Pancreatic Adenocarcinoma: Trends in Epidemiology, Risk Factors, and Outcomes.

There has been a significant increase in pancreatic cancer globally despite advancements in diagnostic and therapeutic interventions when compared with other cancers, which have been declining. Mortality continues to be high among those with pancreatic cancer. The increase in pancreatic cancer cases is attributed to the modifiable and nonmodifiable risk factors described in this review. Further studies are needed to solidify the association between other risk factors and pancreatic cancer. Health initiatives aimed at smoking cessation and obesity may reduce the incidence of pancreatic cancer. Screening for early detection in those with genetic predisposition have already been implemented. Further studies are needed on the role of risk-modifying agents in reducing the risk of pancreatic cancer.

There are several types of pancreatic cancer, but the most common type is pancreatic adenocarcinoma, which comprises 90% of all pancreatic cancer diagnoses. This review focuses on the epidemiology of adenocarcinoma of the pancreas and includes studies limited to pancreatic adenocarcinomas or studies of all pancreatic cancer combined, where the vast majority of patients have adenocarcinoma of the pancreas. Globally, pancreatic cancer is the 12th most common cancer and the 7th most common cause of cancer-related death. 1 By 2025, it is projected to be the third leading cause of cancer-related deaths in Europe. 2 Pancreatic cancer accounts for 2.6% of all new cancer diagnoses, 4.7% of all new cancer deaths, 1 and 12.7% of all deaths due to gastrointestinal-related cancers in the world. 3 In the United States, it accounts for 3.2% of all new cancer cases and 8% of all cancer deaths, and a lifetime risk of 1.7. 4 The incidence is highest in Western Europe with rates of 8.6 per 100,000, followed by Northern America at 8.0 per 100,000, Central and Eastern Europe at 7.5 per 100,000, and Northern Europe at 7.4 per 100,000. The incidence is lowest in Southeast Asia, with rates of 1.3 per 100,000. 5 In the United States, the incidence of pancreatic cancer has been increasing at a rate of 1% per year. Pancreatic cancer is the fourth leading cause of death in men and women. 6 The increasing trend is largely due to longer life spans, urbanization of countries worldwide, and exposure to increased modifiable risk factors. Countries with a very high and high human developmental index (HDI) and gross domestic product per capita have a higher number of pancreatic cancers. 5 , 7 Regions with very high and high HDI had age-standardized rates of 7.9 and 4.6 per 100,000, respectively, compared with 1.8 and 1.2 per 100,000 in regions with low and medium HDI, respectively. 5 Increased prevalence of lifestyle risk factors, increased availability of imaging, heightened health awareness, and longer life expectance have led to the higher incidence and mortality of pancreatic cancer. 8 The incidence of pancreatic cancer increases with age ( Fig. 1 ). Despite the median age of pancreatic cancer being 71 years in the United States, there has been an increase in incidence in younger patients. The average annual percent change in pancreatic cancer incidence increased with decreasing age, from 0.77% (95% confidence interval [CI], 0.57–0.98) for ages 45 to 49 years to 2.47% (1.77–3.18) for ages 30 to 34 years, and 4.34% (3.19–5.50) for ages 25 to 29 years. 9 Although this is partly due to the increase in younger patients, there is also an increase in the prevalence of modifiable risk factors in this population. Mortality from pancreatic cancer also varies depending on the part of the world and is highest in Western Europe, North America, and Central and Eastern Europe and lowest in Middle Africa and South Central Asia. 5 Mortality is highest in regions with very high and high HDI. 5 Survival rates for pancreatic cancer remain low despite recent improvements in overall 5-year survival from 2% from 1975 to 1979 to 10% in the United States and Europe in 2019. 7 The low survival rates are due to 52% of cases being diagnosed as metastatic disease (stage IV) with a 5-year survival rate of 3% in the United States, whereas only 11% of cases being diagnosed as stage I or localized cancer, with a 5-year survival rate of 41.6% 4 ( Figs. 2 and 3 ). Among patients who undergo surgical resection, the 5-year survival rate is approximately 15% to 25%. 