A novel approach to engineering three-dimensional bladder tumor models for drug testing | Scientific Reports

Scientific Reports volume 14, Article number: 26883 (2024) Cite this article

Metrics details

Bladder cancer (BCa) poses a significant health challenge, particularly affecting men with higher incidence and mortality rates. Addressing the need for improved predictive models in BCa treatment, this study introduces an innovative 3D in vitro patient-derived bladder cancer tumor model, utilizing decellularized pig bladders as scaffolds. Traditional 2D cell cultures, insufficient in replicating tumor microenvironments, have driven the development of sophisticated 3D models. The study successfully achieved pig bladder decellularization through multiple cycles of immersion in salt solutions, resulting in notable macroscopic and histological changes. This process confirmed the removal of cellular components while preserving the native extracellular matrix (ECM). Quantitative analysis demonstrated the efficacy of decellularization, with a remarkable reduction in DNA concentration, signifying the removal of over 95% of cellular material. In the development of the in vitro bladder cancer model, muscle invasive bladder cancer patients’ cells were cultured within decellularized pig bladders, yielding a three-dimensional cancer model. Optimal results were attained using an air–liquid interface technique, with cells injected directly into the scaffold at three distinct time points. Histological evaluations showcased characteristics resembling in vivo tumors derived from bladder cancer patients’ cells. To demonstrate the 3D cancer model’s effectiveness as a drug screening platform, the study treated it with Cisplatin (Cis), Gemcitabine (Gem), and a combination of both drugs. Comprehensive cell viability assays and histological analyses illustrated changes in cell survival and proliferation. The model exhibited promising correlations with clinical outcomes, boasting an 83.3% reliability rate in predicting treatment responses. Comparison with traditional 2D cultures and spheroids underscored the 3D model’s superiority in reliability, with an 83.3% predictive capacity compared to 50% for spheroids and 33.3% for 2D culture. Acknowledging limitations, such as the absence of immune and stromal components, the study suggests avenues for future improvements. In conclusion, this innovative 3D bladder cancer model, combining decellularization and patient-derived cells, marks a significant advancement in preclinical drug testing. Its potential for predicting treatment outcomes and capturing patient-specific responses opens new avenues for personalized medicine in bladder cancer therapeutics. Future refinements and validations with larger patient cohorts hold promise for revolutionizing BCa research and treatment strategies.

Bladder cancer (BCa) is four times more prevalent in men than in women, with global incidence and mortality rates of 9.5 and 3.3 per 100,000 among men, and 2.4 and 0.9 for women, respectively1. In Canada, BCa ranks as the fifth most common cancer, with an estimated 13,400 new cases and 2,600 deaths in 20232,3. Non-muscle-invasive bladder cancer (NMIBC) constitutes 75% of all bladder tumors, characterized by preserved detrusor muscle and invasion limited to outer, mucosal, and submucosal layers. Although NMIBC generally has a favorable prognosis, its high recurrence rates contribute to its status as the most expensive cancer to treat8. Muscle-invasive bladder cancer (MIBC), accounting for the remaining 25% of cases, poses a greater risk of metastasis and requires more aggressive management, often involving neoadjuvant chemotherapy before radical cystectomy4,5,6,7.

The challenge in BCa treatment lies in predicting the efficacy of neoadjuvant chemotherapy, as identifying chemo-sensitive cancer is intricate4,5,6,7. Without reliable predictive biomarkers, treatment decisions rely on clinical evidence and expert opinion, highlighting the critical need for personalized approaches. Although some predictive tumor models exist, they fall short in meeting all requirements.

Traditional 2D cell cultures, while valuable for understanding molecular pathways, lack the complexity to predict responses to therapy effectively8,9,10. In contrast, 3D cultures, such as spheroids and organoids, better mimic the tumor microenvironment (TME), incorporating diverse cell types and extracellular matrices11,12. Patient-derived 3D organoids from BCa tissue exhibit promising genetic and histological correlations, proving useful for molecular investigations and preclinical drug testing, albeit with limitations in representing immune and stroma cells for immunotherapy studies13,14,15,16.

Patient-derived tumor xenografts (PDTX) currently stand as the most studied predictive tumor model, offering a closer in vivo environment resembling the original tumor conditions17,18,19,20. Despite their advantages, PDTX models are labor-intensive, require substantial animal resources, and are time-consuming and expensive, limiting their application primarily to research studies.

Recognizing the need for economical, reliable, and high-throughput models, this project proposes the development of a 3D in vitro patient-derived BCa tumor model. This innovative model utilizes decellularization, a biomedical engineering process isolating the extracellular matrix (ECM) of a tissue without its cells. The resulting non-immunogenic ECM scaffold can be reseeded with the host’s cells, promoting migration and differentiation. By ensuring complete decellularization, this model holds promise for predicting responses to therapies in a manner that is both clinically relevant and expedient, ultimately offering a personalized and efficient preclinical platform for BCa treatment outcomes.

Animal organ retrieval was conducted in accordance with the institutional guidelines, approved by the Institutional Review Board of UBC (Vancouver, British Columbia, Canada; A22-0119). A total of 3 animals were euthanized under the corresponding primary research protocol and used for the experiments. Animals were euthanized under the corresponding primary research protocol. Immediately after euthanizing the pig, the abdominal cavity was opened with a midline incision, the bladder was identified and removed. The bladder was cut into multiple small pieces (1 cm x 1 cm) for faster decellularization, that were immediately placed in cold PBS solution.

