The increasing burden of early-onset pancreatic and biliary tract cancers: a review of risk factors

Abstract

The rising incidence of early-onset pancreatic and biliary tract cancers (PBTCs) challenges conventional assumptions that these malignancies primarily affect older populations. Although early-onset PBTCs have a similarly poor prognosis to later-onset cases, emerging evidence suggests that unique or mixed genetic, environmental, and lifestyle factors contribute to their distinct etiologies. This review synthesizes current data on the epidemiology, risk factors, and molecular features of early-onset (in individuals <50 years of age) pancreatic ductal adenocarcinoma and biliary tract cancers. Recent studies indicate a significant increase in early-onset PBTC incidence, with variations by sex, geographic region, and genetic predisposition. Hereditary factors, including BRCA1/2, CDKN2A, and mismatch repair gene mutations, contribute to familial clustering, while other nonhereditary risk factors such as obesity, smoking, pancreatitis, and diabetes seem to disproportionately impact younger patients.

Additionally, environmental exposures, including air pollution, pesticides, and endocrine-disrupting chemicals, are suggested to contribute to carcinogenesis, while changes in microbiota may cause inflammation, immune modulation, and treatment resistance. As the burden of early-onset PBTCs grows, refining screening strategies and integrating microbiome-based risk assessment into biobanking strategies will be critical. Future research should focus on age-stratified genetic risk profiling, microbiota-driven interventions, and personalized therapeutic approaches to improve early detection and patient outcomes.

Highlights

• •Early-onset PBTCs are rising, challenging the notion that these cancers mainly affect older populations.
• •Genetic, environmental, and lifestyle factors may contribute to early-onset PBTC development.
• •Environmental toxins and microbiota changes may fuel carcinogenesis and treatment resistance.
• •Improved screening, microbiome-based risk assessment, and personalized therapies are key to better outcomes.

Introduction

The rising incidence of pancreatic and biliary tract cancers (PBTCs) in individuals aged <50 years (i.e. early onset) marks a significant shift in epidemiological trends. Traditionally considered diseases of older populations, these malignancies are now more frequently diagnosed in younger patients, suggesting that early-onset PBTCs may have distinct etiologies, clinical presentations, and molecular characteristics compared with their late-onset counterparts.1, 2, 3, 4, 5, 6 Although early-onset PBTCs share a poor prognosis with later-onset cases, emerging evidence suggests that they represent distinct disease entities rather than merely earlier presentations of the same malignancy. This review synthesizes current data on the epidemiology, genetic and nongenetic risk factors, environmental exposures, and the role of the microbiota in early-onset pancreatic ductal adenocarcinoma (PDAC) and biliary tract cancers (BTC), to provide insights for both clinicians and researchers.

Epidemiology of early-onset PBTCS

PDAC is the sixth leading cause of cancer-related mortality worldwide, with an overall 5-year survival rate of ∼11.5% and markedly lower survival rates (<5%) in metastatic disease.¹ Global estimates from 2022 indicate that PDAC accounted for ∼510 992 new cases and 467 409 deaths, with a slight male predominance.² Projections indicate that PDAC could become the second-leading cause of cancer-related mortality by 2030, with annual incidence increases of 2.4% in females and 1.2% in males.^3,7,8 Although the median age at diagnosis is typically 65-70 years, early-onset PDAC (EOPC)—defined as diagnosis before 50 years of age—represents a small but growing subset (<5% of all cases).^2,8,9

Regional data highlight a rapid increase in PDAC incidence among younger individuals. For instance, between 2008 and 2017, French women aged 30-49 years experienced an average annual incidence increase of 7.74%, while United States women in the same age group showed a 2.31% rise; in contrast, increases among men were 1.46% in France and 0.49% in the USA.^4,5 Furthermore, United States data indicate that PDAC incidence in women aged <55 years rose at a rate of 1.93% per year [95% confidence interval (CI) 1.57% to 2.28%] compared with 0.77% in men—with the most dramatic increases among those aged 15-34 years.⁵ Similar trends have been observed in the UK, where the incidence among females from birth to 24 years old increased by 208%, and 34% in those aged 25-49 years.¹⁰ Despite representing a small fraction of PDAC cases, EOPC has a significant impact on potential years of life lost (PYLL). In some European countries, early-onset disease may account for as much as 40% of the total PYLL attributable to PDAC.⁹

