Monitoring of tumor growth and vascularization with repetitive ultrasonography in the chicken chorioallantoic-membrane-assay

The chorioallantoic-membrane (CAM)-assay is an established model for in vivo tumor research. Contrary to rodent-xenograft-models, the CAM-assay does not require breeding of immunodeficient strains due to native immunodeficiency. This allows xenografts to grow on the non-innervated CAM without pain or impairment for the embryo. Considering multidirectional tumor growth, limited monitoring capability of tumor size is the main methodological limitation of the CAM-assay for tumor research. Enclosure of the tumor by the radiopaque eggshell and the small structural size only allows monitoring from above and challenges established imaging techniques. We report the eligibility of ultrasonography for repetitive visualization of tumor growth and vascularization in the CAM-assay. After tumor ingrowth, ultrasonography was repetitively performed in ovo using a commercial ultrasonographic scanner. Finally, the tumor was excised and histologically analyzed. Tumor growth and angiogenesis were successfully monitored and findings in ultrasonographic imaging significantly correlated with results obtained in histological analysis. Ultrasonography is cost efficient and widely available. Tumor imaging in ovo enables the longitudinal monitoring of tumoral development, yet allowing high quantitative output due to the CAM-assays simple and cheap methodology. Thus, this methodological novelty improves reproducibility in the field of in vivo tumor experimentation emphasizing the CAM-assay as an alternative to rodent-xenograft-models.

Scientific Reports | (2020) 10:18585 | https://doi.org/10.1038/s41598-020-75660-y www.nature.com/scientificreports/ According to the Russell's and Burch's "Principles of Humane Experimental Technique" the reasonable use of the CAM-assay contributes to the refinement of animal experiments by minimizing pain and suffering of animals. As this methodology diversifies the applicability of the CAM-assay as a replacement for homologue rodent experiments, this model therefore meets the ethical obligations to reduce, refine, and replace (3Rs) animal usage in tumor research.
Specifically, this study evaluates the suitability of both single time as well as repetitive ultrasonography of tumors grown on the CAM for the quantification of tumor size, tumor growth and the evaluation of tumor vascularization. Comparative dropout rates were determined. Moreover, results obtained in the ultrasonographic measurements were compared to the corresponding histological specimen using immunohistochemical analysis for the determination of tumor size and tumor vascularization. Finally, an exemplary comparative mathematical analysis of the costs of the in ovo tumor model and the in vivo tumor model was conducted.

Methods
Eggs and tumor development. White Leghorn hens' eggs were placed horizontally in an incubator (Brutmaschinen-Janeschitz GmbH, Hammelburg, Deutschland) at 37.5 °C. To allow exposure of the CAM by detaching the membrane from the eggshell, on day 3 of incubation, 6 ml albumen was removed by aspiration with a sterilized syringe. The shell was subsequently opened with sterilized scissors, parts of the shell were removed and the CAM was exposed. After opening the eggshell, the aperture was covered with PARAFILM® (Bemis Company Inc., Neenah, Wisconsin, USA) to avoid evaporation. Cultivation of the liver cancer HuH7 tumor cells 38 on the CAM started by day 7.
One day before placement onto the CAM the HuH7 tumor cells were harvested by tryptic digestion from the cell culture flask, counted with a Neubauer counting chamber, and distributed in 1.5 ml tubes (5 Mio. cells per egg). After centrifugation at 1400 rpm for 10 min the supernatant was removed, and the cell pellet was subsequently suspended in 20 µl of ice cooled Matrigel™ (Corning™, Brumath, France) similar to procedures previously published 20,27 . The five million cells were then incubated for 30 min on a 6-well plate at 37.5 C (Greiner, bio-one International GmbH, Kremsmünster, Austria) until the Matrigel™ had hardened to a firm consistency. The 3D cell culture was then covered with culture medium, and finally incubated overnight.
