Introduction
Colorectal cancer (CRC) is one of the most commonly diagnosed and represents approximately 8% of all cancer deaths due to the late stage at diagnosis when treatments are ineffective. The primary prevention of CRC is therefore very important for high-risk patients. The average lifetime risk for developing sporadic CRC is about 1 in 23, but there are millions at 3-5 fold higher risk, including those with recurrent polypectomy and those with first-person relatives with an early diagnosis (Roos et al., 2019). The typical recommendations for these high-risk patients include lifestyle modifications that have a marginal effect, and colonoscopy with polypectomy to detect and remove precancerous polyps (Oruc et al. , 2019; Yao et al. , 2017). Colonoscopy is the most effective approach to CRC prevention, but there are many cultural, geographic and economic barriers that limit the availability of this invasive procedure (Sachdev et al. , 2019; Wagner et al. , 2019; Zhao et al. , 2019). Chemoprevention agents are needed for CRC, but apart from NSAIDs that have significant side effects (Veettil et al., 2017), nothing has been approved for this purpose.
A large body of preclinical evidence has highlighted the potential use of cGMP-elevating agents to prevent colon cancer (Islam et al. , 2018). Uroguanylin and guanylin are peptide hormones that trigger intracellular cGMP production by activating epithelial guanylyl cyclase C receptors (GC-C) (Forte, 2004). In vivo studies using knockout mice with defective cGMP signaling have shown that cGMP controls intestinal homeostasis by slowing epithelial turnover (Li et al. , 2007; Steinbrecher et al. , 2002; Wang et al. , 2012). GC-C knockout mice also exhibit increased tumorigenesis in both AOM andApc Min/+ mouse models of colon cancer (Li et al., 2007; Lin et al., 2010). Moreover, increasing cGMP in wild type mice using exogenous GC-C agonists inhibits tumorigenesis inApc Min/+ mice (Chang et al., 2017; Shailubhai et al., 2000; Sharman et al., 2018). Phosphodiesterase 5 is expressed in the intestinal epithelium, and antagonizes the effects of GC-C agonists by degrading cGMP (Bakre et al. , 2000). Phosphodiesterase 5 inhibitors (PDE5i) such as sildenafil amplify the effect of endogenous GC-C agonists and suppress proliferation in the colon (Sharman et al., 2018; Sharman et al., 2017; Wang et al., 2014). Landmark studies showed that a clinically relevant dose of PDE5i can block tumorigenesis by 50% in both AOM/DSS and Apc Min/+ mouse models of CRC (Islam et al., 2017; Sharman et al., 2018). Sildenafil reduced polyp multiplicity in these studies but did not affect the size of the initiated polyps, indicating that the central effect of cGMP is to suppress tumor initiation (Browning, 2019; Islam et al. , 2018).
Two recent large cohort retrospective studies suggest that PDE5i might also prevent CRC in humans. The first report used the Swedish Hospital Discharge Register to show that polypectomy patients who had taken PDE5i had 36% less incidence of CRC compared to those who had never been exposed to PDE5i (Huang et al. , 2019). The second report interrogated the Veterans Affairs Database, and found that patients taking PDE5i over the long term had up to 50% reduction in CRC incidence (Lilley et al. , 2020). Prospective clinical studies are needed to support the repurposing PDE5i for CRC chemoprevention. Conventional PDE5i (e.g. sildenafil, tadalifil) that are most widely prescribed for erectile dysfunction, have also been safely used to treat chronic diseases including pulmonary arterial hypertension and benign prostate hyperplasia (Cohen et al. , 2019; Hatzimouratidis, 2014). However, side effects of these drugs such as headache, flushing and dyspepsia, would be less tolerated by otherwise healthy patients using them to reduce their risk of CRC. Moreover, due to common predisposing factors, many high-risk CRC patients also take nitroglycerin to treat ischemic heart disease (Boden et al. , 2015). Due to drug-drug interactions, these patients could not benefit from PDE5i for CRC prevention. PDE5i that avoid systemic delivery would therefore be a major advantage as an agent for CRC chemoprevention. The present study has designed an analog of sildenafil that exhibits poor membrane permeability, and we show that it avoids systemic delivery while retaining the anti-proliferative ability of PDE5i in the colon epithelium.