Downstaging as a direct patient benefit in cancer screening
Paul F Pinsky, Elyse M LeeVan, Wendy S Rubinstein, Christos Patriotis, Lori MinasianBackground
The standard metric to assess the benefit of cancer screening is reduction in cancer-specific mortality. However, patients may benefit from downstaging of disease due to screening, even among those who would have survived late-stage disease absent screening.
Methods
We developed a model of stage and survival in the absence and presence of screening. Within each stage/survival category without screening, there are transition probabilities to various stage/survival and mode of detection (screen detected or not) categories under screening. The model includes rules for allowable transitions; for example, stage and survival are unchanged for non-screen detected cancers, and survival status changing from dying to surviving requires downstaging. A two-stage (early/late) model was fit to National Lung Screening Trial (NLST) data. The low-dose computed tomography (LDCT) arm was considered the screening arm and the chest radiograph arm the control (non-screening) arm.
Results
There were 1032 and 993 lung cancer cases in the NLST LDCT and control arm, respectively. 15.2% of control arm cases were late-stage (III/IV) survivors. Their rate of transitioning to early-stage (I/II) survivors under screening was 18.4%, resulting in 2.8% (15.2% * 18.4%) of all cases transitioning from late-stage survivors under no screening to early-stage survivors with screening. 47.1% of late-stage control arm cases did not survive; their rate of transitioning to early-stage survivors under screening was 21.5%, resulting in 10.1% (47.1% * 21.5%) of all cases under no screening with that transition.
Conclusions
Modeling the effects of screening on stage and survival can help elucidate a more holistic set of screening benefits.