Upon placement on the CAM a well vascularized spot was selected and the CAM was carefully incised using a single-use-scalpel (Feather, Dr. Junghans Medical GmbH, Bad Lausick, Germany) over a distance of approximately 0.5 cm opening the upper cellular layer. This incision has shown to have a positive effect on tumor inoculation as well as vessel ingrowth into the tumor without having relevant effects on the dropout rate. The 3D culture was subsequently placed onto the incision and 20 µl of Matrigel™ was pipetted onto the culture for protection against cell desiccation and to further immobilize the tumor on the CAM. The eggs were then incubated as mentioned above.
Ultrasonography. Ultrasonographic evaluations were performed by highly experienced sonographic examiners (i.e. DEGUM Level III) to ensure the best possible investigational quality. Starting from day 12 of incubation (5 days after tumor inoculation) the GE Healthcare Ultrasound LOGIQ E9 (GE Healthcare Little Chalfont, UK) 15 MHz linear transducer was used in the B-Mode (Gain 35) for ultrasonographic imaging (Fig. 1). Instead of ultrasound gel the space between the CAM and shell opening was filled with an average of 4 ml NaCl 0.9% to allow transduction of ultrasound waves. Tumors were then visualized in both longitudinal and transversal axes to enable a three-dimensional quantification of the tumor size. The respective image was frozen using the "Freeze" function and the tumor length, width and thickness was measured and documented.
Color-duplex-sonography was carried out using the same methodology while the built-in Duplex mode enabled visualization of the vessels within the tumor. Video sequences were saved for offline analysis.
For repetitive measurements, the same procedure was carried out on days 12, 13, and 14 respectively. The NaCl 0.9% solution was removed after each measurement using an electrical pipette (INTEGRA Biosciences GmbH, Biebertal, Germany) with a 10 ml sterile tube (Greiner CELLSTAR® serological pipette, Greiner AG, Bischofsheim, Germany).
Immunohistochemistry. After completion of the study protocol the embryo was sacrificed by decapitation. The CAM bearing the tumor was then excised with sterilized surgical scissors and placed onto filter paper stripes. The longitudinal and transversal axis were marked on the paper, the tumor-bearing CAM was transferred into a plastic cassette (Carl Roth GmbH + Co. KG, Karlsruhe, Germany), immobilized and put into a formalin solution (4%) (VWR International bvba, Leuven, Belgium) for 24 h. Afterwards, the plastic cassette was removed from the 4% formalin solution, washed three times with purified water for 20 min each, and incubated in isopropanol solution with increasing concentrations (80%/90%/100%) for 1 h each. The cassette was then washed with purified water and incubated in xylene (AppliChem GmbH, Darmstadt, Germany) for 24 h. Each specimen was imbedded in paraffin and cut into 5 µm slides with a microtome (Leica CM1900, Leica Biosystems Nussloch GmbH, Nussloch, Germany) according to the marking previously placed on the slides.
For HE stains, Paraffin was removed from the slide and the specimen was incubated in Mayer's Hemalum Roth (Carl Roth GmbH + Co. KG, Karlsruhe, Germany) for 5 min. Subsequently, each slide was again washed in purified water and incubated in eosin (Merck KGaA, Darmstadt, Germany) for 2 min. Slides were then incubated in isopropanol solution with increasing concentrations (80%/90%/100%) for 2 min each and xylene for 10 min. Finally, slides were prepared for microscopy by embedding the specimen in Eukitt® (Sigma-Aldrich, St. Louis, Missouri, USA). Immunohistochemical

Data management and off-line analysis. Tumor diameters, measured in ultrasonographic images
were transferred into EXCEL sheets (Microsoft Corp., Redmond, WA, USA). For determination of tumor vascularization, the video files of color-duplex-ultrasonography were exported and the presence and intensity of intratumoral vessels on the CAM was rated in a three step rating system (0 = no intratumoral vascularization, 1 = moderate intratumoral vascularization, 2 = intense intratumoral vascularization) by two blinded otorhinolaryngological specialists (J.E., B.E.), experienced in clinical ultrasonographic diagnostics independently. Ratings were then evaluated for concordance and reevaluated in case of discrepancy until a clinical consensus was reached between both investigators. Results were again inserted into EXCEL sheets. The images of the HE-stained tumors were used to determine tumor size by measuring the diameters in both sagittal and transversal planes by laboratory personnel blinded to the results of the ultrasonographic imaging. To evaluate tumor vascularization Alpha-SMA staining was used. Similarly, to ultrasonographic imaging, the amount of intratumoral vascularization was quantified in a three-step rating system (0 = no intratumoral vascularization, 1 = moderate intratumoral vascularization, 2 = intense intratumoral vascularization).
