1. Requirement of Non-Invasive Imaging Modalities to Confirm and Evaluate Tumor Progression in an Intravesical Murine Model of BCa

As tumor formation and progression occur within the bladder of intravesical models, BCa is not readily detectable. However, an intravesical model is essential for a clinically relevant investigation of novel therapies. In earlier studies, intravesical BCa tumors were detected and monitored by physical examination such as bladder palpitation or clinical symptoms including hematuria, significant weight loss, and behavioral changes ^[1][2]. Upon confirmation of such masses or symptoms, it is very likely that the cancer is in its late stages and is likely fatal, making longitudinal investigation of the efficacy of BCa therapies nearly impossible. Another issue that arises during in vivo BCa studies is the need to confirm cancer presence to guarantee balanced randomization of experimental groups. Tumor-take rates for any intravesical model fall short of 100%, thus previous studies required a cohort of animals that would be sacrificed for the purpose of confirming the presence of BCa ^[3]. However, these few mice do not sufficiently represent what is occurring in the remainder of the experimental group. Thus, accurate, non-invasive assessment of BCa tumors has become a primary objective for developing intravesical murine models. Established modalities that reveal and evaluate the occurrence of BCa within murine models include bioluminescence imaging (BLI), micro-ultrasound imaging (MUI), magnetic resonance imaging (MRI), and positron emission tomography (PET).

2. Bioluminescence Imaging (BLI)

BLI allows for the visualization of BCa cells through the detection of light emission using an in vivo imaging system. BCa cell lines are first transfected with a luciferase vector and then inoculated into the bladder wall of a murine model. Subsequent intravascular or intraperitoneal administration with a D-luciferin substrate stimulates its oxidation by luciferase, converting D-luciferin to oxyluciferin, resulting in the emission of green light at 550–570 nm. Importantly, this reporter system is oxygen- and ATP- dependent. Therefore, only metabolically active cancer cells are capable of producing light. The optimal time to wait after D-luciferin administration for BCa visualization ranges from 10 to 18 min ^{[4][5][6][7][8]}.

As genetic engineering of the cells is required, this imaging modality is limited to an orthotopic BCa model. In studies detecting intravesical BCa tumor presence in vivo, BLI confirmed tumor presence as early as 4 days post-inoculation ^[9]. Other studies detected tumor presence at 5 to 15 days ^{[5][7][10][11][12][13][14]} while another was as late as 33 days ^[5]. This discrepancy may be due to the sensitivity of the BLI system or the number of living cells that remained after inoculation. Whether the inoculated cells were human or mouse-derived did not appear to affect the detection time, however, the mouse strain did. For example, one study utilizing mice with severe combined immunodeficiency (SCID; lacking both T and B lymphocytes) visualized bioluminescence on day 4 ^[10], whereas another study utilizing the double mutant, SCID-beige, did not visualize bioluminescence until 33.9 ± 18.3 days ^[5]. Both studies utilized the same cell line and the latter study inoculated more than double the number of cells than the former study, ruling out the cell line and inoculated cell number as potential causes of this discrepancy. Interestingly, no clinical signs of BCa were reported at any of these time points. Only one study reported a false negative with one out of 12 mice possessing a small tumor cell nest that did not express bioluminescence, but was confirmed by histology and immunohistochemistry at necropsy ^[12].

Moreover, the quantity of orthotopically implanted luciferase-expressing human BCa cells has been shown to positively correlate with bioluminescence intensity ^[4]. This can be attributable to the increase in metabolically active cells and thus higher ATP production. Cell line-dependent coefficients of determination (R²) ranged from 0.95 to 0.99, which was corroborated by another study ^[6]. In the same study, the authors utilized ultrasound to demonstrate a positive correlation between the tumor size and bioluminescence, with an R² value of 0.97 ± 0.02. This finding was also corroborated by another study ^[12]. In a preceding study, using an orthotopic xenograft model, the authors reported varying tumor size and bioluminescence intensity, which correlated overtime with R² values ranging from 0.75 to 0.92 ^[9]. However, this was followed by a decline in bioluminescence, alluding to a limitation of BLI. The same research group also observed that bioluminescence correlated differently to tumor size based on the xenograft type, attributing this to the presence or absence of hypoxia and necrosis ^[6]. Several studies have also demonstrated a loss of bioluminescence strength, with BCa progression. For instance, Scheepbouwer et al. ^[8] observed an increase in bioluminescence from days 5-7 to days 18–29 followed by a plateau until days 29–31, whereafter the mice were sacrificed. Investigation of the stability of long-term, orthotopic implantation with BLI demonstrated positive but relatively decreased activity over a 2-month period ^[7].