10 , 11 A large study involving multiple national databases, including the Surveillance, Epidemiology and End Results (SEER) and national cancer registries of Netherlands, Belgium, Norway, and Slovenia, provided age-stratified survival rates based on the stage of disease. It was found that the 3-year survival rate of those with stage I to II disease was 20% to 34% if age was less than 60 years, 14% to 25% if age was between 60 and 69 years, and 9% to 13% if age was greater than or equal to 70 years, whereas the survival rates for those with stage III to IV disease were 2% to 5%, 1% to 2%, and less than 1%, respectively. Alternatively, 3-year survival after resection for stage I to II disease was 23% to 39% if age was less than 60 years, 16% to 31% if age was between 60 and 69 years, and 17% to 30% if age was greater than or equal to 70 years, whereas the survival rates for those with stage III to IV disease were 23% to 39%, 16% to 31%, and 17% to 30%, respectively. 11 There exists gender and racial differences in pancreatic cancer rates. The age-standardized incidence of pancreatic cancer is 5.7 per 100,000 in men and 4.1 per 100,000 in women. 7 There has been a recent increase in incidence rates of pancreatic cancer in women and men, which is likely due to an increase in life expectancy and increase prevalence of obesity. 12 In the United States, since 2000, incidence rates from pancreatic cancer have been increasing at a rate of 0.9% annually in men and 0.8% in women. 4 Since 2000, the death rate from pancreatic cancer has been increasing at a rate of 0.3% annually in men and 0.2% in women. 4 Cancer occurrence, diagnosis, and outcome vary substantially between racial and ethnic groups. This variation is likely attributable to socioeconomic barriers, access to care, exposures to risk factors, barriers to cancer prevention and early detection, and treatment. In the United States, age-adjusted incidence rates were higher in blacks at 15.8 per 100,000 compared with that in whites at 13.7 per 100,000, in Hispanics at 12.4 per 100,000, and in Asians at 10.5 per 100,000 in 2018. 4 Although understudied, the increased risk in blacks has been suggested to be primarily due to differences in the prevalence of established risk factors such as smoking, high body mass index (BMI), and diabetes, 13 as well as the decreased likeliness of being evaluated by a health professional and be enrolled in clinical trials in pancreatic cancer. 14 , 15 Nevertheless, although survival rates of most cancers are usually lower for blacks compared with whites, that of pancreatic cancer is the same for the 2 groups. 6 Data on pancreatic cancer among Hispanics are limited. However, based on the SEER database, there has been an increase in pancreatic cancer among Hispanics since 1974. 16 The few published data have conflicting results, with 1 showing worse outcomes 17 and others showing better outcomes among Hispanics compared with non-Hispanics. 18 Studies on pancreatic cancer in Asians have shown heterogeneity within this group. In a study evaluating the incidence of pancreatic cancer in the California Cancer Registry, the investigators found that Japanese and Koreans have pancreatic cancer risk as high as non-Hispanic whites, whereas Chinese, Filipinos, and South Asians have a similar low risk as non-Hispanic American Indians/Alaska Natives. 19 In addition, a recent study found higher morbidity after resection in Asians and individuals reporting Hispanic ethnicity. 20 Cigarette smoking is a well-established risk factor for pancreatic cancer. 21 – 25 A meta-analysis assessing the impact of smoking on the risk of pancreatic cancer reported a 48% relative increase in ever-smokers compared with never-smokers and an excess risk of 82% in current smokers compared with 17% in former smokers. Risk decreased for increasing time from cessation, reaching the level of nonsmokers after 20 years. 26 The elevated risk of mortality in smokers diagnosed with pancreatic cancer is independent of alcohol use, BMI, and history of diabetes. 27 Smoking has also been associated with a younger onset of pancreatic cancer. 28 Although there has been a decrease in the prevalence of cigarette smoking in much of Europe and North America, 29 the prevalence rates in Asia 30 and other parts of the world remain high. The association between diabetes and pancreatic cancer is intricate. Diabetes is both a risk factor and a consequence of pancreatic cancer. In a large cohort study of young and middle-aged subjects who attended routine health screening programs in Korea, individuals with diabetes had a 2-fold risk increase for pancreatic cancer mortality. 