BTCs, which include intrahepatic and extrahepatic cholangiocarcinoma along with gallbladder cancer, are less common than PDAC. With an annual incidence of ∼1.3 cases per 100 000 individuals, ∼8000 new cases are diagnosed annually in the USA.¹¹ Despite their lower overall incidence, BTCs are frequently diagnosed at advanced stages, with a median overall survival of 10-12 months.^12,13 In France, the median age at BTC diagnosis is 72 years for men and 78 years for women.¹⁴ However, certain BTC subtypes—such as intrahepatic cholangiocarcinoma and gallbladder cancer—are increasing at an especially fast rate in younger patients, with annual rises of 8% and 2%, respectively.¹⁵ These findings have led to the hypothesis that early-onset PBTCs may be biologically distinct entities with unique risk factors (Figure 1) and clinical trajectories.^16,17

Figure 1.

Known and suspected risk factors of early-onset pancreatic and biliary cancers. Gray: modifiable risk. Blue: non-modifiable risk. Green: environmental risk. PFAS, Per- and polyfluoroalkyl substances.

Clinical risk factors

Hereditary factors

Hereditary factors play a crucial role in early-onset PDAC risk. Familial aggregation is observed in ∼5%-10% of PDAC cases, and numerous studies have identified germline mutations in key cancer susceptibility genes as significant contributors.^1,2 Mutations in BRCA1, BRCA2, and CDKN2A—as well as alterations in DNA mismatch repair genes associated with Lynch syndrome, along with mutations in ATM, PALB2, STK11, and TP53—are commonly observed in hereditary pancreatic cancer.^16,17 For example, the Pancreatic Cancer Genetic Epidemiology (PACGENE) study reported that 7.4% of familial PDAC cases harbor pathogenic mutations in BRCA1, BRCA2, or CDKN2A.¹⁸ Findings from the United States National Familial Pancreas Tumor Registry suggest that relatives of EOPC patients have a markedly higher risk, particularly in families with a documented young-onset case.^16,19 In one study involving 450 EOPC patients, 31.9% of those tested were found to have pathogenic variants, most commonly in BRCA1/2, PALB2, and CHEK2.19, 20, 21, 22 Recent research further highlights that both genetic mutations and nongenetic factors contribute to early-onset pancreatic cancer risk.²³

For BTCs, the data regarding hereditary predisposition remain limited. Given the shared embryological origin of the pancreatic head and the distal bile duct, familial factors may also contribute to BTC risk. However, the prevalence of identifiable germline mutations in BTC is low, and current evidence does not support routine genetic testing for BTC. Notably, microsatellite instability-high tumors—often associated with Lynch syndrome—appear more frequently in early-onset BTC (4.1% versus 2.4% in later-onset cases), although this difference has not consistently reached statistical significance.¹⁹

Environmental and lifestyle risk factors in PDAC

Multiple population-based studies have demonstrated that several nonhereditary risk factors are more strongly associated with early-onset PDAC (Table 1). Established risk factors such as cigarette smoking, obesity, diabetes, tall stature, and non-O blood type have been found to have stronger associations in younger individuals compared with older patients.^24,25 A genome-wide association study (GWAS) further indicated that a weighted genetic risk score correlates robustly with PDAC in early-onset cases.^21,22,24,25 Moreover, demographic factors such as male sex and black race have been consistently linked with a higher risk of PDAC among younger populations.

Table 1.