Bladders were collected as described above, pieces were immersed in Dulbecco Modified Eagle Medium (DMEM) (Fisher Scientific, 13,345,364, PA, USA) with 50 nM Latrunculin B (Cayman Chemical, CAS 76343–94-7, Michigan, USA) and incubated for 120 min at 37 °C. After incubation, tissues were washed with dH2O, then immersed in sterile 0.6 M potassium chloride solution (Fisher Scientific, AC424090010, PA, USA) at room temperature (RT) with agitation on a shaker for 2 h. The tissues were then immersed into sterile 1 M potassium iodine solution (Sigma-Aldrich, 03,124, Darmstadt, Germany) at RT with agitation on a shaker for 2 h, followed by immersion in sterile dH2O at room temperature overnight. The above-described steps were considered one full cycle. Each cycle was repeated 10–12 times depending on the dimensions of the tissue to ensure proper decellularization. Finally, tissues were then incubated in DNase I (1 kU/mL; Sigma-Aldrich, D4513, MO, USA) for 120 min and washed in dH2O for 2 days with daily water changes to remove the remaining reagents. Decellularized tissue was stored at -80 °C for further work.

Both native and decellularized tissue were cut into 25 mg pieces prior to DNA extraction procedures. The small pieces were digested individually using Proteinase K (QIAGEN, 19,133, Hilden, Germany) at 56 °C with agitation until they were completely lysed. DNA was purified using the DNeasy Blood & Tissue Kit (QIAGEN, 69,504, Hilden, Germany) according to the manufacturer’s instructions. Afterward, both native and decellularized tissue’s DNA extracts were quantified spectrophotometrically using NanoDrop technology (Thermo Scientific, ND-2000, MA, USA).

A small piece of the native and decellularized bladder tissue was fixed with 4% paraformaldehyde (Electron Microscopy Sciences, 15,710, PA, USA) in 0.1 M PIPES (Sigma-Aldrich, P6757, Darmstadt, Germany), followed by a second fixation in 2.5% glutaraldehyde (Electron Microscopy Sciences, 16,020, PA, USA) in 0.1 M PIPES buffer at RT. Subsequently, the tissue was fixed in 1% Osmium Tetroxide (OsO4) (Electron Microscopy Sciences, 19,150, PA, USA) in 0.1 M PIPES buffer at pH 6.8, at RT. After fixation, the tissues were rinsed with double distilled water (ddH2O). After rinsing, the specimen underwent dehydration at RT through a graded ethanol–water series, followed by three rounds of 100% ethanol (Electron Microscopy Sciences, 15,056, PA, USA). Critical point drying with CO2 (Tousimis Samdri®-795 critical point dryer) was then applied for 1 h to ensure complete dehydration, and the specimen was mounted on an aluminum stub using sticky carbon tapes. Finally, the specimen on the stub was then coated with a thin layer of Au/Pd coating (2 nm thickness) using the Leica EM MED020 Coating (Centre for High-Throughput Phenogenomics, UBC, Vancouver, Canada). Images were recorded with a Helios NanoLab 650 Focused Ion Beam SEM (Centre for High-Throughput Phenogenomics, UBC, Vancouver, Canada).

Human urinary bladder transitional cell carcinoma (UM-UC3) were cultured in Minimal Essential Medium (MEM) (Gibco, 11,095,080) supplemented with 10% Fetal bovine serum (FBS) (Gibco, 12,483,020). All media used in this study were supplemented with 1% Antibiotic–Antimycotic (Gibco, 15,240–062, ON, Canada) to prevent bacterial and fungal contamination. All cells used were cultured at 37° C in a 5% CO2 incubator and mycoplasma contamination was tested at regular intervals for each cell line or primary cell. When cells were confluent, they were passaged by incubation at 37 °C for 3–4 min with 0.25% trypsin (Gibco, 25,200,056, ON, Canada), centrifuged at 1,200 RPM for 5 min and resuspension in medium. Cells were stored long term in Bambanker (NIPPON Genetics, BB01, Düren, Germany) at a concentration of one million cells per mL and in liquid nitrogen.

Cells were trypsinized, washed, and centrifuged to form a pellet. The pellet was then resuspended in Matrigel (Corning, 354,234, ME, USA) to generate spheroids at a density of 1,000 cells per μL of Matrigel. A 20 μL aliquot of the cell suspension was plated in each well of a 48-well plate. The plate was incubated at 37 °C for 30 min to allow the Matrigel to solidify. After solidification, culture media was gently added. The media was supplemented with 10% R-Spondin1 Conditioned Medium from Cultrex HA-R-Spondin1-Fc 293 T Cells (R&D Systems 3710-001-01) and 5 μM Y-27632 (Enzo, ALX-270–333-M001). The spheroids were maintained in a humidified incubator at 37 °C with 5% CO2 for five days, with media being changed daily, before initiating treatment.

Human BCa tissue was obtained from the Vancouver General Hospital, Vancouver, BC from 17 BCa patients. All patients provided informed consent, and this study was approved by UBC Clinical Research Ethics Board Chair (Protocol #H19-00,814). Tumor tissue was utilized to secure a heterogeneous cell population, which includes both cancer cells and other cell types, such as immune and stromal cells. Tissue was washed 3 times with sterile PBS and 10% Antibiotic-Antimycotic (Ghibco, 15,240–062, ON, Canada), then minced with a scalpel and incubated with Collagenase A (Sigma Aldrich, 10,103,586,001) and Dispase II (Gibco, 17,105–041) for 2 h at 37 °C with occasional agitation. After this, cells were filtered through a 40um cell strainer. Cells were washed with sterile 1X PBS, and then incubated for 15 min at 37 °C with TrypLE express enzyme (Thermo fisher, 12,605–028). Cells were washed once with 1X PBS and incubated for 5 min with 1X Red Blood Cell (RBC) Lysis Buffer (ABCam, ab204733). The cellular pellet was resuspended in a 1:1:1ratio mixture of REBM (Lonza, CC-3190), EBM-2 (Lonza, CC-3202), and DMEM (Gibco, 11,965,092) media, and then cultured at 37 °C in 5% CO₂.

