that originate from pluripotent stemcells by means of self-organization, and are capable ofsimulating the architecture and functionality of nativetissues and organs [25, 51]. Organoids permit bothin vitro and in vivo investigations, and represent one ofthe latest innovations in development of models to re-capitulate the physiologic processes of whole organisms.Organoids have been successfully generated from pri-mary tumors of the breast, colon, pancreas, and prostate[23, 25]. These tumor-derived organoids have emergedas pre-clinical models that have the potential to predictpersonalized response to treatment. For example, a livingbiobank of tumor organoids from patients with meta-static gastrointestinal cancer was demonstrated to recap-itulate the therapeutic responses of these patients toanti-cancer agents in clinical trials [29].
Use of PDX models for precision medicineMounting evidence demonstrates that PDX models havethe potential to predict effectively the efficacy of bothconventional and novel anti-cancer therapeutics, sug-gesting that these models could be employed in “co-clin-ical trials” [52]. Thus, investigations in vivo and inclinical trials could be performed in parallel in order toidentify therapeutic target molecules, even in rare typesof cancer. Particularly, the concept of “co-clinical trials”incorporates patient selection strategies based on mo-lecular abnormalities or the identification of the machin-eries of resistance to anti-cancer agents, for thedevelopment of precision medicine aimed at personaliz-ing anti-cancer treatment. In this context, PDX modelscan be also used as an “avatar model [53],” in whichPDX models obtained from cancer patients enrolled in aclinical trial can be treated with the same therapy ad-ministered to the patient, thereby permitting identifica-tion of novel biomarkers of sensitivity or resistance tothe anti-cancer treatment(s) of interest (Fig. 1).
Colorectal adenocarcinomaAcquired resistance of tumors to chemotherapeutics is amajor reason for treatment failure [54, 55], and the intro-duction of novel drugs or combinational therapy permitsselection of effective therapeutic strategies for second-linetreatment. Based on this concept, Misale et al. treated ad-vanced colorectal adenocarcinoma (CRC) patients withanti-EGFR antibodies as single agents after the initial re-sponse to EGFR inhibitors resulted in disease relapse dueto emerging resistance [12]. Thus, Misale et al. investi-gated how acquisition of resistance to EGFR-targetedtherapies can be reduced in CRC tumor cells using the
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 3 of 16
PDX model. In vitro genetic screening and functional in-vestigations identified that dual blockade of EGFR andMEK prevents acquired resistance, and Misale et al. per-formed experiments in vivo using PDXs derived from aCRC patient carrying a quadruple wild-type gene profile(BRAF, KRAS, NRAS, and PIK3CA), which recapitulatesthe expression profile of patients sensitive to anti-EGFRantibodies. While treatment of PDX models with theMEK inhibitor pimasertib alone only slightly reducedtumor growth, treatment with the EGFR inhibitor cetuxi-mab effectively reduced cancer proliferation by more than70% [12]. Notably, the subsequent regrowth of these tu-mors suggested resistance to drug re-challenge, which issimilar to clinical findings. By contrast, combination treat-ment with cetuximab and pimasertib induced a completeresponse, in which tumor tissues remained undetectablefor more than 6 months. These findings suggested thatcombination treatment is highly likely to inhibit develop-ment of resistant tumors with intra-tumoral heterogeneity,and highlighted the utility of PDX models in establishinghighly effective treatment regimens.Okazawa et al. established green fluorescent protein
(GFP)-labelled CRC PDX-derived organoids to detect
spontaneous micrometastatic lesions [56]. Micrometastases,which are cancer cell depositions smaller than 2 mm, haveoften been overlooked in the pathohistological analysis ofsections from affected distant organs in experimental mur-ine models, because they may not be detected with this ap-proach. Therefore, Okazawa et al. developed a protocol toefficiently transduce GFP lentivirus into PDX-derived CRCorganoids in three-dimensional culture prior to transplant-ation, enabling highly sensitive detection of micrometas-tases in distant organs such as the liver and lungs (Fig. 2).Using this technology, Okazawa et al. employed a PDXmodel to demonstrate that lung micrometastases could bedetected in the majority of engrafted mice 3 months aftertransplantation. Moreover, liver metastases were detectedwithin 1 month after injection of organoids into the spleen.Further, the injection of CRC organoids into the rectal sub-mucosa of immunocompromised mice resulted in meta-static dissemination into the lungs, but not into the liver,likely through the inferior vena cava.
Pancreatic cancerAlthough the progression of pancreatic ductal adenocar-cinoma (PDAC) is driven by constitutive activation of
Fig. 1 Identification of optimal therapeutics using PDX mouse clinical trials. PDX models are potentially useful when the optimal course oftreatment cannot be readily determined for individual patients. For instance, in the illustration, there are three patients (A-C) with gastric cancer,who hope to receive treatment with the novel therapy drug X if its therapeutic efficacy is proven. In this case, it would be time-consuming andrequire significant clinical risk to compare the therapeutic response to conventional drugs and the new drug X without “co-clinical trials.” Whilexenografts derived from patient A respond to drug X, xenografts derived from patient C respond to conventional treatments, but not drug X(step 1). Contrastingly, patient B–derived xenografts partially respond to both therapeutics. This pre-clinical screening by an avatar model ishelpful to determine which treatment would have the optimal outcome in each patient (step 2)
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 4 of 16
RAS/RAF/MEK/ERK signaling, MEK1/2 inhibition withtrametinib is not clinically effective in PDAC patients. Arecent study revealed that trametinib-induced autophagyflux increases the survival of PDAC cells subjected toMEK1/2 inhibition [57]. Combination treatment withtrametinib and chloroquine increased apoptotic celldeath in a PDX model derived from a distant neck me-tastasis that was refractory to the conventional FOLFOXregimen. It is widely accepted that chloroquine obstructsautophagy flux, increasing ubiquitination, p62/SQSTM1activation, and LC3-II accumulation [58]. Furthermore,the growth of orthotopic and/or subcutaneous PDX tu-mors in NOD/SCID mice was synergistically inhibitedby combination treatment with trametinib and chloro-quine [57]. The efficacy of this modality was also dem-onstrated in PDX models using gain-of-functionalmutated NRAS-driven malignant melanoma andBRAF(V600E)-driven colorectal cancer. Notably, thetherapeutic efficacy of this combination was superior tothat of conventional anti-PDAC drugs such as gemcita-bine. The compensatory activation of protective autoph-agy through the ULK1/AMPK/LKB1 axis appears to beresponsible for susceptibility to trametinib in the pres-ence of chloroquine.
Brain tumorsSingh et al. reported that SOX2 expression is markedly el-evated in PDX derived from glioblastoma multiforme(GBM) [59]. They established PDX models derived from20 GBM patients, comprising different subtypes of
putative GBM driver oncogenes such as EGFR, MET, andPDGFRA. Remarkably, all four IDH1 mutant PDX linesalso exhibited high expression levels of SOX2. High SOX2expression was present in GBM PDX that were driven bydifferent oncogenes, suggesting that multiple oncogenicsignaling pathways are likely to converge to drive expres-sion of this pluripotency transcription factor. While PDXtumors maintained the putative oncogenic mutations ofseveral receptor tyrosine kinases (RTKs) in the parentaltumors, the significant increase in SOX2-expressing GBMcells may reflect the functional state of the subset of GBMcells capable of driving tumor growth in the mouse brainrather than genetic differences between the original pa-tient tumor and PDX model. Singh et al. demonstratedthat the transcriptional regulatory network comprised ofSOX2, OLIG2, and ZEB1 is independent of upstreamRTKs, and is capable of driving glioma initiating cells.Mithramycin is an antibiotic isolated from Streptomycesplicatus that binds to specific DNA regions to inhibit tran-scription of specific genes. Notably, mithramycin down-regulated SOX2 and ZEB1 in sarcoma cells [60], andeffectively inhibited growth of a SOX2-positive cell popu-lation that propagates medulloblastoma [61]. Administrat-ing mithramycin in vivo markedly reduced expression ofSOX2, OLIG2, and ZEB1, which coincided with dramaticreductions in tumor growth [59]. Taken together, thesestudies employing PDX models strongly suggest the im-portance evaluating the efficacy of combining mithramy-cin treatment with chemotherapy and radiation therapy infuture investigations.
Fig. 2 High-sensitivity detection of distant micrometastases of PDX-derived organoids by GFP transduction. After a primary colorectaladenocarcinoma (CRC) diagnosed as a moderately differentiated type was surgically resected, CRC cells were subcutaneously implanted into NOGmice to establish a PDX model. PDX tissue was treated with collagenase to obtain tumor cell suspension. CRC organoids were then establishedfrom PDX tissue and expanded in three-dimensional culture using the artificial extracellular matrix after infection with GFP lentivirus. These GFP-labelled organoids implanted orthotopically revealed distant micrometastases in the lungs within 3 months [56]
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 5 of 16
Breast carcinomaPDX models maintain the essential properties of the ori-ginal patient tumors, including metastatic tropism, sug-gesting their physiological relevance for study of humancancer metastasis [4]. In immunodeficient mice, PDXsspontaneously metastasize many of the same organs af-fected in the original patient. In addition, mesenchymalstem cells (MSCs) in the PDX model enhance tumorgrowth rates by promoting angiogenesis, decreasing ne-crosis, and increasing blood volume, which would con-tribute to the observed increase in tumor growth.Lawason et al. demonstrated using the PDX model thatprogression to high metastatic burden is associated withincreased proliferation and Myc expression, which canbe attenuated by the treatment with cyclin-dependentkinase (CDK) inhibitors [8]. In this study, the mostmetastatic PDX had the highest percentage of cancerstem-like basal primary tumor cells, while the leastmetastatic PDX had the lowest. This suggests that pri-mary tumors contain a rare subpopulation of stem-likecells, and that the relative abundance of these cells couldcorrelate with metastatic potential. Thus, Lawason et al.used PDX models to propose a hierarchical model formetastasis, in which metastases are initiated by cancerstem-like cells, which proliferate and differentiate to pro-duce advanced metastatic disease.
