PDX Models in Theranostic Applications: Generation and Screening for B Cell Lymphoma of Human Origin

Abstract

This MIB guide briefly summarizes the generation of patient-derived xenografts (PDXs) and highlights the importance of validating PDX models for the presence of B cell lymphoma of human origin before their use in radiotheranostic applications. The use of this protocol will allow researchers to learn different methods for screening PDX models for Epstein-Barr virus (EBV)-infected B cell lymphoma.

Introduction

The evaluation of new radiotheranostic agents in cancer research requires the use of preclinical models that best mimic the development of tumors and their microenvironment. Patient-derived xenografts (PDX) are a valuable preclinical model in the studies of cancer imaging and therapy as they are composed of a heterogeneous population of cancer cells [1, 2]. The establishment of PDX models involves transplanting freshly resected human cancer tissue into mice [3]. PDX models stand out for their ability to accurately mirror the characteristics of cancer and are considered one of the most promising approaches to identify prognostic biomarkers, investigate the impact of intratumor heterogeneity on tumor progression, and assess the efficacy of new drugs [4]. The stromal and immune environment of the original human tissue sample may also be present in the PDX model, as well as the histologic and genetic characteristics of the original patient tumor. Due to the shared characteristics between PDX models and patient tumors, PDX models are extensively used in biodistribution and therapy studies testing novel radiotheranostics [5–9].

PDX models have been used in several studies exploring radiotheranostics for both imaging and treatments of various tumor types. The tumor fragments can be implanted subcutaneously, orthotopically, or heterotopically, including inoculation in the intracapsular fat pad, the anterior compartment of the eye, or beneath the renal capsule [3, 10]. Subcutaneous PDX models are extensively utilized in theranostic approaches since they are easy to implant and monitor tumor growth. Orthotopic implantation has technical challenges and demands more time, often requiring expertise in surgery techniques and ultrasound examinations or other approaches to confirm tumor presence. The advantage of an orthotopic model over a subcutaneous model lies in preserving the microenvironment more closely. Orthotopic models better approximate the “natural” setting of human tumors and may enhance metastatic incidence during xenograft growth, making it a consideration when investigating the application of theranostics in tumor metastasis [10]. Other studies have shown that orthotopically implanted tumors exhibit higher re-transplantation rates, faster growth, and increased aggressiveness when compared with subcutaneously implanted tumors [11]. The choice between a subcutaneous or an orthotopic model depends on the specific plan at hand, and in the context of theranostics applications, researchers should carefully consider the available options that best align with their study objectives. Zhou et al. used an orthotopic PDX from a patient with pancreatic cancer to determine the ability of a theranostic nanoparticle to both image and treat pancreatic cancer [5]. In this work, a combination of an anti-human insulin-like growth factor 1 (IGF-1) antibody and magnetic iron oxide nanoparticles with anthracycline doxorubicin was successful at imaging the PDX via optical and magnetic resonance imaging techniques and showed therapeutic efficacy by reducing the tumor size [5]. In another study performed by Zhang et al., a nanobody targeting integrin-associated protein CD47 was labeled with gallium-68 and zirconium-89 for positron emission tomography and lutetium-177 (¹⁷⁷Lu) for therapy and single-photon emission computerized tomography [6] in a PDX model of gastric cancer. Rao et al. have demonstrated that human epidermal growth factor receptor 2 (HER2)-expressing PDX models respond to HER2-targeted radioimmunotherapy using ¹⁷⁷Lu-labeled anti-HER2 trastuzumab [9]. Other studies have also evaluated radioimmunotherapy in different PDX models, such as Tully et al., using a PDX model of small cell lung cancer, with medium expression of delta-like ligand 3 (DLL3) [8]. In this study, Tully et al. showed that targeting DLL3 with a ¹⁷⁷Lu-labeled anti-DLL3 antibody reduced tumor size and, in some cases, led to a complete treatment response. These studies collectively show the important role of PDX models in advancing the development and assessment of novel theranostic strategies for a diverse range of cancers.