Relationship between potential risk factors and early onset of PDAC and BTC

Category	Risk	Tumor location	Type of study	Authors	Number of patients	Measure of association	Comment
Hereditary factors
	Germline mutation	PDAC	Pancreatic Cancer Genetic Epidemiology	Petersen et al.	379 familial PDAC	NA	7.4% of familial PDAC cases harbor pathogenic mutations in BRCA1, BRCA2, or CDKN2A
	Germline mutation	PDAC	United States National Familial Pancreas Tumor Registry	Beeghly-Fadiel et al.	1407 PDAC	NA	31.9% of EOPC tested had pathogenic variants
Environmental and lifestyle risk factors in PDAC
	Obesity	PDAC	United States case-control study	Li et al.	841 PDAC and 754 controls	P < 0.001	Younger age at diagnosis in case of obesity at age 20-49 years
		PDAC	United States case-control study	McWilliams et al.	1954 PDAC and 3278 control	P = 0.52	No interaction of obesity with age at PDAC onset (<45 years versus <60 years)
	Cigarette smoking	PDAC	United States case-control study	Rulyak et al.	251 individuals of 28 families	HR 3.7, 95% CI 1.8-7.6	Smokers developed PDAC 10 years younger than nonsmokers
		PDAC	Retrospective Chinese cohort	Huang et al.	331 PDAC	P = 0.026	Mean PDAC diagnosis age of 59.0 years for smokers compared with 61.5 years for nonsmokers
		PDAC	Retrospective Chinese cohort	Jiang et al.	1789 PDAC, 156 EOPC	P = 0.024	Higher proportions of ever smokers among younger PDAC patients (≤45 years versus >45 years)
		PDAC	Retrospective US cohort	Salem et al.	516 PDAC	P = 0.03	Higher proportions of ever smokers among younger PDAC patients (≤50 years versus ≥70 years)
		PDAC	Retrospective Japanese cohort	Eguchi et al.	36145 PDAC, 526 EOPC	P = 0.293	No difference in smoking exposure between EOPC and non-EOPC
		PDAC	Retrospective Romanian cohort	Bunduc et al.	148 PDAC	P = 0.06	No different proportion of ever smokers among younger PDAC patients (≤45 years versus ≥45 years)
		PDAC	Retrospective Japanese cohort	Ohmoto et al.	908 PDAC, 17 >40 years old	P = 0.475	No difference in smoking exposure between EOPC and non-EOPC
		PDAC	United States case-control study	McWilliams et al.	1954 PDAC and 3278 control	P = 0.20	No interaction between smoking and age
		BTC	Retrospective French cohort	Lebeaud et al.	1256 BTC, 188 EOBTC	P = 0.0001	Higher proportion of active smokers in EOBTC
	Physical activity	PDAC	European case-control study	Noor et al.	88 PDAC and 3970 controls	HR 4.0, 95% CI 1.1-14.3	Physical inactivity before the age of 60 years was associated with a significantly increased risk of PDAC compared with active individuals
	Alcohol consumption	PDAC	Retrospective Chinese cohort	Huang et al.	331 PDAC	P = 0.022	Younger age at diagnosis in case of alcohol consumption
		PDAC	Retrospective Romanian cohort	Bunduc et al.	148 PDAC	P = 40.01	Higher proportions of people consuming alcohol among younger PDAC patients (≤45 years versus ≥45 years)
		PDAC	United States case-control study	McWilliams et al.	1954 PDAC and 3278 controls	≥26 g/day: HR 2.18, 95% CI 1.17-4.09	No interaction between alcohol and age at PDAC onset (<45 years versus <60 years), P = 0.20
		PDAC	Retrospective Chinese cohort	Jiang et al.	1789 PDAC, 156 EOPC	P = 0.697	No significant differences in alcohol consumption were noted between patients diagnosed with PDAC at younger ages
		PDAC	Retrospective Japanese cohort	Eguchi et al.	36145 PDAC, 526 EOPC	P = 0.623 (frequency) and P = 0.8530 (quantity)	No differences between EOPC and non-EOPC (≤40 years versus ≥40 years)
	Pancreatitis	PDAC	Retrospective United States cohort	Dzeletovic et al.	2573 PDAC	P = 0.005	Patients with a pancreatitis history were younger at diagnosis than those without
		PDAC	Pooled analysis if 10 case-control studies	Duell et al.	4674 PDAC, 10 703 controls with history of pancreatitis	P = 0.006	Patients younger than 65 years were more likely to have a history of pancreatitis than older patients
		PDAC	Cohort study	Augustine et al.	98 PDAC, 82 tropical pancreatitis	P = 0.0035	Patients with tropical pancreatitis were diagnosed at a significantly younger age (47.5 years) compared with those with PDAC alone
		PDAC	Retrospective Romanian cohort	Bunduc et al.	148 PDAC	P = 0.59	No significant differences in chronic pancreatitis frequency between younger age (≤45 years) and older age (>45 years) patients
		PDAC	Retrospective Japanese cohort	Eguchi et al.	36145 PDAC, 526 EOPC	P = 0.600	No differences between those diagnosed before 40 years of age and those aged ≥40 years
		PDAC	United States case-control study	McWilliams et al.	1954 PDAC and 3278 controls	HR 2.66, 95% CI 0.19-37.9	No significant correlation with diagnoses before age 45 years after adjustments
	Pancreaticobiliary maljunction	BTC	Retrospective Japanese cohort	38	774 EOBTC	P < 0.001	Higher proportion of people with pancreaticobiliary maljunction in EOBTC
	Diabetes	PDAC	Retrospective United States cohort	Gupta et al.	149 PDAC and 36 482 controls with diabetes	P = 0.07	No trend in rate ratio between increasing age and rate of PDAC in case of diabetes
		PDAC	Prospective Italian cohort	Dugnani et al.	296 PDAC	P = 0.8	No different mean age at diagnosis in case of diabetes
		PDAC	United States case-control study	McWilliams et al.	1954 PDAC and 3278 controls	HR 0.85, 95% CI 0.25-2.93	No interaction between diabetes and age at diagnosis
	Cirrhosis	BTC	Retrospective French cohort	Lebeaud et al.	1256 BTC, 188 EOBTC	P = 0.2299	No different proportion of underlying cirrhosis in EOBTC
		BTC	Retrospective United States cohort	Pappas et al.	847 BTC, 124 BTC	P = 0.29	No different proportion of underlying cirrhosis in EOBTC
	Cholangitis	BTC	Retrospective United States cohort	Pappas et al.	847 BTC, 124 BTC	P < 0.001	Higher proportion of primary sclerosing cholangitis in EOBTC (≤50 years versus ≥50 years of age)
	HBV, HCV	BTC	Retrospective United States cohort	Pappas et al.	847 BTC, 124 BTC	P = 0.55 and P = 0.64	No different proportion of HBV and HCV in EOBTC

