Induction and characterization of neoplastic bladder tumors in a transgenic porcine model

Abstract

Background

Large animal models of bladder cancer are lacking.

Objective

This study aimed to develop and characterize a transgenic porcine model of bladder cancer (BC) using Oncopigs expressing Cre-inducible KRAS^G12D and TP53^R167H mutations.

Methods

Eleven female Oncopigs underwent tumor induction via three cystoscopic inoculation procedures: Procedure I (N = 3, 1 inoculation/pig), chemical dissolution of the glycosaminoglycan layer with N-Dodecyl-β-d-Maltoside DDM followed by adenoviral Cre-recombinase (AdCre) instillation; Procedure II (N = 4, 3 inoculation/pig), mechanical mucosal denudation followed by AdCre instillation; and Procedure III (N = 4, 3 inoculation/pig), cystoscopy-guided submucosal injection of AdCre. Animals were clinically monitored throughout follow-up (14–28 days). Tumor development was assessed on cystoscopy and ultrasonography, and pathologically, immunohistochemically (IHC), and genomically characterized.

Results

All pigs remained clinically healthy. Tumors developed at 59% (16/27) of inoculation sites: nine (33%) were neoplastic and seven (26%) were inflammatory. Procedure I achieved 100% neoplastic tumors and produced both non-muscle invasive (71%) and muscle-invasive (29%) tumors. Procedure II achieved 50% neoplastic tumors, all of which were muscle invasive (100%). Procedure III generated only inflammatory tumors. Histologically, neoplastic tumors were pathologically interpreted as urothelial cell carcinomas with sarcomatoid differentiation, with IHC confirming the presence of both epithelioid and sarcomatoid features with abundant mixed leukocytic infiltrates. Genomic analyses verified Cre-induced alterations alongside other mutations seen in human BC.

Conclusions

We herein demonstrate an efficient and reproducible method for developing autochthonous neoplastic bladder tumors in Oncopigs that resemble human bladder cancer of varying stages. This large animal model facilitates the evaluation of novel surgical and intravesical therapies in BC.

Keywords

oncopig bladder cancer transgenic animals viral cell transformation neoplastic cell transformation swine

Introduction

With over 500,000 new cases and 200,000 deaths annually, urothelial carcinoma of the bladder (BC) remains a significant global health burden.¹ Approximately 75% of new diagnoses present with non-muscle invasive disease (NMIBC),² highlighting the ongoing need for novel bladder-preserving therapeutics. With a rapidly expanding field of locoregional therapeutics, reliable pre-clinical models for testing novel approaches, alongside evaluating their potential interaction with tumor immune microenvironments and immunotherapies, are necessary.

Currently, 95% of preclinical studies in BC rely on rodent models, which, while genetically tractable and cost-efficient, often fail to recapitulate the anatomical, immunological, and pathophysiological complexities of human disease. Indeed, translational success from rodent models is low, with <5% of preclinically successful therapeutics progressing beyond early-phase clinical trials.^3–5 Large animal models, particularly pigs, present a promising solution. Pigs are physiologically similar to humans and have anatomically comparable urinary tracts that can encompass human surgical instruments, making them particularly valuable for evaluating locoregional therapeutics.⁶ There are however currently no porcine models of BC.

The Oncopig is a transgenic porcine cancer model harboring Cre-inducible KRAS^G12D and TP53^R167H mutations, whereby autochthonous tumors may be induced with exposure to adenoviral Cre-recombinase (AdCre).⁷ KRAS is an isoform of the oncogene, RAS, and is mutated in 25% of all human malignancies.⁸ RAS pathway mutations are seen in 15–40% of BC, with a higher prevalence seen in low-grade NMIBC.⁹ TP53 is the most commonly mutated gene in BC, occurring in over half of tumors, and is associated with aggressive and muscle-invasive (MIBC) disease.^10,11 We have previously demonstrated that percutaneous inoculation of AdCre in the liver,¹² pancreas,¹³ and lungs¹⁴ of Oncopigs led to the development of site-specific poorly differentiated carcinomas that were histologically similar to human cancers. We thus hypothesized that bladder tumors may be induced in Oncopigs using a similar process.

We developed bladder tumors by exposing Oncopig bladders to AdCre via three endoscopic procedures. The efficacy and safety of each method, and histopathology of induced tumors were assessed.

Materials and methods

The conduct of this prospective study was approved by our Institutional Animal Care and User Committee (IACUC Protocol #: 11-02-004). All animal research conducted at our institution complies with the Guide for the Care and use of Laboratory Animals. This study was reported in accordance with The ARRIVE guidelines 2.0 (Appendix A).¹⁵ The primary objective of this study was to establish a standardized, reproducible, methodology for inducing neoplasms which originate from both superficial and muscular layers of the bladder.

Study design

Eleven Oncopigs were obtained from the University of Illinois, Department of Animal Science and Sus Clinicals Inc. (Chicago, IL, USA). This number of pigs were chosen to allow for at least three animals per procedure to assess reproducibility, based on prior success rates of 30–80% in liver, lung, and pancreatic models.^12–14 Only female Oncopigs were used in this study to facilitate practical cystoscopic access.

