Short-term circulating tumor cell dynamics in mouse xenograft models and implications for liquid biopsy

Abstract

Motivation Circulating tumor cells (CTCs) are widely studied using liquid biopsy methods that analyze single, fractionally-small peripheral blood (PB) samples. However, little is known about fluctuations in CTC numbers that occur over short timescales in vivo, and how these may affect accurate enumeration from blood samples. Methods We recently developed an instrument called ‘diffuse in vivo flow cytometry’ (DiFC) that allows continuous, non-invasive counting of rare, green fluorescent protein expressing CTCs in large deeply-seated blood vessels in mice. Here, we used DiFC to study short-term changes in CTC numbers in multiple myeloma and Lewis lung carcinoma xenograft models. We analyzed 35- to 50-minute data sets, with intervals corresponding to approximately 1, 5, 10 and 20% of the PB volume, as well as changes over 24-hour periods. Results For rare CTCs, the use of short DiFC intervals (corresponding to small PB samples) frequently resulted in no detections. For more abundant CTCs, CTC numbers frequently varied by an order of magnitude or more over the time-scales considered. This variability far exceeded that expected by Poisson statistics, and instead was consistent with rapidly changing mean numbers of CTCs in the PB. Conclusions Because of these natural temporal changes, accurately enumerating CTCs from fractionally small blood samples is inherently problematic. The problem is likely to be compounded for multicellular CTC clusters or specific CTC subtypes. However, we also show that enumeration can be improved by averaging multiple samples, analysis of larger volumes, or development of new methods for enumeration of CTCs directly in vivo.