Lung cancerChen et al. recently demonstrated an unexpected plasti-city and interaction of lung squamous cancer cells(LSCCs) with the tumor microenvironment [62]. Over-expression of SOX2 in the TUM622 cell line, which wasestablished from a PDX model, enhances spheroid-forming potential and drives a hyperplastic to dysplasticalteration in acinar phenotype, in which apical-basal cellpolarity is disrupted, and solid non-invasive spheroidsare formed. Remarkably, the presence of CAFs inhibitsSOX2-induced dysplasia and restores an acinar-likephenotype, but TUM622 cells appear to exhibitepithelial-mesenchymal transition (EMT) at the invasivefront towards CAFs, thereby forming “teardrop”-shapedstructures [62, 63]. Indeed, CAF-secreted stromal cell-derived factor-1 (SDF-1) promoted EMT and the acqui-sition of stemness in LSCCs [64]. Although the majorityof LSCCs were positive for E-cadherin and only a smallpopulation were positive for Vimentin and SOX2, thesefactors showed considerable heterogeneity in TUM622-derived spheroids [62]. Because there were cells positivefor both E-cadherin and Vimentin, it is likely that partialEMT occurs in spheroids, the PDX model and the ori-ginal tumor [62, 65].Single cancer cell migration, also known as mesenchy-
mal migration, is characterized by fibroblast-like morph-ology, but effective metastasis of cancer cells can occur
without complete loss of epithelial morphology orcomplete acquisition of mesenchymal morphology. Can-cer cells undergoing mesenchymal migration areenriched at the invasive front in vivo, consistent withprevious findings that partial EMT is involved in collect-ive tumor migration [65, 66]. Leader cells expressingmesenchymal-like or basal epithelial traits are located atthe front of the follower epithelial cancer clusters, anddrive their collective migration in response to microenvi-ronmental cues. SOX2 appears to induce the commit-ment and differentiation of TUM622 cells to thesquamous lineage instead of regulating epithelial/mesen-chymal plasticity [62, 67]. SOX2 preferentially interactswith the transcription factor p63, as opposed to the tran-scription factor OCT4 in LSCCs, which is the preferredSOX2-binding partner in embryonic stem cells [68].FGFR1 accelerates tumor development without forcingcells toward a particular tumor subtype. By contrast,SOX2 appears to be critical in driving cells toward anaggressive and penetrant LSCCs phenotype [67, 69]. Fur-thermore, CAF-derived CD81-positive exosomesmobilize Wnt11 produced by breast carcinoma cells,thereby activating β-catenin-independent Wnt planarcell polarity (PCP) signaling, which promotes lung me-tastasis [70]. Chen et al. suggested that the β-catenin-dependent canonical Wnt signaling pathway and cancerstem cell marker SOX2 synergistically induce acinarmorphogenesis of TUM622 cells in three-dimensionalculture [62]. Further investigations are warranted to de-termine how SOX2 overexpression and co-culture withCAFs affects the β-catenin-independent PCP pathwaywhen spheroid cells derived from this PDX model oflung cancer acquire an invasive phenotype via partialEMT.
Malignant melanomaBoth intrinsic and acquired therapy resistance remains aserious challenge for the management of BRAF(V600E)-mutant malignant melanoma. Many resistant cases ex-hibit reactivation of MAPK and PI3K signaling in thepresence of MAPK-targeting agents. De novo lipogenesisis emerging as a central player in multiple oncogenicprocesses. Constitutive activation of the lipogenic path-way in tumor tissue is required for the synthesis of phos-pholipids, which function as essential building blocks ofmembranes and promote cell growth and proliferation.In vitro, a marked decrease in de novo lipogenesis wasobserved in all BRAF(V600E)-mutant therapy-sensitive,but not therapy-resistant, cell lines in the presence ofthe BRAF inhibitor vemurafenib [71]. Remarkably, BRAFinhibition induced only a moderate decrease in expres-sion of sterol regulatory element-binding protein-1(SREBP-1), a master regulator of lipid metabolism, anddid not significantly affect lipogenesis in therapy-
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 6 of 16
resistant melanoma cells. To assess the therapeutic po-tential of these findings, Talebi et al. investigated the im-pact of SREBP-1 inhibition in an in vivo pre-clinicalBRAF(V600E)-mutant melanoma model. Talebi et al. se-lected a PDX that responded poorly to BRAF inhibitorsalone [71, 72]. SREBP-1 contributes to the anti-tumorresponse induced by BRAF inhibition, and SREBP-1 in-hibition sensitizes therapy-resistant melanoma cells toMAPK-targeting therapy. In vivo analysis of oxidativestress revealed that while fatostatin or vemurafenib treat-ment alone did not significantly increase lipid peroxida-tion, combined vemurafenib/fatostatin treatment greatlyenhanced lipid peroxidation [71]. Notably, SREBP-1 in-hibition enhances the sensitivity to BRAF inhibitors in apre-clinical PDX model of melanoma [71].
Ovarian cancerChoi et al. established PDX models derived from serousadenocarcinoma and chemoresistant carcinosarcoma toinvestigate the therapeutic potential of itraconazole, anorally bioavailable anti-fungal drug that inhibits the en-zyme lanosterol 14α-demethylase [73]. Previous reportsidentified itraconazole as a potent antagonist of theHedgehog (Hh) signal pathway, which is different fromthe pathway involved in the inhibitory effect of this drugon fungal sterol biosynthesis [74]. Systemically adminis-tered itraconazole suppresses Hh pathway activity andmedulloblastoma growth in a mouse allograft modelsimilar to other Hh pathway antagonists, and surpris-ingly, itraconazole inhibits the Hh pathway at serumlevels comparable with those of patients undergoinganti-fungal therapy [74, 75]. This is a typical example ofdrug repurposing, in which the anti-cancer properties ofmedications otherwise administered for non-malignantdisorders serve to develop new treatments strategies forcancer [58]. Mechanistically, itraconazole antagonizesthe Hh pathway component Smoothened (Smo) througha mechanism distinct from that of cyclopamine andother known Smo antagonists, and prevents ciliary accu-mulation of Smo mediated by Hh stimulation. Both theHh and mammalian target of rapamycin (mTOR) signal-ing pathways are associated with angiogenesis in thetumor microenvironment. In ovarian cancer PDXmodels, addition of itraconazole to paclitaxel signifi-cantly enhanced therapeutic efficacy compared withpaclitaxel monotherapy. Expression of CD31 andVEGFR2 (angiogenesis markers), Gli1 (hedgehog signal-ing downstream molecule), and S6K1 (mTOR pathway)were all decreased in tumors treated with paclitaxel anditraconazole combination therapy [73]. Itraconazole hassynergistic effects when used in combination withpaclitaxel-based chemotherapy, which is one of the mosteffective available chemotherapeutics for ovarian cancer.
Table 2 provides the summary of the latest importantpapers using PDX models and organoids.
Future challenges of PDX modelsWhile the incorporation of PDX models in cancer re-search brings some exciting improvements, PDXs haveimportant limitations that must be addressed to improvethe availability of PDX models for translational researchand pre-clinical investigations, including “co-clinical tri-als.” Issues that must be addressed or standardized to fa-cilitate wide use of PDX models include (a) the site forimplantation of original tumor fragments, (b) timecourse for PDX tumor tissue generation, (c) the engraft-ment rate, (d) replacement of human stroma with mur-ine stroma, (e) failure to evaluate the immune system,and (f) the challenging problems of Matrigel.
Optimal implantation siteIt is important to define the best engraftment site (sub-cutaneous, subrenal capsule, or orthotopic implantation)in each tumor type of interest. Most of the publishedstudies using PDX models have relied on surgical speci-mens, which naturally provide large quantities of tumortissues. Although much effort has been expended in es-tablishing PDX models of cholangiocarcinoma and headand neck squamous cell carcinoma, acquiring a sufficientvolume for transplantation is prohibitive due to thesmall size of the original tumor. Thus, suitable methodswith smaller samples obtained by biopsy or fine-needleaspirations should be established.
Time course of PDX tumor tissue generationDelay between the engraftment period in mice and opti-mal treatment schedules for patients is a limiting factorfor the use of PDX models in real-time personalizedmedicine applications. Developing a PDX model for pre-clinical study generally requires 4–8 months with severaltumor passages, which is much longer than patients canordinarily wait to commence treatment. The second-and third-generation PDX passages require only 10 daysto form a palpable xenograft. Due to the time require-ment to generate PDX models, some groups are usingshort-term single-cell suspension and short-term culturein organoid models to evaluate sensitivity to potentialtreatments.
Tumor engraftment rateThe engraftment failure rate is still high for some cancertypes and phenotypes, which is a major obstacle to wide-spread PDX use. It is thus essential to improve tumorengraftment rates to an acceptable level, up to 60–70%.Most importantly, patients with breast and renal cancerwhose tumors successfully engraft have the worst prog-nosis [4, 5, 103], which strongly suggests a selective
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 7 of 16
Table
2Recent
investigations
ofsolid
tumorsusingPD
Xmod
els
Tumor
type
Reference
Animal
Site
Metastasis
Colon
aden
ocarcino
ma
Misaleet
al.[12]
NOD-SCID
mice
Subcutaneo
usLiver
PDXmod
elsde
rived
from
aqu
adruplewild-type(KRA
S,NRA
S,BRAF,and
PIK3CA
)colorectaltumor
aresensitive
toEG
FRblockade
alon
e.EG
FR/M
EKcombinatio
nblockade
issupe
riorto
treatm
entwith
anti-EG
FRantib
odies,or
tode
liveringaMEK
inhibitorwhe
nresistance
tocetuximab
orpanitumum
abhasalreadyde
velope
d.
Tumor
type
Reference
Animal
Site
Metastasis
Colon
aden
ocarcino
ma
Fujiiet
al.[24]
NOGmice
Subren
alcapsuleand
spleen
Liver
Fujiiet
al.establishe
dacolorectaltumor
organo
idlibrary
comprised
of55
organo
idsde
rived
from
52tumorsand43
patients,includ
inghype
rplasticpo
lyps
andsessile
serrated
aden
oma/po
lyps.Patho
histolog
ical
subtypes
aswellasdifferentiatio
nhierarchiesof
CRC
swerecell-intrinsically
conservedregardless
ofen
vironm
ent(i.e.,inpatients,in
vitro,
andin
PDXmod
els).PDXmod
elsestablishe
dby
transplantationinto
the
kidn
eysubcapsulesof
immun
ocom
prom
ised
micede
velope
dthesize
ofen
graftedsubren
alCRC
sthat
secretenichefactors,includ
ingp3
8-MAPK,TGF-β,
andEG
F.In
contrastto
therobu
sten
graftm
entefficiency
insubren
alcapsules,the
metastatic
capacity
ofspleen
-injected
CRC
organo
idswas
diverse.
Tumor
type
Reference
Animal
Site
Metastasis
Colon
aden
ocarcino
ma
Fior
etal.[17]
Zebrafish
Subcutaneo
usNon
e
Fior
etal.g
enerated
andtreatedfivezebrafish-basedPD
X(zPD
X)mod
elsof
coloncancer
derived
from
different
patients,andtreatedzPDXs
with
theFO
LFOXregimen
over
3days.TwoPD
Xsrespon
dedto
treat-
men
t,as
indicatedby
increasedcaspase3cleavage
.The
setw
osensitive
zPDXs
correspo
nded
topatientsin
who
mCEA
levelsremaine
dstable6mon
thsaftersurgerywith
outrelapse.Con
trastin
gly,am
ongthe
threezPDXs
inwhich
FOLFOXwas
noteffective,thecorrespo
ndingpatientsde
velope
dincreasing
CEA
levelsandclinicaleviden
ceof
relapse.Furthe
rmore,Fior
etal.investig
ated
thepred
ictiveeffect
oftheEG
FRinhibitorCetuxim
abin
combinatio
nwith
FOLFIRIreg
imen
,finding
that
resistantzPDXmod
elsarede
rived
from
tumorswith
mutations
ineither
BRAF
orKRAS.