Tumor induction protocols and follow-up schedules are summarized in Figure 1. Each Oncopig underwent one induction procedure. All animals had comprehensive examinations including bloodwork (complete blood count [CBC] and complete metabolic panel [CMP]), kidney and bladder ultrasound (establish baseline radiographic normal anatomy, rule out obstruction), and cystoscopy (evaluate normal anatomy, rule out abnormalities) prior to AdCre inoculation (Day 0) and during follow up. All tumor inoculation procedures, cystoscopies, and imaging studies were performed by one urologist (A.A.) under sedation or general anesthesia; findings were interpreted via consensus by two urologists (A.A., J.A.C.).

Figure 1.

Tumor induction and study protocol.

Follow-up and euthanasia

Periodic routine bloodwork and daily clinical examinations monitored all animals for signs of general health deterioration, pain, and infection throughout follow-up period. Cystoscopy and ultrasonography were undertaken throughout follow-up to monitor tumor growth, progression, and for signs of upper urinary tract obstruction.

Euthanasia, via intravenous pentobarbital sodium (87 mg/kg) and phenytoin sodium (11 mg/kg), was performed at days 14, 21, or 28 as defined a priori based on prior Oncopig studies in other organs (pancreas, liver, lung),^12,14,16 unless otherwise indicated due to deteriorating health condition of the animals as determined by the RARC.

Inoculation procedures

The Oncopig genome contains an inducible LOX-P cassette (TP53 ^R167H and KRAS^G12D) across all cell types within the entire organism, including urothelial cells – cells which overlay in seven layers to constitute the urothelium, the innermost mucosal layer of the bladder.¹⁷ Notably, this luminal surface of urothelium is shielded from exogenous pathogens and toxins, including viral vectors, by a dense glycosaminoglycan (GAG) layer.¹⁸ We thus developed three distinct cystoscopic procedures to overcome the natural physical barriers of the bladder mucosa and inoculate the urothelial and deeper bladder layers with an AdCre viral vector (Ad5CMVCre-eGFP; University of Iowa Viral Vector Core, Iowa City, IA, USA). All procedures were performed with cystoscopic assistance by one urologist with over three years of clinical experience (A.A.).

Procedure I (GAG dissolution followed by AdCre instillation) [N = 3]

We hypothesized that chemical dissolution of the GAG layer followed by passive inoculation with AdCre would produce superficial bladder tumors akin to non-muscle invasive BC. The bladder was first emptied, and the pig was positioned in supine sharp Trendelenburg, making the posterior inferior wall (inoculation region) the lowest gravitational point within the bladder. The bladder was then partially insufflated with air and laterally compressed, creating a small dependent pit for inoculation. Next, a 10-cc mixture of AdCre CMV (10⁹-pfu, IOWA CORE) and N-Dodecyl-β-d-Maltoside (DDM; 68-mg (0.05% W/V), Sigma Aldrich) was instilled under vision to the inoculation region using a Williams needle (Cook Medical). DDM, a sugar-like surfactant, was used to dissolve the GAG layer and directly expose urothelial cells to the viral vector.¹⁹ After one hour of incubation, the bladder was emptied and irrigated with three successive washes with 50-ml of normal saline.

Procedure II (mucosal denudation followed by AdCre instillation) [N = 4]

We hypothesized that mechanical mucosal denudation may lead to a higher proportion of muscle-invasive bladder tumors akin to MIBC. The pig was placed in supine Trendelenburg, and the bladder was emptied. Next, the bladder was filled with 100-cc of normal saline and the urothelium was superficially denuded using cold-cup biopsy forceps (Storz) once each on the lateral walls and once on the posterior wall. Subsequently, the bladder was emptied, and a 100-cc solution containing 10⁹-pfu (N = 3) or 10¹¹-pfu (N = 1) AdCre was instilled via a 14 Fr Foley catheter. After an hour of incubation, the bladder was emptied and irrigated with three successive washes with 100-cc of normal saline.

Procedure III (cystoscopic submucosal injection) [N = 4]

We hypothesized that direct cystoscopic injection of AdCre into the bladder submucosa/muscle may lead to a higher proportion of muscle-invasive or locally advanced bladder tumors. The pig was placed in supine Trendelenburg position and its bladder emptied. The bladder was then filled with 100-cc of normal saline. Next, an InjeTAK needle (catalog# DIS199, injeTAK Adjustable Tip Needle, 4.8Fr, 35 cm) was used to submucosally inject 1.0-cc of buffered AdCre solution (1.2–1.5 × 10⁸ [N = 3]; 6.2 × 10⁸ [N = 1]) in either calcium chloride and gelatin foam (N = 1) or plant starch (N = 3). Each bladder was injected thrice, once each on the lateral walls and once on the posterior wall. The bladder was emptied and irrigated with three successive washes with 100-cc of normal saline.