In red: positive association. In blue: no positive association. Gray: modifiable risk. Blue: non-modifiable risk. Green: environmental risk.

BTC, biliary tract cancer; CI, confidence interval; EOBTC, early-onset biliary tract cancer; EOPC, early-onset pancreatic ductal adenocarcinoma; HBV, hepatitis B virus; HCV, hepatitis C virus; HR, hazard ratio; PDAC, pancreatic ductal adenocarcinoma.

Obesity

Obesity is a well-recognized risk factor for PDAC. The Global Burden of Diseases study has underscored that smoking, alcohol use, and high body mass index (BMI) are major contributors to pancreatic cancer mortality.^2,11 An epidemiologic review by the International Agency for Research on Cancer specifically identified obesity as a key risk factor for early-onset PDAC. In a United States case-control study that included 841 PDAC patients and 754 controls, individuals who were overweight (BMI 25-29.9) or obese (BMI ≥30) between the ages of 14 and 49 years had an increased risk of developing PDAC—and experienced cancer onset 2 to 6 years earlier than those with normal weight.²⁶ Notably, the effect of obesity was particularly pronounced in men and smokers. In contrast, a multivariable analysis of 1954 patients reported that while obesity was significantly associated with PDAC in those aged under 60 years [odds ratio (OR) 1.28, 95% CI 1.08-1.52], the association was not statistically significant in patients <45 years of age (OR 1.13, 95% CI 0.67-1.90), suggesting age-related differences.^22,24,25

Cigarette smoking

Cigarette smoking is one of the most consistently implicated risk factors for PDAC.²³ Multiple studies report that smokers (tobacco users) are diagnosed with PDAC ∼10 years earlier than nonsmokers. For example, a United States case-control study by Rulyak et al. found that the mean age at diagnosis was 59.6 years for smokers versus 69.1 years for nonsmokers.²⁷ Similarly, a retrospective cohort study from China²⁴ (N = 331) observed a mean diagnosis age of 59.0 years in smokers compared with 61.5 years in nonsmokers. When focusing specifically on early-onset PDAC, only two out of seven studies showed statistically significant associations between smoking and younger age at diagnosis. Other studies—including a United States nested case-control study and pooled analyses—did not reveal significant differences after adjusting for confounders such as family history, diabetes, sex, smoking intensity, alcohol use, and BMI.27, 28, 29, 30 A GWAS further indicated that smoking may have a pronounced effect in early-onset PDAC (OR 2.92, 95% CI 1.69-5.04).^20,31

Physical activity

Although fewer studies have investigated physical activity in relation to PDAC risk, available evidence suggests that physical inactivity is associated with an increased risk of the disease. In a multivariate case-control analysis from the European Prospective Investigation of Cancer (EPIC)-Norfolk Study, physical inactivity before the age of 60 years was associated with a hazard ratio (HR) of 4.0 (95% CI 1.1-14.3) for PDAC compared with active individuals.³¹ However, this association was not observed among those who became inactive after the age of 60 years. Another large prospective cohort study reported that physical activity was linked to an increased risk of PDAC in individuals aged ≤70 years (HR 1.37, 95% CI 1.11-1.69), whereas no significant association was found in those aged >70 years (HR 0.98, 95% CI 0.83-1.15).³²

