Untargeted longitudinal analysis of a wellness cohort identifies markers of metastatic cancer years prior to diagnosis
It’s clear that catching cancer early improves survival. In this study we leverage our deep phenotyping analysis to see if there were markers detectable in blood that could alert physicians that cancer is developing well in advance of current practices. Through a commercial wellness program, we had evaluated over 1,000 proteins in the blood of several thousand people. Ten participants developed cancer during the course of the study. We were able to go back in time to look at their blood samples pre-diagnosis to find a few proteins that spiked up to 2 years in advance of the cancer diagnosis, suggesting these proteins might be good biomarkers for early cancer detection in the future.
- This study evaluated the blood of 10 cancer patients, compared to 69 controls from a longitudinal wellness study
- We identified one protein called CEACAM5 as being highly elevated in 3 of the 10 individuals who developed cancer (compared to individuals who did not develop cancer) as much as 26 months in advance of the cancer diagnosis—this was true across several time points for one lung cancer, one breast cancer and one pancreatic cancer patient, suggesting this could be a good early and non-specific cancer (pan-cancer) marker that should be tested further for this utility
- We identified another protein called CALCA that was elevated in an individual later diagnosed with pancreatic cancer, 16 months prior to the diagnosis
- A third protein, called ERBB2 spiked in two metastatic breast cancer patients, 4 and 10 months prior to diagnosis
- Deep phenotyping, analyzing thousands of proteins and metabolites, could provide much better tools for early diagnosis of disease