IMI ConcePTION core data elements for pregnancy pharmacovigilance studies using primary source data collection methods

Strategies for correct diagnosis, treatment evaluation and recurrence prediction are important for the prognosis and mortality rates among cancer patients. In spite of major improvements in clinical management, gastrointestinal stromal tumors (GISTs) can still be deadly due to metastasis and recurrences, which confirms the unmet need of reliable follow-up modalities. Tumor-specific secreted, shed or leaked proteins (collectively known as secretome) are considered promising sources for biomarkers, and suitable for detection in biofluids. Herein, we stimulated cell secretion in the imatinib-sensitive GIST882 cell line and profiled the secretome, collected as conditioned media, by using a shotgun proteomics approach. We identified 764 proteins from all conditions combined, 51.3% being predicted as classically/non-classically secreted. The protein subsets found were dependent on the stimulatory condition. The significant increase in protein release by the classical pathway was strongly associated with markers already found in other cancer types. Furthermore, most of the released proteins were non-classically released and overlapped to a high degree with proteins of exosomal origin. Imatinib pre-treatment radically changed these secretory patterns, which can have clinical implications when investigating biomarkers in imatinib-treated versus non-treated GIST patients. Our results show, for the first time, that GISTs contain a secretome signature. In the search for suitable biomarkers in the more complex GIST patient samples, this study aids in the understanding of basic GIST secretome characteristics.