31 Studies evaluating the Mayo Clinic population initially indicated that up to 1% of newly diagnosed diabetic patients develop pancreatic cancer within 3 years of their diabetes diagnosis. 32 However, studies conducted within the VA system have shown a lower risk of less than 0.3% within 3 years of diagnosis. 33 A recent study found that those with new-onset diabetes, defined as less than or equal to 4 years, had an age-adjusted risk of 2.97 (95% CI, 2.31–3.82), whereas the risk of those with longstanding diabetes, defined as greater than 4 years, was 2.16 (95% CI, 1.78–2.60) compared with those without diabetes. 34 However, the time interval from diagnosis of new-onset diabetes to pancreatic cancer has been a topic of debate, with earlier studies showing stronger associations between diabetes and pancreatic cancer when the diabetes was less than or equal to 4 years 35 and a more recent study suggesting within the first year of diagnosis. 36 Obesity, defined by BMI greater than or equal to 30, has increased substantially worldwide between 1975 and 2016. 37 Obesity is associated with a cascade of other conditions, such as hyperglycemia, insulin resistance, and hypercholesterolemia. Elevated cholesterol has been shown to have a linear dose-response relation to the risk of pancreatic cancer, with 1 meta-analysis indicating that the risk of pancreatic cancer increased by 8% with 100 mg/d of cholesterol intake. 38 , 39 Several other studies have also shown an increased risk of pancreatic cancer with increased BMI. 40 – 42 Temporal onset of changes in BMI has also been studied in the risk of pancreatic cancer. Some studies have suggested that weight gain after age 50 years is associated with an increased risk of pancreatic cancer, 43 whereas others suggest in early life. 44 Obesity among children has grown substantially. A population-based study of 51,505 children found that the most rapid weight gain was between 2 and 6 years of age, and 90% of children who were obese at the age of 3 years were overweight or obese in adolescence. 45 A recent study of data from 20 pooled prospective cohorts suggests that excess body weight during early adulthood (ages 18–21 years) could be a more important influence on pancreatic cancer risk than weight gain later in life. 46 Alternatively, weight loss has also been implicated but likely as a result of an underlying pancreatic cancer rather than as a risk factor. In a large cohort study, Yuan and colleagues 34 found that compared with those with no weight loss, participants who reported a 0.45- to 1.8-kg weight loss had an age-adjusted hazard ratio (HR) for pancreatic cancer of 1.25 (95% CI, 1.03–1.52), those with a 2.25- to 3.6-kg weight loss had an age-adjusted HR of 1.33 (95% CI, 1.06–1.66), and those with more than an 3.6-kg weight loss had an age-adjusted HR of 1.92 (95% CI, 1.58–2.32). In addition, those with new-onset diabetes accompanied by weight loss of 0.45 to 3.6 kg had an HR 3.61 (95% CI, 2.15–6.10) and those with weight loss of more than 3.6 kg had an HR of 6.75 (95% CI 4.55–10.00). The risk was further increased in older individuals, those with healthy weight before weight loss, and those with unintentional weight loss. Co-occurrence of these symptoms should be recognized by clinicians because this group may benefit from early detection strategies. Alcohol has a variety of effects on the pancreas, and the effects are dose dependent. A pooled analysis of data from the PanC4 found a statistically significant increased risk of pancreatic cancer among heavy alcohol drinkers defined by 9 or more drinks/d compared with drinkers taking less than 1 drink/d (odds ratio [OR], 1.6; 95% CI, 1.2–2.2). 47 A meta-analysis of 19 prospective studies found that light (0–12 g/d) to moderate (≥12–24 g/d) alcohol intake had little or no effect on the risk of pancreatic cancer, whereas high intake (≥24 g/d) was associated with an increased risk (Relative Risk (RR), 1.15; 95% CI, 1.06–1.25). 48 Other studies have corroborated these findings. 49 , 50 Analysis of the NIH-AARP Diet and Health study and American Cancer Society Prevention Study II reported a relative risk of 1.32 to 1.45 in pancreatic cancer among those who drink more than 3 drinks/d. 51 , 52 Other studies have indicated that binge drinking increases the risk of pancreatic cancer by 2.5 times. 