For the longitudinally cut histological slides the approximated two-dimensional tumor area could be calculated by using the ellipsis formula (A = Pi x d1[sagittal] x d2[transversal]).
To further allow comparability of size quantification by ultrasonography with the size determined by histological analysis, the tumor area in ultrasonography was determined by insertion of the longitudinal and sagittal diameters in the above-mentioned ellipsis formula.
Success rates of tumor development were determined by calculating the percentage of solid tumors on the CAM on day 11 of incubation in relation to the number of eggs incubated on day 0. Accordingly, success rates regarding the possibility of ultrasonographic imaging in all three dimensions were calculated by dividing the eggs in which application of ultrasonography was possible by the number of CAMs with solid tumors. GraphPad Prism™ (GraphPad Software, Inc., La Jolla, CA, USA). Besides column statistics (mean, median, range, standard deviation, standard error, Gaussian distribution and confidence intervals 95) comparative analysis was carried out as follows: As most data sets regarding tumor volume and vascularization did not show a Gaussian distribution, correlation between tumor sizes determined in ultrasonography as well as histology was determined using the Spearman-correlation. For longitudinal analysis of tumor growth as well as changes in vascularization determined in ultrasonography the Friedman repeated measures test was used to determine whether volumes significantly differed between the days of observation.
Differences in survival after repetitive ultrasonography were determined using the log-rank test as well as the Gehan-Breslow-Wilcoxon test.
Cost analysis. For comparing the costs of the CAM assay to rodent tumor models, the average costs for mice represented by the Crt:NU(NCr)-Foxn1nu/nu line (male, 6 weeks old) for a duration of 14 days were calculated. As source data the median prices for chicken eggs of the main distributer in the specific region (Bio-Aufzucht LSL Rhein-Main GmbH, Dieburg, Germany) as well as mice according to the 3 available distributers for nude mice in the specific region (Janvier Labs, Paris, France; Charles River Wiga GmbH, Sulzfeld, Germany; Envigo RMS GmbH, Roßdorf, Germany) were taken into account. Costs for transport as well as the gross running costs, provided by Translational Animal Research Center, University Medical Center of the Johannes Gutenberg-University Mainz, Mainz, Germany were considered. The cost of keeping animals in our facility is on average the price of equivalent German facilities known to us. An amount of 75 eggs as well as 25 mice was taken as the average sample size. Fig. 2 due to an extensive experience with this model we were able to retrospectively analyze a very high number of eggs (n = 1197) used for different studies in our lab. Of these eggs, upon opening of the eggshell 73.91% (± 8.90%) (n = 866) were fertilized and vital. After opening of the eggshell of the 866 eggs opened, 67.96% (± 10.68%) (n = 609) survived and were colonized with a tumor on day 7. Regarding the success rate of inoculation of (HUH7) tumors on the CAM, without any intervention till day 11, 50.39% (± 11.46%) of these 1197 eggs showed viability as well as sufficient ingrowth of the transplanted tumor localized within the observational window and suited for further investigation.

CAM-assay and tumor development. As shown in
For this specific study regarding ultrasonographic analysis 189 eggs were randomly selected. Repetitive ultrasonography was performed on 36 eggs. 54.00% (± 29.47%) of the randomly selected eggs had a tumor positioning on the CAM eligible for sufficient ultrasonographic imaging and measuring of all three tumor diameters respectively. On 30 eggs, color-duplex-ultrasonography was performed additionally.
