1 BACKGROUND AND EPIDEMIOLOGY

Since cystic fibrosis (CF) was first characterized in 1938 as a fatal neonatal disease,¹ stepwise medical advancements have dramatically changed the expected outlook for people with the disease. In 2010, the US median life expectancy was 39 years, and for those born that year, it was expected to be 56 years assuming mortality trends held,² with similar numbers in the United Kingdom.³ In 2012, the advent of modulators that restore the function of the defective CF transmembrane conductor regulator (CFTR) protein ushered in a new era of targeted therapy. Initial data suggest patients may live an additional 18 years or more,⁴ with a 1% reduction in annual mortality from CFTR modulation alone.⁵ The most recent data from the 2021 CF Foundation Annual Report anticipated patients born in 2021 had a median predicted life expectancy of 65 years.⁶

With greater than 58% of people with CF over the age of 18,⁶ CF is increasingly becoming a disease of adults.⁷ Increased attention is needed for diseases of aging among this patient population. One of the most notable “late” complications of CF is the development of gastrointestinal (GI) cancers, the most prevalent of which is colorectal cancer (CRC). This observation has consistently been made across the world in large population-based studies using registry data. In 1996, a landmark study bolstered anecdotal observations of increased risk of GI malignancies in CF, noting an odds ratio (OR) of 9.3 for small and large bowel cancers among a European CF population (95% confidence interval [CI]: 3.5−25).⁸ Most concerningly, these malignancies were identified at young ages, with an OR of 20.2 (95% CI: 6.1−67.0) for patients in their third decade of life.⁸ This was mirrored by similar findings in another European study where the standardized incidence ratio (SIR) for CRC was 4.5 (95% CI: 2.0−9.9).⁹

Longitudinal follow-up from studies have confirmed this association (Table 1). The 10-year follow-up data in the United States showed a SIR of 7.4 (95% CI: 3.7−13.2) for colon cancer in pretransplant patients,¹⁰ while the 20-year follow-up data showed a SIR of 6.2 (95% CI: 4.2−9.0) in patients who had not undergone transplant.¹¹ Similar numbers were noted in a Swedish study examining 35 years of data through national registries (SIR 4.8, 95% CI: 0.6−17.3), although the association did not reach significance.¹³ Despite growing medical awareness and the implementation of more aggressive screening programs, the most recent UK data has not shown a change in incidence trends (SIR 5.0, 95% CI: 3.2−6.9).¹² In the United States, among those over the age of 30 years, a total of 43 CRCs detected by screening have been reported to the CF Foundation Patient Registry from 2019 to 2021 (personal communication with CF Foundation Patient Registry staff). In a meta-analysis combining large registry-based studies with studies looking at CF patients after transplant, Yamada et al. found a pooled SIR for CRC of 10.9 (95% CI: 8.4−14.1).¹⁴

This high pooled SIR is driven by the incidence of CRC among the posttransplant CF population, with one study estimating an SIR as high as 30.1 (95% CI: 15.8−52.2; Table 2).¹¹ The striking incidence of cancer in the transplanted CF population appears to be driven by both posttransplant biology and the pathophysiology of CF, as patients who were transplanted before the age of 50 had a 5-year absolute risk of CRC of 0.3% compared to 0% of lung transplant patients without CF.¹⁶ In a large US-based study comparing rates of CRC between different types of organ transplant recipients, the SIR was approximately five times higher for patients with CF undergoing lung transplant versus patients with primary sclerosing cholangitis (PSC) and inflammatory bowel disease (IBD) undergoing liver transplant (SIR 27 [95% CI: 14.8−45.3] vs. 5.69 [95% CI: 4.0−7.9]).¹⁵ This increased risk represents a major priority within the transplant CF population, where CRC may cause 12% of all deaths and should be largely preventable with adequate screening regimens.¹⁷