53 The underlying mechanism for the risk of pancreatic cancer with alcohol is likely its association with pancreatitis and the generation of toxic metabolites that lead to pancreatic carcinogenesis. 54 Pancreatitis is both a modifiable and nonmodifiable risk factor, depending on the cause. Smoking, alcohol, hypertriglyceridemia, and genetic mutations are common causes of pancreatitis. There are 2 types of pancreatitis. Acute pancreatitis is an acute inflammation of the pancreas triggered by one or more factors. Acute pancreatitis is a risk factor for pancreatic cancer and can also be a symptom of underlying pancreatic cancer. A meta-analysis of retrospective and prospective studies evaluating the association between acute pancreatitis and pancreatic cancer risk found an effect estimate of 2.07 (95% CI, 1.36–2.78) during a 10-year follow-up. Further subgroup analysis showed a relative risk of 7.81 (95% CI, 5.00–12.19) for pancreatic cancer in patients with acute pancreatitis compared with healthy controls. The strongest association was found within the first year after acute pancreatitis (effect estimate, 23.47 [95% CI, 3.26–43.68]) and diminished over time, such that after greater than 10 years following acute pancreatitis there was only a slightly increased OR of 1.17 (95% CI, 0.78–1.57). 55 Similar findings were observed in data from the Danish National Patient Registry, Civil Registration System, and Danish Cancer Registry. 56 The relative hazard of pancreatic cancer in individuals with acute pancreatitis was 19.3 (95% CI, 14.6–25.4) within the 2 years following the pancreatitis diagnosis but continued to be elevated (adjusted relative hazard, 2.02; 95% CI, 1.57–2.61) for patients followed beyond 5 years. The investigators also adjusted for alcohol- and smoking-related conditions and Charlson Comorbidity Index and excluded patients with chronic pancreatitis (CP). This study found that those with idiopathic cause of acute pancreatitis were associated with the highest risk of pancreatic cancer with an adjusted HR of 2.52 (95% CI, 1.83–3.47). 56 CP is defined as progressive inflammatory and fibrotic changes leading to irreversible structural damage, causing impairments in endocrine and exocrine functions. Patients with CP have a cumulative risk for pancreatic cancer of 1.8% (95% CI, 1–2.6) and 4% (95% CI, 2–5.9) 10 and 20 years after diagnosis, respectively. 57 A more recent meta-analysis found that although CP increases the risk of pancreatic cancer by 16-fold, the risk decreases over time. The risk after 5 years of diagnosis was 7.9-fold (95% CI, 4.26–14.66) and decreased to 3.5-fold (95%C 1.69–7.38) at 9 years suggesting closer follow-up in the first few years following CP diagnosis. 58 There are several suggested mechanisms of pancreatitis increasing the risk of pancreatic cancer. Obstruction of the pancreatic duct, such as by a tumor, is a risk factor for the development of acute pancreatitis. Alcohol consumption and smoking are risk factors for acute pancreatitis and CP as well as pancreatic cancer. Chronic inflammation increases the possibility of development of mutations within cells, most commonly KRAS mutations, which are found in pancreatic intraepithelial neoplasia, one of the most common precursor lesions to pancreatic cancer. 59 The immune system’s role in the development of pancreatic cancer is of increasing interest. Individuals with a personal history of allergies have been shown to be protective against pancreatic cancer. 59 Studies have found that those with asthma, nasal allergies, hay fevers, and other related symptoms have a lower risk of pancreatic cancer. 60 – 63 A recent exhaustive analysis found an inverse relationship between pancreatic cancer risk, severity, and duration. Furthermore, an increased asthma severity was associated with a reduced risk of pancreatic cancer. 60 The mechanism is unknown, but the underlying hypothesis is that individuals with active immune systems may have increased tumor immunity as a result of sustained elevated levels of IgE, 60 and the ability to survey the pancreas with the innate and adaptive antigen-specific immune system such as NK, NK-T, γδ, and αβ T cells. 64 , 65 Genetic factors and gene-environmental interactions have also been shown to play a role. Atopy-related genetic variants have been associated with reduced pancreatic cancer risk. 65 However, further studies are needed for us to understand the underlying mechanism. The microbiome has been extensively studied in its role in inflammation and carcinogenesis, with the speculation that the microbiota plays an essential role in activating, training, and modulating the host immune response. 