Pathologic analysis

Following euthanasia, all animals underwent necropsy by the same urologist that performed the inoculations (A.A.). The bladder excised en bloc and was bivalved anteriorly to photo-document and grossly measure tumor characteristics. Adjacent organs, including uterus, perivesical fat, regional pelvic lymphadenopathy and bowel were grossly inspected for any abnormalities or metastatic deposits; any suspicious tissue was resected for microscopic evaluation. Representative tumor tissue samples from the bladder, regional lymph nodes, and suspicious adjacent tissue were fixed in 10% neutral buffered formalin fixative for ≥3-days and sectioned for histological evaluation. Selected bladder tissues were processed in alcohol and xylene, embedded in paraffin, sectioned at 5 μm, and stained with hematoxylin and eosin (H&E). A subset of selected tumor samples underwent immunohistochemistry (IHC) straining for p53, high molecular weight keratin CK 34BE1, pan-cytokeratin AE1/AE3 (CK AE1/AE3), pan-cytokeratin WSS (CK WSS), cytokeratin 5 (CK 5), cytokeratin 7 (CK 7), cytokeratin 8/18 (CK 8/18), cytokeratin 14 (CK 14), vimentin, CD3 and IBA1 in which the chromogen was 3,30-diaminobenzidine tetrachloride (DAB) and the counterstain was hematoxylin (Supplementary Table 1). IHC staining was validated in appropriate control swine tissues. Inoculation success was defined as the development of neoplastic tumors. The development of inflammatory tumors was not considered a successful inoculation, as we did not perform sham control inoculations to definitively rule out the absence of bland inflammatory tumor development simply due to mechanical manipulation. Slides were interpreted by a board-certified veterinary pathologist (S.M.) and a human pathologist with expertise in urothelial cancer (H.A.); tumor grade, depth and invasion of neoplastic tumors were evaluated in accordance with the 8^th Edition of the American Joint Committee on Cancer Tumor, Node, Metastasis (AJCC-TNM) system.²⁰

Genomic characterization

Detailed methodology on DNA extraction and characterization are provided in the Appendix B. Select tumors from FFPE section were identified for genomic DNA extraction. Whole genome sequencing (WGS) of matched tumor and normal DNA was performed and aligned to pig genome sequence. Mutations were called using the standard Mutect2 pipeline. Mutations with a depth of ≥5 supporting tags and >2% variant allele frequency were considered as so-called high confidence.

All detected variants were double-checked manually.

Statistical analysis

Descriptive numerical data were reported as medians with ranges; categorical data were reported as frequencies. No formal statistical analyses were undertaken as the intent of this study was to develop a reproducible technical methodology for tumor induction in Oncopig bladders, instead we quantify and qualify our findings for each procedure employed.

Results

Oncopig characteristics

Eleven female Oncopigs with a median age of 19-weeks (range: 12, 27) and a median weight of 46-kg (range: 22, 59) were included. A total of 27 inoculations were performed, three (one inoculation per Oncopig, N = 3) using Procedure I, 12 (three inoculations per Oncopig, N = 4) using Procedure II, and 12 (three inoculations per Oncopig, N = 4) using Procedure III (Table 1). Each Oncopig underwent only one procedure and one general anesthetic and received one dose of perioperative antibiotic prophylaxis.

Table 1.

Animal characteristics and procedure details.

Oncopig ID	Weight(kg)	Inoculation age(weeks)	Inoculation method	AdCre dose(PFU)	Treatment sites (No.)	Euthanasia time(days from inoculation)
Procedure I (GAG dissolution followed by AdCre instillation) [N = 3]
1	41.4	19	10cc solution containing N-Dodecyl-β-d-Maltoside (68 mg) and AdCre instilled to dependent portion of bladder; incubate for 1 h	1.0 × 10⁹	1	14
2	58.5	24		1.0 × 10⁹	1	21
3	40.9	19		1.0 × 10⁹	1	21
Procedure II (mucosal denudation followed by AdCre instillation) [N = 4]
4	23.5	19	mucosal denudation x3 on posterior and lateral bladder, followed by instillation of 100cc saline solution containing AdCre; incubate for 1 h	1.0 × 10⁹	3	14
5	48.7	24		1.0 × 10⁹	3	21
6	54.0	19		1.0 × 10⁹	3	21
7	46.4	12		1.0 × 10¹¹	3	28
Procedure III (cystoscopic submucosal injection) [N = 4]
8	21.5	16	1.0-cc of buffered AdCre solution in calcium chloride and gelatin foam submucosally injected x3 (posterior, right lateral, left lateral walls)	1.5 × 10⁸	3	14
9	45.8	27	1.0-cc of buffered AdCre solution in plant starch submucosally injected x3 (posterior, right lateral, left lateral walls)	6.2 × 10⁸(HT)	3	21
10	50.5	16		1.2 × 10⁸	3	21
11	24.4	12		1.2 × 10⁸	3	28

Abbreviations: KG, kilograms; PFU, plaque-forming units; GAG, glycosaminoglycan; AdCre, adenovirus Cre-recombinase; HT, high titer

Inoculation success

Overall, inoculations lead to tumor development at 16 of 27 sites (59%), nine (33%) were neoplastic and seven (26%) were inflammatory in nature (Table 2). Histologic characteristics of tumors are described in detail in the Tumor Pathology section.

Table 2.

Pathological outcomes and tumor characterizations.