Tumor
type
Reference
Animal
Site
Metastasis
Colon
aden
ocarcino
ma
Okazawaet
al.[56]
NOGmice
Ortho
topic
Lung
s
Pieces
ofresected
CRC
sweresubcutaneo
uslyim
plantedinto
NOGmiceto
gene
rate
thePD
Xmod
el.The
organo
idcells
werethen
extractedfro
mthePD
Xmod
elfortissuecultu
re,and
CRC
organo
idswere
infected
with
GFP
lentiviru
s,allowinghigh
lysensitive
visualizationof
micrometastases(Fig.2).Notably,lun
gmicrometastaseswerede
tected
inthreeou
tof
four
miceexam
ined
2.5mon
thsafterorthotop
icinjectionof
GFP-labe
lledPD
X-de
rived
organo
ids.Theim
plantatio
nof
organo
idsinto
therectalsubm
ucosaof
NOGmiceresultedin
metastatic
dissem
inationinto
thelung
s,bu
tno
ttheliver,p
resumablythroug
htheinferio
rvena
cava.
Tumor
type
Reference
Animal
Site
Metastasis
Pancreaticcancer
Zhou
etal.[76]
Nud
emice,SC
IDmice
Ortho
topic
Non
e
Becauseinsulin
grow
thfactor
1receptor
(IGF1R)
ishigh
lyexpressedin
both
pancreaticcancer
cells
andstromalfib
roblasts,Z
houet
al.d
evelop
ednano
particleswith
recombinant
human
IGF1
conjug
ated
tomagne
ticiro
noxidecarrying
anthracyclinedo
xorubicin(IG
F1-IO
NP-Dox),andde
mon
stratedan
enhanced
therapeutic
effect
comparedwith
conven
tionalD
oxtreatm
entin
theorthotop
icpancreaticdu
ctal
aden
ocarcino
ma(PDAC)PD
Xmod
el.T2-weigh
tedmagne
ticresonanceim
aging(M
RI)revealed
system
icde
liveryof
IGF1R-targeted
Dox
afteradministrationof
IGF1-IO
NP-Dox.N
otably,n
on-spe
cific
uptake
ofIGF1-
IONP-Dox
inthespleen
didno
tcauseapop
tosis,as
demon
stratedby
lack
ofactivecaspase3.
Tumor
type
Reference
Animal
Site
Metastasis
Pancreaticcancer
Witkiewiczet
al.[77]
NSG
mice
Subcutaneo
usNon
e
PDXmod
elsof
PDACen
ableprecisiontherapy,as
resistance
toMEK
inhibitorsisparado
xically
associated
with
compe
nsatoryAkt
sign
alingactivation[78].W
itkiewiczet
al.usedthismod
elto
demon
strate
that
combinatio
ntherapywith
tram
etinib
(MEK
inhibitor)anddasatin
ib(tyrosinekinase
Srcinhibitor)sign
ificantlysupp
resses
PDXtumor
proliferatio
n.Furthe
rmore,theBcl-2
inhibitorABT737andcheckpoint
kinase
in-
hibitorAZD
7762
indu
cesatherapeutic
respon
sein
coop
erationwith
conven
tionalanti-cancerdrug
ssuch
asdo
cetaxeland
gemcitabine
.
Tumor
type
Reference
Animal
Site
Metastasis
Pancreaticcancer
Rajeshkumar
etal.[79]
Nud
emice
Subcutaneo
usNon
e
PDXmod
elsof
PDACrespon
dmorerobu
stlyto
mito
chon
drialcom
plex
Iinh
ibito
rs(phe
nformin
andmetform
in)than
toothe
rmetabolicinhibitors,including
aglutam
inaseinhibitor,atransaminaseinhibitor,and
anautoph
agyinhibitor.Aminoacidsandmetabolitesinvolved
inglycolysissuch
aslactatearede
creasedby
complex
Iinh
ibito
rs,w
hileoxidized
glutathion
eisincreased.
Thereisno
correlationbe
tween
phen
form
inrespon
seandge
netic
abno
rmalities
inTP53,PTEN,SMAD
4,andKRAS.A
lthou
ghph
enform
inhasno
tbe
enappliedclinicallydu
eto
theriskof
lacticacidosis[80],thisstud
ysugg
eststheclinicalprom
ise
ofthisbigu
anideagen
t.
Tumor
type
Reference
Animal
Site
Metastasis
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 8 of 16
Table
2Recent
investigations
ofsolid
tumorsusingPD
Xmod
els(Con
tinued)
Cho
lang
iocarcinom
aGarciaet
al.[14]
SCID
mice
Subcutaneo
usNon
e
Five
PDXmod
elsof
cholangiocarcino
maexhibitediden
ticalKRAS
mutations
tothoseof
thetumorsfro
mwhich
they
werede
rived
.Ind
eed,
theon
coge
nicKRAS
mutationfre
quen
tlyoccursas
apo
intmutationin
codo
n12,and
thesemutations
resultin
constitutiveactivationof
thePI3K
pathway
andRA
S/RA
F/MEK/ERK
axis,w
hich
prom
otecancer
cellsurvivalandproliferatio
n[81].C
ell-cycleregu
latory
proteins,including
Chk1andE2F1,w
ereselectivelydo
wn-regu
latedby
BETproteininhibitorJQ
1in
JQ1-sensitive
PDXmod
els.WhileJQ
1failedto
affect
c-Myc
expression
inJQ
1-insensitive
PDXmod
els,aBETinhibitordo
wn-
regu
latedc-Myc
inJQ
1-sensitive
PDXtumors,which
was
accompanied
bydo
wn-regu
latio
nof
downstream
transcrip
tionaltarge
tssuch
asChk1andE2F1.
Tumor
type
Reference
Animal
Site
Metastasis
Glioblastomamultiforme
Leeet
al.[82]
Nud
emice
Ortho
topic
Non
e
Leeat
al.d
evelop
edGBM
recurren
tPD
Xmod
elsindu
cedby
temozolom
ide,in
which
theim
med
iate
peak
stabilizatio
nof
HIF1α
inPD
Xcells
afterexpo
sure
tochem
othe
rapy
isexpe
cted
torepresen
tan
essential
step
intheconversion
ofno
n-cancer
stem
cells
into
undifferentiatedcancer
stem
cells
(CSC
s).Inbo
thGBM
6(classical,M
GMThype
rmethylated),and
GBM
43(prone
ural,M
GMTun
methylated)
PDXmod
els,HIF1α
levelswereincreasedin
theCD133-po
sitiveCSC
popu
latio
nbo
thpo
st-the
rapy
andat
diseaserecurren
ce.
Tumor
type
Reference
Animal
Site
Metastasis
Glioblastomamultiforme
Sing
het
al.[59]
NOD-SCID
mice
Ortho
topic
Non
e
Ectopicco-expressionof
SOX2
,OLIG2,andZEB1
transformstumor-sup
pressor-de
ficient
astrocytes
into
glioma-initiatingcells
intheabsenceof
anup
stream
RTKon
coge
ne.A
mon
gthethreetranscrip
tionfactors,
SOX2
expression
issign
ificantlyup
regu
latedin
PDXGBM
s.Histone
H3lysine
27residu
e(H3K27)acetylationwas
presen
tin
morethan
90%
ofthePD
XSO
X2bind
ingregion
sin
thethreeanalyzed
patient
GBM
specim
ens,which
indicatestheseregion
sas
activecis-regu
latory
elem
entsin
patient
GBM
.
Tumor
type
Reference
Animal
Site
Metastasis
Glioma
Fack
etal.[83]
NOD-SCID
mice
Ortho
topic
Non
e
Fack
etal.app
liedin
situ
metabolicprofiling
andLC
-MSon
brainsections
ofgliomaPD
Xandhu
man
gliomasamples
with
andwith
outisocitratede
hydrog
enase1(ID
H1)
mutations.M
assspectrom
etry
imaging
(MSI)andLC
-MSanalysisof
orthotop
icIDH-m
utated
gliomaPD
Xmod
elsrevealed
IDH-spe
cific
adaptivemechanism
sin
metabolicpathways.Notably,cystathionine
-β-synthaseexpression
isano
velp
rogn
ostic
fac-
torin
theoligod
endrog
lialg
liomasubtype.
Tumor
type
Reference
Animal
Site
Metastasis
Med
ulloblastoma
Garne
ret
al.[84]
Nud
emice
Ortho
topic
Non
e
Threegrou
p3med
ulloblastomaPD
Xmod
elswereestablishe
dto
evaluate
thetherapeutic
efficacyof
FTY720,anim
mun
osup
pressant
that
hascurren
tlybe
enapproved
fortreatm
entof
multip
lesclerosis[85].
FTY720
activates
thetumor
supp
ressor
proteinph
osph
atase2A
,and
thus
treatm
entof
human
med
ulloblastomaPD
Xcells
with
FTY720
arrested
thecellcycleat
theG1ph
aseandindu
cedcaspase-de
pend
ent
apop
totic
cellde
ath.
Tumor
type
Reference
Animal
Site
Metastasis
Breastcarcinom
aLawsonet
al.[8]
NOD-SCID
mice
Ortho
topic
Lung
s,bo
nemarrow,liver,b
rain
Inthisstud
y,themostmetastatic
PDXmod
el(HCI-010)exhibitedthehigh
estpe
rcen
tage
ofbasal/stem-like
tumor
cells,w
hiletheleastmetastatic
mod
el(HCI-002)hadthelowest.Thissugg
eststhat
prim
ary
tumorscontainarare
subp
opulationof
stem
-like
cells,and
that
therelativeabun
danceof
thesecells
correlates
with
metastatic
potential.Highe
r-bu
rden
metastatic
cells
enteredthecellcycle,expressing
lower
levelsof
quiescen
ceanddo
rmancy-associatedge
nesandhigh
erlevelsof
cell-cycle-prom
otingge
nes,includ
ingCD24,C
DK2,M
MP1,and
MYC
,which
have
been
associated
with
reactivationafterdo
rmancy.D
inaci-
clib,a
CDKinhibitorthat
indu
cesapop
tosisin
high
MYC
-expressingcancer
cells
viasynthe
ticlethality
[86,87],sign
ificantlyinhibitsthemetastatic
potentialo
fPD
Xmod
elsde
rived
from
drug
-naïve
patients.