66 Periodontal disease and oral microbiome have been shown to have increased the risk of pancreatic cancer. 67 , 68 Studies evaluating Helicobacter pylori in the risk of pancreatic cancer have been inconsistent. 69 – 71 Studies evaluating fecal microbiota have been limited. 69 Distinct intratumor microbiome has also been shown to promote pancreatic cancer through mutations of protumorigenic genes, inflammation, immune suppression, and treatment resistance. 72 Proteobacteria, Fusobacterium nucleatum , Bacteroidetes, and Firmicutes have shown to dominate the microbiome of pancreatic cancer. 72 – 76 Up to 10% of patients with pancreatic cancer report having a close relative with pancreatic cancer, 77 , 78 and heritability estimates for pancreatic cancer range from 21% for array-based heritability 79 to 36% in twin studies. 80 Both rare high-risk genetic variants and common variants have been associated with pancreatic cancer. Many of the high-risk genetic variants are associated with established hereditary cancer syndromes. These and the associated risk of pancreatic cancer are detailed in the following sections and Table 1 . Individuals with a family history of pancreatic cancer have an increased risk. Still, the risk is variable and depends on the number of family members affected and their relationship with affected family members. Individuals with a single close relative have approximately a 1.8 to 2.5 fold increases risk for pancreatic cancer. 81 – 83 Familial pancreatic cancer (FPC) is defined as kindred where there has been a diagnosis of pancreatic cancer in at least 2 close relatives (parent and child or sibling pair). At-risk relatives in FPC families who have at least 1 close relative (sibling, parent, or child with pancreatic cancer) have a 6.8-fold (95% CI, 4.5–9.8) increased risk of pancreatic cancer, 84 , 85 whereas those with 3 or more close relatives with pancreatic cancer have a 17-fold increased risk (95% CI, 7.3–33.5). 81 Approximately 15% to 20% of patients with a family history of pancreatic cancer will have a pathogenic genetic variant identified as the likely basis for their increased risk of pancreatic cancer when undergoing multigene clinical genetic testing; this is in contrast to 5% to 10% of all patients with pancreatic cancer (unselected for family history). Hereditary breast-ovarian cancer has been associated with an increased risk of pancreatic cancer. Several gene mutations associated with DNA repair have been implicated, including BRCA1, BRCA2, PALB2, and ATM . Individuals with pathogenic variants in the BRCA1 gene have up to a 3-fold increased risk of pancreatic cancer with a lifetime risk of up to of 3.6%. 86 Pathogenic variants in BRCA1 have been reported in 0.35% to 1% of patients with pancreatic cancer. 87 – 89 Pathogenic variants in the BRCA2 gene are more common in patients with pancreatic cancer, occurring in 2% to 7% of patients with pancreatic cancer unselected for family history and up to 16% of patients from families with 3 or more pancreatic cancers. The risk of pancreatic cancer in BRCA2 pathogenic variant carriers has been estimated to be 4- to 6-fold higher than in noncarriers. 90 – 93 The PALB2 protein, like its binding partner BRCA2 , acts to repair double-stranded DNA breaks, and pathogenic variants in the PALB2 gene are associated with an increased risk of pancreatic cancer. Pathogenic variants in PALB2 are uncommon, with prevalence estimates less than 1% in patients with pancreatic cancer, but large population studies are showing risk estimates ranging from 2.5 to 3.5. 94 , 95 Pathogenic variants in the ATM gene, named due to the syndrome that occurs among individuals with homozygous germline pathogenic variants in the gene, are present in 1% to 3% of patients with pancreatic cancer. 87 , 96 – 98 The risk of pancreatic cancer in ATM pathogenic variant carriers has been estimated to be 9.5% (95% CI, 5.0%–14.0%) by age 80 years, with an overall relative risk of 6.5 (95% CI, 4.5–9.5) compared with noncarriers. 99 Inherited pathogenic variants in the CDKN2A gene, commonly associated with familial atypical multiple mole melanoma syndrome, are often associated with a high risk of melanoma, respiratory tract tumors, soft tissue tumors, basal and squamous cell carcinomas, and pancreatic cancer. 100 Pancreatic cancer is the second most common cancer after melanoma in these patients, with a 12- to 38-fold increased risk and a lifetime risk of 17%. 