Exploring the utility of exosomal HER receptor dimerization as a resistance mechanism in NSCLC

Background: Personalized therapy in lung cancer has improved following the introduction of EGFR tyrosine kinase inhibitors. However, virtually all patients develop resistance leading to disease relapse and decreased patient survival. With the advances in research, several mechanisms have been identified to be involved in the development of resistance, including activation of ErbB3/HER3. HER3 is a key dimerization partner of EGFR and plays an important role in tumorigenesis and cancer progression. As a result, anti-HER3 therapies have been developed and clinical trials with monoclonal antibodies against HER3 are ongoing in NSCLC in combination with other therapies ('standard of care'). The influence of concomitant therapies such as platinum chemotherapy on EGFR-HER3 heterodimerization remains largely unexplored. We hypothesize that in a subgroup of NSCLC patients, cisplatin treatment will impact on the 'dependence' of the tumor on HER3, due to its disruptive effect on the EGFR-HER3 dimer interaction. Methods: We developed a protocol for immunostaining and fluorescence lifetime imaging microscopy (FLIM) of patient-derived exosomes to determine the ability to characterize the EGFR-HER3 dimer from clinical samples of patients with advanced NSCLC. Blood samples were collected prior to initiation of chemotherapy (cisplatin) and 3 weeks after. Biochemical and microscopy analyses were performed on purified exosomes from the paired serum samples of patients. Results: The presence and degree of the EGFR-HER3 dimer in the serumderived exosomes of NSCLC patients was variable and was significantly reduced in a proportion of patients following treatment with Cisplatin. These results were reproduced in cellline derived exosomes treated under the same conditions. Conclusions: Our observations may have important clinical implications for the use of anti-HER3 monoclonal antibodies in NSCLC patients, with regard to scheduling of cisplatin treatment. They can also allow the creation of a multivariate predictive model of combining exosomal ErbB receptor activation with other biomarkers of disease to prospectively predict response to chemotherapy and targeted agents as well as time to metastatic relapse.