Tumor
type
Reference
Animal
Site
Metastasis
Breastcarcinom
aEvanset
al.[13]
Nud
emice
Ortho
topic
Non
e
Morethan
25PD
Xmod
elsde
rived
from
triple-neg
ativebreastcancer
(TNBC
)tissues
variedin
theextent
ofPI3K
andMAPK
activation.PD
Xswerealso
heteroge
neou
sin
theirsensitivity
tochem
othe
rape
uticagen
ts;
whilePI3K,m
TOR,andMEK
inhibitorsrepressedgrow
thbu
tdidno
tcausetumor
regression
,the
PARP
inhibitortalazoparib
caused
drastic
regression
infiveof
12PD
Xs.N
otably,fou
rou
tof
fivetalazoparib
-sensitive
mod
elsdidno
tharbor
germ
lineBRCA
1/2mutations,b
utseveralh
adsomaticalteratio
nsin
homolog
ousrepairpathways,includ
ingATM
deletio
nandBRCA
2alteratio
ns.
Tumor
type
Reference
Animal
Site
Metastasis
Breastcarcinom
aYu
etal.[88]
NOD-SCID
mice
Subcutaneo
usNon
e
Toelucidatethemechanism
of5-Aza-2'-d
eoxycytid
ine(decitabine
)actio
non
TNBC
,Yuet
al.investig
ated
DNAmethyltransferases
(DNMTs)expression
levels,w
hich
werecorrelated
with
respon
seto
decitabine
inbreastcancer
organo
idmod
elsbasedon
PDXtumorsde
rived
from
TNBC
patientsrecruitedto
aprospe
ctivene
oadjuvantstud
yof
anthracycline-
andtaxane
-based
chem
othe
rapy.O
rganoids
andPD
Xmod
elsex-
pressing
elevated
levelsof
DNMTproteins
weremoresensitive
tode
citabine
than
thoseexpressing
low
levelsof
DNMTs,sug
gestingthat
DNMTmay
beapred
ictivebiom
arkerfortreatm
entrespon
seto
decitabine
inTN
BC.Thiseffect
was
med
iatedby
decitabine
-indu
cedub
iquitin
ationandlysosomalde
gradationof
allthree
DNMTs,w
hich
was
depe
nden
ton
theE3
ligaseTN
Freceptor-associatedfactor
6(TRA
F6).Decitabine
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 9 of 16
Table
2Recent
investigations
ofsolid
tumorsusingPD
Xmod
els(Con
tinued)
treatm
entof
PDXtissues
also
increasedTRAF6transcrip
tion.
Tumor
type
Reference
Animal
Site
Metastasis
Breastcarcinom
aXiao
etal.[9]
NSG
mice
Subcutaneo
usNon
e
p21protein–
activated
kinase
2(PAK2)isactivated
bylow
C-terminalSRCkinase
levels,w
hich
drives
estrog
en-in
depe
nden
ttumor
grow
thin
patientswith
estrog
enreceptor
(ER)–p
ositive
breastcancer
resistantto
endo
crinetherapy.Using
aPD
Xmod
elde
rived
from
ER(+)/Pg
R(+)/HER2(-)invasive
ductalcarcinom
a,Xiao
etal.con
firmed
that
combinatio
ntreatm
entwith
thePA
K2inhibitorFRAX5
97andan
ERantago
nistsyne
r-gisticallysupp
ressed
breasttumor
grow
th.
Tumor
type
Reference
Animal
Site
Metastasis
Breastcarcinom
aIkbaleet
al.[89]
NSG
mice
Ortho
topic
Lymph
node
s,visceralmetastasis
Com
paredwith
treatm
ent-naïveTN
BC–d
erived
PDX,
chem
oresistant
TNBC
–derived
PDXhigh
lyexpressedWnt10B-relatedmolecules,including
non-ph
osph
orylated
activeβ-catenin,
Axin2
,CD44,and
HMGA2.ICG-
001,acano
nicalW
ntsign
alinginhibitor,redu
cedtumor
grow
thandlymph
node
metastatic
burden
.The
combinatio
nof
doxorubicinandICG-001
efficientlyrepressedlung
dissem
inationaftertailvein
injectionof
tumor
cells
that
haddissociatedfro
mchem
oresistant
TNBC
PDX.
Thissyne
rgistic
effect
was
med
iatedby
Bcl-2
downreg
ulation.
Tumor
type
Reference
Animal
Site
Metastasis
Lung
cancer
Weede
net
al.[90]
NSG
mice
Subcutaneo
usNon
e
PDXmod
elsde
rived
from
lung
squamou
scellcarcinom
aen
abledWeede
net
al.toiden
tifyFG
FR1mRN
Alevelrathe
rthan
FGFR1am
plificatio
nas
aprecisepred
ictorof
respon
seto
anFG
FRtyrosine
kinase
inhibitor.Seventeenpe
rcen
tof
PDXs
evaluated(sixou
tof
36cases)exhibitedFG
FR1am
plificatio
n,andFG
FRinhibitio
nde
creasedcellproliferatio
nanden
hanced
differentiatio
n,which
decreasedtumor
grow
thrate
andon
lymod
estly
increasedapop
totic
cellde
ath.
Treatm
entwith
theFG
FRinhibitorBG
J398
blockedbo
thAKT
andERKsign
aling,
andcombinatio
ntherapywith
BGJ398
andaPI3K
inhibitorwas
bene
ficial
onlyin
PDXs
inwhich
FGFR
inhibitio
ndidno
talterPI3K
sign
aling.
Tumor
type
Reference
Animal
Site
Metastasis
Lung
cancer
Drapkin
etal.[91]
NSG
mice
Subcutaneo
usNon
e
Becausecollectionof
circulatingtumor
cells
(CTC
s)en
ablesno
n-invasive
serialtum
orsampling,
Drapkin
etal.establishe
dmorethan
30PD
Xmod
elsof
smallcelllun
gcancer
(SCLC
)de
rived
from
noton
lybiop
sytissues
butalso
from
CTC
s.SC
LCPD
Xmod
elsretain
stablege
nomesomaticalteratio
nsbe
tweeninitialPD
Xmod
elge
neratio
nandserialp
assage
s.Surprisingly,CTC
-derived
PDXmod
elsreflect
theevolving
chem
o-therapysensitivitiesof
theoriginaltumor
atthetim
eof
CTC
collection.
Etop
osideresistance
ispo
sitivelycorrelated
with
activationof
theMyc-associatedge
nesign
ature,which
isconsistent
with
results
obtained
usingthege
neticallyen
gine
ered
mou
semod
elof
SCLC
[92,93].In
comparison
toothe
rmalignancies,SC
LCPD
Xmod
elsexhibitincreasedclon
alho
mog
eneity
andge
nomicstability.
Tumor
type
Reference
Animal
Site
Metastasis
Lung
cancer
Gon
get
al.[94]
NOD-SCID
mice
Subcutaneo
usNon
e
Inge
neral,patientswith
wild
type
EGFR
(EGFRwt)do
notrespon
dto
treatm
entwith
EGFR
tyrosine
kinase
inhibitors(TKIs),w
hilemostpatientswith
EGFR
activatingmutations
initiallyrespon
dto
EGFR
TKIs,b
utinevitablyde
velopsecond
aryresistance
toTKItreatmen
t.Simultane
ousinhibitio
nof
EGFR
andTN
Fpreven
tsde
velopm
entof
thisacqu
iredEG
FRresistance.N
otably,com
binatio
ntreatm
entwith
EGFR
TKI
(erlo
tinib)plus
thalidom
idewas
high
lyeffectivein
inhibitin
gtumor
grow
thin
anEG
FRwtPD
Xmod
el,w
hileEG
FRinhibitio
nor
thalidom
idealon
ewas
ineffective.
Tumor
type
Reference
Animal
Site
Metastasis
Lung
cancer
Che
net
al.[62]
Nud
emice
Subcutaneo
usNon
e
TheTU
M622celllineestablishe
dfro
mPD
Xsof
lung
squamou
scellcarcinom
ahasincreasedsphe
roid-fo
rmingcapacity
dueto
overexpression
ofthestem
cellfactor
SOX2
,and
show
sahype
rplasticto
dysplastic
change
inits
acinar
phen
otype,in
which
apical-basalcellpo
larityisdisrup
ted.
Wnt/β-caten
insign
alingandSO
X2,w
hich
contrib
uteto
norm
allung
developm
ent,areinvolved
inacinar
morph
ogen
esisof
TUM622
cells
inthree-dimen
sion
alcultu
res.Che
net
al.rep
orteden
hanced
epith
elial/m
esen
chym
alplasticity
ofTU
M622cells
afterincorporatingcancer-associatedfib
roblasts(CAFs)into
asphe
roid
cultu
resystem
,and
expand
ingthissystem
tomod
elno
ton
lycancer
cellEC
M,b
utalso
cancer
cell-CAFinteractions
durin
gtumorigen
esis.C
AFs
antago
nizedon
coge
nicSO
X2to
restoretheform
ationof
luminalstructures
andpro-
moteinvasion
.
Tumor
type
Reference
Animal
Site
Metastasis
Malignant
melanom
aHirata
etal.[95]
NSG
mice
Subcutaneo
usLung
s
Intravitalimagingof
BRAF-m
utantmelanom
acells
containing
anERK/MAPK
biosen
sorrevealed
how
thetumor
microen
vironm
entaffectsrespon
sesto
BRAFinhibitio
nby
PLX4
720.Inde
ed,BRA
F-mutantmelanom
acells
respon
dto
PLX4
720he
teroge
neou
sly,andBRAFinhibitio
nactivates
CAFs,leading
toFA
K-de
pend
entmelanom
asurvivalsign
aling.
Fibron
ectin
-richmatrices
with
3-12
kPaelastic
mod
ulus
aresufficien
tto
in-
duce
PLX4
720tolerance.ThePD
Xmod
elrevealed
that
co-in
hibitio
nof
BRAFandFA
Kabolishe
sERKreactivationin
tumor
stroma,leadingto
moreeffectivecontrolo
fBRAF-m
utantmelanom
a.
Tumor
type
Reference
Animal
Site
Metastasis
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 10 of 16
Table
2Recent
investigations
ofsolid
tumorsusingPD
Xmod
els(Con
tinued)
Malignant
melanom
aKrep
leret
al.[96]
NSG
mice
Subcutaneo
usBrain,
lung
s
Tofacilitatetheadvancem
entof
pre-clinicalin
vivo
mod
eling,
Krep
leret
al.establishe
dmorethan
450PD
Xmod
els.Half(55%)of
allanalyzedsamples
wereBRAF
hotspo
tmutants,20%
wereNRA
Smutants,7%
wereNF1
mutants,2%
wereKITmutants,1.4%
wereGNAQ
/GNA11mutants,and
18%
wereWT,allo
fwhich
wereconsistent
with
theTC
GAdatabase.The
simultane
ousinhibitio
nof
MDM2andERKishigh
lyeffect-
ivein
PDXs
refractoryto
BRAFinhibitors.
