The incidence and mortality rates of lung cancer have consistently increased in accordance with global patterns of malignant tumor (Sharma, 2022). Despite continuous advancements and refinements in lung cancer treatment, the pursuit of effective prevention methods to mitigate the risk of developing this disease remains paramount (Kordiak et al., 2022). Observational studies have suggested a potential association between statin use and a reduced risk of lung cancer (Khurana et al., 2007; Yang et al., 2015; Kwon et al., 2019). However, a case–control study conducted within the UK population reported an increased risk of lung cancer associated with long-term statin exposure (Vinogradova et al., 2011). Furthermore, recent Mendelian randomization (MR) studies have challenged the notion of a causal relationship (Carter et al., 2020; Min et al., 2023). Investigations of lipidomic profiles revealed that the use of statins, PCSK9 inhibitors, and fenofibrate, instead of NPC1L1 inhibitors, results in a reduction in plasma ceramide and sphingomyelin levels (Ng et al., 2014; Tarasov et al., 2014; Ng et al., 2015; Croyal et al., 2018). Nonetheless, findings regarding the impact of statins on plasma sphingosine-1-phosphate (S1P) levels have exhibited inconsistency (Egom et al., 2013; Therond and Chapman, 2022). Preclinical investigations have elucidated the significant role of sphingolipid metabolism in the pathogenesis of lung cancer (Ogretmen, 2018; Meng et al., 2021; Tang et al., 2023), although the precise relationship remains unclear (Lin et al., 2022). Among the sphingolipids, ceramide and S1P have garnered particular attention within lung cancer studies (Goldkorn et al., 2013). Notably, ceramidases play a pivotal role in the conversion of ceramide to S1P (Pyne and Pyne, 2010). While a case–control study has uncovered associations between higher concentrations of plasma sphingosine-1-phosphate and ceramides and the increasing risk of lung cancer, these findings contradict the preclinical results that suggest that ceramides promote apoptosis in lung cancer cells (Ogretmen and Hannun, 2004; Carpinteiro et al., 2008; Alberg et al., 2013). Evidence on how lipid-lowering drugs (LLDs) may influence the risk of lung cancer by modulating sphingolipid metabolites remains elusive, along with the potential underlying mechanisms.

To explore the intricate relationships among LLDs, sphingolipid metabolites (including plasma ceramide, ceramidase, and S1P), and the risk of lung cancer, we employed a drug-target MR analysis (Holmes et al., 2021) to address a critical knowledge gap and provide valuable insights into potential chemoprevention in lung cancer.

In this drug-target MR analysis, we assessed the effects of LLDs on lung cancer risk via six LLD targets (APOB, APOC3, HMGCR, LPL, NPC1L1, and PCSK9), comprising 12 lipid-lowering pathways. Our findings revealed that APOB inhibition decreased the lung cancer risk in ever-smokers and SCLC. APOC3 inhibition decreased the LUAD risk but increased the SCLC risk. Additionally, HMGCR inhibition increased the risk of SCLC, whereas the LPL agonist decreased the risk of SCLC.

We delved deeper into the mediating mechanisms involving sphingolipid metabolites within these processes. Our investigation revealed that inhibiting APOB, APOC3, and HMGCR increased plasma neutral ceramidase levels. Notably, we observed that neutral ceramidase played a protective role in various forms of lung cancer, including overall lung cancer, lung cancer in ever-smokers, LUAD, and LUSC. These findings highlight the potential therapeutic significance of neutral ceramidase as a mitigating factor in the development and progression of various types of lung cancer. In addition, we found that APOB and PCSK9 inhibition led to a reduction in plasma ceramide levels, while the LPL agonist increased plasma ceramide and S1P levels. Nonetheless, no substantiated evidence has emerged to establish a causal link between ceramides and the risk of lung cancer.