Oncopig ID	Euthanasia day	Tumor	Pathology	Size at necropsy (cm)				Neoplastic tumor vs inflammatory tumor
Oncopig ID	Euthanasia day	Tumor	Pathology	Left Lesion	Posterior Lesion	Right Lesion	Other location	Left Lesion	Posterior Lesion	Right Lesion
Procedure I (GAG dissolution followed by AdCre instillation) [N = 3]
1	14	Yes	3x Ta	2.6	1.0	–		NPT	NPT	NPT
2	21	Yes	1x Ta	1.0	–	–	Urethra	NPT	–	–
3	21	Yes	1x Ta; 2x T2a	2.1	1.4	3.0	Right lateral superior	NPT	NPT	NPT
Procedure II (mucosal denudation followed by AdCre instillation) [N = 4]
4	14	Yes	3x T2a; 1x T2b	0.2	0.4	–		NPT	NPT	–
5	21	Yes	3x T2a	3.0	1.0	3.0		NPT	NPT	NPT
6	21	No	–	–	–	–		–	–	–
7	28	Yes	1x T2b	1.0	–	–		IFT	–	–
Procedure III (cystoscopic submucosal injection) [N = 4]
8	14	No	–	–	–	–		–	–	–
9	21	Yes	1x T2a; 1x T2b	–	–	1.5		–	–	IFT
10	21	Yes	1x T2a; 2x T3	–	0.3	2.0		–	IFT	IFT
11	28	Yes	2x T3	–	3.0	1.0	Uterus	–	IFT	IFT

Abbreviations: NPT, neoplastic tumor; IFT, inflammatory tumor; GAG, glycosaminoglycan; AdCre, adenovirus Cre-recombinase.

In Procedure I, one Oncopig was euthanized at 14-days and the other two at 21-days. All three Oncopigs developed neoplastic tumors (3/3, 100% inoculation success) with each tumor site developing 1–3 tumors for a total of seven neoplastic tumors. Tumors were staged (per AJCC-TNM) as both non-invasive (pTa/T1, in 5 of 7 [71%]) and muscle-invasive (pT2, in 2 of 7 [29%]) phenotypes. Muscle invasive tumors only developed in swine euthanized 21-days post-inoculation.

In Procedure II, one Oncopig was euthanized at 14-days, two at 21-days, and one at 28-days. Three Oncopigs developed tumors: two developed neoplastic tumors (2/4, 50% inoculation success), while one developed a solitary inflammatory tumor; one animal did not develop any tumors. Of the two animals that developed neoplastic tumors, each inoculation site developed 3–4 tumors for a total of seven neoplastic tumors. All neoplastic tumors were staged (per AJCC-TNM) as muscle-invasive (pT2, in 7 of 7 [100%]) and were present at both 14- and 21-days post-inoculation. No animals developed non-muscle invasive neoplastic tumors.

In Procedure III, one Oncopig was euthanized at 14-days, two at 21-days, and one at 28-days. Three Oncopigs developed tumors, all of which were inflammatory in nature and non-neoplastic; one animal did not develop any tumors. Each pig developed 1–3 tumors, and tumors were present at all time points. Inflammatory tumors were not staged. One inflammatory tumor was identified beyond the bladder, in the uterus, likely resulting from a needle that inadvertently penetrated beyond the bladder and injected AdCre into the uterus (Supplementary Figure 1).

Ultrasonographic and cystoscopy findings

Ultrasonography of the bladder and kidneys did not reveal any abnormal findings throughout the course of the study. None of the pigs developed hydronephrosis or visible tumors on ultrasonography. On cystoscopic evaluation, tumors demonstrated similar characteristics to those encountered in humans in both neoplastic and inflammatory settings (Figure 2). All tumors demonstrated interval growth over the study period without evidence of spontaneous tumor regression. Nearly all lesions initially appeared consistent with flat inflammatory changes (Day 7). With Procedures I and II, By the time of euthanasia (Day 14–28), most lesions were multifocal, with a raised nodular or papillary appearance; some demonstrated an appearance in keeping with inflammatory edematous blebs. No tumors showed overt carpet-like erythematous changes as typically seen with urothelial carcinoma in situ.

Figure 2.

Cystoscopy and necropsy visual evaluation of tumors. Legend: Procedure 1: [A1] Day 0 induction with bladder filled with air, instillation of DDM and AdCre; [A2] Day 7 demonstrating papillary edema, multifocal neoplastic changes; [A3] Day 21 necropsy demonstrating multifocal erythematous papillary neoplasms. Procedure 2: [B1] Day 0 site of denudation in one location; [B2] Day 7 demonstrating scar and multifocal papillary edematous changes; [B3] Day 21 necropsy demonstrating multifocal erythematous papillary neoplasms. Procedure 3: [C1] Day 0 site of injection showing submucosal bleb; [C2] Day 7 demonstrating persistent localized unifocal edematous change; [C3] Day 21 necropsy demonstrating hemorrhagic unifocal edematous neoplasm.

