Abstract:
In the recent past of virotherapy, the characteristic trait of virus-induced oncolysis to trigger an immunogenic stimulus against malignant cells was utilized for new immunovirotherapeutic combination approaches, thereby overcoming preexisting resistances. Aim of this dissertation thesis was to investigate such a novel multimodal therapeutic strategy for treatment of colorectal carcinoma (CRC). In an in vitro model of human CRC cell lines from the NCI-60 tumor cell panel, it was tested whether immune checkpoint inhibitors (ICI) could achieve a therapeutic gain, combined with an oncolytic measles vaccine virus expressing green fluorescent protein (MeV-GFP) and with NK cells or peripheral blood mononuclear cells (PBMC).
Sulforhodamine B (SRB) cytotoxicity assays were performed to uncover preexisting resistances of the respective monotherapies in the different tumor cell lines. In two human CRC cell lines, HT29 (susceptible to measles-induced oncolysis) and SW-620 (exhibiting an intermediate resistance to measles-induced oncolysis), infection with MeV-GFP achieved a multiplicity of infection (MOI)- and time-dependent reduction of tumor cell mass, whereas HCT-15 tumor cells were observed to be highly resistant to MeV-GFP-induced oncolysis. Moreover, monotherapeutic treatment of HT29 and HCT-15 with nivolumab (targeting PD-1) or atezolizumab (targeting PD-L1) did not reduce tumor cell viability in the absence of immune cells.
FACS analysis of PD-L1 expression on CRC cell lines was conducted to firstly estimate the basal expression of this immune checkpoint ligand, whereupon tumor cells were measles-infected and, in a second step, the influence of infection on PD-L1 expression was investigated. Showing different degrees of basal PD-L1 expression, infection with both MeV-GFP and MeV-SCD, a suicide gene-enhanced measles virotherapeutic coding for Super-cytosine deaminase (SCD), increased PD-L1 expression in all three human CRC cell lines. In terms of MeV-SCD-infection, expression of PD-L1 further rose with augmentation of MOI for all three tested CRC cell lines and also with time of culture after infection for two out of three tested tumor cell lines. Furthermore, expression rates of the immune checkpoint receptor PD-1 on CD56 positive NK cell populations from four different healthy donors were investigated, resulting in less than 1 % of PD-1 positive cells. This result can be assessed as a poor precondition to take place in a direct PD-1 / PD-L1 interaction.
In a next step, we tested for augmented anti-tumor efficacy under the influence of our combination treatment and immune cell coculture, using the real-time tumor cell growth and viability xCELLigence analysis. Thereby, the combination of immune cell coculture with measles infection could already show increased therapeutic effects in comparison to the respective monotreatments, albeit unfortunately, this effect could not be further strengthened by additional application of ICI (nivolumab and / or atezolizumab).
Finally, antiviral effects of immune checkpoint blockade were examined: Neither in a viral growth curve model, nor in viral titrations after immune cell coincubation, an influence of ICI on replication and spread of MeV-GFP in CRC cell lines could be found.
To summarize, the upregulation of PD-L1 on human CRC cells via MeV-infection correlates with a promising therapeutic setting for combining ICI with measles-based virotherapy. However, in vitro xCELLigence analysis under immune cell coculture could not reveal a therapeutic gain of our immunovirotherapeutic approach. Nonetheless, considering the limited possibilities of an in vitro model of the complex human immune system, our therapeutic regimen should be further investigated in an immunocompetent mouse model of CRC and, even more, in the context of early clinical trials (i.e. phase I/II studies).