Alcohol consumption

The relationship between alcohol consumption and PDAC risk is complex and appears to differ by age. A retrospective cohort study from China reported that PDAC patients who regularly consumed alcohol were diagnosed at a significantly younger age (57.96 years versus 61.37 years, P = 0.022).²⁰ Similarly, a Romanian cohort study observed that 43% of patients <45 years reported alcohol consumption, compared with only 5% among older patients.³⁰ However, a multivariable analysis of eight pooled case-control studies (N = 1954) found that although alcohol intake exceeding 26 g/day increased PDAC risk in both age groups, the overall risk did not differ significantly between patients aged <45 years and those <60 years of age (P = 0.20).^21,25,29

Pancreatitis and cholangitis

A history of pancreatitis is an established risk factor for PDAC; its specific association with early-onset disease, however, remains debatable. Several studies have reported that PDAC patients with a history of pancreatitis are diagnosed at a younger age. For example, a retrospective cohort study from India involving 98 PDAC cases found that patients with tropical pancreatitis were diagnosed at a mean age of 47.5 years compared with 61.5 years for those without (P < 0.001).³³ Similarly, a United States study of 2573 participants reported that individuals with a history of pancreatitis were diagnosed at a younger mean age (63 years versus 65 years, P = 0.005).³⁴ A pooled analysis of 10 case-control studies further noted that a history of pancreatitis lasting >2 years was more prevalent among patients aged <65 years, suggesting an association with earlier PDAC onset.³⁵ Conversely, data from a Romanian cohort and a large Japanese Cancer Registry (N = 36 145) did not show significant differences in the frequency of chronic pancreatitis between younger and older PDAC patients.^29,30 Moreover, a multivariable analysis indicated that while long-standing pancreatitis was associated with PDAC diagnosis before age 60 years, its association with diagnosis before age 45 years was not statistically significant—possibly due to small sample sizes.²²

Primary sclerosing cholangitis is also a risk factor for BTC. In the study of Boonstra et al.³⁶ that included patients with primary sclerosing cholangitis, median age at BTC diagnosis was 47 years with a range of 21-87 years. Median time between PSC primary sclerosing cholangitis diagnosis and BTC diagnosis was 6 years (range 0-36 years).

Diabetes

Although diabetes mellitus and insulin resistance are known risk factors for PDAC, the available evidence does not indicate that diabetes confers a disproportionately higher risk in younger patients compared with older individuals.^{22,27,35,37,38}

Nonhereditary risk factors in BTC

Compared with PDAC, fewer studies have systematically examined nonhereditary risk factors for early-onset biliary tract cancers (EOBTC). In a Japanese study of 774 patients <50 years of age, pancreaticobiliary maljunction was the most common congenital anomaly associated with EOBTC, identified in 10.6% of cases—and in nearly 39% of patients aged <30 years.³⁹ Other factors that have been associated with BTC include choledochal cysts, cholelithiasis, hepatitis B virus infection, and a history of other cancers; these associations, however, tend to vary according to tumor location.⁴⁰ Additional exposures such as liver fluke infection, chronic liver disease, inflammatory biliary conditions, obesity, and tobacco/alcohol use have also been suggested as contributors to EOBTC, although robust age-specific data are still emerging.41, 42, 43

Environmental risk factors

Environmental risk factors encompass all external physical, chemical, biological, and work-related exposures that may impact human health.⁴⁴ A recent emphasis on ‘uninformed and unconsented exposure’ has highlighted the challenge of quantifying exposures that occur without individuals’ explicit consent and that vary over time, geography, and socioeconomic status. While exposures such as tobacco smoking are well characterized, many unconsented exposures—particularly chemical contaminants—are more difficult to assess.⁴⁴

Early-life exposures and the DoHAD concept

Early-life exposure to environmental agents may predispose individuals to cancer later in life through epigenetic modifications and gene–environment interactions. The developmental origins of health and disease (DoHAD) hypothesis posits that adverse exposures during critical developmental periods can permanently reprogram cellular functions, thereby increasing susceptibility to chronic diseases such as PDAC and BTC.^44,45 Although conditions like diabetes and obesity are known risk factors for these cancers, the contribution of early developmental programming to early-onset PBTCs remains an active area of research.⁴⁶ Since early-onset PBTCs represent only 5%-10% of cases diagnosed before age 50 years, gathering robust epidemiological evidence to support the DoHAD hypothesis remains challenging.^2,11