Interestingly, the photo emission intensity has also been demonstrated to be lower in murine models compared to in vitro models ^[4]. This disparity was attributed to interference of the abdominal wall. This is supported by preceding studies where the absorption of light by tissue, dark fur, pigmented skin, and BCa stage have also been indicated as additional potential causes, initiating further investigation ^[6][12]. These interferences may contribute to the inability to detect cancer cells immediately after inoculation. As previously mentioned, Black et al. ^[6] investigated the role of hypoxia and necrosis in the relationship between bioluminescence and tumor size using 253J-BV and KU7 xenografts; cell lines now discovered to be cross contaminated with HeLa. Nonetheless, KU7 tumors were found to have substantially more hypoxic and necrotic tissue compared to 253J-BV tumors, contributing to the R² values of 0.39 and 0.90 for bioluminescence and tumor burden, respectively. They speculated that the loss of vascular differentiation in the KU7 tumors may have resulted in the inadequate delivery of luciferin to the entire tumor.

In a subsequent study, Jurczok et al. ^[12] denoted the loss of bioluminescence in large tumors to be due to necrosis and hemorrhage, confirmed by histological analysis, which contributed to the large tumor volume but lower bioluminescence. As further confirmation, immunohistochemical staining demonstrated substantially lower levels of luciferase in these areas compared to vital tumor areas. In conclusion, BLI is a reliable modality to confirm and quantitively evaluate early-stage BCa presence, but can only qualitatively assess the BCa presence in advanced stages.

3. Micro-Ultrasound Imaging (MUI)

MUI utilizes soundwaves to produce 2-dimensional images of organs and other tissues. Compared to conventional ultrasound, which uses a frequency of 2–10 MHz, MUI uses shorter spatial pulses, resulting in a mean frequency of 40 MHz, higher resolution, and higher clarity images. For example, resolutions of 80 μm in the lateral plane and 40 μm in the axial plane were achieved ^[3][15][16], whereas others have achieved a 30 μm resolution ^[4]. To enhance the image quality in a murine model, the abdomen is shaved, depilatory cream is used to remove fine hairs, and a high viscosity ultrasound gel is applied. These modifications allow for imaging of both superficial and invasive tumors where the tumor volume can be calculated using MUI with the formula:

π/6 × length × width²

MUI has a short acquisition time of approximately 5 min per mouse ^[16], further underscoring its attractiveness for non-invasive imaging. Furthermore, MUI is independent of tumor origin, whether orthotopically, carcinogen, or genetically induced, making it suitable for use in conjunction with all intravesical BCa models. However reported disadvantages of MUI include the lack of 3-dimensional imaging ^[17] and a reliance on a skilled operator to detect BCa tumors, as well as the restricted availability of this technology ^[17][18].

The mean BCa tumor detection time in a syngeneic orthotopic model using MUI was found to be 10 days, ranging from 8 to 17 days, whereas that for the clinical symptoms of BCa was 20.8 days, ranging from 14 to 28 days ^[15][16]. Subsequent studies were able to detect orthotopic BCa tumors as early as 4 ^[9], 7 ^[19] and 11 days ^[8][15][20], which did not appear to be influenced by the cell line or mouse strain. Conversely, a more recent study was unable to detect tumors in a syngeneic orthotopic model until the thirteenth day and mice began to die on the sixteenth day from clinical symptoms ^[18].

A MUI validation study imaged a syngeneic orthotopic model before and after BCa cell inoculation ^[16]. At each imaging session, mice were sacrificed to confirm tumor presence or absence, as indicated by MUI. The tumor detection rate was 87%, with 13 out of 15 mice being correctly identified as having BCa. All mice possessed non-muscle invasive BCa and only one mouse developed clinical symptoms prior to tumor detection. Moreover, stereochemistry was used to confirm the tumor location and volume, with the latter producing a correlation coefficient of 0.97. The smallest tumor to be detected by MUI in this study was approximately 0.52 mm³. Interestingly, Scheepbouwer et al. ^[8] detected tumors as small as 0.4 mm³, while Chan et al. ^[15] detected tumors between 1 and 3 mm in diameter, further underscoring the sensitivity of MUI.

