Cancer is the leading cause of death in Korea, and the mortality rate of this disease continuously to increase annually (1). As the cancer incidence continues to rise, the importance of cancer prevention is being emphasized. Cancer treatment is expensive as well as developing effective anticancer drugs. Thus, if cancer is successfully prevented, the overall medical cost can be reduced. Moreover, it is also challenging to plan cancer prevention strategies through clinical trials in terms of its duration and cost. Cancer prevention clinical trials take more than 5-10 years to complete and usually require thousands of participants. The estimated cost for large clinical trials involving more than 10,000 people is approximately $100 to $200 million (2). Despite decade-long efforts to find effective cure, candidates for anticancer drugs are usually discontinued during the phase 3 of the clinical trials due to problems, such as efficacy and toxicity (3). As the results of these trials do not always lead to successful cancer prevention strategies, there is an urgent need for identifying alternative drug therapies effective in preventing cancer. Drug repurposing is the process of searching for new indications for drugs that already exist in the market (4). Since this method is based on previously accumulated research and development data, the new drug development process can be accelerated, cutting costs at the same time (5). Recently, many studies on drug repurposing are being conducted based on genome, phenome, and insurance claim data (6).

Inflammation is a critical part in the pathogenesis of cancer, and the correlation of high levels of cysteinyl leukotrienes (CysLT) and CysLT1 receptor (CYsLTR) with various types of cancer have been reported several times in in-vitro studies (7–12). CYsLTR antagonists (LTRAs), including montelukast, pranlukast, and zafirlukast, have been widely used for treating asthma, allergic rhinitis, or chronic cough (13), and are the most promising chemoprevention options to target CysLT signaling in cancer. In addition to CysLT1 signaling, montelukast essentially induces apoptosis in cancer cells while zafirlukast is found to be involved in the cancer cell cycle (14, 15). Moreover, the role of LTRAs could also be associated with cancer metastasis, showing cell migration and invasion were suppressed in glioblastoma cells (14), colon cancer cells (16), skin cancer cells (17), and 5-FU-resistant colon cancer cells (18). However, the chemopreventive effects of LTRAs described above are all reported in in-vitro studies. Thus, it is still questionable whether the same effects can be observed in people taking LTRAs, especially since only limited randomized clinical trials (RCTs) and observational studies for humans are available (19). To observe the cancer-preventing effects of LTRAs in humans, a long-term follow-up study with a sufficiently large cohort size is essential. Therefore, using insurance claim data, we aimed to evaluate whether LTRAs have cancer prevention effects in a real-world setting by observing patients who took LTRAs.

Our study analyzed patients who are using LTRAs through a long follow-up study. To our knowledge, this is the first research that consider the demographic information, co-medications, underlying comorbidities, and the patients’ physical examination data, including smoking status, alcohol intake, and BMI, while using a sufficiently large sample size. Our study results found that LTRA use was associated with an overall decreased risk of cancer. In addition, by dividing the dose and period of LTRA use into several subgroups, our study could identify the amount of dose and duration that may significantly lower the risk of cancer.

A previous cohort study also showed that the use of LTRAs significantly decreased the overall cancer risk, specifically for lung, colorectal, and breast cancer (25). The same trends were also found in our study; however, the magnitude of the reduced risk was smaller than the previous reported study. This result can be attributed to differences in the sample size and in the use of various covariates. In the work of Tsai et al., the number of patients after the propensity score matching was 25,110 (4,185 in the taking group, 20,925 in the non-taking group), which was much smaller than the 188,906 participants in our study. In addition, their study did not consider the variables related to lifestyle (e.g., smoking status, alcohol intake), which are major risk factors of cancer.

We found that the use of LTRAs had a significant preventive effect on overall cancers, which was consistent with other previous findings. Many studies have reported that LTRAs are effective not only for treatment (14, 15, 26, 27), including cancer metastasis (14, 16, 17, 28), but also for prevention (11, 25, 29–33), so it seems that LTRAs can be used in various stages of cancer. First, LTRAs inhibit the growth and/or induce apoptosis of a large series of human cancer cell lines. LTRAs inhibit growth of glioblastomas cells, by decreasing expression of B-cell lymphoma 2 (Bcl-2) protein and reducing the phosphorylation of extracellular signal-regulated kinase 1/2 (14). In breast cancer cells, apoptosis was also induced (15). A similar mechanism was found in colon cancer. In addition to significant reductions in cell proliferation, adhesion and colony formation, the induction of cell cycle arrest and apoptosis were observed in a dose-dependent manner (26). Montelukast induced down-regulation of Bcl-2, up-regulation of Bcl-2 homologous antagonist/killer, and nuclear translocation of apoptosis-inducing factors in lung cancer cells (27). Second, LTRAs could inhibit metastasis of cancer by preventing tumor cell migration through both cerebral and peripheral capillaries (14). Matrix metallopeptidase-9 (MMP-9) degrades extracellular matrix proteins and was increased in colon cancer patients. The MMP-9 expression and activity were reduced by montelukast (16). LTRAs inhibited epidermal growth factor-induced T cell lymphoma invasion and metastasis inducing protein 1 expression in skin cancer cells (17). There seems to be a difference in roles of preventive mechanisms within the LTRAs. Pranlukast can inhibit tumor cell migration through both the brain and peripheral capillaries, whereas montelukast inhibits tumor cell migration only in the peripheral capillaries (28). The preventive effect of LTRAs has been reported in several in-vitro and in-vivo studies for certain cancers, including colorectal (29), gastric (11), and pancreatic cancer (30). A previous cohort study also showed similar results, reporting that the risks of breast, colorectal, and liver cancers were significantly reduced (25). However, a non-significant association between lung cancer and LTRA use was found in our study, while other groups have reported its cancer risk reduction effect (25, 31). Three studies also showed that LTRAs reduced the risk of metastatic lung cancer, but not of lung cancer itself (28, 32, 33).