To create the cancer model, first decellularized bladders were cut with a punch biopsy into 5 mm diameter pieces and placed on cell culture inserts (Fisher Scientific, 08–771, PA, USA), and then, these inserts were placed in 6-well plates (Corning, CL-S3516, ME, USA) to create an air–liquid interface. The muscularis propria of the bladder was partially removed in order to expose the stroma for the recellularization. A total of 1.5 × 106 UM-UC3 cells or BCa patient cells were injected into the decellularized pig bladder, at 3 different time points (Day 0, 3 and 5). Tissues were incubated for 9 days more with corresponding growth media (3 mL of each well) at 37° C with 5% CO2. The media was changed for fresh media every other day. On day 14, the recellularized tissues would be fixed in 10% formalin and embedded in paraffin, and 5 μm-thick sections were obtained for H&E and immunohistochemistry (IHC) evaluations.

Cells were seeded in 96-well plates at 4000 cells/well at 37° C and 5% CO2. After 24 h, different concentrations of drugs (Cisplatin (Cis), Gemcitabine (Gem) and combination) were added and media alone was used as a control. Cells were incubated in treatment media for 72 h. Cell viability was measured with MTS reagent (Sigma-Aldrich, MO, USA) in 200 μl of fresh media (1:20 ratio) incubated at 37 °C, in 5% CO2, and plate readings were taken at 60 min at 490 nm (BioTek, VT, USA). Each experiment had 3 technical replicates.

Spheroids were evaluated for cell viability using luminescence at day 0 (pre-treatment) and day 6 (post-treatment) following the addition of 50 µL of CellTiter-Glo® 3D Cell Viability Assay (Promega, Madison, WI, G9681) to each well. The plates were then agitated on a shaker for 5 min and incubated for an additional 25 min on a rocking platform at room temperature before luminescence was measured using a Tecan Infinite M200 Luminometer. All the tests were conducted in triplicate and standard deviations were reported.

All tissues were fixed overnight in neutral buffered formalin (10% formalin) (Fisher Scientific, 22–046-361, PA, USA) and then transferred to 70% ethanol prior to paraffin embedding. Formalin-fixed, paraffin-embedded samples were sectioned into 5 μm thickness, placed on glass slides (Fisher Scientific, 12–550-15, USA). After the tissues were deparaffinized and dehydrated, slides were washed with distilled water, hematoxylin solution, Gill No.2 (Sigma Aldrich, GHS232, Darmstadt, Germany) for 3–5 min and rinsed with tap water followed by immersion in Shandon bluing reagent (Thermo Fisher, 6,769,001, MA, USA) for 30 s. The sections were again rinsed in tap water and stained with Eosin Y-solution (Millipore Sigma, 1,098,441,000, Darmstadt, Germany) for 30 s. We detected apoptotic cells of the tumor model with a TUNEL staining. After dewaxing and dehydrating, paraffin sections were stained with a TUNEL assay kit (Abcam, ab206386, Cambridge, MA, USA) according to the manufacturer’s instructions. For Ki-67 labelling, sections were stained with Ki-67 Monoclonal Antibody 1:1000 (eBioscience™13–5698-82). All coverslips were mounted using CitosealTM XYL (Thermo Fisher Scientific, 8312–4). The slides were then scanned with an Aperio Digital Whole Slide Scanner (Leica Biosystems).

Drugs were selected based on current application in the clinic for BCa treatment Cis (Cisplatin Injection 100 mg/100 ml, Teva Standard) and Gem (Gemcitabine injection 2 g/52.6 m). For this study, both drugs were obtained from BC Cancer Hospital, Vancouver, British Columbia, Canada as a donation.

A robust preclinical model system must reflect the genetic heterogeneity of tumors to guide optimal treatment, meaning it should retain the genetic alterations of the primary tumor . To assess the genetic stability of organoid lines in culture, as well as in orthotopic xenografts and xenograft-derived organoids, deep sequencing is typically performed to compare their mutational profiles42. To evaluate genetic alterations in our 3D bladder cancer model derived from patient cells, we isolated DNA at days 7, 14, and 21. Prior to DNA library preparation, tissue samples were fragmented to a median target insert size of 200 bp using the Covaris M220 focused-ultrasonicator. Sequencing libraries were constructed from 50 or 100 ng of fragmented input DNA using the KAPA HyperPrep kit and IDT xGen CS adapters. Libraries were pooled and hybridized to a custom KAPA HyperDesign target capture panel spanning 60 bladder cancer-associated genes and 3000 evenly distributed SNPs. Sequencing was performed on the Illumina NovaSeq 6000 using a 2 × 150 bp S4 kit. Analysis was performed using previously published custom in-house bioinformatic tools21. For DNA extracted from unfixed cells, a minimum of 5 supporting reads and a minimum VAF of 1% were required for somatic variant calling. For FFPE-derived samples, 8 supporting reads and 5% VAF were used.

We performed statistical analysis using Prism GraphPad software version 8. We used student–test to compare two variables. One-way ANOVAs were used for multiple comparisons, followed by Bonferroni post hoc testing or 2-way ANOVA when comparing experimental multiple groups. Quantitative data were expressed as means ± SD when relevant. P values of < 0.05 were considered statistically significant.

Bladder tissue decellularization was successfully achieved in pig bladders through ten cycles of immersion in salt solutions, resulting in a noticeable macroscopic color change (Fig. 1A,B). Histological analysis using H&E staining confirmed the preservation of the microarchitecture typical of a bladder, with intact urothelium, muscle layer, and vascular structures. Notably, all cellular components were effectively removed. Scanning electron microscopy (SEM) further validated complete cell removal, revealing an undisturbed extracellular matrix (Fig. 1A–D).

Macroscopic visualization and histological examination of native (A1, A2, A3) and decellularized (B1, B2, B3) bladders. (A1) Macroscopic native pig bladder. (B1) Macroscopic decellularized pig bladder. H&E stains: native (A2) and decellularized (B2) bladder with 4X magnification. (A3) Native bladder square with 20X magnification. (B3) Decellularized bladder square with 20X magnification. Scale bars in panels A2, A3, B2 and B3 represent 100 μm. Structural analysis of native and decellularized bladders. SEM images: urothelium of native bladder (C1) (250X) and (C2). Basement membrane (lamina propria) of decellularized bladder (D1) (250X) and (D2) (500X). DNA Quantification of native and decellularized bladders. Native bladders have mean DNA concentration of 213.8 ng/μL (n = 4) and decellularized bladders have mean DNA concentration of 4.65 ng/μL (n = 4). Statistical analysis was performed using the unpaired t test, p < 0.0001 (E).