Further refinement of MUI includes the use of ultrasound contrast agents. Microbubbles, or nanobubbles of gas, are administered intravenously and expand and compress in response to ultrasonic waves, enhancing contrast and allowing for the visualization and quantification of superficial tumor vasculature. Chan et al. ^[15] utilized this method to effectively target the expression of vascular endothelial growth factor receptor 2 (VEGFR2), a regulator of endothelial migration and proliferation, in a syngeneic orthotopic model. In addition to the modifications mentioned previously, an elastic band was applied over the lower abdomen to prevent respiration and bowel movements from creating artifacts but without inhibiting blood flow. This study reported the performance time of high-contrast MUI to last approximately 60 min per mouse, which is a substantial change compared to MUI alone. A subsequent study successfully used contrast- enhanced ultrasound to visualize the perfusion status of an orthotopic xenograft model ^[9]. An elastic was also placed over the abdomen, mimicking the previously mentioned study. Many advancements have been made to ultrasound imaging, contributing to its success as a rapid, multiuse, non-invasive imaging modality for intravesical BCa in murine models.

4. Magnetic Resonance Imaging (MRI)

MRI produces high-resolution 3-dimensional images of anatomical structures via a strong magnetic field and radio waves. Several studies have investigated the feasibility of MRI as a method for the identification and quantitative analysis of intravesical murine BCa tumors. In a carcinogen induced BCa murine model, MRI was used to calculate the bladder wall and tumor area ^[17]. In this study, MRI produced T1 and T2 weighted images with ~100 μm spatial resolution in less than 10 min per mouse. Bladder wall measurements, representative of tumor burden, were calculated by subtracting the area of inner lumen from the area of the outer edge of the bladder using single axial images. The area of the bladder wall was found to be strongly associated with tumor stage (p = 0.0003) despite being weakly correlated with ex vivo bladder weight (rs = 0.39). Interestingly, the bladder wall area was also strongly associated with both non-muscle invasive and muscle-invasive diseases, whether grouped as individual stages (p = 0.003) or invasiveness (p = 0.002). An acknowledged limitation of this study was the use of one axial image in the calculation of the bladder wall area. Multiple sections are recommended for a more complex analysis. However, the authors claimed that one image was sufficient in demonstrating the usefulness of MRI quantification and the ease of image processing and analysis. A subsequent study also revealed a significant positive correlation between tumor diameter, as determined by MRI, and tumor stage ^[21]. As further evidence for the usefulness of MRI as a tumor quantification method, Black et al. ^[6] found the correlations of MRI and autopsy for tumor volume and weight to be 0.96 and 0.92, respectively, in an orthotopic xenograft model.

To enhance the MRI acquisition capabilities, Kikuchi et al. ^[21] investigated various contrast agents and their ability to optimize contrast to noise in a syngeneic orthotopic model. They found that the intravesical instillation of 50 [mu]l Gd-diethylenetetramine pentaacetic acid with 50 [mu]l air prior to imaging, compared to air, undiluted and diluted Gd-diethylenetetramine pentaacetic acid alone, resulted in better delineation of tumors from the bladder wall. In the same study, the authors imaged mice on days 10, 14, 17, and 24 to investigate the MRI tumor detection capabilities. MRI produced T1-weighted images with a 1.5 mm thickness ^[21]. On day 10, 14 tumors were identified by MRI whereas 17 tumors (all non-muscle invasive) were identified by pathology. The false negatives were attributed to the tumor sizes being less than 1 mm in diameter. On day 14, nine tumors were identified by MRI, however, two of these abnormalities were confirmed to be hyperplasia and catheter-induced trauma during BCa cell implantation ^[21]. By day 14, tumor detection rate by MRI was 86.4%. Prior to this point, no clinical signs were detectable in any of the mice, demonstrating the ability of MRI to detect early-stage BCa. On day 17, MRI imaged all bladders to be tumor-filled, while the MRI identified tumors invading the bladder muscle layer on day 24, which was later confirmed by histology ^[21]. In addition to tumor detection, the accuracy of MRI for determining tumor size was also investigated. MRI-based measurements strongly correlated with the actual tumor size (r2 = 0.977, p < 0.001). The smallest tumor size detected was 1.5 mm in diameter, which they attributed to the use of a second set of transverse images shifted by 1 mm. This study demonstrates the ability to accurately detect and quantify both early and late-stage BCa in an optimized MRI approach ^[21].

Further advancements in MRI for BCa assessment include the use of mesoporous silica nanoparticles (MSNs) in conjunction with fluorescence ^[14]. Sweeney et al. ^[14], attracted to the non-toxicity and tolerability of MSNs, covalently bound fluorescein molecules to MSNs and injected them into a syngeneic orthotopic model. Compared to non-labelled tumors, MSN-labelled BCa tumors displayed more intricate features, which were later confirmed by histology to be areas of faster tumor growth and high cell density. This study demonstrates the combination of MRI and MSNs to provide more detailed images of tumors, allowing for the detection of defining BCa features.