Tumor size. In ultrasonographic analysis the median tumor volume on the CAM determined on day 14 was 0.075 cm 3 (± 0.072 cm 3 ). Accordingly, the calculated two-dimensional tumor size was 0.69 cm 2 (± 0.4355 cm 2 ). Measurement of tumor size in histology determined an average tumor size of 0.096 cm 2 (± 0.052 cm 2 ). For both The colored timeline shows the day of hatching after incubation (d0-d14). Breeding of the fertilized eggs started at day 0. Eggs were opened at day 3 and in vitro cultured tumors transferred on the CAM at day 7. Solid tumors were than analyzed from days 12 to 14 using ultrasonography. Of the 1197 eggs evaluated, 74% (n = 866) were fertilized and showed viability day 3 50% (609 eggs) successfully inoculated a tumor at day 11 not dropping out due to luxation, insufficient ingrowth or death of the embryo. Of the 186 eggs with solid tumors evaluated for ultrasonographic imaging, 100 eggs (54%) were suitable for sufficient ultrasonographic visualization. Tumor vascularization. Tumor vascularization could be visualized in 81.0% of eggs on day 14 using colorduplex-ultrasonography. 9.5% did not show any intratumoral vascularization, 28.6% were ranked as moderate intratumorally vascularized, and 61.9% showed an intense intratumoral vascularization. Analysis of histological slides showed vessel formation within the tumor tissue in 90.5% of all tumors analyzed. Comparable to the results obtained by ultrasonography 9.5% did not show any intratumoral vascularization. In 45% of cases the intratumoral vascularization was ranked as moderate and 47% of evaluated slides showed an intense intratumoral vascularization. Comparative analysis with the Spearman test regarding the intensity of vascularization within the tumor tissue correlated significantly (r = 0.65, P[two-tailed] < 0.0001) between ultrasonography and histological analysis (Fig. 4).  Cost calculation. Calculation of costs regarding different methodologies is difficult due to diverse prices for animals and housing in different institutions. Furthermore, additional costs like tumor cells, tools for preparation, housing or infrastructure may differ depending on the specific experimentation setup. Therefore, costs for cages, incubators, tumor cells, cell-medium or further equipment were not taken into consideration. www.nature.com/scientificreports/ Furthermore, since we currently do not perform any experimentation with subcutaneously implanted tumors in rodents in our laboratory, inoculation rates from published data using liver cancer cell lines [38][39][40][41][42][43] were utilized as comparative data. As shown in Table 1 median costs for a hen's egg were approximately 2.1 € while a single nude mouse (Crt:NU(NCr)-Foxn1nu/nu) has an average cost of 53.8 € per animal. Of all the eggs incubated after shipment, 53% show successful tumor inoculation on day 11 after incubation. Although solid data regarding ingrowth of subcutaneously implanted tumors in scientific articles is scarce and has shown a large variety ranging from below 50% 42 to 92% 41 . The median inoculation rate was estimated around 77.5% (median value considering literature analysis).

Repetitive ultrasonography measurement.
Taking these numbers into consideration and adding the running costs like food and housing for two weeks into consideration, the average price for an egg bearing a tumor on the CAM is 4.1 € while a mouse with a subcutaneous tumor will cost approximately 69.5 € per animal (Table 1).

Discussion
In recent years, the CAM-assay has been established for many scientific purposes. In tumor experimentation however, rodent experiments with subcutaneously implanted xenografts are still considered the "gold standard" regarding the evaluation of tumor treatment. In our working group the CAM-assay is frequently used for tumor experimentation. As shown in Fig. 2, 50.39% (± 11.46%) of the eggs incubated with tumors showed viability as well as sufficient ingrowth of the transplanted tumor on day 11. Tumor take rates reported in scientific literature show a large variety and range from around 45% 44 to much higher inoculation rates of 70-80% 20, 32, 45 depending on the specific tumor entity used for experimentation. Unfortunately, in many studies "true" dropout rates are not sufficiently reported and only positive ingroates or treatment effects are documented. In our experience, differences in dropout rates are likely to be attributable to varying fertilization rates or possible blunt damages to the egg during transportation. During experimentation, main reasons for dropout include insufficient ingrowth, luxation or movement of the tumor on the CAM as well death of the chicken embryo. Furthermore, the eligibility for ultrasonographic imaging is heavily determined by the placement of the tumor on the CAM. Approximately 54.0% (± 29.47%) of the eggs bearing an inoculated tumor on the CAM meet this criterion and ultrasonographic imaging and measuring is possible in all three dimensions (Fig. 2).