While PDX models accurately replicate tumor composition, their applicability across all tumor types may be limited. Furthermore, there have been instances of unexpected occurrences of B cell lymphoma of human origin emerging in several PDX models [12, 13], including our own research in gastric carcinomas [7]. The occurrence of B cell lymphoma of human origin in PDXs has been linked with Epstein-Barr virus (EBV) ribonucleic acid (RNA) within lymphoma cells, as demonstrated by in situ hybridization [13]. Considering that over 90% of the global population is seropositive for EBV, and given EBV’s association with post-transplant lymphoproliferative disorders, particularly B cell lymphoma in immunocompromised humans, it is crucial to consider its potential impact before tissue implantation into animal hosts. Latent EBV infection in PDX transplants can lead to the infection of B lymphocytes, potentially leading to lymphoma, which might affect biodistribution and efficacy of novel radiotheranostics (Fig. 1). Therefore, validation of PDX models as they are used in experiments involving radiotheranostics, is a valuable practice.

Fig. 1.

Mice implanted with PDXs have a relatively high risk of spontaneous lymphoma which could cover up accurate results due to the distinct drug uptake and sensitivities between lymphoma and other tumors. Figure at the top: PDX without EBV infiltration. Figure at the bottom: PDX with EBV infiltration

Step 1: Screening of B Cell Lymphoma of Human Origin Before Tissue Implantation

The validation methods described below are valuable in assessing tumor samples to be used in the generation of PDXs, however, it remains unclear whether they are suitable for predicting lymphomagenesis at the initial examination stages of the patient sample before xenotransplantation.

Step 1.1 Sample Collection

• Tumor samples are obtained from a patient during a surgical resection or biopsy.

• After tumor sample removal, the sample should be immediately stored in the appropriate collection media (e.g. Roswell Park Memorial Institute RPMI-1640 containing penicillin-streptomycin (pen-strep) and anti-fungal amphotericin B).

• The tissue sample can then be screened for diagnosis of cancer type by a pathologist.

Notes: At this stage, it is difficult to determine if the tissue sample will develop EBV-infected B cell lymphoma in the mouse. However, the methods described in Step 1.2 have been proposed to evaluate the risk of lymphomagenesis in xenografts.

Step 1.2 Screening for EBV-infected Lymphoma in Patient Samples Before Xenotransplantation

While numerous tests have been explored for this stage, none have demonstrated predictive efficacy for lymphoma development in mice. The tests listed below have been reported, and it is the decision of the researcher to determine their potential utility in their models/analyses, especially considering the limited quantity of available human tissue for research purposes.

• Specific markers of lymphocytes can be screened by immunohistochemistry (IHC), including cluster of differentiate 20 (CD20, a B cell marker) or CD45 (leukocytes common antigen, a pan-leukocyte marker) [14]. Previous studies have shown that EBV-induced lymphomas in PDXs exhibited a CD45⁺, CD20⁺, panCK⁻ IHC staining pattern [14].

Limitation:

The presence of cells positive for the markers above does not indicate a diagnosis of lymphoma as they are expressed by reactive B cells that may be present in the tumor microenvironment. IHC must be interpreted in the context of morphology to differentiate reactive lymphocytic infiltrated from lymphoproliferative processes. IHC at this stage is not usually effective in predicting occurrence of lymphoma in PDXs, as the lymphoproliferative lesions arise after the PDX tissue is implanted in mice and reports of EBV-induced induced lymphoma in PDXs in which the patient specimens were assessed have not demonstrated histologic evidence of lymphoma or other atypical lymphocytic infiltrates in the patient specimens. This is likely because EBV latently infected B-cells are present in a very small number and are not histologically distinct from normal B cells in patients, causing them to be indistinguishable using these markers [15, 16].

Other Approaches:

Positive polymerase chain reaction (PCR) based amplification of BamHI W region has also been used to predict EBV transfection before xenotransplantation [17]. However, we do not recommend this assay at this step since this method has shown promise in a single study, and further studies are needed to determine the potential and applicability on a broader scale. PCR-based amplification of BamHI W region has been explored since BamHI W region is a major internal repeat in the EBV genome and it consists of two regions: a promoter and W1W2 exon domains. Specifically, the promoter causes B cell transformation by inducing EBV proteins (Epstein-Barr virus nuclear protein 2, EBNA2, and Epstein-Barr nuclear antigen leader protein, EBNA-LP). W1W2 exons encode the EBNA-LP protein, important for B cell lymphoma transformation by EBV.