Several MR studies have explored the potential preventive effects of LLDs on malignancies (Carter et al., 2020; Liu et al., 2021a; Min et al., 2023); however, these investigations have primarily focused on statins and have not included stratified analysis of different subtypes of lung cancer. In contrast to previous MR studies, our study expands the range of LLD targets under consideration. In our positive control analysis, we observed that ANGPTL3 inhibition did not achieve statistical significance. This finding aligns with the results reported by Wang et al. (2021). We made the decision not to conduct further analysis for ANGPTL3 inhibition in subsequent MR analysis.

In contrast to previous MR analysis of the effects of statins on lung cancer risk, our study reveals novel findings that HMGCR inhibition increases the risk of SCLC, as does APOC3 inhibition. This aligns with the findings reported by Vinogradova et al., where the link between statins and heightened lung cancer risk persisted even after adjusting for cardiovascular factors. However, their investigation lacked a subgroup analysis focusing on lung cancer types (Vinogradova et al., 2011). Moreover, evidence hints at the potential for long-term statin use to heighten cancer risk in women (Friedman et al., 2008). The mechanistic underpinnings behind the potential of statins in increasing SCLC risk remain elusive. We posited that this effect might stem from off-target consequences associated with prolonged statin use (Liu et al., 2021b; Jiang et al., 2023).

Immunohistochemical assessments of lung cancer tissue samples have unveiled lower protein expression of APOC3 in SCLC compared to normal lung tissue. This discrepancy might propose a pathway through which APOC3 inhibitors heighten the vulnerability to SCLC (Shi et al., 2016). Furthermore, an investigation under the UK’s Early Access to Medicines Scheme reported instances of lung cancer metastasis after the prolonged utilization of APOC3 inhibitors (volanesorsen) (Jones et al., 2023). Our findings strongly advocate for heightened vigilance concerning the potential for APOC3 inhibitors to increase lung cancer risk.

While alipogene tiparvovec acts as an LPL agonist, its limited market presence and high cost (Senior, 2017) have left the landscape of observational inquiries into its long-term cancer risk ambiguous. Nonetheless, our study propounds a shielding effect of LPL agonists against SCLC susceptibility. Moreover, Cerne et al. (2007) reported diminished LPL gene expression within lung cancer tissues as opposed to normal tissues, potentially reinforcing the notion of LPL agonists mitigating lung cancer risk.

The observed causal relationships between APOB inhibition and reduced lung cancer risk in ever-smokers and LUAD risk are particularly noteworthy, as neutral ceramidase plays a significant mediating role in these relationships. In recent years, there has been increasing research into the role and mechanisms of neutral ceramidase in malignancies (Coant and Hannun, 2019). Preclinical studies have demonstrated that inhibiting neutral ceramidase can prevent the occurrence and progression of colon cancer (Garcia-Barros et al., 2016), whereas inhibiting acid ceramidase can inhibit the proliferation of non-small-cell lung cancer and enhance its sensitivity to cisplatin (Yildiz-Ozer et al., 2018; White-Gilbertson et al., 2019). However, observational studies have presented contrasting findings, with high levels of acid ceramidase being associated with improved prognosis in breast and ovarian cancers (Ruckhaberle et al., 2009; Hanker et al., 2013; Sanger et al., 2015). The conflicting results from clinical observational studies and preclinical studies suggest that ceramidases may have pleiotropic effects on the occurrence, treatment, and prognosis of cancers. These findings highlight the complex and multifaceted nature of ceramidases in cancer-related processes. In our MR analysis, we identified a causal relationship between increased neutral ceramidase and a reduced risk of lung cancer in ever-smokers. This suggests that neutral ceramidase may play a critical role in the development of smoking-induced lung cancer, with APOB inhibition potentially acting as a protective factor. Further research is warranted to elucidate the precise mechanisms underlying their effects and to reconcile the disparities observed among different types of cancer.