Different working groups have published data describing the evaluation of tumor growth using the CAMassay. In many publications however, tumor growth is only measured two-dimensionally or estimated using surrogate parameters like bioluminescence 32 . Furthermore end-point-analysis by means of size measuring 46 , cell count in flow cytometry of digested tumors, or weight measurements 47 49 in ovo. Since only a singular time point was analyzed in all studies mentioned above, repetitive longitudinal monitoring of tumor size and monitoring of tumor growth has not been previously described.
As shown in Fig. 5 we were able to repetitively analyze tumor size and vascularization using a commercial ultrasonographic scanner for human use. As ultrasound is commonly used in clinical diagnosis, many scientific institutions attached to hospitals have access to this imaging technique. Additionally, operating costs are rather low, the methodology is time-efficient and there is no need to apply potentially harmful contrast agents.
With regard to the assessment of the tumor size using ultrasonographic imaging compared to the histological sectional image, we expected great inaccuracies due to deviations in the sectional plane, artefacts and shrinkage during the preparation of histological slides. Nevertheless, tumor size in a two-dimensional sectional plane significantly correlated between ultrasonographic imaging and histological analysis (Fig. 3). The average plane of the tumors was much lower in histological analysis (13%) when compared to ultrasonographic measurements. Hence, we assume an intense shrinkage of the specimen during the fixation process. This is further supported by the fact that photographic documentation of the tumors showed diameters similar to ultrasonographic measurements far exceeding the measurements obtained from histological specimen (Fig. 3). The accuracy of the ultrasonographic measurements was further evaluated indicating an inaccuracy of approximately 1% compared to caliper measurements ( Supplementary Information S2). A positive correlation was also determined between the three-dimensional tumor volume in ultrasonography and the tumor area determined in the histological slide (r = 0.48, P[two-tailed] = 0.0059) ( Supplementary Information S3). By implication, we strongly believe that tumor size can be sufficiently monitored using ultrasonographic imaging.
Additionally, repetitive visualization of tumor vascularization further allows the monitoring of tumor perfusion (Fig. 4). This is immensely useful since therapeutical effects on intratumoral vessels (e.g. during radiotherapy or after application of anti-angiogenic drugs) can be longitudinally quantified 40,50 .
Using repetitive ultrasonographic imaging, tumor angiogenesis, resembled by an increase in tumor vascularization also was visualized (Fig. 5). Similar to the findings reported by Huang et al. 49 tumor vascularization also correlated with the findings obtained in histological analysis. Using a three step rating system for intratumoral Table 1. Kindly provide caption for the Table. 1 Resembled by the Crt:NU(NCr)-Foxn1nu/nu strain. 2 Median costs/ egg (including transport costs) of the main distributer of Hens-Eggs in the specific region (Bio-Aufzucht LSL Rhein-Main GmbH, Dieburg, Germany). 3 Median Costs for a male, six weeks old Crt:NU(NCr)-Foxn1nu/ nu mouse according to the 3 available distributers for nude mice in the specific region (Janvier Labs, Paris, France; Charles River Wiga GmbH, Sulzfeld, Germany; Envigo RMS GmbH, Roßdorf, Germany). 4 Gong et al. 39 , Huang et al. 40 , Robertson et al. 41 , Xu et al. 42 and Zhang et al. 43 . 5 Resembled by the average running costs for SCID mice in our specific institution (Translational Animal Research Center, Gutenberg University Mainz). www.nature.com/scientificreports/ vascularization, we were able to show a significant correlation (r = 0.65, P[two-tailed] = 0.0001) between findings obtained in ultrasonography and histological analysis (Fig. 4). Even though increase in tumor-size of this magnitude was not expected (Fig. 5), other working groups have reported similar courses after penetration of the tumor by new blood vessels 51 . These findings are further supported by the increase of detectable vascularization in color-duplex-ultrasonography indicating a progredient supply of oxygen and nutrients within the tumor (Fig. 5).