Step 2: Implanting PDX Tissues Into Immunocompromised Mice

Once genetic profiling of the tumor tissue sample has been completed, the PDX can be implanted into immunocompromised mice. Conventional PDX models are typically established using immunocompromised mice, including athymic nude mice, severe combined immunodeficiency (SCID) mice, non-obese diabetic-severe combined immunodeficiency (NOD-SCID) mice, NOD-SCID−/IL2gamma-receptor null (NSG) mice, BALB/cRag2 null/IL2gamma-receptor null (BRG) mice, and Rag-2 null/Jak3 null (BRJ) mice. However, not all PDXs develop in the less immunocompromised mice strains and, therefore, the successful development of PDXs often requires the use of highly immunodeficient animals such as NSG mice that lack functional T and B lymphocytes as well NK cells and have defective cytokine signaling (BRG/BRJ > NSG > NOD/SCID > SCID > nude). Important to note is that the EBV-transformed B lymphocyte in primary tumors often outgrows after implantation, especially in NSG mice [13]. The reason is that EVB-infected B cells are in a latent state in the human body, and when tissues are transplanted into severely immunodeficient mice, these cells are prone to be activated to form lymphomas due to the lack of functional immune cells.

We recommend that researchers engage with their specific PDX cores or suppliers to determine the most suitable mouse model. This choice should not only align with the tumor’s growth characteristics but also align with the researcher’s project goals.

Step 2.1 Transporting a PDX Sample

• PDX sample is placed in a bucket filled with ice.

• Vials used for PDX transporting should be sterile and contain collection media specific to the PDX being used.

• PDX tumor is segmented into 1 mm³ for implantation purposes.

Step 2.2: PDX Inoculation (subcutaneous)

• Tumor implantation should take place in a proper procedure room and should be conducted according to institutional guidelines and animal protocol. Appropriate anesthesia methods should be used. Additionally, immunocompromised mice can develop C. bovis infection which can impede tumor growth, and therefore, it is important to perform all steps in a sterile environment.

• Mice’s hair is shaved using an electric razor so that the skin can be seen in the area of tumor implantation - dorsal flank or shoulder (the step of hair removal can be skipped if using a nude mouse).

• The exposed skin where the tumor will be implanted is cleaned with betadine followed by 70% ethanol.

• The skin is grasped and raised using sterile forceps.

• A 2–3 mm incision is made in the skin, creating a small pocket under the skin using iris scissors.

• The tumor fragments are removed from the vial and 2 fragments are placed into the skin pocket. To improve tumor take, use forceps to submerge the PDX tissue in Matrigel Matrix before subcutaneous inoculation.

• The incision is held closed using forceps and sealed with a small drop of GLUture. Alternatively, wound clips or sutures can be used to close the skin incision.

• Once the procedure is complete, the mouse is placed back into their cage and monitored until they wake up from the anesthesia.

• After implantation, different PDX tumors will take different amounts of time to grow (ranging from a few days to several months) and animals should be monitored throughout this process.

Notes: While we present a summary of the procedure for establishing a subcutaneous PDX model to be used in theranostic studies, we recommend referring to references by Jung et al. [18] and Moy et al. [19], which provide comprehensive details on the method. Additionally, we emphasize the significance of utilizing PDX models at the same passage number for conducting theranostic studies.

While this section describes subcutaneous PDX generation, orthotopic implantation of PDXs can also be performed as described above. Instances of unexpected lymphomagenesis in PDXs have predominantly been reported in subcutaneous models, likely due to their more widespread utilization and the relative ease of their development compared to orthotopic models.

Step 3: Screening of B Cell Lymphoma of Human Origin After Tissue Implantation

PDX models can not be expanded in vitro, and therefore, the sample must be implanted into an immunodeficient mouse (see Step 2). Once the tumor has grown to 1–2 cm³ [20], the tumor can be removed, sectioned, and xenografted into new mice (Fig. 1). However, tumor composition may change over multiple passages [21]. Therefore, validation of PDX models and monitoring of possible lymphoma development when they are initially generated, and at later times as they are passaged in mice and used in experiments, is a valuable practice.

