166 Description of radiological growth patterns in screen-detected lung cancers: The ASCENT study cohort

Abstract

Introduction Lung cancer screening (LCS) with low-dose CT (LDCT) scans tracks tumour growth from small nodules to histologically proven cancers. The current growth model, based on the modified Schwartz formula, assumes an exponential growth pattern, but it is uncertain whether this applies to most screen-detected lung cancers (SDLC) across multiple time-points. The ASCENT study (NCT04204499) aims to characterise genomic landscape driving variations in tumour growth rates in surgically resected SDLC. Here, we describe and quantify radiological growth patterns of SDLCs enrolled in ASCENT [Formula presented] Methods Surgically resected SDLCs manifesting as nodules with three or more measurement timepoints before resection were included. Volumes or mass, as appropriate, were used to size tumours. To model exponential growth, linear regression was used on log-transformed measurements to calculate tumour doubling time from all available timepoints. The R2 coefficient of determination and visual inspection assessed the goodness of fit and alignment with the original data, to define different growth patterns. Results In 62 solid nodules and 30 subsolid nodules, the median volume and mass at baseline were 79.1 mm3 (IQR: 24.7-236.3 mm3) and 127.3 mg (IQR: 37.6-315.3 mg), respectively. Nodules were followed for a median of 532 days (IQR: 381-798.3 days) before lung cancer surgery. Among the SDLCs, 87.1% of solid nodules and 70.0% of subsolid nodules exhibited an excellent or good fit (R2 >0.8) to the exponential growth pattern. Nodules not following the exponential pattern showed distinct but transient phases of growth, transitioning from fast to slow or slow to fast (Figure 1). Conclusions The exponential growth model reflects the growth pattern of the majority of SDLC manifesting as solid and subsolid nodules. However, some tumours transition between slow and fast phases; these may be influenced by factors such as genetic mutations and the tumour's immune microenvironment. Future analyses within the ASCENT study will explore the molecular and immune landscape underpinning these observations. Disclosure No significant relationships.