In contrast to radiological imaging like CT, ultrasound represents a cost-efficient alternative without the effects of ionizing radiation. In addition, CT scans evaluating soft tissue regions, highly depend on the application of contrast agents. Contrary to CT, MRI does not require radiation or contrast agents. Yet, the increased time and cost factor as well as impaired imaging resolution due to movement of the embryo are obvious methodological limitations of compared to ultrasonography 48 .
3-5 ml of NaCl on the CAM are necessary for transduction of ultrasound waves to the CAM. As the fluid itself or the higher hydrostatic pressure might influence the experimental protocol, dropout rates were determined for all eggs undergoing repetitive ultrasonographic imaging as well as untreated eggs. As shown in Fig. 5 no significant differences in the dropout rate were detected during the respective timeframe.
Obviously, quality of imaging could be improved by using high resolution imaging like UMI 49 . However, ultra-high frequency imaging systems represent highly specialized and costly research equipment which is not available in most institutions. In contrast, commercial ultrasonographic scanners are pretty much omnipresent in clinical patient care.
The comparison of tumor size between ultrasonography and histological slides has obvious drawbacks like folding artefacts, shrinkage of tissue and deviations in the sectional plane. As tumors inoculated in the CAM often show extensive hematomas as well as intense ingrowth of the CAM's stromal cells into the tumor tissue, increase of the macroscopic tumor may not be attributable to growth of tumor tissue itself but rather by a bidirectional infiltration of tumor tissue and CAM and accumulation of fluid by means of hematoma 27 . In rodents however, due to stromal cell invasion 52 measurement of tumor size also bears similar chances of inaccuracy of measurements of subcutaneously implanted tumors. Although subcutaneously implanted tumors do rarely infiltrate the adjacent tissue e.g. muscle and stroma, measuring with calipers-based measurement bares the possibility of inaccuracy due to deviations in transtumoral measuring axis and inter-operator variability 53,54 . Mouse models have a widespread availability, fast reproduction rate as well as relatively low housing costs compared to larger laboratory animals. Furthermore, the availability of multiple knock out and transgenic lines makes them the favorable animal model for most researchers. According to the German Ministry of Food and Agriculture, in 2017, 1.37 million mice were used for animal experimentation in Germany alone, making them by far the most frequently used vertebrate species in research today. Xenografts are usually implanted subcutaneously. In most tumor experiments three-dimensional size monitoring is realized using calipers or advanced imaging like CT, MRI, or bioluminescence detection 55 . Unfortunately, any evaluation of the tumor size in rodents results in stress or discomfort for the animal through repetitive handling, movement restriction, intravenous/intraperitoneal application or sedation. In contrast, in the CAM-assay neither the CAM nor the embryo are nociceptive. This not only makes in ovo experimentation ethically more justifiable but reduces the influence of factors like pain and stress on the experimental outcome. The short observational window may be considered as an obvious disadvantage of in ovo experimentation as the hatching of the chicken on day 21 will ultimately terminate the investigational timeframe. In contrast, rodent experiments allow much longer therapeutic and observational timeframes. Due to ethical concerns and the occurrence of nonspecific inflammatory reactions after 15 days of incubation 2 , we only evaluated tumor growth in the CAM-assay till day 14 after incubation. However, other working groups have published research evaluating tumors grown in ovo up to day 20 of incubation 56 . Additionally, tumors on the CAM grow much faster than equivalent tumors subcutaneously implanted into mice. Hence, vessel ingrowth 51 as well as tumor metastasis were successfully determined using the CAM-assay 57 allowing a sufficient monitoring of tumor development, regardless of the limited observational time frame. Aside from Hu et al. who observed equivalent tumor growth patterns and metastatic behavior for renal carcinoma cells in the CAM-assay and in mice 56 , further working groups were able to demonstrate parallels regarding the tumor size for different tumor entities in comparative experiments 46,57 . In