Quantitative assessment of DNA concentration showed a significant reduction in the decellularized bladder compared to the native bladder (4.65 ng/μl vs. 213.8 ng/μl, p < 0.0001 by Student’s t-test). This indicates that more than 95% of DNA was successfully eliminated during the decellularization process (Fig. 1E).

The development of an in vitro BCa model involved successful growth of UM-UC3 cells within the decellularized pig bladder, following protocol optimization for the creation of a three-dimensional cancer model. After two weeks in culture, H&E staining images revealed characteristics resembling an in vivo tumor derived from UM-UC3 cells, exhibiting dense cell growth with large nuclei and prominent nucleoli.

Optimal results were achieved using an air–liquid interface technique (Fig. 2A). Cells were injected directly into the scaffold at three different time points (Day 0, 3, and 5), enabling their proliferation and migration throughout the entire tissue. This dynamic process is visually represented in both longitudinal and transversal slides of the tissue (Fig. 2B,C).

Bladder cancer model using a decellularized pig bladder. (A1) Protocol used to recreate a 3D cancer model using a decellularized pig bladder as a scaffold with UM-UC3 cells. Hematoxylin and eosin staining of a recellularized pig bladder transverse section with 2X magnification (B1), 20X (B2) and 40X (B3). Hematoxylin and eosin staining of a recellularized pig bladder longitudinal section with 2X magnification (C1), 20X (C2) and 40X (C3).

To demonstrate the effectiveness of our 3D cancer model as a drug screening platform, we treated the BCa prototype with Cis, Gem, and a combination of both drugs. Subsequently, we conducted a cell viability assay and histological analyses to showcase changes in cell survival and proliferation. In 2D culture, Cis concentrations ranging from 0.1 to 100 μM, with a 72-h exposure period, exhibited an IC50 value of 0.8783 μM (Fig. 3A). For Gem, concentrations ranged from 0.001 to 10 μM, yielding an IC50 value of 10.57 nM (Fig. 3B). When evaluating drug response in BCa spheroid cells exposed to drugs in 2 cycles of 72 h each, Cis concentrations ranged from 0.5 to 200 μM, and the IC50 remained comparable to the 2D culture at 0.8389 μM (Fig. 3C). However, Gem doses, ranging from 0.001 to 1 μM, resulted in an IC50 almost four times higher than in 2D culture, at 40.57 nM (Fig. 3D,E).

UM-UC3 cell viability in standard culture methods. Dose–response curves between different cisplatin (A) and gemcitabine (B) concentrations and cell viability of UM-UC3 cells in 2D culture. Dose–response to different doses of cisplatin (C) and gemcitabine (D) using UM-UC3 spheroids as a 3D culture. Cell viability rates were determined by using MTS assay. The results are expressed as the percentage of the viability rate of control cells. All values are represented as the means ± SD of three independent experiments. (E) UM-UC3 spheroids were treated with increasing doses of cisplatin and gemcitabine, which correlates with graphs C and D. Spheroids were made of 20,000 cells per spheroid. Bars = 1000 μm (4X objective lens).

Using our developed 3D BCa model for drug testing (Fig. 4A), cells were treated with Cis at doses ranging from 0.5 to 25 μM and Gem from 0.005 to 5 µM, over three cycles of 72 h each. As demonstrated in Fig. 4B 1–3, low concentration of both therapies showed up to a 43.8% reduction in luminescence compared to the control. This was confirmed with histological evaluation with H&E staining revealed a decrease in cellularity with increasing Cis dose. Ki67 staining for cell proliferation demonstrated a reduction corresponding to the increase in Cis dose. In contrast, there was a significant increase in TUNEL staining with rising Cis concentrations (Fig. 4C).

Drug testing of 3D Bladder cancer model. (A) Scheme of the protocol used to recreate a 3D cancer model using a decellularized pig bladder as a scaffold with UM-UC3 cells, and the treatment time points. (B) Bioluminescence assay to measure the viability of UMUC3 cells in the 3D cancer models treated with cisplatin (B1), gemcitabine (B2) and a combination (B3), that correlates with the histology. (C) Representative Histology of 3D cancer tumors receiving no treatment, and 0.5, 1.0, 5.0 and 25 μM of cisplatin. The first column shows H&E staining, the second Ki67, and the last column is a TUNEL assay. (D) Representative Histology of 3D cancer tumors receiving no treatment, and 0.005, 0.005, 0.5 and 5 μM of gemcitabine. The first column shows H&E staining, the second Ki67, and the last column is a TUNEL assay. (E) Representative Histology of 3D cancer tumors receiving no treatment, and different dose combinations of cisplatin and gemcitabine. The first column shows H&E staining, the second Ki67, and the last column is a TUNEL assay. Big boxes bars = 100 μm (10X objective lens) and small boxes Bars = 25 μm (40X objective lens).

Cell viability measurements using luminescence showed a statistically significant reduction as Cis concentrations increased. Similar trends were observed with Gem alone, with statistically significant reduction in viability above 0.05 µM (Fig. 4D). Combination treatment with both drugs, mimicking clinical usage, resulted in a significant decrease in cell viability as the combined drug dose increased (Fig. 4E). These findings highlight the potential for a synergistic effect of the drug combination as seen in patients. This observation explains why gemcitabine and cisplatin are often administered in combination to reduce the required dosage and in return minimize side effects.

To assess the genomic relationship between the original tumour and our model, we performed parallel DNA sequencing of FFPE tissue taken from the original implanted tumour, and cells sampled from the model on day 7, 14 and 21. The median deduplicated depth of coverage was 625 × among FFPE samples, and 654 × among cell samples. Of the 17 BCa patient samples profiled, all but one was negative for tumor material, suggesting a selective propagation of benign supportive cells during passaging outside the model. However, samples derived from patient Px_14 did contain detectable somatic alterations that increased in allele frequency in a stepwise manner from time point 7 to 21 (Fig. 5). Detected mutations were 100% concordant with the source material from the patients’ original tumour.