Despite these efforts, there have not been reports on any definite association between LTRAs and a specific type of cancer yet. The pathogenesis of cancer appears to be multifactorial, and such findings may have arisen due to differences in the study samples, study designs, or statistical methods. Tsai et al. (2016) also showed that the use of LTRAs was an independent protecting factor for overall cancers, reporting an HR of 0.31 (95% CI: 0.24–0.39). The magnitude of reduced risk was found to be smaller in our study (HR 0.85, 95% CI = 0.83–0.87), which might be due to larger sample size and the use of additional covariates. For instance, the patient’s smoking status had a high HR range, 1.16 (against liver cancer) to 1.67 (against lung cancer), implying that the smoking covariate is a large proportion in our cox proportional hazard regression model.

In our study, the analysis of dose and duration of LTRAs use is noteworthy. Most LTRA prescriptions (99.6%) provided for the patients in this study were low (<0.5 DDD) or intermediate (0.5 ≤ DDD < 1.0), and only a few proportions were high (0.4%). Our results showed that overall cancer risk was rapidly lowered when LTRAs were used in high doses. In the duration analysis, >3 years of LTRA use correlated with a much lower HR for cancers. LTRAs are usually considered as safe during long-term administration even at doses substantially higher than the recommended dose (34). Therefore, this suggests that future studies should consider a higher dose and longer duration when prescribing LTRAs to be able to secure its anti-cancer property without having to worry about its side effects. However, recently, neuropsychiatric events were reported in post-marketing surveillance and resulted in safety alert in 2008 and a black box warning in 2020. Additionally, conflicting reports on the association between LTRAs and neuropsychiatric events have been published (35, 36). Therefore, it is necessary to pay attention to these precautions. The results of our study could also be used in the design of clinical trials. For instance, RCTs have been conducted with zileuton, a 5-lipoxygenase inhibitor that shares a similar mechanism with LTRAs, as an adjuvant agent to conventional chemotherapy for lung cancer patients (37). With this, new and improved RCTs can be conducted using LTRAs as an addition to existing anticancer therapies.

In our subgroup analysis, notable results were also observed in specific patient groups. The greater preventive effects of LTRAs in lowering the risk of cancer were observed in the following: in men, patients aged >65 years, patients with a history of smoking or are currently smoking, and those with high CCI scores. Considering that men, aged patients, smoking, and the presence of various comorbidities are well-known risk factors, LTRAs may contribute to lowering the cancer risks in patients with these particular characteristics. For the design of realistic and feasible clinical trials, the selection of specific patient groups with the above-mentioned risk factors may be beneficial and more effective.

There are several limitations encountered in our study. Due to the nature of the real-world data, the purpose of prescribing LTRAs to the patients was not for cancer prevention. Moreover, our study does not include an active comparator, and therefore it may be susceptible to selection bias. However, to reduce bias, as many variables were collected and matched to minimize the differences between groups. But note that there may still be some residual confounding after bias reduction. It was impossible to specify the stage/subtype of cancer because the disease information provided by the ICD-10 code was limited. We also suggest that some caution should be exercised when interpreting our results. There have been several studies showing that the use of LTRAs are also effective in reducing the risk of lung cancer, but the results in our study were not statistically significant (25, 31). Considering that baseline comorbidities, such as asthma, can have a significant effect on the occurrence of lung cancer (38, 39), this study may not have completely ruled out the effects of other comorbid diseases on cancer because it used CCI score as an indirect measure of various disease severity. Likewise, our study used a retrospective cohort design and not all information are included and available in the KNHIS data. Therefore, although we adjusted for all possible confounders, there still might be residual confounding factors present during our analyses.

The findings of our study suggest that the use of LTRAs was associated with a decreased risk of overall cancer. The high dose and long duration of LTRA use correlated with the lowered risk. The greater preventive effects of LTRAs were also observed in patients with specific risk factors related to sex, age, smoking, and the presence of comorbidities. As LTRAs have not yet been used for the prevention or treatment of cancer, our findings could be used for developing a new chemo-regimen or in designing feasible RCTs. For future studies, further research is needed to elucidate the specific mechanism and clinical significance of our results.

The data analyzed in this study is subject to the following licenses/restrictions: Data that can view all the records of a patient are difficult to share due to the policy of the NHIS. It can only be viewed in anonymized form when analyzed. Therefore, if there is a request for original data, the statistical data obtained after the desired statistical processing on the server will be shared. Requests to access these datasets should be directed to National Health Insurance Service, nhiss.nhis.or.kr.

The studies involving human participants were reviewed and approved by Institutional review board (IRB) of Seoul National University (IRB No. E1901/003-004). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

HJ contributed to the conception and design of the study, data acquisition, analysis and interpretation of results, drafting, and revision of the manuscript. JO and I-WK contributed to the conception and design of the study, analysis and interpretation of results, and revision of the manuscript. All authors contributed to the article and approved the submitted version.

This study was supported by the National Research Foundation of Korea grant funded by the Korean government (MSIT) (no. NRF-2018R1A2B6001859).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.