Pathologic evaluation

Pathologic evaluation revealed three distinct lesions: neoplastic tumors, inflammatory tumors, and non-tumor reactive changes. Neoplastic tumors were composed of atypical cells forming two patterns: epithelioid cells arranged in branching cords located at the superficial regions of the tumor near the urothelium, and sarcomatoid round and spindle-shaped cells forming poorly organized sheets and streams located throughout the tumors. On IHC, neoplastic cells displayed immunoreactivity for cytokeratins (pan-cytokeratin WSS and AE1/AE3, high molecular cytokeratin 34βE12, and cytokeratin 5, 8/18 and 14) and vimentin, compatible with epithelial and mesenchymal differentiation, respectively (Figure 3). These tumors were thus interpreted as urothelial cell carcinoma with sarcomatoid variant. Most neoplastic tumors were strongly reactive for p53 on IHC, contained an abundance of mixed inflammatory infiltrates (composed of CD3+ T lymphocytes, Iba1+ macrophages and giant cells, neutrophils, and eosinophils), and a moderate density of tumor associated blood vessels with endothelial cells positive for Von Willebrand factor (Supplementary Table 2).

Figure 3.

Histopathological and immunohistochemical representative slides.

Inflammatory tumors displayed a marked inflammatory infiltrate comprised of CD3+ T lymphocytes, Iba1+ macrophages and giant cells, neutrophils, and eosinophils, alongside rare small clusters of atypical cells with intense p53 nuclear activity, representing <1% of the tumor volume. Non-tumor reactive changes were composed of a marked inflammatory infiltrate on a background of fibrovascular stroma, without evidence of neoplastic markers on H&E or IHC.

The urothelium overlying all tumors (neoplastic or inflammatory) displayed reactive changes including hyperplasia and mucinous metaplasia, but no preneoplastic or neoplastic changes, and rare or absent nuclear p53 staining. Samples with non-tumor reactive changes demonstrated normal urothelium or urothelial hyperplasia +/- minimal lymphocytic inflammatory infiltrates and edema of the lamina propria.

All bladders were intact at necropsy and none demonstrated evidence of local tumor progression beyond the bladder serosa. No pigs experienced tumor growth leading to ureteral obstruction, lymph node or distant metastases. No tumors developed in the upper urinary tracts.

Genomic characterization

Neoplastic tumors were genomically characterized in four Oncopigs (N = 2 from Procedure I, and N = 2 from Procedure II). Somatic genomic alterations in TP53 and KRAS (i.e., induced driver mutations in Oncopigs) were confirmed across all analyzed neoplastic tumors (Supplemental Figure 2). Additional somatic alterations were also identified, 87% of which were missense mutations. Mutations profiles were generally confined to each Oncopig, with few shared mutations (Supplemental Figure 3). Oncopig neoplastic tumors demonstrated at least one mutation in 10 of 17 (59%) genes frequently mutated in human urothelial cancer (Supplement Figure 4), suggesting relative genomic similarity to human UCB.

Safety

No animals required early euthanasia on the basis of health concerns. One Oncopig demonstrated distress with decreased activity and oral intake with a low-grade fever (Tmax 104.7 F) at Day 10 following inoculation with Procedure II. Bloodwork at the time demonstrated a leukocytosis with (34.0 K/uL) with moderate neutrophilia (24.0 k/uL) consistent with an infectious/inflammatory process without evidence of acute kidney injury. Based on the clinical features and history of recent cystoscopic instrumentation, a presumptive diagnosis of bacterial cystitis was made. The animal was treated with a 10-day course of oral antibiotics and recovered to baseline function with normalized bloodwork by post-treatment day 7.

Discussion

We successfully characterized a large animal model for developing intravesical bladder tumors in pigs, using transgenic Oncopigs carrying Cre-inducible KRAS^G12D and TP53^R167H mutations. Tumor inductions were evaluated with varying degrees of success for the development of inflammatory and neoplastic tumors. Inoculations were safe, straightforward, site-specific and reproducible, and lead to tumor development within a week, with muscle-invasion seen as early as 14 days. Histological and IHC studies confirmed both non-muscle invasive and muscle-invasive bladder tumors of both mesenchymal and epithelial origin that histologically and genomically resembled human urothelial cell carcinoma with sarcomatoid differentiation; inflammatory tumors demonstrated benign growths with inflammatory infiltrate.

The Oncopig offers the only model for rapid induction of reproducible urothelial cancers in large animals with physiological similarities to humans. While canine models have demonstrated spontaneous bladder tumor development, their late onset (typically after 10 years) and variability in tumor development render them impractical.²¹ In contrast, our study was able to consistently induce bladder tumors within two weeks of AdCre inoculation, reducing need for prolonged animal housing and optimizing efficiency for testing novel therapeutics. Moreover, tumor inoculation was relatively simple and without need for additional procedures or significant technical demand; the procedures studied herein may be performed by any urologist or technician trained in basic cystoscopic technique. Cystoscopically, neoplastic tumors appeared as submucosal nodules or papillary lesions and inflammatory tumors appeared as inflammatory blebs, similar to that seen in humans. In contrast to murine models, the Oncopig bladder tumor model allows investigators to use and study technologies that can be directly used in humans including novel transurethral surgical instruments, anti-tumor injectable and intravesical agents, and photodynamic-therapy based approaches. In addition, tumors were induced in Oncopigs autochthonously and without immunosuppression, allowing for tumor evolution, stromal and immune interactions, and locoregional spread to occur in vivo in a manner biologically resembling humans. Oncopig BC cell lines exhibit treatment responses and gene expression profiles closely mirroring those of human urothelial cancer when exposed to standard chemotherapeutic agents, underscoring the model's translational relevance for preclinical therapeutic testing.²² This model may facilitate the study of the complex interplay between tumors and their local microenvironment, as well as to test responses to novel systemic immunotherapies including checkpoint inhibitors in an immunocompetent setting. The tumors induced herein are also uniquely positioned to study emerging KRAS-targeted therapies including novel KRAS^G12C/G12D-inhibitors.²³ Taken together, the Oncopig bladder cancer model may help decrease the translational gap wherein many therapies that succeed in rodents fail clinically.^4,5

