A short‐term in vivo model for Merkel Cell Carcinoma

Abstract In vivo tumor models are essential for studying the biology of cancer, identifying tumor targets and evaluating antitumor drugs. Considering the request for the minimisation of animal experiments and following the “3R”‐rule (“replacement,” “refinement,” “reduction”), it has become crucial to develop alternative experimental models in cancer biology. Several studies have already described the avian chorioallantoic membrane (CAM) model as an alternative to rodents, suitable to investigate growth, progression and metastasis of various types of cancer. In the present work, we grafted three Merkel cell carcinoma (MCC) cell lines onto the avian CAM and monitored tumor growth and development of solid tumor nodules. Morphology of xenograft was characterised histologically and immunohistochemically. Our results demonstrate CAM assay as a useful tool to study MCC pathophysiology.


| BACKG ROU N D
Merkel Cell Carcinoma (MCC) is a rare, highly aggressive neuroendocrine tumor of the skin with poor prognosis that typically occurs in elderly and immunosuppressed patients. [1] The MCC is characterised by the presence of cytokeratin 20 (CK-20) and neuroendocrine granules. The outcome of immune surveillance suggested viral carcinogenesis, which was indeed demonstrated in the majority of cases. [2] UV radiation exposure is an additional epidemiologic risk factor for MCC. [3] Due to the development of immune checkpoint inhibitors, a new therapeutic window opened for MCC patients. [4][5][6] Recently, treatment with three humanized antibodies, namely avelumab, pembrolizumab and nivolumab targeting PD-L1/PD-1 pathway have shown durable responses in MCC patients, and avelumab has been approved by the FDA for the treatment of advanced MCC. [7] Even though a panel of well characterised MCC cell lines is available, [8][9][10] the use of these cells in 2D culture systems is of limited value for translation into clinical settings. Therefore, we aimed to establish an in vivo model for MCC using the CAM assay to investigate growth and proliferation properties of onplanted MCC tumors. The CAM model is a time-and cost-effective drug screening system that was successfully used to characterise growth, proliferation and metastasis in a number of other cancer entities. [11][12][13]

| QUE S TI ON A S KED
We investigated whether the CAM system is suitable as a short-term in vivo model for MCC to study tumor growth, proliferation and angiogenesis.

| E XPERIMENTAL DE S I G N
We grafted three MCPyV-positive MCC cell lines (MKL-1, PeTa, WaGa), that clearly differ in their phenotype and growth behaviour in cell culture in vitro, onto the CAM and monitored proliferation as well as development of solid tumor nodules and characterised their morphology as shown schematically in Figure S1. We performed three independent experiments with four onplants for each cell line.

| RE SULTS
All cell lines formed tumors within 3 days after transplantation; progression of tumor formation was monitored by photo-documentation throughout the incubation period ( Figure S2A). MKL-1, PeTa and WaGa tumors revealed a reproducible growth pattern. MCC cells developed into solid nodules from day 3 after transplantation; vascularisation steadily progressed, and avian vessels developed radially towards the tumors ( Figure 1A-C, Figure S2A). On day 5 posttransplantation, the tumor area (mm2) was determined by ImageJ ( Figure   S2B), and the number of vessels surrounding the xenografts was counted manually ( Figure S2C) according to Ribatti et al, 2006. [14] We did not observe any significant difference within MCC cell lines with respect to tumor area or macroscopic blood vessels.  Figure 1G-I, arrows). [15] Furthermore, using immunohistochemistry method (IHC), the sections were analysed for the expression of MCC marker CK-20, MCPyV-LT antigen and the proliferation marker Ki-67 ( Figure 2). Expression of specific neuroendocrine tumor markers such as chromogranin-A (CGA) and synaptophysin (p-38) are shown in Figure S3. [16,17] CK-20 was present in all MCC xenografts in a typical dot-like perinuclear staining pattern (Figure 2A-C). [18] The MCPyV-LT antigen was detected in all three MCC cell lines. The IHC showed nuclear staining of LT antigen in MKL-1 ( Figure 2D) and cytoplasmic staining for PeTa and WaGa ( Figure 2E-F) due to the differences in truncating mutation in LT antigen. [19] This staining of

| CON CLUS ION
We here demonstrated that the CAM system can be used as an experimental in vivo tool that reproduces tumor-stroma interaction, angiogenesis and growth in MCC. Our data indicate that the CAM could represent a valuable preclinical model suitable to study MCC biology and drug response.

ACK N OWLED G EM ENTS
We thank SFL Technologies (Stallhofen, Austria) for Olympus microscopes used in this study.

CO N FLI C T O F I NTE R E S T
The authors have declared no conflicting interests.