Among patients with CF, in an Australian study, the mean age at diagnosis with CRC was 49 compared to 69 for the general population in Australia.¹⁸ Among transplanted CF patients the median age at diagnosis is lower still at 43 with a median time from transplant to diagnosis of 6 years.¹⁶ As a corollary, individuals with CF have an increased prevalence of adenomatous polyps, the predominant precursor lesions to CRC. The prevalence of adenomas among CF patients undergoing screening colonoscopy from age 40−49 ranged from 26% to 49% and was 35% in a younger patient population posttransplant,¹⁷ compared to 11% among matched controls.^18-20 This preponderance of precancerous lesions among CF patients under 50 is comparable to non-CF patients in their nineties.²⁰ Among younger patients (aged 30−39) undergoing diagnostic colonoscopies, the prevalence of polyps was still as high as 15%.²⁰

In addition to forming polyps at younger ages, patients with CF have a higher rate of advanced polyps (size ≥1 cm, villous features, high-grade dysplasia, or adenocarcinoma on histology) with prevalence ranging from 17% to 32% compared to 6% among controls.^18-20 The percentage of patients with a significant polyp burden is also elevated, with as many as 23% having three or more polyps.²⁰ On surveillance, the prevalence of polyps is as high as 86%.^{19, 20} There is also a potential missed adenoma rate in this patient population, with 47% of those with negative exams having polyps on rescreening,²⁰ although this could also represent a short dwell time for polyp and CRC formation in CF.

This striking prevalence of polyps at younger ages and lower median age at CRC diagnosis has led to the concept that CF should be considered a hereditary cancer syndrome. This idea is further supported by the association of more severe mutations and CF phenotypes with malignancy. For example, among patients with homozygous delF508 mutations, the SIR for all GI cancers is 5.2 (95% CI: 1.9−11.3), and for patients with failure to thrive/malnutrition at the time of CF diagnosis, the SIR is 7.9 (95% CI: 3.8−14.5).¹⁰ In contrast, patients with non-delF508 mutations have a SIR of 2.5 (95% CI: 0.1−13.9).¹⁰ This is in line with data showing that patients with delF508 have an increased risk of forming polyps (OR 3.8 on univariate analysis).²⁰ However, CFTR does not appear to be a classic tumor suppressor gene as first-degree relatives of patients with CF have not been shown to have an increased risk of CRC.¹³

The increased prevalence of malignancy among patients with CF may be a complex, multi-hit phenomena as evidenced by the significant association with transplant status (SIR 30.1, 95% CI: 15.8−52.2).¹¹ This is further supported by the observation that CF patients have a risk of both distal and proximal colon cancer.^{15, 16} In contrast, the risk of rectal cancer in patients with CF does not appear to be elevated to the same degree.¹¹

2 PATHOPHYSIOLOGY/PURPORTED MECHANISMS

Several potential mechanisms have been proposed to explain the elevated risk of CRC in CF, including the potential role of CFTR as a tumor suppressor gene, intestinal inflammation, and alterations in the microbiome. These factors may potentially act in concert to promote the development of colorectal polyps, as hypothesized in Figure 1. The notion that CFTR acts as a tumor suppressor is supported by the development of both small and large bowel cancers in CFTR knockout mice.²¹ Than et al. found that loss or inactivation of CFTR resulted in increased tumor growth, enhanced cell proliferation, and reduced apoptosis (cell death) in the intestinal cells. Furthermore, this study explored the underlying mechanisms through which CFTR exerts its tumor suppressor effects. It identified various signaling pathways and molecular interactions involving CFTR that regulate cell cycle control, DNA repair mechanisms, and the suppression of oncogenic processes. This study highlights the therapeutic potential of targeting CFTR to inhibit tumor growth and promote cancer regression in the intestine.²¹ It also appears that in sporadic CRC, CFTR plays a role in cancer phenotype with improved survival among patients with higher levels of CFTR expression within malignant cells.²¹ Interestingly, colonic expression of CFTR is highest in the cecum, with progressively lower expression distally.²²