Analyses of TCGA and UROMOL datasets have shown RAS pathway alterations, including KRAS, to be enriched in NMIBC (10–20% mutation frequency), while TP53 mutations predominate in MIBC (50–60%).^9,10 These alterations are typically mutually exclusive, supporting distinct molecular trajectories of bladder cancer pathogenesis. Accordingly, the Oncopig model, which encompasses both KRAS and TP53 mutations, may capture features across both NMIBC and MIBC and recapitulate key biological pathways seen in human urothelial cancer.

This is the first model of the Oncopig where differentiated urothelial carcinomas of sarcomatoid variant were established. Other models utilizing the Oncopig such as soft tissue sarcomas, liver, pancreas and lung only showed poorly or un-differentiated sarcomatoid tumors with inflammatory infiltration, and the cell of origin could not be identified with certainty.^12–14^,24 In the present study, while the tumors were histologically poorly differentiated, the presence of epithelioid cords in the superficial regions near the urothelium, and the expression of cytokeratins in these cords and in some of the sarcomatoid atypical cells in deeper regions, strongly suggest that the neoplastic population originated from the urothelium. The sarcomatoid population most likely represent epithelial to mesenchymal transition of the neoplastic urothelial cells, but we cannot rule out that this component may have originated from the transformation of mesenchymal cells, such as smooth muscle cells or fibroblasts, in the bladder wall, as the viral vector did not specifically target urothelial cells.²⁵

We investigated three cystoscopy-based approaches to tumor induction, including GAG dissolution followed by AdCre instillation (Procedure I), mechanical mucosal denudation followed by AdCre instillation (Procedure II), and direct submucosal injection of AdCre (Procedure III). While Procedure I yielded the highest inoculation success with all animals developing neoplastic bladder cancers, only 30% were muscle-invasive and only encountered 3-weeks post-inoculation. In contrast, while the inoculation success rate was lower with Procedure II (50%), all neoplastic bladder cancers that developed were muscle-invasive with onset as early as 2-week post-inoculation. Taken together, selective implementation of these procedures and post-inoculation surveillance intervals may be utilized to study novel locoregional therapies in non-muscle invasive or muscle-invasive disease, based on the interests of prospective investigators.

Procedure III only resulted in inflammatory nodules, similar to that of other Oncopig organs wherein injection-based tumor inductions have been undertaken.^13,14 Potential explanations for these observations may relate to the method of vector delivery, vector titre, or vehicle medium. It is possible that direct submucosal AdCre injection led to a more robust immune response compared to its topical administration. Indeed we and others have demonstrated a marked inflammatory component in pulmonary, intramuscular, hepatic and pancreatic tumors after injection-based inductions in Oncopigs, composed of an infiltrate consisting of multiple types of leukocytes, including cytotoxic and regulatory T cells and only rare neoplastic cells.^26,27 Moreover, the higher AdCre vector titers or the potential vector medium (plant starch or porcine gelatine), may have further aggravated a strong immune response. Taken together, it is conceivable that in this setting, a localized strong immune response may more effectively identify and control early neoplastic changes, ultimately inhibiting tumor development and leaving only inflammatory tissue identifiable at final pathology. While immune infiltration was also observed in neoplastic tumors in Procedures I and II, it may not have been sufficiently robust to overcome tumor proliferation thereby leading to cancerous neoplastic growth.

Several limitations of our work warrant acknowledgement. First, while rapid tumor induction was achieved over study period, longer follow-up is needed to understand tumor biology as it relates to progression, metastasis, or regression. Indeed, some tumors have been noted to spontaneously regress in the lung Oncopig model after 2-weeks.¹⁴ It should be noted that in our study all induced tumors remained stable or grew during follow-up. Second, this model is not urothelium-specific, with AdCre indiscriminately transducing all exposed cell types. This may in part explain the sarcomatoid components we observed (from mesenchymal transformation) and the development of inflammatory pseudotumors, if off-target cells are infected. Refining gene delivery with urothelium-specific promotors/vectors may lead to higher yields of pure urothelial cancers. Additionally, the inducible mutations carried in Oncopigs (KRAS, TP53), while encountered to varying degrees in human urothelial cancer, represents only a small focus of the heterogenous mutational landscape of human urothelial cancer and may not capture the molecular diversity of human bladder cancer.²⁸ Third, our study was only able to adequately differentiate malignant vs benign tumors at final pathologic evaluation. Thus prospective studies may warrant cystoscopic tissue sampling to confirm the presence of neoplastic tumors prior to interventions, particularly if induction success is imprecise (i.e., Procedure II). Fourth, our experimental design did not include a control group with scrambled AdCre placement. The absence of tumor formation in a subset of animals serves as a partial internal control. In addition, external validation confirming AdCre-induced tumors in other Oncopig models of liver, lung, and pancreas, as well as genomic analyses on four neoplastic urothelial tumors in our study confirming induced KRAS^G12D and TP53^R167H mutations, supports inoculation-driven tumor induction. Lastly it must be noted that large animal models such as the Oncopig can initially carry a large financial burden, although this is expected to decline with further protocol standardization and broader adoption.