5. Positron Emission Tomography (PET)

PET, dissimilar to the modalities described above, evaluates the metabolic activity of organs and tissues, providing information on the physiology and anatomy in addition to quantitative cancer detection. Molecules naturally used by the organs or tissues of interest are tagged by a radioactive atom, producing a radiotracer, allowing for the detection and evaluation of diseases such as cancer. Molecular targets for imaging cancers are derived from pathways or proteins typically overexpressed in cancer cells or tumors. As such, a diverse selection of radiotracers exists, giving PET its high specificity. Current radiotracers in cancer research include ⁶⁸Ga-fibroblast activation protein inhibitor (FAPI), ⁶⁸Ga-ligand-prostate specific membrane antigen (PSMA), ¹⁸F- and ¹¹C-choline and ¹¹C-methionine, for example. The most widely used radiotracer is 2-¹⁸F-fluoro-2-deoxy-glucose (¹⁸F-FDG). ¹⁸F-FDG is a glucose analog and works based on the premise that cancer cells, which are more metabolically active than non-cancerous cells, will uptake glucose at a higher rate. This uptake is detected by PET, thus identifying cancerous regions. The attractiveness of PET is attributed to its ability to identify genomic aberrations and the dysregulation of proteins ^[22], which cannot be detected by the previously described imaging modalities. Unfortunately, the evaluation of BCa with PET is difficult due to the accumulation of ¹⁸F-FDG within the bladder from renal excretion, which obstructs the delineation of the tumor from the bladder wall and hinders the detection of BCa ^[23].

This limitation currently makes PET an unattractive option for BCa imaging. However, its unique ability to detect phenotypic changes has encouraged several studies to investigate modifications to potentially improve detection in intravesical murine BCa models. For example, Mahendra et al. ^[24] investigated the efficacy of two isomers of ¹⁸F- fluoro-alpha-metylphenylalanine (¹⁸F-FAMP), L-2-¹⁸F-FAMP, and D-2-¹⁸F-FAMP, in an orthotopic xenograft model. Following one hour of intravenous injection, the biodistributions of both radiotracers demonstrated an accumulation in BCa tumors ^[24]. L-2-¹⁸F-FAMP demonstrated significant accumulation compared to ¹⁸F-FDG, whereas D-2-¹⁸F-FAMP showed a noticeably higher but non-significant accumulation compared to ¹⁸F-FDG. PET imaging at 1 h post-injection also demonstrated a clear visualization of all three radiotracers. Interestingly, at 3 h post-injection, the visualization of tumors was clearer, which was attributed to their fast elimination rate including rapid blood clearance and low renal accumulation ^[24].

Concurrently, Pereira et al. ^[19] investigated the efficacy of using galectin-targeted imaging in an orthotopic xenograft model. Following one hour of the intravesical administration of ¹⁸F-labeled galactodendritic unit 4, accumulation was significantly higher in BCa tumor-bearing mice compared to non-tumor bearing mice, with SUV_mean values of 43.5 ± 4.2 and 2.0 ± 0.4, respectively ^[19]. These findings demonstrate the selectivity of galectin-targeted radiotracers in imaging intravesical BCa cells. Nevertheless, this study utilized a high galectin-expressing cell line and acknowledged the need for the further investigation of the nonspecific binding of the radiotracer. Neither study utilized PET to confirm the presence of BCa before commencement, rather, bladder palpation and MUI were used, respectively. Thus, PET has been demonstrated to be potentially useful to monitor BCa tumors, only after detection by an alternative modality.

Immuno-PET is another branch of PET imaging that utilizes antibodies for the specific targeting of molecules expressed by cancer cells and tumors. In an orthotopic xenograft model, detection of the epidermal growth factor receptor (EGFR)-expressing BCa cells by the radioimmunoconjugate [⁸⁹Zr] Zr-DFO-panitumumab was investigated ^[20]. Imaging after 72 h after intravenous administration, [⁸⁹Zr] Zr-DFO-panitumumab injection resulted in tumor size-dependent accumulation, with a correlation co- efficient of 0.99 exhibiting the quantitative capability of PET. Interestingly, these mice demonstrated inconsistencies in their EGFR protein expression and tumor sizes ^[20]. Thus, it is speculated that the strong correlation is caused by the non-specific binding of the antibody due to enhanced permeability and retention effects.

Although imaging with PET in intravesical BCa models is limited due to the high renal excretion activity, studies have continued to investigate promising solutions to apply the advantages of PET to the detection and evaluation of BCa.

This entry is adapted from the peer-reviewed paper 10.3390/cancers15082381