Air pollution

Air pollution—encompassing particulate matter (PM), black carbon, and carbon monoxide—is a recognized carcinogen, particularly in relation to lung cancer.^44,47 A recent meta-analysis by Pritchett et al. provided inconclusive evidence for an association between air pollution and the risk of PDAC or BTC, partly because few studies have explored age-related effects.⁴⁸ One study from Los Angeles, however, demonstrated that high levels of PM2.5 were associated with an HR of 3.59 (95% CI 1.60-8.06) for PDAC, with an overall HR of 1.61 (95% CI 1.09-2.37) for PM exposure.⁴⁹ Although associations between air pollution and hepatocellular carcinoma have been reported,^50,51 few studies have directly addressed the impact on BTC, and data regarding maternal exposure remain limited.^46,50,52

Chemical exposure

Exposure to chemical agents, including pesticides, polychlorinated biphenyls (PCBs), and benzene, has been suggested as a potential factor in the development of early-onset PBTC. Many of these chemicals persist in the environment and may accumulate in soil, water, and human tissues. Some chemicals act as endocrine disruptors, which is particularly concerning given the endocrine functions of the pancreas.⁵³ Epidemiological studies employing age-matched designs have demonstrated dose–response relationships linking organochlorine substances such as trans-Nonachlor and dichloro-diphenyl-trichloroethane (DDT) to an increased risk of PDAC.54, 55, 56 In contrast, although hypotheses have been proposed linking pesticide exposure (for example, via estrogenic pathways involving glyphosate) to BTC risk, epidemiological confirmation remains pending.⁵⁷

Per- and polyfluoroalkyl substances (PFAS)

PFAS represent a new class of environmental contaminants that have garnered attention in recent decades. With >1400 compounds used in diverse industrial applications—from packaging to waterproofing—PFAS are characterized by their resistance to degradation.^54,55 They are ubiquitously present in the environment and human serum, with half-lives ranging from 1.8 to 8 years.⁵⁸ Although some studies have begun to explore the association between PFAS serum levels and PDAC risk, the evidence remains inconclusive, particularly with respect to age-dependent effects. Given their widespread occurrence and potential toxicity, further research is needed to fully elucidate the role of PFAS in early-onset PBTC.59, 60, 61, 62

The microbiota and its role in carcinogenesis

General considerations

The human microbiota, particularly within the gastrointestinal tract, plays an essential role in maintaining host homeostasis. Under normal conditions, the gut microbiota contributes to the integrity of the intestinal barrier, regulates drug metabolism, and modulates immune responses.^63,64 Similarly, the oral microbiota is critical for maintaining oral and systemic health, and disruptions in its composition have been linked to chronic inflammation and immune dysregulation.⁶⁵ When the balance of these microbial communities is disturbed—a state known as dysbiosis—it can contribute to carcinogenesis through multiple mechanisms.⁶⁶

Gut microbiota and carcinogenesis

Dysbiosis in the gut microbiota may promote tumor initiation and progression by increasing the production of pro-inflammatory bacterial toxins and metabolites, altering local and systemic immune responses, and disrupting host metabolism.⁶⁷ Although much of the research on dysbiosis has focused on colorectal cancer, emerging data suggest that similar microbial alterations may be present in PDAC and BTC.⁶⁸ For instance, distinct microbial signatures have been identified in early-onset colorectal cancer compared with later-onset cases, suggesting that early-onset PBTCs might also harbor unique microbiome profiles.69, 70, 71, 72

