Lung cancer is the leading cause of cancer-related deaths in the United States. Non-small cell lung cancer (NSCLC) comprises of 85 % - 90 % of lung cancer. In 2024, an estimated 234,580 new cases of lung cancer will be diagnosed, and 125,070 people will die from the disease [1], [2].

Statins are a class of commonly prescribed medications used for the prevention of cardiovascular diseases. Statins function by inhibiting the production of endogenous cholesterol and blocking protein prenylation [3], [4]. The reduced cholesterol could lead to decreased cell proliferation, enhanced apoptosis, and inhibited angiogenesis [5], [6], all of which are potential anticancer properties with possible benefits in cancer prevention [7]. In population studies, the use of a statin has been shown to be associated with improved survival for common cancers, including lung cancer [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. Results from two meta-analysis studies of lung cancer also demonstrated the improved overall survival and cancer-specific survival associated with statin use [18], [19], [20], [21]. However, the published studies were subjected to some limitations. First, the timing of statin use relative to lung cancer diagnosis was not clearly defined in a number of published studies [15], [21]. Results on drug use and clinical outcomes could be seriously biased without considering the timing. Specifically, immortal time bias, commonly found in studies of post-diagnosis drug use and clinical outcomes [22], [23], distorts the association between post-diagnosis drug use and survival outcome towards a false inverse association, an artifact of protective effect of drug use, because patients can use drug only when they are alive. Moreover, although the published meta-analysis supported the protective effect, the pooled risk estimates in the meta-analysis were derived from some individual studies impacted by the immortal time bias, thus, the meta-analysis results were also subjected to the bias. Third, population-based studies performed in the U.S. general population had limited capacity to address how the differences in healthcare access could influence the findings because barriers to health care, such as socioeconomical status as reflected by the insurance status, affect both the use of prescribed medication and cancer outcomes in the U.S. population [24], [25]. Such effects on research are presumably reduced in a study in a health system that provides universal care.

The U.S. military health system (MHS) provides beneficiaries with universal access to health care with no or little out-of-pocket cost [26], [27.], and therefore is an unique resource for assessing the relationship between statin use and cancer outcomes. As the beneficial effects of statins were originally observed from clinical trials in which patients started statin therapy after their cancer diagnosis, population studies based on real-world data focusing on post-diagnosis use would provide more evidence on whether statin use after cancer diagnosis is associated with survival among cancer patients. In this study, we investigated the relationship between statin use and survival of MHS beneficiaries with NSCLC. We employed time-dependent survival analyses to control for the immortal time bias.