When grafting PDX samples into mice, it is important to consider using an early passage to ensure better reflection of the patient-derived tumor microenvironment and avoid possible contaminations, such as EBV-transformed B lymphocytes.

Step 3.1 Preserving PDX at Early Passages by Banking of Cryopreserved Early Passage Fragments

• When a PDX is established, it is important to preserve fragments of early passage numbers so that they can be used in future studies. Institutions should have established biobanks for PDX samples to be stored properly. We recommend that researchers work with the PDX core at their institution to establish proper protocols for storing and maintaining PDX samples.

Step 3.2: Monitoring PDX Sample for EBV-infected B Lymphocytes

Several assays have been tested at this step and below is what we recommend based on our experiences.

In the initial steps, the pathologist assesses the hematoxylin and eosin (H&E) stained sections to identify the expected tumor and check for any evidence of lymphoma by cell morphology.

• Tissues are fixed in 10% neutral buffered formalin, processed in alcohol and xylene and infiltrated with paraffin in an automated tissue processor.

• Tissues are then embedded in a paraffin block, sectioned at 5 μm thickness, mounted on glass slides, and stained with H&E and IHC.

• IHC is performed on a Leica Bond RX automated stainer using Bond bulk reagents (Leica Biosystems, Buffalo Grove, IL), and a polymer detection reagent kit (DS9800, Novocastra Bond Polymer Refine Detection, Leica Biosystems). The chromogen is 3,3 diaminobenzidine tetrachloride (DAB), and sections are counterstained with hematoxylin. Details for each target are shown in the table below (Table 1).

• If the expected tumor is observed and with no evidence of lymphoma, continue using the PDX to Step 3.3 (no additional testing such as IHC is required in this case). We would recommend repeating this step at each PDX passage. However, if there is evidence of lymphoma (Fig. 2) then proceed to the next step [15].

• If histopathologic findings compatible with lymphoma are observed on H&E (atypical round cell neoplastic infiltrate), a presumptive diagnosis of lymphoma is made and can be confirmed by ancillary tests if necessary (see below). But practically, the pathologist can be almost certain it is lymphoma from the H&E alone. To confirm that the lymphoma is of human origin and EBV related, then additional tests are required (see below). For cost and time efficiency, H&E finding alone is in practice sufficient to make the decision to exclude this PDX from further use in theranostic applications. Ancillary tests below are optional.

Table 1.

Information on targets and antibodies for IHC

Target	Reacts with human target	Reacts with murine target	Epitope retrieval	Primary antibody (vendor, catalog number, dilution)	Post-primary antibody (vendor, catalog number, dilution)	Secondary antibody (vendor, catalog number, dilution)
Pancytokeratin	Yes	Yes	Heat induced at pH 6.0	Dako, Z0622, 1:500	Not applicable	Leica Biosystems, DS9800 kit, reagent #3, no dilution performed.
Human CD45	Yes	No	Heat induced at pH 6.0	Dako, M0701, 1: 100	Abcam, ab133469, 1: 1000	Leica Biosystems, DS9800 kit, reagent #3, no dilution performed.
Human CD19	Yes	No	Heat induced at pH 6.0	Leica, PA0843, no dilution performed	Abcam, ab133469, 1: 1000	Leica Biosystems, DS9800 kit, reagent #3, no dilution performed.
Human CD20	Yes	No	Heat induced at pH 6.0	Dako, M0755, 1: 1000	Abcam, ab133469, 1: 1000	Leica Biosystems, DS9800 kit, reagent #3, no dilution performed.

Fig. 2.

Photomicrographs of a carcinoma (upper row) and a diffuse large B cell lymphoma of human origin (lower row) that occurred at the subcutaneous PDX implantation site of a human gastric carcinoma in an NSG mouse. The carcinoma is composed of atypical epithelioid cells forming nests supported by a moderate amount of fibrovascular stroma. The atypical cells display membranous and cytoplasmic immunoreactivity for pancytokeratin (epithelial cell marker) while no reactivity is observed for human CD45 (panleukocyte marker) and human CD20 (B cell marker). The lymphoma consists of atypical round cells forming sheets associated with a minimal fine fibrovascular stroma. The atypical cells display membranous immunoreactivity for human CD45 and human CD20 and absence of reactivity for pancytokeratin. Scale bars: 20 μm. Source: Patrícia M. R. Pereira, Komal Mandleywala, Sébastien Monette, Melissa Lumish, Kathryn M. Tully, Mike Cornejo, Audrey Mauguen, Ashwin Ragupathi, Marissa Mattar, Yelena Y. Janjigian, Jason S. Lewis. Caveolin-1 temporal modulation enhances antibody drug efficacy in heterogeneous gastric cancer, Nature Communications, 2022, 13, 2526. Creative Commons license: https://creativecommons.org/licenses/by/4.0/