Conclusion

In summary, this study utilized multiple cystoscopy-based procedures to develop an in vivo large animal model of bladder cancer using transgenic Oncopigs. Neoplastic tumors were rapidly and reproducibly induced, and grossly, histologically, and genetically resembled human urothelial cancers with sarcomatoid differentiation. This large animal model, given its anatomical and physiological similarities to humans, coupled with the ability to perform clinically relevant procedures and test novel therapeutics in an immunocompetent setting, is a valuable tool for advancing bladder cancer research and therapy development.

Prior Presentation: SUO 2023, AUA 2023, GU ASCO 2025

Supplemental Material

sj-docx-1-blc-10.1177_23523735261439691 - Supplemental material for Induction and characterization of neoplastic bladder tumors in a transgenic porcine model

Supplemental material, sj-docx-1-blc-10.1177_23523735261439691 for Induction and characterization of neoplastic bladder tumors in a transgenic porcine model by Andreas Aulitzky, Viranda H Jayalath, Abraham Meyerson, Sanaz Firouzi, Daniel V Rodriguez, Rebecca Dubrovsky, Laura Alvim, Caoimhe Ryan, Rand Wilcox Vanden Berg, Christopher Cheleuitte-Nieves, Alexandre Doudt, Lennert Eismann, Natasha Kudinova, Kwanghee Kim, Hikmat Al-Ahmadie, Sebastien Monette, Shahrokh F Shariat and Jonathan A Coleman in Bladder Cancer

Footnotes

Acknowledgements

The authors thank Margaret Reilly, Joann Phifer, Yuliana Miranda, Stephanie Harris-Ash, and Brian Cuevas who helped with animal experiments; we thank Jacqueline Candelier who helped with necropsy.

ORCID iDs

Andreas Aulitzky

Abraham Meyerson

Daniel V Rodriguez

Rebecca Dubrovsky

Laura Alvim

Christopher Cheleuitte-Nieves

Alexandre Doudt

Hikmat Al-Ahmadie

Ethical considerations

The conduct of this prospective study was approved by the Institutional Animal Care and User Committee at Memorial Sloan Kettering Cancer Center (IACUC Protocol #: 11-02-004). This study was reported in accordance with The ARRIVE guidelines 2.0 (Appendix A).

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Author contributions

Conceptualization: Andreas Aulitzky, Jonathan A. Coleman, Kwanghee Kim

Formal analysis: Andreas Aulitzky, Viranda H. Jayalath, Sebastien Monette, Sanaz Firouzi, Hikmat Al-Ahmadie

Funding acquisition: Jonathan A. Coleman

Investigation: Andreas Aulitzky, Viranda H. Jayalath, Rand Wilcox Vanden Berg, Rebecca Dubrovsky, Christopher-Cheleuitte-Nieves, Abraham Meyerson, Daniel V. Rodriguez, Laura Alvim, Caoimhe Ryan, Natasha Kudinova, Alexandre Doudt, Lennert Eismann

Methodology: Andreas Aulitzky, Jonathan A. Coleman, Sharokh Shariat

Project administration: Andreas Aulitzky, Viranda H. Jayalath, Jonathan A. Coleman

Resources: Kwanghee Kim, Rebecca Dubrovsky, Laura Alvim, Caoimhe Ryan

Supervision: Jonathan A. Coleman

Visualization: Andreas Aulitzky, Viranda H. Jayalath, Abraham Meyerson, Sanaz Firouzi, Daniel Rodriguez

Writing – original draft Viranda H. Jayalath, Andreas Aulitzky, Abraham Meyerson, Sanaz Firouzi, Daniel V. Rodriguez

Writing – review & editing, Viranda H. Jayalath, Andreas Aulitzky, Abraham Meyerson, Sanaz Firouzi, Daniel V. Rodriguez, Rebecca Dubrovsky, Laura Alvim, Caoimhe Ryan, Rand Wilcox Vanden Berg, Christopher Cheleuitte-Nieves, Alexandre Doudt, Lennert Eismann, Natasha Kudinova, Kwanghee Kim, Hikmat Al-Ahmadie, Sebastien Monette, Sharokh Shariat, Jonathan A. Coleman

Funding

The authors disclosed the receipt of the following financial support for the research, authorship, and publication of this article: This work was supported by the Thompson Family Foundation; National Institutes of Health/National Cancer Institute Cancer Center support grant [P30 CA008747].

Declaration of conflicting interest

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr Sharokh Shariat is an Editorial Board Member of this journal, but was not involved in the peer-review process nor had access to any information regarding its peer-review.

Data availability

The data is available to bona fide researchers who request it from the authors. Data used in this experiment can be found at Memorial Sloan Kettering Cancer Center.