Tumor
type
Reference
Animal
Site
Metastasis
Malignant
melanom
aTalebi
etal.[71]
Nud
emice
Subcutaneo
usNon
e
SREBP-1protectsvemurafen
ib-resistant
melanom
acells
from
lipid
peroxidatio
n.Fatostatin
treatm
entalon
eof
PDXMEL006inhibitstumor
grow
thmorepo
tentlythan
vemurafen
ib.Impo
rtantly,com
bine
dvemura-
fenib/fatostatin
co-treatmen
thasagreateranti-tumor
effect
than
either
mon
othe
rapy
regimen
.
Tumor
type
Reference
Animal
Site
Metastasis
Malignant
melanom
aEinarsdo
ttiret
al.[97]
NOGmice
Subcutaneo
usNon
e
PDXmod
elsde
rived
from
metastatic
melanom
awereestablishe
dto
assess
heteroge
neou
srespon
sesaftertreatin
gtumorswith
karonu
dib,
which
inhibitstheoxidized
nucleo
tide-sanitizingen
zymeMTH
1[98].
Com
parison
ofthemutationprofile
betw
eenrespon
segrou
psrevealed
that
karonu
dibhasacytotoxiceffect
inmelanom
aPD
Xmod
els,irrespe
ctiveof
themutationstatuses
ofthemostcommon
driver
gene
sin
melanom
a.Im
portantly,h
ighexpression
ofABC
B1isapo
tentialresistancebiom
arker.
Tumor
type
Reference
Animal
Site
Metastasis
Headandne
cksquamou
scellcarcinom
a(HNSC
C)
Grasset
etal.[99]
NMRI-nud
emice
Subcutaneo
usNon
e
ThePD
Xmod
elrevealed
that
theLO
XinhibitorBA
PNattenu
ated
ECM
remod
elingandtumor
stiffne
sswith
collage
nIb
undles
byredu
cing
EGFR
tyrosine
kinase
activity
ofcancer
cells.The
Ca2
+channe
lblockers
phen
ylalkylamineverapamilandno
ndihydropyrid
inediltiazem
,usedforthetreatm
entof
hype
rten
sion
andarrhythm
iaforde
cade
s,areeffectivein
PDXmod
elsforthepreven
tionof
collectivetumor
cellinvasion
bysign
ificantlydo
wnreg
ulatingph
osph
orylated
myosinlight
chain2.Mechanistically,tum
or-derived
ECM
stiffne
ssandactivated
EGFR
sign
alingen
hances
intracellularCa2
+concen
trationmed
iatedby
theL-type
Ca2
+channe
lCav1.1in
squamou
scancer
cells.
Tumor
type
Reference
Animal
Site
Metastasis
Ovariancancer
Liuet
al.[100]
Nud
emice,NSG
mice
Intraperito
nealspace
Periton
eald
isseminationsuch
asom
entum,liver,p
ancreas,bo
wel,spleenand
diaphragm
Liuet
al.establishe
dPD
Xmod
elsof
ovariancarcinom
ade
rived
from
floatingtumor
cells
intheascitesof
irradiatednu
demice,andascitesin
establishe
dPD
Xmod
elswerethen
implantedintraperito
neallyinto
NSG
mice.Freshascites-de
rived
tumor
cells
from
thesePD
Xtumor-bearin
gNSG
miceweretransfectedwith
lentiviru
sen
coding
fireflyluciferase
andmChe
rryforbiolum
inescenceim
aging(BLI)of
PDXmod
els.Co-
hortsof
NSG
micewith
luciferized
PDXtumorsweretreatedwith
carbop
latin
,orpaclitaxel,either
asamon
othe
rapy
orin
combinatio
n,followed
byweeklyBLIm
easuremen
ts,the
results
ofwhich
positivelycorre-
latedwith
thoseof
plasmaCA125andcell-fre
eDNAassays.N
otably,PDXmod
elsde
rived
from
platinum
-refractorypatientsde
mon
stratedsign
ificant
resistance
tocarbop
latin
.
Tumor
type
Reference
Animal
Site
Metastasis
Ovariancancer
Kim
etal.[101]
NSG
mice
Ortho
topic
Non
e
Kim
etal.investig
ated
thetherapeutic
effectsof
aPA
RPinhibitor(PARPi;olaparib),an
ATR
inhibitor(ATRi;AZD
6738),andaCHK1
inhibitor(CHK1i;MK8776)
forthetreatm
entof
BRCA
-mutanthigh
-grade
serous
ovariancancer
(HGSO
C)cells.A
lthou
ghPA
RPiand
CHKias
sing
leagen
tsmod
estly
supp
ressed
tumor
grow
th,the
additio
nof
ATRi/C
HK1itoPA
RPiinBRCA
-mutantovarianPD
Xmod
elssign
ificantlyde
creased
tumor
volumecomparedwith
sing
le-age
nttherapies.Remarkably,sign
ificant
differences
wereob
served
inrespon
sivene
ssto
thePA
RPi-A
TRiand
PARP
i-CHK1icom
binatio
ns.W
hilePA
RPi-C
HK1icom
binatio
ntreat-
men
tsign
ificantlyincreasedPD
Xtumor
supp
ressionover
sing
le-age
nttreatm
ents,PARPi-A
TRicom
binatio
ntreatm
entsign
ificantlyincreasedtheincide
nceof
PDXtumor
regression
.
Tumor
type
Reference
Animal
Site
Metastasis
Ovariancancer
Geo
rgeet
al.[18]
NSG
mice
Ortho
topic
Diaph
ragm
PDXmod
elsde
rived
from
BRCA
-mutantHGSO
Cspatholog
icallyexhibitedsolid,p
seud
oend
ometrio
id,and
transitio
nalcellcarcino
mas,referredto
asSETfeatures.Because
approxim
ately50%
ofHGSO
Cpatients
have
defectsin
homolog
ousrecombinatio
n(i.e.,loss-of-fu
nctio
nalm
utations
ofBRCA
1/2)
[102],blocking
cellcyclecheckpointscouldpo
tentially
indu
cesynthe
ticlethality
inHGSO
C.Ind
eed,
PDXmod
elsof
BRCA
2-mutantHGSO
Cexhibitedhigh
erlevelsof
phosph
orylated
CHK1
than
BRCA
-intact
HGSO
C.PET
imagingof
PARP-1
with
thePA
RPinhibitoranalog
ue[18 F]FTTwas
perfo
rmed
inPD
Xmod
els,which
revealed
that
PARP
inhibitortreatm
entresultedin
tumor
supp
ressionbu
tno
tcompletetumor
regression
,sim
ilarto
therespon
seob
served
inclinicalsettings.
Tumor
type
Reference
Animal
Site
Metastasis
Ovariancancer
Cho
ietal.[73]
Nud
emice
Subren
alcapsule
Non
e
Cho
ietal.examined
thetherapeutic
effect
oftheantifun
galitracon
azolein
PDXmod
elsof
serous
aden
ocarcino
maandcarcinosarcoma.Com
binatio
ntherapywith
paclitaxeland
itracon
azolesign
ificantly
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 11 of 16
Table
2Recent
investigations
ofsolid
tumorsusingPD
Xmod
els(Con
tinued)
inhibitedtumor
grow
thandsupp
ressed
angiog
enesis.C
ombinatio
ntherapywas
moreeffectivein
decreasing
CD31,VEG
Freceptor,G
li1-m
ediatedHh,
andmTO
Rsign
alingthan
paclitaxelm
onothe
rapy.
Tumor
type
Reference
Animal
Site
Metastasis
Ovariancancer
Kond
rashovaet
al.(2018)
[49]
NOGmice
Subcutaneo
usand
orthotop
icNon
e
Therespon
seof
PDXmod
elsto
thePA
RPinhibitorrucaparib
largelyde
pend
son
BRCA
1prom
oter
methylatio
nzygo
sity.PDXmod
elswerede
rived
from
12HGSO
Cpatients,tenof
who
mwerechem
othe
rapy-
naïve,andtw
oof
who
mhadreceived
multip
lelines
oftherapy.Ahe
terozygo
ustherapeutic
respon
seto
rucaparib
was
observed
amon
gBRCA
1/2-mutantHGSO
CPD
Xmod
els.Ko
ndrashovaet
al.o
bservedaltered
zygo
sity
ofBRCA
1methylatio
n,characterized
byho
mozygosity
inthechem
o-naïvepatient-derived
samples
andhe
terozygo
sity
inthepreviouslytreatedHGSO
Cpatient–d
erived
samples.Lossof
homozygou
sBRCA
1methylatio
nislikelyrespon
siblein
partforacqu
iredrucaparib
resistance.
Tumor
type
Reference
Animal
Site
Metastasis
Bladde
rcancer
Leeet
al.(2018)[50]
NOGmice
Ortho
topic
Muscleinvasion
Leeet
al.d
evelop
edan
optim
ized
metho
dology
toconvertbladde
rtumor
organo
idlines
into
orthotop
icPD
Xmod
elswith
upto
80%
efficiencyusingultrasou
nd-guide
dim
plantatio
nof
organo
idsbe
tweenthe
bladde
rurothe
lium
andlaminaprop
ria.Tum
orevolutioniscommon
inorgano
ids,even
intheabsenceof
drug
treatm
ent.Patient-derived
bladde
rcancer
organo
idsmaintaine
dhe
teroge
neity,leading
toclon
alevolution.
Notably,the
basalo
rganoidph
enotypereversiblychange
dinto
theluminalph
enotypein
PDXmod
els,likelydu
eto
cellularplasticity.PDXmod
elsretain
drug
respon
sesto
tram
etinib
(MEK
inhibitor)
andge
mcitabine
.
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 12 of 16
pressure towards more aggressive phenotypes. It is pos-sible that as the passages of the PDX model proceed,proliferative, and highly metastatic clones are selected toestablish the next generation of PDXs.
Replacement of human stroma with murine stromaIntratumoral heterogeneity can be influenced by tumor-extrinsic factors in the microenvironment, includingmurine host cells. After 3–5 passages, when PDX modelscan be used for drug screening, tumor-associated stromaare almost completely replaced by murine-derived ECMand fibroblasts. This new murine stroma is likely tocause drastic changes in paracrine regulation of thetumor as well as in physical properties such as intersti-tial fluid pressure, which disturb drug distribution [104].Importantly, some cytokines are species-specific. Asstated previously, human-derived IL-2 stimulates theproliferation of murine T cells at similar concentrations,whereas mouse IL-2 stimulates human T cells at a sig-nificantly lower efficacy [45]. Furthermore, the T cellstimulating potential of IL-4 seems to be species specific.IL-15 binds to human or mouse IL-15 receptor α withequal high affinity [47]. Surely human-derived IL-15 canfunction on murine cells; human NK cells are poorlysensitive to murine IL-15 [48].
Failure to evaluate the immune systemOne key requirement in establishing PDX models is theneed to use immunodeficient host strains for tumor en-graftment and propagation. These mice lack functionalelements of the immune system, so that foreign tumortissues cannot be rejected. For example, NSG mice lacknatural killer cells and both B and T lymphoid cells.Therefore, the contribution of the host immune systemto responses to conventional chemotherapeutics cannotbe assessed. Currently available PDX models are also un-able to accurately assess immunotherapies such as vac-cines and immune modulators, and drugs that activatethe anti-tumor immune system.