Loss of CFTR expression in CF also induces an intra-intestinal inflammatory milieu, with many patients having significantly elevated fecal calprotectin and small bowel lesions, including areas of villous blunting and erosions/ulcerations.^23-25 The extent of inflammation can be profound, with levels approaching those of active IBD (e.g., median calprotectin >500) and with mucosal lesions that could mimic significant NSAID enteropathy or Crohn's disease.^{23, 24} These findings have been associated with pancreatic insufficiency²⁴ and inversely associated with lung function,²³ both of which are also predictors of severe disease. CF is also associated with abnormal gut permeability.²⁵ This combination of increased inflammation and permeability has been postulated to promote tumorigenesis through increasing exposure of colonocytes to microbial toxins.²⁶

Another prominent explanation for the increased risk of CRC in CF is that it may be a product of “systemic dysbiosis.” Within the first few months of life, significant alterations in microbial diversity are seen among infants with CF even before antibiotic exposures.²⁷ Microbiome studies have shown a reduction in Bifidobacterium and Bacteroides spp. relative to healthy controls and enrichment of Proteobacteria such as E. coli and Enterobacter as well as species such as Fusobacterium, which have been linked to CRC in adults.^{27, 28} As a result of these changes, CF patients have an increased abundance of proinflammatory bacteria species and increased metabolism of short-chain fatty acids, which are partially responsible for protecting the GI mucosa.^{27, 28} Studies have shown that probiotics and antibiotics in CF patients can reduce calprotectin levels,^{29, 30} lending credence to a causal role for microbiome changes in stimulating inflammation. Moreover, microbiome changes are ameliorated in patients with milder phenotypes, as suggested by pancreatic sufficiency²⁸; milder phenotypes are also correlated with a lower risk of CRC, as previously discussed. This may, in part, be due to worsening dysbiosis in patients with fat malabsorption and a high-fat diet, which is selected for a proinflammatory microbiome.³¹

There is also plausible evidence that reduced CFTR expression directly sensitizes epithelial cells within the GI tract to activation of oncogenic pathways when exposed to an inflammation-enriched environment. In a study³² using CRC-derived cell lines, CFTR knockout resulted in an exuberant production of IL-6 and IL-8 when cells were exposed to inflammatory signaling (via TNF and IL-1β) mediated through MAPK-ERK and NF-κB. These pathways have been heavily implicated in CRC growth and metastasis.^{33, 34} Moreover, loss of CFTR results in reduced bicarbonate secretion and intracellular alkalinization within intestinal stem cells (ISCs), which upregulates Wnt/β-catenin signaling and causes modest hyperproliferation.³⁵ Over decades, this growth advantage increases the risk of malignancy. This observation could explain the tendency of CFTR knockout mice to develop small bowel cancers, while humans primarily develop CRC as mice and humans have the bulk of their ISCs in these respective segments of the GI tract.³⁵

The mechanism for increased prevalence of CRC among lung transplant patients with CF is not completely understood. Prior studies have shown that much of the increased prevalence of CRC in post-lung transplant patients is driven by secondary conditions that increase the risk of CRC independently, such as CF, PSC, and IBD.¹⁵ This risk may be worsened by immunosuppression shifting balance from effector T-cell activity to regulator T-cell (Treg) activity which promotes immune evasion by tumor cells, as increased Treg levels predict higher risk of CRC.³⁶ However, the degree of immunosuppression does not appear to be clearly correlated with CRC risk.^{15, 36} Studies have shown that cyclosporine and azathioprine increase the risk of proximal colon cancer in posttransplant CF patients (OR: 1.53, 95% CI: 1.05–2.23).¹⁵ Cyclosporine increases TGF-β production in tumor cells which promotes growth and immune evasion while azathioprine impairs DNA synthesis and increases risk of mutations.³⁶ Synergy between pretransplant risk for CRC, immunosuppression, and carcinogenic immunosuppressive agents may explain the heightened risk for CRC after transplant in CF.