Notes: Additionally, we recommend monitoring organ enlargement as an indicator of lymphoma. The mice that develop EBV-induced lymphoma have a rapid metastasis of lymphoma. This can be seen by the enlargement of various organs including lymph nodes, liver, and spleen [15, 22].

Step 3.3: Confirmation of Lymphoma (Optional)

This step involves screening samples for B cell lymphoma markers by IHC [15]. IHC for CD45 (leukocyte marker) and B cell marker (CD20 has been used most often, other B cell markers such as CD19 can also be used) are expected to be positive. Note that lymphoma of mouse origin can occur so it is important to use anti-human antibodies to ensure that EBV-induced lymphoma from the original human sample is detected.

Notes: To confirm that the lymphoma is EBV-related, markers of EBV can be screened using IHC, including EBNA2 and latent membrane protein 1 (LMP1) [15], RNA in situ hybridization probed for EBER and LMP1 [16], quantitative and droplet digital PCR screening for EBV-specific genes (BamHI-W and Epstein-Barr virus nuclear protein 1) [16, 17], and gene signature [23]. Yet, in practical terms, when screening PDXs for theranostic applications, we skip this step once we confirm the presence of lymphoma.

Step 3.4: PDX Expansion

If the engrafted tumor is growing and is found to be free of lymphoma, then the PDX is reimplanted into new mice for expansion. It is important to continue monitoring PDX samples for EBV at each passage.

Step 3.5: Eliminating Lymphoma-Positive PDX Samples

If samples are confirmed to contain a lymphoma by any method in Step 3.2, it is recommended to discard the PDX (Fig. 3) as this could prevent accurate research due to the distinct uptake and sensitivity of the radiotheranostic agent between lymphoma and other tumors (Fig. 1).

Fig. 3.

Schematic representation of tissue sample collection and screening for B cell lymphoma before PDX implantation (Step 1), PDX implantation (Step 2) and screening for B cell lymphoma after PDX implantation (Step 3)

Notes: Although strategies have been reported to prevent the expansion of EBV-induced lymphoma, we do not recommend those. As an example, rituximab (an anti-CD20 monoclonal antibody) was shown to prevent the formation of B cell lymphomas in a cohort of ovarian cancer PDXs [14]. Rituximab would also eliminate normal (non-transformed) B cells, resulting in alterations of the tumor microenvironment, which may impact experimental results.

Conclusions

The utilization of PDXs as models for testing new radiotheranostic agents represents a promising avenue in translational cancer research. However, it is crucial to recognize that PDXs are complex models, necessitating rigorous validation and continuous monitoring to ensure the production of robust and reliable research outcomes. To maximize the utility of PDXs in research exploring new radiotheranositc agents, it is essential for the imaging community to engage the expertise of pathologists in characterizing PDX models and identifying any unexpected issues that may arise during data acquisition. Further guidance on this matter is detailed in this MIB guide. Here, we highlighted the need to determine the unexpected occurrence of B cell lymphoma of human origin that has been observed in several PDXs. However, it is important to note that B cell lymphoma is only one of the unexpected issues that can occur in a PDX sample, therefore consulting with a pathologist to perform histopathology as a quality assurance assessment of PDXs is crucial. Performing histopathology serves the dual purpose of not only excluding lymphoma but also identifying other potential issues that may interfere with research in PDX models, such as xenogeneic graft-versus-host disease [24, 25], tumors of mouse origin [26], and the transmission of infectious agents through the PDX, including lactate dehydrogenase-elevating virus [27], mouse kidney parvovirus [28], and reovirus [29].