Disadvantages of MatrigelCellular interactions with the ECM profoundly alter notonly gene expression pattern but also tumor cell behav-ior [105]. Tight ECM regulation in the tumor micro-environment is lost during engraftment, and tissuearchitecture degrades as PDX model tissue undergoespassages. Furthermore, Matrigel is frequently used to in-crease the engraftment rate in PDX models, since thepresence of growth factors in Matrigel can favor engraft-ment of primary tumor cells. However, it should be rec-ognized that this is a murine basement membraneextract [106], and Matrigel from murine sources can bea source of infectious murine viruses. In fact, treatmentof primary cancer cells with lactate dehydrogenase-
elevating virus-contaminated Matrigel in early studiesmay account for the reported instances of contamin-ation, and for this reason, some groups produced theirown Matrigel in order to avoid potential viral contamin-ation [107].
ConclusionsPDX models mimic not only the pathohistological andgenetic/epigenetic features of original tumor tissues butalso therapeutic responses to anti-cancer treatments.Therefore, PDX models have the potential to predict in-dividual responses to drugs and treatments, thereby fa-cilitating personalized medicine. Furthermore, many ofthe mechanisms underlying acquired drug resistance inmany tumor types have been elucidated using PDXmodels. Although the tumor stroma, including CAFs, isreplaced by mouse stroma during xenograft passages,the heterogeneity of the tumor microenvironment andthe metastatic potential of the tumor cells are main-tained in both subcutaneous and orthotopic PDXmodels. Notably, biofluorescence imaging of PDXmodels derived from organoids is a highly sensitivemethod for detecting micrometastatic lesions. Furtherinvestigations are necessary to develop strategies toevaluate the effects of immune-checkpoint inhibitors,because PDX models can only be established in im-munocompromised mouse strains.
AbbreviationsCAFs: Cancer-associated fibroblasts; CRC: Colorectal adenocarcinoma;CTC: Circulating tumor cell; ECM: Extracellular matrix; EMT: Epithelial-mesenchymal transition; GBM: Glioblastoma multiforme; GFP: Greenfluorescent protein; HGSOC: High-grade serous ovarian cancer;Hh: Hedgehog; IDH: Isocitrate dehydrogenase; mTOR: Mammalian target ofrapamycin; PDAC: Pancreatic ductal adenocarcinoma; PDXs: Patient-derivedxenografts; RTKs: Receptor tyrosine kinases; SREBP-1: Sterol regulatoryelement-binding protein-1; TNBC: Triple-negative breast cancer
AcknowledgementsI would like to thank Dr. Orimo A. of Juntendo University and Dr. MotoharaT. of Kumamoto University for advice and useful comments that helpedfinalize this review article.
Author’s contributionsGJY searched the literature and wrote the manuscript. The author read andapproved the final manuscript.
FundingThis review article was financially supported by the Japan Society for thePromotion of Science (19K23898).
Availability of data and materialsThe datasets used and/or analyzed during the current study are availablefrom the corresponding author on reasonable request.
Ethics approval and consent to participateNot applicable.
Consent for publicationGJY agrees to the content of this review paper and to being thecorresponding author.
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 13 of 16
Competing interestsThere are no competing interests to be addressed.
Received: 17 October 2019 Accepted: 13 November 2019
References1. Kirschbaum A, Geisse NC, Sister TJ, Meyer LM. Effect of certain folic acid
antagonists on transplanted myeloid and lymphoid leukemias of the Fstrain of mice. Cancer Res. 1950;10(12):762–8.
2. Hutchinson L, Kirk R. High drug attrition rates--where are we going wrong?Nat Rev Clin Oncol. 2011;8(4):189–90.
3. Bertotti A, Migliardi G, Galimi F, Sassi F, Torti D, Isella C, Cora D, DiNicolantonio F, Buscarino M, Petti C, et al. A molecularly annotated platformof patient-derived xenografts (“xenopatients”) identifies HER2 as an effectivetherapeutic target in cetuximab-resistant colorectal cancer. Cancer Discov.2011;1(6):508–23.
4. DeRose YS, Wang G, Lin YC, Bernard PS, Buys SS, Ebbert MT, Factor R,Matsen C, Milash BA, Nelson E, et al. Tumor grafts derived from womenwith breast cancer authentically reflect tumor pathology, growth, metastasisand disease outcomes. Nat Med. 2011;17(11):1514–20.
5. Sivanand S, Pena-Llopis S, Zhao H, Kucejova B, Spence P, Pavia-Jimenez A,Yamasaki T, McBride DJ, Gillen J, Wolff NC et al: A validated tumorgraftmodel reveals activity of dovitinib against renal cell carcinoma. Sci TranslMed 2012, 4(137):137ra175.
6. Hidalgo M, Amant F, Biankin AV, Budinska E, Byrne AT, Caldas C, Clarke RB, de JongS, Jonkers J, Maelandsmo GM, et al. Patient-derived xenograft models: an emergingplatform for translational cancer research. Cancer Discov. 2014;4(9):998–1013.
7. Tentler JJ, Tan AC, Weekes CD, Jimeno A, Leong S, Pitts TM, Arcaroli JJ,Messersmith WA, Eckhardt SG. Patient-derived tumour xenografts as modelsfor oncology drug development. Nat Rev Clin Oncol. 2012;9(6):338–50.
8. Lawson DA, Bhakta NR, Kessenbrock K, Prummel KD, Yu Y, Takai K, Zhou A, EyobH, Balakrishnan S, Wang CY, et al. Single-cell analysis reveals a stem-cell programin human metastatic breast cancer cells. Nature. 2015;526(7571):131–5.
9. Xiao T, Li W, Wang X, Xu H, Yang J, Wu Q, Huang Y, Geradts J, Jiang P, Fei T,et al. Estrogen-regulated feedback loop limits the efficacy of estrogenreceptor-targeted breast cancer therapy. Proc Natl Acad Sci U S A. 2018;115(31):7869–78.
10. Reyal F, Guyader C, Decraene C, Lucchesi C, Auger N, Assayag F, De Plater L,Gentien D, Poupon MF, Cottu P, et al. Molecular profiling of patient-derivedbreast cancer xenografts. Breast Cancer Res. 2012;14(1):R11.
11. Zhao X, Liu Z, Yu L, Zhang Y, Baxter P, Voicu H, Gurusiddappa S, Luan J, SuJM, Leung HC, et al. Global gene expression profiling confirms themolecular fidelity of primary tumor-based orthotopic xenograft mousemodels of medulloblastoma. Neuro Oncol. 2012;14(5):574–83.
12. Misale S, Bozic I, Tong J, Peraza-Penton A, Lallo A, Baldi F, Lin KH, Truini M,Trusolino L, Bertotti A, et al. Vertical suppression of the EGFR pathway preventsonset of resistance in colorectal cancers. Nat Commun. 2015;6:8305.
13. Evans KW, Yuca E, Akcakanat A, Scott SM, Arango NP, Zheng X, Chen K,Tapia C, Tarco E, Eterovic AK, et al. A population of heterogeneous breastcancer patient-derived xenografts demonstrate broad activity of PARPinhibitor in BRCA1/2 wild-type tumors. Clin Cancer Res. 2017;23(21):6468–77.
14. Garcia PL, Miller AL, Gamblin TL, Council LN, Christein JD, Arnoletti JP, HeslinMJ, Reddy S, Richardson JH, Cui X, et al. JQ1 induces DNA damage andapoptosis, and inhibits tumor growth in a patient-derived xenograft modelof cholangiocarcinoma. Mol Cancer Ther. 2018;17(1):107–18.
15. Topp MD, Hartley L, Cook M, Heong V, Boehm E, McShane L, Pyman J,McNally O, Ananda S, Harrell M, et al. Molecular correlates of platinumresponse in human high-grade serous ovarian cancer patient-derivedxenografts. Mol Oncol. 2014;8(3):656–68.
16. Nunes M, Vrignaud P, Vacher S, Richon S, Lievre A, Cacheux W, Weiswald LB,Massonnet G, Chateau-Joubert S, Nicolas A, et al. Evaluating patient-derivedcolorectal cancer xenografts as preclinical models by comparison withpatient clinical data. Cancer Res. 2015;75(8):1560–6.
17. Fior R, Povoa V, Mendes RV, Carvalho T, Gomes A, Figueiredo N, Ferreira MG.Single-cell functional and chemosensitive profiling of combinatorialcolorectal therapy in zebrafish xenografts. Proc Natl Acad Sci U S A. 2017;114(39):E8234–43.
18. George E, Kim H, Krepler C, Wenz B, Makvandi M, Tanyi JL, Brown E, ZhangR, Brafford P, Jean S, et al. A patient-derived-xenograft platform to studyBRCA-deficient ovarian cancers. JCI Insight. 2017;2(1):e89760.
19. Sato T, Stange DE, Ferrante M, Vries RG, Van Es JH, Van den Brink S, VanHoudt WJ, Pronk A, Van Gorp J, Siersema PD, et al. Long-term expansion ofepithelial organoids from human colon, adenoma, adenocarcinoma, andBarrett's epithelium. Gastroenterology. 2011;141(5):1762–72.
20. Gao D, Vela I, Sboner A, Iaquinta PJ, Karthaus WR, Gopalan A, Dowling C,Wanjala JN, Undvall EA, Arora VK, et al. Organoid cultures derived frompatients with advanced prostate cancer. Cell. 2014;159(1):176–87.
21. Boj SF, Hwang CI, Baker LA, Chio II, Engle DD, Corbo V, Jager M, Ponz-Sarvise M, Tiriac H, Spector MS, et al. Organoid models of human andmouse ductal pancreatic cancer. Cell. 2015;160(1-2):324–38.
22. van de Wetering M, Francies HE, Francis JM, Bounova G, Iorio F, Pronk A,van Houdt W, van Gorp J, Taylor-Weiner A, Kester L, et al. Prospectivederivation of a living organoid biobank of colorectal cancer patients. Cell.2015;161(4):933–45.
23. Clevers H. Modeling development and disease with organoids. Cell. 2016;165(7):1586–97.
24. Fujii M, Shimokawa M, Date S, Takano A, Matano M, Nanki K, Ohta Y,Toshimitsu K, Nakazato Y, Kawasaki K, et al. A colorectal tumor organoidlibrary demonstrates progressive loss of niche factor requirements duringtumorigenesis. Cell Stem Cell. 2016;18(6):827–38.
25. Li M, Izpisua Belmonte JC. Organoids - preclinical models of human disease.N Engl J Med. 2019;380(6):569–79.
26. Smith RC, Tabar V. Constructing and deconstructing cancers using humanpluripotent stem cells and organoids. Cell Stem Cell. 2019;24(1):12–24.