Supplemental material

Supplemental material for this article is available online.

References

Sung

Ferlay

Siegel

, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021; 71: 209–249.

Burger

Catto

Dalbagni

, et al. Epidemiology and risk factors of urothelial bladder cancer. Eur Urol 2013; 63: 234–241.

Christopher

Alavijeh

Palmer

, et al. The role of mouse tumour models in the discovery and development of anticancer drugs. Br J Cancer 2019; 121: 101–108.

Sun

Gao

, et al. Why 90% of clinical drug development fails and how to improve it?. Acta Pharmaceu Sinica B 2022; 12: 3049–3062.

Mak

Evaniew

Ghert

. Lost in translation: animal models and clinical trials in cancer treatment. Am J Transl Res 2014; 6: 114–118.

Swindle

Makin

Herron

, et al. Swine as models in biomedical research and toxicology testing. Vet Pathol 2012; 49: 344–356.

Schook

Collares

, et al. A genetic porcine model of cancer. PLoS One 2015; 10: e0128864.

Pylayeva-Gupta

Grabocka

Bar-Sagi

. RAS Oncogenes: weaving a tumorigenic web. Nat Rev Cancer 2011; 11: 761–774.

Jebar

Hurst

Tomlinson

, et al. FGFR3 And Ras gene mutations are mutually exclusive genetic events in urothelial cell carcinoma. Oncogene 2005; 24: 5218–5225.

10.

Robertson

Kim

Al-Ahmadie

, et al. Comprehensive molecular characterization of muscle-invasive bladder cancer. Cell 2017; 171: 540–556 e25.

11.

Shariat

Bolenz

Karakiewicz

, et al. P53 expression in patients with advanced urothelial cancer of the urinary bladder. BJU Int 2010; 105: 489–495.

12.

Gaba

Elkhadragy

Boas

, et al. Development and comprehensive characterization of porcine hepatocellular carcinoma for translational liver cancer investigation. Oncotarget 2020; 11: 2686–2701.

13.

Boas

Nurili

Bendet

, et al. Induction and characterization of pancreatic cancer in a transgenic pig model. PLoS One 2020; 15: e0239391.

14.

Ghosn

Elsakka

Petre

, et al. Induction and preliminary characterization of neoplastic pulmonary nodules in a transgenic pig model. Lung cancer (Amsterdam, Netherlands) 2023; 178: 157–165.

15.

Sert

Ahluwalia

Alam

, et al. Reporting animal research: Explanation and elaboration for the ARRIVE guidelines 2.0. PLoS Biol 2020; 18: e3000411.

16.

NS P

, et al. Induction of pancreatic neoplasia in the KRAS/TP53 Oncopig - PubMed. Dis Model Mech 2023; 16: dmm049699.

17.

, et al. Grading of Urothelial Carcinoma and The New “World Health Organisation Classification of Tumours of the Urinary System and Male Genital Organs 2016”. Eur Urol Focus 2019; 5: 457–466.

18.

Hurst

. Structure, function, and pathology of proteoglycans and glycosaminoglycans in the urinary tract. World J Urol 1994; 12: 3–10.

19.

Ramesh

Memarzadeh

, et al. Identification of pretreatment agents to enhance adenovirus infection of bladder epithelium. Mol Ther 2004; 10: 697–705.

20.

Amin

Greene

Edge

, et al. The Eighth Edition AJCC Cancer Staging Manual: Continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging. CA Cancer J Clin 2017; 67: 93–99.

21.

de Brot

Robinson

Scase

, et al. The dog as an animal model for bladder and urethral urothelial carcinoma: comparative epidemiology and histology. Oncol Lett Aug 2018; 16: 1641–1649.

22.

Segatto

Simões

Bender

, et al. Frontiers | Oncopig bladder cancer cells recapitulate human bladder cancer treatment responses in vitro. Front Oncol 2024; 14: 1323422.

23.

Tang

Kroemer

Kang

. KRAS-targeted therapies in cancer: novel approaches and overcoming resistance. BMJ Oncology 2025; 4: e000946.

24.

Schachtschneider

Liu

Mäkeläinen

, et al. Oncopig Soft-Tissue Sarcomas Recapitulate Key Transcriptional Features of Human Sarcomas. Sci Rep 2017; 7. doi:10.1038/s41598-017-02912-9

25.

Xiao

Mohseny

Hogendoorn

PCW

, et al. Mesenchymal stem cell transformation and sarcoma genesis. Clin Sarcoma Res 2013; 3: 10.

26.

Overgaard

Principe

Schachtschneider

, et al. Frontiers | genetically induced tumors in the oncopig model invoke an antitumor immune response dominated by Cytotoxic CD8β+ T cells and differentiated γδ T cells alongside a regulatory response mediated by FOXP3+ T cells and immunoregulatory molecules. Front Immunol 2018; 9: 1301.

27.

Fenner

. Immune infiltration associated with outcomes. Nature Rev Urology 2022; 19: 256.

28.

Segatto

Bender

Seixas

, et al. Perspective: humanized pig models of bladder cancer. Front Mol Biosci 2021; 8: 681044.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.04 MB