Genetic validation Representative graph of a cancer model from patient. The most characteristic gene mutations keep after 21 days in culture.

We acquired tissue samples from 17 patients with BCa, encompassing both NMIBC and MIBC cases (Table 1). Utilizing early passages of patient tumor cells, we successfully developed a 3D cancer model using a decellularized pig bladder as a scaffold, and subsequently subjected it to genetic validation (Fig. 5). The recreation of patients’ BCa from all 17 samples demonstrated varying results in terms of the number and distribution of cells within the scaffold.

Out of the 17 patients, 13 were identified as MIBC, and these were treated with Cis and the combination of Cisplatin and Gemcitabine (Cis/Gem). Notably, a positive response to the drug combination was observed in 3 patients (Fig. 6A), while the remaining patients did not exhibit a response (Fig. 6B).

MTS Assay of Bladder Cancer 3D tumor model. Representative histology of 3D cancer tumors in two patients receiving no treatment, and a standard dose of Cis and combination Cis/Gem. The first column shows H&E staining, the second Ki67, and the last column is a TUNEL assay. Patient Px_01 respond to treatment (C) while Px_03 does not respond (D). Big boxes bars = 100 μm (10X objective lens) and small boxes Bars = 25 μm (40X objective lens). MTS assay to measure the viability of patient cancer cells in the 3D cancer models treated with Cis and Cis/Gem, in all MIBC cases. Three patients showed a sadistically significant decrease in viability with treatment (A), and the ten other patients did not show response to treatment (B).

A positive correlation in treatment response was identified in 5 out of the 6 patients, providing an 83.3% reliability rate in our BCa model. Histological confirmation of the results was evident in the viability assay. A representative patient responding to therapy displayed a decrease in cells in the H&E stain when treated with Cis and Cis/Gem, accompanied by an increase in TUNEL positive cells (Fig. 6C). Conversely, a representative patient with no drug response showed similar cellularity in the H&E stain (Fig. 6D). These findings underscore the potential utility of our 3D BCa model in predicting treatment responses and validating them against clinical outcomes.

Using human BCa cells from the 6 patients with clinical outcomes, we conducted a comparative analysis of the response to Cis and Cis/Gem treatment in 2D culture, spheroids, and our 3D model. Early passages of patient cells were cultured with varying concentrations of Cis, ranging from 0 to 10 μM for 2D culture and 0 to 200 μM for spheroids. The drug dose–response curves exhibited a statistically significant decrease in cell viability in all patients when increasing the Cis dose for both models (Fig. 7A,B).

Drug dose–response curve in 2D and 3D models. Drug dose–response curves for Cis (A-B) and combination Cis/Gem (C-D) using bladder cancer tumor cells from patients. Patient cells were cultured in a monolayer (2D culture) and spheroids (3D culture) and then treated with four cisplatin doses for 72 h for 2D culture and 2 cycles of treatment of 72 h each. Percentage of viability values for cisplatin and gemcitabine treatments at 72 h for 2D culture and 6 days for 3D culture. Response to treatment was considered when there is a statistically significant change with respect to the control with no treatment. Results are expressed as mean ± SD (n = 3).

In the case of Cis/Gem combination treatment at low doses, anticipating a synergistic effect, a significant decrease in cell viability was observed with increasing Cis dose in all patients (Fig. 7C). A similar response was noted with Cis/Gem, with the exception of one patient who did not exhibit a significant reduction in cell viability (Fig. 7D). To assess the reliability of our BCa model, treatment responses were compared with clinical outcomes in 6 patients (Table 2). Unfortunately, data from other patients were not available due to therapy refusal, incomplete follow-up, or other reasons.

When comparing the results across different models, our 3D model demonstrated higher reliability at 83.3%, as compared to spheroids at 50%, and 2D culture at 33.3%. These findings underscore the enhanced predictive capacity and utility of our 3D BCa model in assessing treatment responses compared to traditional 2D culture and spheroid models.

This study introduces a 3D patient-derived bladder cancer (BCa) model using decellularized pig bladders as scaffolds, advancing the field by providing a novel platform for drug screening that more closely mirrors the in vivo tumor environment. Bladder cancer represents a significant health challenge globally, particularly in identifying effective neoadjuvant chemotherapy options due to the disease’s complex biology. By implementing a three-dimensional scaffold, we attempt to recapitulate the intricate tissue architecture and microenvironment essential for accurate drug response prediction, addressing limitations observed in traditional 2D cultures.

Since Eiraku et al. pioneered 3D aggregation culture with cerebral cortex tissue from embryonic stem cells in 2009, there has been increasing recognition of the value of in vitro models that better replicate the native tissue structure and interactions22. Organoid models derived from patient tumors have become essential for simulating histopathological and genetic features of cancers, yet their exclusion of immune and stromal components limits their clinical relevance16,23,24. Although bladder cancer organoids show strong reliability in capturing mutation profiles (e.g., PIK3CA, FGFR3, TP53), these models often demonstrate heterogeneous responses to treatment and reduced chemotherapy sensitivity, likely due to the absence of surrounding cellular components or limited drug accessibility25,26,27.

While patient-derived xenografts (PDX) retain much of the original tumor’s histology and genetic profile, they pose practical challenges with low success rates and high costs. Establishing a PDX typically takes months and requires substantial resources, limiting their scalability for drug screening28,29,30,31. To bridge these gaps, our model integrates a natural extracellular matrix (ECM) from decellularized pig bladders to better simulate the tumor microenvironment, preserving the structural integrity of the bladder while providing an optimized scaffold for bladder cancer cells32,33. The result is a model that more accurately reflects the in vivo conditions, which we demonstrated by establishing a reliable 3D growth of UM-UC3 BCa cells, showing dense proliferation and histological features consistent with tumors34.