27. Weeber F, van de Wetering M, Hoogstraat M, Dijkstra KK, Krijgsman O,Kuilman T, Gadellaa-van Hooijdonk CG, van der Velden DL, Peeper DS,Cuppen EP, et al. Preserved genetic diversity in organoids cultured frombiopsies of human colorectal cancer metastases. Proc Natl Acad Sci U S A.2015;112(43):13308–11.
28. Pauli C, Hopkins BD, Prandi D, Shaw R, Fedrizzi T, Sboner A, Sailer V, AugelloM, Puca L, Rosati R, et al. Personalized in vitro and in vivo cancer models toguide precision medicine. Cancer Discov. 2017;7(5):462–77.
29. Vlachogiannis G, Hedayat S, Vatsiou A, Jamin Y, Fernandez-Mateos J, Khan K,Lampis A, Eason K, Huntingford I, Burke R, et al. Patient-derived organoidsmodel treatment response of metastatic gastrointestinal cancers. Science.2018;359(6378):920–6.
30. Guo W, Giancotti FG. Integrin signalling during tumour progression. Nat RevMol Cell Biol. 2004;5(10):816–26.
31. Lu P, Takai K, Weaver VM, Werb Z: Extracellular matrix degradation andremodeling in development and disease. Cold Spring Harb Perspect Biol2011, 3(12).
32. Desgrosellier JS, Cheresh DA. Integrins in cancer: biological implications andtherapeutic opportunities. Nat Rev Cancer. 2010;10(1):9–22.
33. Hynes RO. Integrins: versatility, modulation, and signaling in cell adhesion.Cell. 1992;69(1):11–25.
34. Lelievre SA. Contributions of extracellular matrix signaling and tissuearchitecture to nuclear mechanisms and spatial organization of geneexpression control. Biochim Biophys Acta. 2009;1790(9):925–35.
35. Jinka R, Kapoor R, Sistla PG, Raj TA, Pande G. Alterations in cell-extracellular matrix interactions during progression of cancers. Int J CellBiol. 2012;2012:219196.
36. Yoshida GJ, Azuma A, Miura Y, Orimo A: Activated fibroblast programorchestrates tumor initiation and progression; molecular mechanisms andthe associated therapeutic strategies. Int J Mol Sci 2019, 20(9).
37. Mbeunkui F, Johann DJ Jr. Cancer and the tumor microenvironment: a reviewof an essential relationship. Cancer Chemother Pharmacol. 2009;63(4):571–82.
38. Runa F, Hamalian S, Meade K, Shisgal P, Gray PC, Kelber JA. Tumormicroenvironment heterogeneity: challenges and opportunities. Curr MolBiol Rep. 2017;3(4):218–29.
39. Foster DS, Jones RE, Ransom RC, Longaker MT, Norton JA: The evolvingrelationship of wound healing and tumor stroma. JCI Insight 2018, 3(18).
40. LeBleu VS, Kalluri R: A peek into cancer-associated fibroblasts: origins,functions and translational impact. Dis Model Mech 2018, 11(4).
41. Bachman KE, Park BH. Duel nature of TGF-beta signaling: tumor suppressorvs. tumor promoter. Curr Opin Oncol. 2005;17(1):49–54.
42. Batlle E, Massague J. Transforming growth factor-beta signaling in immunityand cancer. Immunity. 2019;50(4):924–40.
43. Stanisavljevic J, Loubat-Casanovas J, Herrera M, Luque T, Pena R, Lluch A,Albanell J, Bonilla F, Rovira A, Pena C, et al. Snail1-expressing fibroblasts inthe tumor microenvironment display mechanical properties that supportmetastasis. Cancer Res. 2015;75(2):284–95.
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 14 of 16
44. Pompili L, Porru M, Caruso C, Biroccio A, Leonetti C. Patient-derivedxenografts: a relevant preclinical model for drug development. J Exp ClinCancer Res. 2016;35(1):189.
45. Mosmann TR, Yokota T, Kastelein R, Zurawski SM, Arai N, Takebe Y. Species-specificity of T cell stimulating activities of IL 2 and BSF-1 (IL 4): comparisonof normal and recombinant, mouse and human IL 2 and BSF-1 (IL 4). JImmunol. 1987;138(6):1813–6.
46. Collins MK. Species specificity of interleukin 2 binding to individual receptorcomponents. Eur J Immunol. 1989;19(8):1517–20.
47. Eisenman J, Ahdieh M, Beers C, Brasel K, Kennedy MK, Le T, Bonnert TP,Paxton RJ, Park LS. Interleukin-15 interactions with interleukin-15receptor complexes: characterization and species specificity. Cytokine.2002;20(3):121–9.
48. Huntington ND, Legrand N, Alves NL, Jaron B, Weijer K, Plet A, CorcuffE, Mortier E, Jacques Y, Spits H, et al. IL-15 trans-presentation promoteshuman NK cell development and differentiation in vivo. J Exp Med.2009;206(1):25–34.
49. Kondrashova O, Topp M, Nesic K, Lieschke E, Ho GY, Harrell MI, ZapparoliGV, Hadley A, Holian R, Boehm E, et al. Methylation of all BRCA1 copiespredicts response to the PARP inhibitor rucaparib in ovarian carcinoma. NatCommun. 2018;9(1):3970.
50. Lee SH, Hu W, Matulay JT, Silva MV, Owczarek TB, Kim K, Chua CW, BarlowLJ, Kandoth C, Williams AB, et al. Tumor evolution and drug response inpatient-derived organoid models of bladder cancer. Cell. 2018;173(2):515–28e517.
51. Lancaster MA, Knoblich JA. Organogenesis in a dish: modelingdevelopment and disease using organoid technologies. Science. 2014;345(6194):1247125.
52. Clohessy JG, Pandolfi PP. Mouse hospital and co-clinical trial project--frombench to bedside. Nat Rev Clin Oncol. 2015;12(8):491–8.
53. Ben-David U, Ha G, Tseng YY, Greenwald NF, Oh C, Shih J, McFarland JM,Wong B, Boehm JS, Beroukhim R, et al. Patient-derived xenografts undergomouse-specific tumor evolution. Nat Genet. 2017;49(11):1567–75.
54. Camidge DR, Pao W, Sequist LV. Acquired resistance to TKIs in solidtumours: learning from lung cancer. Nat Rev Clin Oncol. 2014;11(8):473–81.
55. Restifo NP, Smyth MJ, Snyder A. Acquired resistance to immunotherapy andfuture challenges. Nat Rev Cancer. 2016;16(2):121–6.
56. Okazawa Y, Mizukoshi K, Koyama Y, Okubo S, Komiyama H, Kojima Y, GotoM, Habu S, Hino O, Sakamoto K, et al. High-sensitivity Detection ofMicrometastases Generated by GFP Lentivirus-transduced OrganoidsCultured from a Patient-derived Colon Tumor. J Vis Exp. 2018;136.
57. Kinsey CG, Camolotto SA, Boespflug AM, Guillen KP, Foth M, Truong A,Schuman SS, Shea JE, Seipp MT, Yap JT, et al. Protective autophagy elicitedby RAF-->MEK-->ERK inhibition suggests a treatment strategy for RAS-drivencancers. Nat Med. 2019;25(4):620–7.
58. Yoshida GJ. Therapeutic strategies of drug repositioning targetingautophagy to induce cancer cell death: from pathophysiology to treatment.J Hematol Oncol. 2017;10(1):67.
59. Singh DK, Kollipara RK, Vemireddy V, Yang XL, Sun Y, Regmi N, Klingler S,Hatanpaa KJ, Raisanen J, Cho SK, et al. Oncogenes activate an autonomoustranscriptional regulatory circuit that drives glioblastoma. Cell Rep. 2017;18(4):961–76.
60. Tornin J, Martinez-Cruzado L, Santos L, Rodriguez A, Nunez LE, Oro P,Hermosilla MA, Allonca E, Fernandez-Garcia MT, Astudillo A, et al. Inhibitionof SP1 by the mithramycin analog EC-8042 efficiently targets tumorinitiating cells in sarcoma. Oncotarget. 2016;7(21):30935–50.
61. Vanner RJ, Remke M, Gallo M, Selvadurai HJ, Coutinho F, Lee L, Kushida M,Head R, Morrissy S, Zhu X, et al. Quiescent sox2(+) cells drive hierarchicalgrowth and relapse in sonic hedgehog subgroup medulloblastoma. CancerCell. 2014;26(1):33–47.
62. Chen S, Giannakou A, Wyman S, Gruzas J, Golas J, Zhong W, Loreth C,Sridharan L, Yamin TT, Damelin M, et al. Cancer-associated fibroblastssuppress SOX2-induced dysplasia in a lung squamous cancer coculture.Proc Natl Acad Sci U S A. 2018;115(50):E11671–80.
63. Kalluri R, Weinberg RA. The basics of epithelial-mesenchymal transition. JClin Invest. 2009;119(6):1420–8.
64. Bertolini G, D'Amico L, Moro M, Landoni E, Perego P, Miceli R, Gatti L,Andriani F, Wong D, Caserini R, et al. Microenvironment-modulatedmetastatic CD133+/CXCR4+/EpCAM- lung cancer-initiating cells sustaintumor dissemination and correlate with poor prognosis. Cancer Res. 2015;75(17):3636–49.
65. Saitoh M. Involvement of partial EMT in cancer progression. J Biochem.2018;164(4):257–64.
66. Wang X, Enomoto A, Asai N, Kato T, Takahashi M. Collective invasion of cancer:Perspectives from pathology and development. Pathol Int. 2016;66(4):183–92.
67. Ferone G, Song JY, Sutherland KD, Bhaskaran R, Monkhorst K, Lambooij JP,Proost N, Gargiulo G, Berns A. SOX2 is the determining oncogenic switch inpromoting lung squamous cell carcinoma from different cells of origin.Cancer Cell. 2016;30(4):519–32.
68. Watanabe H, Ma Q, Peng S, Adelmant G, Swain D, Song W, Fox C, FrancisJM, Pedamallu CS, DeLuca DS, et al. SOX2 and p63 colocalize at genetic lociin squamous cell carcinomas. J Clin Invest. 2014;124(4):1636–45.
69. Murray NR, Justilien V, Fields AP. SOX2 determines lineage restriction: modelinglung squamous cell carcinoma in the mouse. Cancer Cell. 2016;30(4):505–7.
70. Luga V, Zhang L, Viloria-Petit AM, Ogunjimi AA, Inanlou MR, Chiu E,Buchanan M, Hosein AN, Basik M, Wrana JL. Exosomes mediate stromalmobilization of autocrine Wnt-PCP signaling in breast cancer cell migration.Cell. 2012;151(7):1542–56.
71. Talebi A, Dehairs J, Rambow F, Rogiers A, Nittner D, Derua R,Vanderhoydonc F, Duarte JAG, Bosisio F, Van den Eynde K, et al. SustainedSREBP-1-dependent lipogenesis as a key mediator of resistance to BRAF-targeted therapy. Nat Commun. 2018;9(1):2500.