Ceramidases play essential regulatory roles in the onset and advancement of various cancer forms (Sanger et al., 2015; Garcia-Barros et al., 2016) and are pivotal in sphingolipid metabolism for converting ceramides into sphingosine, S1P, and fatty acids (Parveen et al., 2019). The sphingolipid metabolism pathway proves indispensable in the metastatic process of lung cancer (Pyne et al., 2018; Coant and Hannun, 2019); its potential as a novel therapeutic target holds significant promise for augmenting the efficacy of tumor treatments (Vijayan et al., 2019), and animal experiments have demonstrated that NPC1L1 inhibition can modulate sphingolipid metabolism (Yamanashi et al., 2020). Although observational studies have indicated an association between statin use and reduced plasma ceramide levels, there is limited information on the impact on ceramidase levels.

In alignment with the findings of a previous study, our study identifies increased plasma ceramide levels as a contributing factor to heightened lung cancer risk among non-smokers, albeit with nominal statistical significance (Alberg et al., 2013). Preclinical studies have shown that exposure to cigarette smoke leads to increased ceramide levels in the lung tissue of mice (Filosto et al., 2011; Goldkorn et al., 2014; Lavrynenko et al., 2020); the contribution of ceramides to the heightened risk of smoking-related lung cancer necessitates prospective investigations for conclusive evidence. Consistent with previous observational studies, we found that PCSK9 inhibition reduced plasma ceramide levels (Tarasov et al., 2014), and APOB inhibition had the same effect. In addition, our results revealed that the inhibition of LPL could increase plasma S1P levels. However, owing to the non-convergence of the IVs for S1P in the lung cancer GWAS dataset, an analysis of the association between S1P and lung cancer risk was precluded. Notably, Alberg et al. (2013) observed an increased risk of lung cancer with increased S1P levels.

The strength of our study resides in its augmentation of LLD targets and lipid-lowering pathways. It amplifies the scrutiny of distinct lung cancer subgroups. More importantly, we delve into the mediating effect of sphingolipid metabolites, thus elucidating the intricate mechanism underpinning how LLD influences lung cancer risk. However, it is important to acknowledge the limitations of our study. First, MR analyses could not be performed on two of the target genes of LLD: ANGPTL3 and PPARA. This limitation arises from the inability of IVs to establish positive control associations or the unavailability of suitable genetic variants. Furthermore, given the intricate pharmacological mechanisms underlying these drugs, our MR analysis was unable to assess their potential off-target effects. Second, IVs employed as proxies for lifetime exposure to LLD may be limited in their ability to capture the effects of short-term drug exposure. Additionally, our study is constrained to examining causal directions and does not enable a precise estimation of dosage effects or the cumulative impact of multiple medications. Third, although the F-statistic for the genetic variants exceeds 10, indicating a low probability of weak instrument bias, statistical power is constrained due to the limited availability of only two SNPs as IVs for acidic and neutral ceramidases (Larsson et al., 2019). Moreover, the relatively small sample size utilized for the assessment of sphingolipids raises concerns regarding potential selection bias. As a result, the interpretation of our findings should be exercised with caution. Fourth, despite sensitivity analyses not yielding statistically significant outcomes, it remains essential to acknowledge the potential impact of confounding variables and horizontal pleiotropy, given the possibility of SNPs residing in a state of weak linkage disequilibrium (r ² < 0.30). Fifth, the absence of GWAS summary statistics for sphingolipids and lung cancer across diverse populations precludes validation. MR studies cannot substitute randomized controlled trials, and further clinical and pharmacoepidemiological studies are essential for triangulating. Sixth, the study population predominantly comprised individuals of European ancestry, which may limit the generalizability of our findings to other racial and ethnic groups.

We elucidated the intricate interplay between LLDs, sphingolipid metabolites, and lung cancer risk. Associations of APOB, APOC3, and HMGCR inhibition and the LPL agonist with distinct lung cancer risks underscore the multifaceted nature of these relationships. The observed mediation effects highlight the considerable influence of neutral ceramidase on lung cancer risk reduction achieved by APOB and APOC3 inhibition.