In evaluating drug response, our 3D model displayed enhanced predictive capacity compared to spheroids and traditional 2D cultures, with accuracy rates of 83.3%, 50%, and 33.3%, respectively. This finding reinforces the significance of 3D structures in providing a more accurate, translationally relevant approach for drug screening, which remains a limitation in 2D systems despite their efficiency, low cost, and accessibility. However, cell lines in 2D culture lack the architectural support and complex interactions of 3D models, leading to differences in drug response35,36,37,38.

Existing organoid studies underscore the potential of 3D in vitro models in predicting treatment responses across cancers. The TUMOROID study, a large-scale clinical investigation, reported over 80% accuracy in predicting colorectal cancer responses to irinotecan, illustrating the reliability of patient-derived organoids (PDOs) for colorectal cancer treatment39. Similarly, Yao et al. demonstrated an 84% predictive accuracy using PDOs for rectal cancer treated with chemoradiation, validating the model’s clinical translatability40. Nonetheless, success rates for creating reliable PDOs vary, particularly in cancers like lung adenocarcinoma, where PDCs yield only a 24% success rate, highlighting the need for improved models in some tumor types41.

Bladder cancer models, such as organoids, have shown varied sensitivity to agents like gemcitabine and cisplatin, although the absence of a patient outcome correlation remains a limitation15. Studies have noted that organoids with specific mutations, such as FGFR3, respond selectively to MEK/ERK inhibitors, indicating potential applications for precision medicine, albeit with variable predictive success42. In this context, our model provides a high degree of reliability, likely due to its ECM scaffold’s structural support, which enhances cell–cell and cell–matrix interactions, critical in therapeutic response.

Innovative PDOs for bladder cancer have included "urinoids," organoids derived from urine samples, which retain histopathological and genetic consistency with the original tumor, expanding possibilities for non-invasive cancer monitoring43. Additional approaches include patient-derived xenograft alternatives, like zebrafish tumor xenografts (ZTX) and the chick chorioallantoic membrane (CAM) model, which show promise in predicting bladder cancer drug responses and may supplement traditional PDXs due to their shorter development time and lower cost44.

Our 3D bladder cancer model addresses the limitations of current in vitro and in vivo models by integrating a natural ECM scaffold that enables a more representative tumor microenvironment. Its high predictive reliability for drug responses underscores its potential as a powerful tool for BCa research and preclinical testing, aligning closely with patient outcomes and offering a scalable, more accessible alternative to animal models.

While the 3D model successfully captures the cellular and extracellular matrix components, it is acknowledged that immune and stromal components—critical for immunotherapy studies—are underrepresented due to the limited number of these cells isolated from tumor tissue and incorporated into the model. Future modifications to incorporate these elements could further enhance the model’s applicability. The study’s reliability assessment is based on a limited number of patient samples. Extending the validation to a larger cohort of patients will strengthen the model’s predictive capabilities and generalizability. The current study focuses on Cis and Gem, commonly used in BCa treatment. Expanding the investigation to include other clinically relevant drugs and combination therapies would provide a more comprehensive understanding of the model’s utility. Delving deeper into the molecular and cellular mechanisms underlying the observed treatment responses could provide valuable insights. Understanding the specific pathways influenced by different drug combinations will contribute to refining treatment strategies.

In conclusion, this study pioneers the development of a 3D BCa model with promising applications in drug screening and personalized medicine. The model’s ability to recreate patient-specific responses and predict treatment outcomes marks a significant step forward in the field of BCa research. Continued advancements in this direction hold the potential to revolutionize preclinical testing, ultimately improving therapeutic outcomes for BCa patients.

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Sung, H. et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer J. Clin. 71(3), 209–249. https://doi.org/10.3322/caac.21660 (2021).

Article CAS Google Scholar

Canadian Cancer Statistics Advisory Committee in collaboration with the Canadian Cancer Society, Statistics Canada and the Public Health Agency of Canada. Canadian Cancer Statistics 2023. Toronto, ON: Canadian Cancer Society; 2023. Available at: cancer.ca/Canadian-Cancer-Statistics-2023-EN

Brenner, D. R. et al. Projected estimates of cancer in Canada in 2022. Can. Med. Assoc. J. 194(17), E601–E607. https://doi.org/10.1503/cmaj.212097 (2022).

Article Google Scholar

Van Rhijn, B. W. G. et al. Recurrence and progression of disease in non–muscle-invasive bladder cancer: From epidemiology to treatment strategy. Eur. Urol. 56(3), 430–442. https://doi.org/10.1016/j.eururo.2009.06.028 (2009).

Article PubMed Google Scholar

Gupta, S., Gill, D., Poole, A. & Agarwal, N. Systemic immunotherapy for urothelial cancer: Current trends and future directions. Cancers 9(12), 15. https://doi.org/10.3390/cancers9020015 (2017).

Article CAS PubMed PubMed Central Google Scholar

Von Der Maase, H. et al. Long-Term survival results of a randomized trial comparing gemcitabine plus cisplatin, with methotrexate, vinblastine, doxorubicin, plus cisplatin in patients with bladder cancer. J. Clin. Oncol. 23(21), 4602–4608. https://doi.org/10.1200/JCO.2005.07.757 (2005).

Article CAS PubMed Google Scholar

De Santis, M. et al. Randomized phase II/III Trial assessing gemcitabine/carboplatin and methotrexate/carboplatin/vinblastine in patients with advanced urothelial cancer who are unfit for cisplatin-based chemotherapy: EORTC study 30986. J. Clin. Oncol. 30(2), 191–199. https://doi.org/10.1200/JCO.2011.37.3571 (2012).

Article CAS PubMed Google Scholar

Bellmunt, J. et al. Phase III trial of vinflunine plus best supportive care compared with best supportive care alone after a platinum-containing regimen in patients with advanced transitional cell carcinoma of the urothelial tract. J. Clin. Oncol. 27(27), 4454–4461. https://doi.org/10.1200/JCO.2008.20.5534 (2009).