The gut–liver axis in BTC

The bidirectional communication between the gut and liver is critical in the pathogenesis of BTC. The portal vein transports gut-derived metabolites and bacterial products to the liver, and the liver, in turn, secretes bile acids that help modulate gut microbiota composition.^68,73 Disruption of the intestinal barrier can allow bacterial endotoxins, such as lipopolysaccharide (LPS), to enter the portal circulation and activate inflammatory pathways in hepatic and biliary tissues. Secondary bile acids, including tauroursodeoxycholic and glycoursodeoxycholic acids, further stimulate the release of pro-inflammatory cytokines [e.g. interleukin (IL)-1β, IL-6, IL-8, monocyte chemoattractant protein-1 (MCP-1), tumor necrosis factor (TNF)-α, and interferon (IFN)-γ], which can promote cholangiocyte proliferation and neoplastic transformation.74, 75, 76, 77 Additional mechanisms linking dysbiosis and BTC include chronic infections—such as those caused by Opisthorchis viverrini and the translocation of LPS in the biliary tract.^78,79 Studies of the gut microbiota in BTC patients have reported variable results, with some investigators observing increased alpha diversity and higher abundances of genera such as Lactobacillus, while others report decreased diversity or no significant differences compared with healthy controls.⁸⁰ The main enriched genera in BTC are Lactobacillus, Bacteroides, Firmicutes, Protebacteria, Verrucomicrobia, Actinomyces, Alloscardovia, Muribaculum, Shigella, and Klebsiella.⁸¹ Proposed microbiome-based signatures—including genera such as Citrobacter, Burkholderia, Faecalibacterium, Klebsiella, Ruminococcus gnavus group, Lactobacillus, Dorea, and Veillonella—have shown promise in differentiating BTC patients from healthy individuals, though heterogeneity among studies remains a challenge.82, 83, 84, 85

Biliary microbiota and BTC

Similar to the intestinal barrier, the biliary mucosa can be disrupted by various factors, including persistent infection, calculi, and chemical exposure, potentially contributing to malignant transformation.⁸⁶ Compared with the fecal microbiota, the bacterial load in bile is thought to be much lower and remains less studied, as bile was long considered sterile. Moreover, bile salts exhibit bactericidal properties, with modified bile salts selectively permitting the survival of commensal bacteria.⁸⁰ Metagenomic sequencing of the biliary mucosa has identified Proteobacteria, Firmicutes, and Bacteroides within the gallbladder system.⁸⁷ However, due to small sample sizes, heterogeneous sequencing techniques, inconsistent sample collection (predominantly stool-based), disease heterogeneity, frequent antibiotic use (especially for biliary infections), and other external factors, establishing a definitive relationship remains challenging.

Microbiota in PDAC

Contrary to previous assumptions that the pancreas was sterile, recent studies have demonstrated that it harbors its own microbial community. A healthy pancreas secretes peptides that help regulate gut microbial diversity and protect against inflammation.^88,89

PDAC and gut microbiota

In PDAC, patients typically exhibit reduced microbial alpha diversity, an over-representation of LPS-producing bacteria, and a depletion of butyrate-producing species.90, 91, 92 Metagenomic analyses have identified specific genera—including Veillonella, Klebsiella, Streptococcus, and Akkermansia—that are enriched in the stool of PDAC patients; these findings have been confirmed by several meta-analyses.⁹³

Tumor microbiota

Recent research on the tumor-associated microbiota in PDAC has revealed that bacterial DNA is present at significantly higher levels in tumor tissues compared with normal pancreatic tissue, challenging the long-held assumption that the pancreas is sterile.^94,95 Helicobacter pylori was the first pathogen identified in PDAC. Enrichment in other species such as Firmicutes and Proteobacteria, along with Akkermansia muciniphila, Lactobacillus, and Bacteroides—were found in PDAC tumors.^96,97 Increased tumor microbial diversity has been linked to improved survival, though findings remain inconsistent across studies.⁹⁸

At the 2024 annual American Society of Clinical Oncology (ASCO) conference, Jayakrishnan et al. demonstrated a higher alpha diversity in EOPC compared with late-onset PDAC patients, along with significant variation in tumor microbiome. In EOPC, tumor tissues were enriched with Enterobacter, Neisseria, and Escherichia genera, whereas Klebsiella and Bacillus genera were predominant in late-onset PDAC tumor tissues.⁹⁹

In the context of treatment, one study provided evidence that intratumor bacteria may mediate resistance to the chemotherapeutic drug gemcitabine.⁹⁴ A comprehensive review highlights the complex interplay between microbiota and pancreatic carcinoma pathogenesis.⁹⁷

Oral microbiota

Poor oral health, including periodontal disease and tooth loss, has also been independently associated with an increased risk of PDAC. Studies using salivary RNA profiling have shown elevated levels of pathogens such as Porphyromonas gingivalis, Fusobacterium, Leptotrichia, and Granulicatella in PDAC patients, suggesting that these organisms could serve as noninvasive biomarkers for early detection.^100,101