87 , 101 Lynch syndrome is an autosomal dominant syndrome most commonly associated with early-onset colorectal cancer and endometrial cancer due to mutations in DNA mismatch repair genes (MSH2, MLH1, MSH6, PMS2, EPCAM). Individuals with Lynch syndrome have a 9- to 11-fold increased risk of pancreatic cancer 102 , 103 with a cumulative risk of 3.7% by the age of 70 years. 103 Peutz-Jeghers syndrome (PJS) is an autosomal dominant hamartomatous polyposis syndrome involving the STK11/LKB1 gene. Pathogenic variants in the STK11 gene are very rare and observed in greater than 0.5% of patients with pancreatic cancer. However, individuals with PJS have an extremely high risk of pancreatic cancer of 36% by age 64 years. 104 As discussed earlier, genetic variants may also contribute to the development of acute pancreatitis and CP, thereby increasing the risk of pancreatic cancer. The risk depends on the genes involved. Hereditary pancreatitis is associated with recurrent acute pancreatitis and CP at a young age, in addition to endocrine and exocrine pancreatic insufficiency. The risk of pancreatic cancer has been estimated to be as high as 30% to 40% by age 70 years in some forms of hereditary pancreatitis, such as in individuals with pathogenic variants in the PRSS1 gene, which result in premature trypsin activation and autosomal dominant pancreatitis. 105 – 107 In addition to pathogenic variants in PRSS1, variation in other genes has also been suggested to be involved in pancreatic cancer risk in the setting of pancreatitis. However, the risk of pancreatitis and pancreatitis cancer is less strong in these families; this includes variation in the pancreatic secretory enzymes CPA1 and CPB1 , particularly variants associated with endoplasmic reticulum stress phenotype, which are more common in patients with pancreatic cancer compared with control populations. 108 , 109 Common genetic variation in the CTRB2 locus has been associated with both pancreatitis and pancreatic cancer. 110 , 111 In addition, genetic variation in the CFTR gene, which encodes the cystic fibrosis transmembrane conductance regulator protein, can cause acute pancreatitis and CP without causing pulmonary issues and has been associated with early-onset CP. However, some studies have also shown an increased risk of pancreatic cancer, but this finding is not consistent. 112 In addition to the high-risk variants associated with pancreatic cancer, as detailed earlier, all of which more than double the risk of pancreatic cancer yet rarely occur in healthy individuals, common genetic variants (that occur in >1%–5% of healthy individuals of all ages) have also been associated and occur more frequently among individuals with pancreatic cancer. The first large-scale genome-wide association study (GWAS) for pancreatic cancer identified that individuals with genetic variants causing non-O versus O blood group were associated with an increased risk of pancreatic cancer. Subsequent studies examining the blood antigens themselves 113 found that those with blood groups A, AB, or B were more likely to develop pancreatic cancer with an adjusted hazards ratio for incident pancreatic cancer being 1.32 (95% CI, 1.02–1.72) for blood group A, 1.51 (95% CI, 1.02–2.23) for blood group AB, and 1.72 (95% CI, 1.25–2.38) for blood group B and that smoking may further increase these risks. 114 , 115 A GWAS identified a variant at the ABO locus 9q34 to pancreatic carcinogenesis. This initial GWAS study was followed by larger, better-powered studies both in European ancestry and in East Asian populations, including a recent, large-scale meta-analysis of 421 patients with European ancestry with pancreatic cancer and 426 controls. 113 To date, GWAS studies have identified primary associations at 1q32.1, 1p36.33, 2p13.3, 3q29, 5p15.33, 7p14.1, 8q21.11 8q24.21, 9q34.2, 13q12.2, 13q22.1, 16q23.1, 17q12, 17q25.1,18q21.32, and 22q12.1 in European ancestry individuals; 6p25.3, 12p11.21, 7q36.2, 13q12.2, 13q221, and 16p12.3 in Japanese populations; and 21q21.3, 5p13.1, 21q22.3, 22q13.32, and 10q26.11 in Chinese populations. 110 , 116 , 117 Subsequent studies of these GWAS data including transcriptome-wide association studies, 118 pathway-based approaches, 119 as well as mendelian randomization studies have provided further insight into the genetic basis of pancreatic cancer and the relationship between the genes underlying the connection between pancreatic cancer and some of the risk factors discussed previously.