Given the manifold mechanisms at play for the pathogenesis of early-onset CRC in CF, treatments directed at individual arms of the disease process carry the risk of being ineffective. It is possible that CFTR modulators, when used early in life, may be able to significantly reduce the risk of future malignancy by targeting these diverse mechanisms that hinge on reduced CFTR activity. Early studies show CFTR modulators may help restore a more normal microbiome and reduce fecal calprotectin levels³⁷ as well as normalize small intestinal pH to promote normal bowel function.³⁸ Further observational studies will be essential to determine the effect of increasing uptake of CFTR modulators on the risk of developing adenomatous polyps and CRC.

3 CURRENT SCREENING RECOMMENDATIONS

The incidence of CRC in pre- and posttransplant CF patients is higher than in the general population.^{16, 17} Immunosuppression associated with transplantation increases cancer risk.^{15, 36} Multiple society guidelines for the screening and surveillance of colon cancer have adjusted recommendations based on the well-established increased risk of CRC in CF. In 2015, the Cystic Fibrosis Foundation (CFF) formed an expert task force of 18 members, including pulmonologists, gastroenterologists, a social worker, a nurse coordinator, a surgeon, an epidemiologist, a statistician, a CF adult, and a parent to perform a comprehensive literature review to provide recommendations for screening. In addition, microsimulation modeling was also conducted based on the MISCAN-Colon model to determine the most cost-effective screening strategies in patients with and without organ transplants.^{39, 40} The task force screening recommendations were later published in 2018 in conjunction with the American Gastroenterological Association (AGA).⁴⁰ According to these recommendations, the preferred screening method is with colonoscopy initiated at age 40 years, subsequent rescreening at 5 years, and surveillance at 3 years or less, depending on individual patient findings. Initial CRC screening in posttransplant patients is recommended at age 30 years within 2 years of transplantation. It should be noted that these follow-up intervals are based on the assumption that the patient undergoes an intensive bowel preparation (which includes 3−4 1 L purgative washes with the last wash completed within 4−6 h before the procedure) with adequate visibility.⁴⁰

There are several salient points tabulated in the 10 consensus recommendation statements, which include colonoscopy as the screening exam of choice with intensive bowel preparation and also special consideration for shared decision-making among the patient, gastroenterologist, and a CF healthcare professional when pursuing colonoscopy. The shared decision-making model helps bring attention to a patient's transplant candidacy, quality of life, and life expectancy. In the clinical setting, there may be situations where the sedation risk from pursuing a colonoscopy outweighs the potential benefit of screening in patients with poor lung functional capacity who are non-transplant candidates. It should be noted that there is currently no lung function threshold below which colonoscopy should not be pursued, but rather recommendations for a shared decision-making model with CF specialist, CF patient, and endoscopist.⁴⁰

The National Comprehensive Cancer Network guidelines share the same recommendations as the AGA guidelines, which are both based on the CFF expert task force.⁴¹ Several other foundations focused on CF patient care do not at present have specific recommendations on colonoscopy screening; these include the European Cystic Fibrosis Society, Cystic Fibrosis Trust (based in the United Kingdom), Cystic Fibrosis Canada, Cystic Fibrosis Australia, Cystic Fibrosis Ireland, and Cystic Fibrosis Worldwide.

4 EVOLVING DATA ON CRC RISKS AND POTENTIAL USE OF NONINVASIVE TESTING

According to recent data in the general population, there have been significant changes in the incidence and death rates of CRC across different age groups. Among individuals aged 65 years and older, there has been a consistent and rapid decline in CRC incidence, with a decrease of 3.3% annually between 2011 and 2016.⁴² Additionally, CRC death rates have shown a similar age-dependent trend, declining by 3% annually in this age group. However, the situation is different for individuals younger than 65 years. The incidence of CRC has actually increased among those aged 50−64 years, with a 1% annual increase observed.⁴² Among individuals under the age of 50, there has been an even greater annual increase in CRC incidence at 2%.⁴² The trend in CRC death rates also varies across age groups. While there has been a slight annual decline of 0.6% in CRC death rates for individuals aged 50−64 years, those under 50 years of age have experienced a concerning annual increase of 1.3% in CRC death rates.⁴² These findings highlight the need for increased attention and targeted interventions in younger age groups to address the rising incidence and death rates of CRC.