our experience, a major disadvantage of experimentation with a chicken model is the very limited commercial availability of chicken specific antibodies. With increasing popularity of this model however, this may change in the near future. Furthermore, detection of the intratumoral vessels is obviously limited by the sensitivity and resolution capacity of commercial ultrasonographic scanners. Rodent models like the dorsal skinfold chamber 58-60 may allow an exposition of tumoral tissue similar to the CAM-assay, yet dropout rates and the severity of surgical intervention far exceed effects of subcutaneous tumor implantation 61 . Obviously, factors like inter-operator variability and a learning curve apply to ultrasonographic imaging as well. However, experienced examiners exist throughout multiple medical disciplines. Furthermore, the very simple methodology and visual identification of the tumor, due to its distinct shape on the CAM, allow a rapid acquisition of the skillset necessary for implementation of this technique. Despite its limitations, the CAM-assay is a versatile model offering a translational significance similar to equivalent rodent experiments for many scientific questions in the field of tumor research.
Chicken embryos are not considered independently living animals in most countries. Use of the CAM-assay is therefore not classified as an animal experiment. Still, obvious ethical concerns regarding the replacement of an animal experiment with another animal experiment remain and have to be addressed carefully. However, due to the lack of nociception in the CAM-assay as well as the embryo until day 14, this methodology is ethically preferable to homologous mouse experiments.
As exemplarily shown in Table 1 the costs for a sufficiently ingrown tumor suitable for experimentation are quite low, emphasizing the significantly increased cost efficiency of the CAM-assay compared to mice-experiments. Even though other institutions might be able to obtain mice for low prices or keep animals at moderate Scientific Reports | (2020) 10:18585 | https://doi.org/10.1038/s41598-020-75660-y www.nature.com/scientificreports/ running costs, the CAM-assay has obvious economic benefits regardless of lower tumor inoculation rates. Furthermore, keeping chicken embryos does not require a governmental approval to keep and breed experimental animals. Therefore, this model can also be used in laboratories that do not have an animal experimental unit. Due to a high biological variation, animal experimentation is influenced by both genotype and environmental conditions resulting in an impaired reproducibility 62,63 . Standardization of experimental setup may be one way to address this problem 64,65 . However, effects of biological variability might heavily influence experimental outcome especially if small sample sizes are utilized. Taking these factors into consideration, the CAM-assay may offer some key advantages. Biological variability due to behavioral aspects and influences like social interaction, stress and anxiety are much less present in hatching chicken eggs due to the limited sensorineural input. Secondly, the rather simple methodology facilitates a good standardization and reproducibility of experimental conditions. This leads to a reduced inter-operator bias and a lower time expenditure compared to rodent-xenograft-models. Especially the lower time expenditure may facilitate experimental efficiency which directly translates to an increased quantitative output. All these advantages might contribute to an increased reproducibility of experimentation especially in comparison to equivalent rodent-xenograft-models.

Conclusion
In conclusion, repetitive ultrasonography is suited for sufficient quantification of tumor size and monitoring of intratumoral vascularization without increased dropout rates. For the first time, tumor growth and tumor angiogenesis have been successfully visualized in ovo. Using the CAM-assay for tumor research has obvious advantages like time-and cost-efficiency as well as widespread availability resulting in a high quantitative output and an increased reproducibility. Therefore, ultrasonographic imaging further diversifies the applicability of the CAM-assay as an alternative to homologue rodent models in the field of tumor experimentation.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. License, which permits use, sharing, adaptation, 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 changes were made. 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://creat iveco mmons .org/licen ses/by/4.0/.