Of note, commensal bacteria such as Lactobacillus, Fusobacterium, and Lepotrichia have been associated with a decreased risk for PDAC. These bacteria may reduce gingival inflammation and inhibit the development of periodontal pathogenic bacteria.¹⁰¹

Despite these advances, variations between study cohorts, differences in sequencing methodologies, and multiple confounding factors have prevented the identification of a universal microbial signature for PDAC and BTC. It is increasingly evident that while dysbiosis plays a critical role in carcinogenesis, microbiota profiles alone are unlikely to serve as stand-alone screening tools; rather, they should be integrated with clinical, genetic, and environmental data for a comprehensive assessment of cancer risk.¹⁰²

Molecular profile

Molecular landscape of PBTCs is not very different between early-onset PBTCs and non-early-onset PBTCs. Recent studies consistently agree on a common alteration for BTC: a higher rate of FGFR2 fusions in EOBTCs, likely due to the enrichment of intrahepatic cholangiocarcinomas (iCCAs) in this young population. The proportion of FGFR2 fusions in EOBTCs varies from 11.7% to 15.7% across different series.^103,104 Regarding other alterations, there is a significantly higher proportion of BRAF and ATM mutations in the American CITY study among EOBTCs (P < 0.05), as well as a higher proportion of NIPBL fusions, though in small numbers (1.1% versus 0%), in another American study by Jayakrishnan.¹⁹ The French series also noted fewer IDH1 mutations (7.8% versus 16.6%), but this finding is not confirmed in other studies. In Tsilimigras’s series using data from The Cancer Genome Atlas Program, rarely mutated genes frequently implicated as oncogenic factors were identified, including CDKN2A, IDH1, BRAF, and FGFR2 in young iCCA patients.¹⁰⁵

In PDAC, the molecular results mainly come from two recent European single-center series by Castet and Rémond,^106,107 showing different molecular findings. In the Vall d’Hebron series, fewer KRAS oncogene mutations are observed in EOPC, which are enriched with potentially actionable alterations according to the European Society for Medical Oncology (ESMO) Scale for Clinical Actionability of molecular Targets (ESCAT) I-IIIA classification, both in the VHIO cohort and validation cohorts (19% versus 14% and 14% versus 8%, respectively). The data from the other French series differ somewhat, with no difference in KRAS mutation rates, but a lower tumor mutational burden and fewer CDKN2A/B alterations.

Emerging insights and future directions

Recent studies have begun to further characterize molecular subtypes and genetic alterations that are specific to EOPC, uncovering novel biomarkers and potential therapeutic targets.^23,25 Advances in multiomics research—including the integration of genomic, transcriptomic, proteomic, and microbiomics data— could offer a comprehensive view of the interplay between genetic predisposition, environmental exposures, and microbial dysbiosis.^19,108 Ongoing epidemiological analyses continue to refine our understanding of risk factors, and emerging multiomics studies underscore the importance of integrating data from multiple sources to elucidate the mechanisms underlying early-onset PBTCs. Furthermore, novel therapeutic approaches and personalized treatment strategies are under development based on these integrated insights, with the ultimate goal of improving patient outcomes, especially considering potential delayed diagnosis in those young patients, that might possibly explain some of the differences between early-onset PBTC and non-early-onset PBTC. Translating these findings into targeted prevention and early detection programs remains a key focus of ongoing research, with the potential to reduce the societal burden of these aggressive malignancies.

Conclusion

The landscape of pancreatic and biliary tract cancers in young patients is undergoing rapid change. Although early-onset PBTCs remain relatively rare compared with those diagnosed in older individuals, their increasing incidence has profound implications for public health due to the significant loss of potential years of life. Hereditary and nonhereditary risk factors such as obesity, cigarette smoking, alcohol consumption, pancreatitis, and diabetes further contribute to disease development.

Environmental exposures add an additional layer of complexity to the etiology of early-onset PBTCs. In parallel, the role of the microbiota—whether through gut dysbiosis, alterations in the oral microbiome, or distinct tumor-associated microbial communities—is increasingly recognized as critical in modulating cancer risk and progression.

Future research should employ integrated multiomics approaches and refined epidemiologic designs, including age-matched studies, to elucidate the intricate relationships between genetics, environmental exposures, and microbial dysbiosis. Standardizing methodologies, particularly in microbiome analyses, is essential for identifying robust and reproducible biomarkers that can ultimately facilitate early detection and inform personalized therapeutic strategies. These comprehensive efforts may help mitigate the burden of early-onset PBTCs and enhance patient outcomes.