Due to the rising incidence of CRC in younger adults within the general population, the United States Preventive Services Task Force has decreased the age at which to begin screening from 50 to 45 years for non-CF individuals.⁴³ As detailed above, the CF Foundation task force recommendation for people with CF who have not had a transplant is to initiate screening at age 40. Ultimately, at this time, the decision to undergo a colonoscopy before the age of 40 years old in a non-transplanted patient with CF is based on personal risk factors and individual circumstances. Because CF is considered a colon cancer syndrome, with time, it is possible that recommendations for initiation of screening in CF may change to an earlier age, as is the case for patients with familial CRC syndromes like familial adenomatous polyposis syndrome and hereditary non-polyposis colorectal cancer syndrome, starting at 10−15 years, and 20−25 years, respectively.⁴⁴

With new screening recommendations starting at an earlier age for the general population and the inherent early age of screening for patients with CRC syndromes, there have been several studies developing less invasive blood-based screening methods often referred to as liquid biopsies, which include analysis of circulating tumor DNA (ctDNA), detection of circulating tumor cells, MicroRNA analysis of the blood, analysis of alterations in cell-free DNA, and identifying blood protein biomarkers. As of now, while these blood-based screening options for CRC hold promise, results are still preliminary, and more studies and clinical trials are needed to validate the accuracy, sensitivity, and specificity of these tests before they can be widely adopted for screening purposes.^45-48 If any of these become validated, it would definitely be worthwhile to evaluate their use in CF.

As of now, as detailed in the screening recommendations developed by the CF Foundation CRC screening task force, there is insufficient data for other forms of CRC screening in the CF population, including stool-based testing (e.g., fecal immunochemical test [FIT]), computed tomography, and flexible sigmoidoscopy.⁴⁰ Additional information is needed to determine the efficacy of alternative screening tests in CF patients because current studies are limited. Cost modeling suggests stool-based testing with FIT may be a cost effective option for screening in CF, but more research in the CF population is needed.³⁹ A multicenter study is ongoing to assess the sensitivity of multi-target stool DNA testing and fecal immunochemical testing in people with CF undergoing screening colonoscopy (NCT05362344). The impact of CFTR modulators on CRC risk in patients with CF is as yet unknown. Although theoretically restoring CFTR function could reduce CRC risk, excessive weight gain, which may occur with CFTR modulator use, could also impact risk.

5 CONCLUSION

In summary, advancements in therapy, such as the advent of CFTR modulators, have resulted in dramatic improvement of life expectancy for people with CF, and there is a burgeoning need for research on diseases of aging in this population. People with CF have a heightened risk of CRC and polyp formation which is accelerated after transplant. Studies suggest this is not a mere association, as knockout of CFTR expression can induce dramatic tumor formation in animal models, and patients with CF have significant luminal inflammation and gut microbial derangements that are expected to promote tumorigenesis. Although it is plausible that long-term use of CFTR modulators may eventually reduce the burden of intestinal cancers among patients with CF, a decrease in the incidence of CRC has not yet been seen in recent studies. The current consensus is supportive of an aggressive screening and surveillance regimen that utilizes colonoscopy only. As a high-risk population, screening colonoscopy should be initiated by age 40 in patients without transplant and every 5 years if screening is negative, and every 3 years or more often for patients with polyps. For patients with organ transplants, screening should be initiated at age 30 if they have had their transplant for 2 or more years, with similar intervals as non-transplanted patients. There are currently ongoing trials to investigate the safety and efficacy of using noninvasive screening tools such as multi-target stool DNA and fecal immunochemical testing in lieu of colonoscopy.

AUTHOR CONTRIBUTIONS

Zain Raza: Writing—original draft. Bianca N. Islam: Writing—original draft. Christine Y. Hachem: Supervision; writing—review and editing; conceptualization. Linda C. Cummings: Writing—review and editing; supervision; visualization; conceptualization.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.