Article CAS PubMed Google Scholar

Wilding, J. L. & Bodmer, W. F. Cancer cell lines for drug discovery and development. Cancer Res. 74(9), 2377–2384. https://doi.org/10.1158/0008-5472.CAN-13-2971 (2014).

Article CAS PubMed Google Scholar

Ding, J. et al. Current animal models of bladder cancer: Awareness of translatability (Review). Exp. Ther. Med. 8(3), 691–699. https://doi.org/10.3892/etm.2014.1837 (2014).

Article CAS PubMed PubMed Central Google Scholar

Oliveira, P. A., Arantes-Rodrigues, R. & Vasconcelos-Nóbrega, C. Animal models of urinary bladder cancer and their application to novel drug discovery. Expert Opin. Drug Discov. 9(5), 485–503. https://doi.org/10.1517/17460441.2014.902930 (2014).

Article CAS PubMed Google Scholar

Rookmaaker, M. B., Schutgens, F., Verhaar, M. C. & Clevers, H. Development and application of human adult stem or progenitor cell organoids. Nat. Rev. Nephrol. 11(9), 546–554. https://doi.org/10.1038/nrneph.2015.118 (2015).

Article CAS PubMed Google Scholar

Wang, S., Gao, D. & Chen, Y. The potential of organoids in urological cancer research. Nat. Rev. Urol. 14(7), 401–414. https://doi.org/10.1038/nrurol.2017.65 (2017).

Article CAS PubMed PubMed Central Google Scholar

Okuyama, H. et al. Involvement of Heregulin/HER3 in the primary culture of human urothelial cancer. J. Urol. 190(1), 302–310. https://doi.org/10.1016/j.juro.2012.12.106 (2013).

Article CAS PubMed Google Scholar

Mullenders, J. et al. Mouse and human urothelial cancer organoids: A tool for bladder cancer research. Proc. Nat. Acad. Sci. 116(10), 4567–4574. https://doi.org/10.1073/pnas.1803595116 (2019).

Article ADS CAS PubMed PubMed Central Google Scholar

Broutier, L. et al. Human primary liver cancer–derived organoid cultures for disease modeling and drug screening. Nat. Med. 23(12), 1424–1435. https://doi.org/10.1038/nm.4438 (2017).

Article CAS PubMed PubMed Central Google Scholar

Lin, Z., Zhang, Q. & Luo, W. Angiogenesis inhibitors as therapeutic agents in cancer: Challenges and future directions. Eur. J. Pharmacol. 793, 76–81. https://doi.org/10.1016/j.ejphar.2016.10.039 (2016).

Article CAS PubMed Google Scholar

Inoue, T., Terada, N., Kobayashi, T. & Ogawa, O. Patient-derived xenografts as in vivo models for research in urological malignancies. Nat. Rev. Urol. 14(5), 267–283. https://doi.org/10.1038/nrurol.2017.19 (2017).

Article PubMed Google Scholar

Lin, D. et al. Metabolic heterogeneity signature of primary treatment-naïve prostate cancer. Oncotarget 8(16), 25928–25941. https://doi.org/10.18632/oncotarget.15237 (2017).

Article PubMed PubMed Central Google Scholar

Daniel, V. C. et al. A primary xenograft model of small-cell lung cancer reveals irreversible changes in gene expression imposed by culture in vitro. Cancer Res. 69(8), 3364–3373. https://doi.org/10.1158/0008-5472.CAN-08-4210 (2009).

Article CAS PubMed PubMed Central Google Scholar

Herberts, C. et al. Deep whole-genome ctDNA chronology of treatment-resistant prostate cancer. Nature 608(7921), 199–208. https://doi.org/10.1038/s41586-022-04975-9 (2022).

Article ADS CAS PubMed Google Scholar

Eiraku, M. et al. Self-organized formation of polarized cortical tissues from ESCs and its active manipulation by extrinsic signals. Cell Stem Cell 3(5), 519–532. https://doi.org/10.1016/j.stem.2008.09.002 (2008).

Article CAS PubMed Google Scholar

Yuan, B. et al. Patient-derived organoids for personalized gallbladder cancer modelling and drug screening. Clin. Transl. Med. 12(1), e678. https://doi.org/10.1002/ctm2.678 (2022).

Article CAS PubMed PubMed Central Google Scholar

Kim, M. et al. Patient-derived lung cancer organoids as in vitro cancer models for therapeutic screening. Nat. Commun. 10(1), 3991. https://doi.org/10.1038/s41467-019-11867-6 (2019).

Article ADS CAS PubMed PubMed Central Google Scholar

Minoli, M. et al. Bladder cancer organoids as a functional system to model different disease stages and therapy response. Nat. Commun. 14(1), 2214. https://doi.org/10.1038/s41467-023-37696-2 (2023).

Article ADS CAS PubMed PubMed Central Google Scholar

Berndt-Paetz, M. et al. Cell line-based human bladder organoids with bladder-like self-organization—a new standardized approach in bladder cancer research. Biomedicines 11(11), 2958. https://doi.org/10.3390/biomedicines11112958 (2023).

Article CAS PubMed PubMed Central Google Scholar

Dunne, L. W. et al. Human decellularized adipose tissue scaffold as a model for breast cancer cell growth and drug treatments. Biomaterials 35(18), 4940–4949. https://doi.org/10.1016/j.biomaterials.2014.03.003 (2014).

Article CAS PubMed Google Scholar

Tentler, J. J. et al. Patient-derived tumour xenografts as models for oncology drug development. Nat. Rev. Clin. Oncol. 9(6), 338–350. https://doi.org/10.1038/nrclinonc.2012.61 (2012).

Article CAS PubMed PubMed Central Google Scholar

Dong, X. et al. Patient-derived first generation xenografts of non-small cell lung cancers: Promising tools for predicting drug responses for personalized chemotherapy. Clin. Cancer Res. 16(5), 1442–1451. https://doi.org/10.1158/1078-0432.CCR-09-2878 (2010).