72. Dewaele M, Tabaglio T, Willekens K, Bezzi M, Teo SX, Low DH, Koh CM,Rambow F, Fiers M, Rogiers A, et al. Antisense oligonucleotide-mediated MDM4 exon 6 skipping impairs tumor growth. J Clin Invest.2016;126(1):68–84.
73. Choi CH, Ryu JY, Cho YJ, Jeon HK, Choi JJ, Ylaya K, Lee YY, Kim TJ, Chung JY,Hewitt SM, et al. The anti-cancer effects of itraconazole in epithelial ovariancancer. Sci Rep. 2017;7(1):6552.
74. Kim J, Tang JY, Gong R, Kim J, Lee JJ, Clemons KV, Chong CR, Chang KS,Fereshteh M, Gardner D, et al. Itraconazole, a commonly used antifungalthat inhibits Hedgehog pathway activity and cancer growth. Cancer Cell.2010;17(4):388–99.
75. Yoshida GJ. Metabolic reprogramming: the emerging concept andassociated therapeutic strategies. J Exp Clin Cancer Res. 2015;34:111.
76. Zhou H, Qian W, Uckun FM, Wang L, Wang YA, Chen H, Kooby D, Yu Q,Lipowska M, Staley CA, et al. IGF1 receptor targeted theranosticnanoparticles for targeted and image-guided therapy of pancreatic cancer.ACS Nano. 2015;9(8):7976–91.
77. Witkiewicz AK, Balaji U, Eslinger C, McMillan E, Conway W, Posner B, MillsGB, O'Reilly EM, Knudsen ES. Integrated patient-derived models delineateindividualized therapeutic vulnerabilities of pancreatic cancer. Cell Rep.2016;16(7):2017–31.
78. Pettazzoni P, Viale A, Shah P, Carugo A, Ying H, Wang H, Genovese G, SethS, Minelli R, Green T, et al. Genetic events that limit the efficacy of MEK andRTK inhibitor therapies in a mouse model of KRAS-driven pancreatic cancer.Cancer Res. 2015;75(6):1091–101.
79. Rajeshkumar NV, Yabuuchi S, Pai SG, De Oliveira E, Kamphorst JJ,Rabinowitz JD, Tejero H, Al-Shahrour F, Hidalgo M, Maitra A, et al.Treatment of pancreatic cancer patient-derived xenograft panel withmetabolic inhibitors reveals efficacy of phenformin. Clin Cancer Res.2017;23(18):5639–47.
80. Bailey CJ, Turner RC. Metformin. N Engl J Med. 1996;334(9):574–9.81. Zou S, Li J, Zhou H, Frech C, Jiang X, Chu JS, Zhao X, Li Y, Li Q, Wang H,
et al. Mutational landscape of intrahepatic cholangiocarcinoma. NatCommun. 2014;5:5696.
82. Lee G, Auffinger B, Guo D, Hasan T, Deheeger M, Tobias AL, Kim JY, AtashiF, Zhang L, Lesniak MS, et al. Dedifferentiation of glioma cells to gliomastem-like cells by therapeutic stress-induced HIF signaling in the recurrentGBM model. Mol Cancer Ther. 2016;15(12):3064–76.
83. Fack F, Tardito S, Hochart G, Oudin A, Zheng L, Fritah S, Golebiewska A,Nazarov PV, Bernard A, Hau AC, et al. Altered metabolic landscape in IDH-mutant gliomas affects phospholipid, energy, and oxidative stress pathways.EMBO Mol Med. 2017;9(12):1681–95.
84. Garner EF, Williams AP, Stafman LL, Aye JM, Mroczek-Musulman E, Moore BP,Stewart JE, Friedman GK, Beierle EA. FTY720 decreases tumorigenesis in group3 medulloblastoma patient-derived xenografts. Sci Rep. 2018;8(1):6913.
85. Chun J, Hartung HP. Mechanism of action of oral fingolimod (FTY720) inmultiple sclerosis. Clin Neuropharmacol. 2010;33(2):91–101.
86. Goga A, Yang D, Tward AD, Morgan DO, Bishop JM. Inhibition of CDK1as a potential therapy for tumors over-expressing MYC. Nat Med. 2007;13(7):820–7.
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 15 of 16
87. Horiuchi D, Kusdra L, Huskey NE, Chandriani S, Lenburg ME, Gonzalez-Angulo AM, Creasman KJ, Bazarov AV, Smyth JW, Davis SE, et al. MYCpathway activation in triple-negative breast cancer is synthetic lethal withCDK inhibition. J Exp Med. 2012;209(4):679–96.
88. Yu J, Qin B, Moyer AM, Nowsheen S, Liu T, Qin S, Zhuang Y, Liu D, Lu SW,Kalari KR, et al. DNA methyltransferase expression in triple-negative breastcancer predicts sensitivity to decitabine. J Clin Invest. 2018;128(6):2376–88.
89. El Ayachi I, Fatima I, Wend P, Alva-Ornelas JA, Runke S, Kuenzinger WL, SilvaJ, Silva W, Gray JK, Lehr S, et al. The WNT10B network is associated withsurvival and metastases in chemoresistant triple-negative breast cancer.Cancer Res. 2019;79(5):982–93.
90. Weeden CE, Holik AZ, Young RJ, Ma SB, Garnier JM, Fox SB, Antippa P, IrvingLB, Steinfort DP, Wright GM, et al. Cisplatin increases sensitivity to FGFRinhibition in patient-derived xenograft models of lung squamous cellcarcinoma. Mol Cancer Ther. 2017;16(8):1610–22.
91. Drapkin BJ, George J, Christensen CL, Mino-Kenudson M, Dries R,Sundaresan T, Phat S, Myers DT, Zhong J, Igo P, et al. Genomic andfunctional fidelity of small cell lung cancer patient-derived xenografts.Cancer Discov. 2018;8(5):600–15.
92. Mollaoglu G, Guthrie MR, Bohm S, Bragelmann J, Can I, Ballieu PM, Marx A,George J, Heinen C, Chalishazar MD, et al. MYC drives progression of smallcell lung cancer to a variant neuroendocrine subtype with vulnerability toaurora kinase inhibition. Cancer Cell. 2017;31(2):270–85.
93. Yoshida GJ. Emerging roles of Myc in stem cell biology and novel tumortherapies. J Exp Clin Cancer Res. 2018;37(1):173.
94. Gong K, Guo G, Gerber DE, Gao B, Peyton M, Huang C, Minna JD, Hatanpaa KJ,Kernstine K, Cai L, et al. TNF-driven adaptive response mediates resistance toEGFR inhibition in lung cancer. J Clin Invest. 2018;128(6):2500–18.
95. Hirata E, Girotti MR, Viros A, Hooper S, Spencer-Dene B, Matsuda M, Larkin J,Marais R, Sahai E. Intravital imaging reveals how BRAF inhibition generatesdrug-tolerant microenvironments with high integrin beta1/FAK signaling.Cancer Cell. 2015;27(4):574–88.
96. Krepler C, Sproesser K, Brafford P, Beqiri M, Garman B, Xiao M, Shannan B,Watters A, Perego M, Zhang G, et al. A comprehensive patient-derivedxenograft collection representing the heterogeneity of melanoma. Cell Rep.2017;21(7):1953–67.
97. Einarsdottir BO, Karlsson J, Soderberg EMV, Lindberg MF, Funck-Brentano E,Jespersen H, Brynjolfsson SF, Bagge RO, Carstam L, Scobie M, et al. Apatient-derived xenograft pre-clinical trial reveals treatment responses and aresistance mechanism to karonudib in metastatic melanoma. Cell Death Dis.2018;9(8):810.
98. Warpman Berglund U, Sanjiv K, Gad H, Kalderen C, Koolmeister T, Pham T,Gokturk C, Jafari R, Maddalo G, Seashore-Ludlow B, et al. Validation anddevelopment of MTH1 inhibitors for treatment of cancer. Ann Oncol. 2016;27(12):2275–83.
99. Grasset EM, Bertero T, Bozec A, Friard J, Bourget I, Pisano S, Lecacheur M,Maiel M, Bailleux C, Emelyanov A, et al. Matrix stiffening and EGFRcooperate to promote the collective invasion of cancer cells. Cancer Res.2018;78(18):5229–42.
100. Liu JF, Palakurthi S, Zeng Q, Zhou S, Ivanova E, Huang W, Zervantonakis IK,Selfors LM, Shen Y, Pritchard CC, et al. Establishment of patient-derivedtumor xenograft models of epithelial ovarian cancer for preclinicalevaluation of novel therapeutics. Clin Cancer Res. 2017;23(5):1263–73.
101. Kim H, George E, Ragland R, Rafial S, Zhang R, Krepler C, Morgan M, HerlynM, Brown E, Simpkins F. Targeting the ATR/CHK1 axis with PARP inhibitionresults in tumor regression in BRCA-mutant ovarian cancer models. ClinCancer Res. 2017;23(12):3097–108.
102. Cancer Genome Atlas Research N. Integrated genomic analyses of ovariancarcinoma. Nature. 2011;474(7353):609–15.
103. Li S, Shen D, Shao J, Crowder R, Liu W, Prat A, He X, Liu S, Hoog J, Lu C,et al. Endocrine-therapy-resistant ESR1 variants revealed by genomiccharacterization of breast-cancer-derived xenografts. Cell Rep. 2013;4(6):1116–30.
104. Junttila MR, de Sauvage FJ. Influence of tumour micro-environmentheterogeneity on therapeutic response. Nature. 2013;501(7467):346–54.
105. Wang CC, Bajikar SS, Jamal L, Atkins KA, Janes KA. A time- and matrix-dependent TGFBR3-JUND-KRT5 regulatory circuit in single breast epithelialcells and basal-like premalignancies. Nat Cell Biol. 2014;16(4):345–56.
106. Chou JL, Shen ZX, Stolfi RL, Martin DS, Waxman S. Effects of extracellularmatrix on the growth and casein gene expression of primary mousemammary tumor cells in vitro. Cancer Res. 1989;49(19):5371–6.
107. DeRose YS, Gligorich KM, Wang G, Georgelas A, Bowman P, Courdy SJ,Welm AL, Welm BE: Patient-derived models of human breast cancer:protocols for in vitro and in vivo applications in tumor biology andtranslational medicine. Curr Protoc Pharmacol 2013, Chapter 14:Unit14 23.
Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.
Yoshida Journal of Hematology & Oncology (2020) 13:4 Page 16 of 16
![Failure of a patient-derived xenograft for brain tumor ... · nude or non-obese diabetic/severe combined immu-nodeficient (NOD/SCID) mice [, 15–18], a concept 7 generally referred](https://img.pdfslide.us/doc/110x75/5fda463fed2f21150206154b/failure-of-a-patient-derived-xenograft-for-brain-tumor-nude-or-non-obese-diabeticsevere.jpg)