Article CAS PubMed Google Scholar

Genta, S., Coburn, B., Cescon, D. W. & Spreafico, A. Patient-derived cancer models: Valuable platforms for anticancer drug testing. Front. Oncol. 12, 976065. https://doi.org/10.3389/fonc.2022.976065 (2022).

Article CAS PubMed PubMed Central Google Scholar

Zhang, Z. et al. Establishment of patient-derived tumor spheroids for non-small cell lung cancer. PLOS ONE 13(3), e0194016. https://doi.org/10.1371/journal.pone.0194016 (2018).

Article CAS PubMed PubMed Central Google Scholar

Soto-Gutierrez, A. et al. Cell delivery: From cell transplantation to organ engineering. Cell Transplant. 19(6–7), 655–665. https://doi.org/10.3727/096368910X508753 (2010).

Article PubMed Google Scholar

Destefani, A. C., Sirtoli, G. M. & Nogueira, B. V. Advances in the knowledge about kidney decellularization and repopulation. Front. Bioeng. Biotechnol. 5, 34. https://doi.org/10.3389/fbioe.2017.00034 (2017).

Article PubMed PubMed Central Google Scholar

Raven, P. A. et al. Development of a murine intravesical orthotopic human bladder cancer (mio-hBC) model. Am. J. Clin. Exp. Urol. 6(6), 245–259 (2018).

PubMed PubMed Central Google Scholar

Kulkarni, G. S. et al. Canadian urological association guideline: Muscle-invasive bladder cancer. Can. Urol. Assoc. J. https://doi.org/10.5489/cuaj.5902 (2019).

Article PubMed PubMed Central Google Scholar

Forbes, S. A. et al. COSMIC (the Catalogue of Somatic Mutations in Cancer): A resource to investigate acquired mutations in human cancer. Nucleic Acids Res. 38(Suppl_1), D652–D657. https://doi.org/10.1093/nar/gkp995 (2010).

Article CAS PubMed Google Scholar

Barretina, J. et al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature 483(7391), 603–607. https://doi.org/10.1038/nature11003 (2012).

Article ADS CAS PubMed PubMed Central Google Scholar

Johnson, J. I. et al. Relationships between drug activity in NCI preclinical in vitro and in vivo models and early clinical trials. Br. J. Cancer 84(10), 1424–1431. https://doi.org/10.1054/bjoc.2001.1796 (2001).

Article CAS PubMed PubMed Central Google Scholar

Ooft, S. N. et al. Patient-derived organoids can predict response to chemotherapy in metastatic colorectal cancer patients. Sci. Trans. Med. 11(513), eaay2574. https://doi.org/10.1126/scitranslmed.aay2574 (2019).

Article CAS Google Scholar

Yao, Y. et al. Patient-derived organoids predict chemoradiation responses of locally advanced rectal cancer. Cell Stem Cell 26(1), 17-26.e6. https://doi.org/10.1016/j.stem.2019.10.010 (2020).

Article CAS PubMed Google Scholar

Kim, S.-Y. et al. Patient-derived cells to guide targeted therapy for advanced lung adenocarcinoma. Sci. Rep. 9(1), 19909. https://doi.org/10.1038/s41598-019-56356-4 (2019).

Article ADS CAS PubMed PubMed Central Google Scholar

Lee, S. H. et al. Tumor evolution and drug response in patient-derived organoid models of bladder cancer. Cell 173(2), 515-528.e17. https://doi.org/10.1016/j.cell.2018.03.017 (2018).

Article CAS PubMed PubMed Central Google Scholar

Viergever, B. J. et al. Urine-derived bladder cancer organoids (urinoids) as a tool for cancer longitudinal response monitoring and therapy adaptation. Br. J. Cancer 130(3), 369–379. https://doi.org/10.1038/s41416-023-02494-6 (2024).

Article CAS PubMed Google Scholar

Kowald, S. et al. Novel zebrafish patient-derived tumor xenograft methodology for evaluating efficacy of immune-stimulating BCG therapy in urinary bladder cancer. Cells 12(3), 508. https://doi.org/10.3390/cells12030508 (2023).

Article CAS PubMed PubMed Central Google Scholar

Download references

This study received a CIHR Project grant # 401512. We would like to express our gratitude to Drs. Peter Black, Dr. Morgan Roberts, Dr. Ali Reza and Htoo Zarni Oo for their insights during the planning of the experiments.

Vancouver Prostate Centre, Department of Urologic Sciences, University of British Columbia, Vancouver, Canada

C. U. Monjaras-Avila, A. C. Luque-Badillo, J. V. M. Bacon, A. W. Wyatt & A. So

Michael Smith Genome Sciences Centre, BC Cancer, Vancouver, BC, Canada

A. W. Wyatt

Department of Urologic Sciences, Faculty of Medicine, University of British Columbia, Vancouver, Canada

A. So

Department of Medicine, Faculty of Medicine, University of British Columbia, H.B. 2660 Oak Street, Vancouver, BC, V6H3Z6, Canada

C. Chavez-Munoz

You can also search for this author in PubMed Google Scholar

C.U.M.A., A.S. and C.C.M. conceived the study and designed the experiments. C.U.M.A. and A.C.L.B. carried out the organ harvesting, cell culture, histology. C.U.M.A. was in charge of developing the 3D model. C.U.M.A., J.B. and A.W. did the genetic validation. C.U.M.A., A.S. and C.C.M. wrote the manuscript. All the authors read and approved the submitted manuscript.

Correspondence to C. Chavez-Munoz.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Monjaras-Avila, C.U., Luque-Badillo, A.C., Bacon, J.V.M. et al. A novel approach to engineering three-dimensional bladder tumor models for drug testing. Sci Rep 14, 26883 (2024). https://doi.org/10.1038/s41598-024-78440-0

Download citation

Received: 30 April 2024

Accepted: 30 October 2024

Published: 06 November 2024

DOI: https://doi.org/10.1038/s41598-024-78440-0

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative