Abstract
Background
A family history of pancreatic cancer is associated with increased pancreatic cancer risk. However, risk estimates for individuals in kindreds with an aggregation of pancreatic cancer (>1 relative) are imprecise because of small samples sizes or potentially impacted by biases inherent in retrospective data.
Objective
The objective of this study is to determine the age-specific pancreatic cancer risk as a function of family history using prospective data.
Methods
We compared pancreatic cancer incidence (n = 167) in 21 141 individuals from 4433 families enrolled in the National Familial Pancreatic Cancer Registry with that expected based on Surveillance Epidemiology and End Results data and estimated the cumulative probability of pancreatic cancer using competing risk regression.
Results
Familial pancreatic kindred members (kindreds with pancreatic cancer in 2 first-degree relatives [FDRs] or a pathogenic variant) had a standardized incidence ratio of 4.86 (95% confidence interval [CI] = 4.01 to 5.90), and sporadic kindred members (kindreds not meeting familial criteria) had a standardized incidence ratio of 2.55 (95% CI = 1.95 to 3.34). Risk in familial pancreatic cancer kindreds increased with an increasing number of FDRs with pancreatic cancer, with a standardized incidence ratio of 3.46 (95% CI = 2.52 to 4.76), 5.44 (95% CI = 4.07 to 7.26), and 10.78 (95% CI = 6.87 to 16.89) for 1, 2, and 3 or more FDRs with pancreatic cancer, respectively. Risk was also higher among individuals with a family history of young-onset (aged younger than 50 years) pancreatic cancer.
Conclusion
Pancreatic cancer risk is strongly dependent on family history, including both the degree of relationship(s) and age of onset of pancreatic cancer in relatives. These risk estimates will help inform the design of early detection studies and the risk and benefit analysis of screening trials.
In 2022, approximately 62 210 individuals will be diagnosed with pancreatic cancer, and 49 830 individuals will die of this cancer (1). Inherited genetics play an important role, with an estimated 21% to 36% of all pancreatic cancers due to inherited factors (2,3). Both pathogenic variants in established pancreatic cancer susceptibility genes (BRCA2, BRCA1, PALB2, ATM, CDKN2A, STK11, PRSS1, and mismatch repair genes) (4-10) as well as common variants identified thorough genome-wide association studies (11) have been shown to play a role. However, the majority of the heritability of pancreatic cancer remains unexplained by these established variants, with only approximately 4%-5% of the phenotypic variation of pancreatic cancer explained by known loci (2,11). Among patients with a family history of pancreatic cancer (2 first-degree relatives [FDRs]) more than 80% have noninformative findings after undergoing germline genetic testing with a hereditary cancer panel (4). In addition to a family history of pancreatic cancer, cigarette smoking, diabetes mellitus, and obesity have been associated with an increased risk of developing pancreatic cancer (12). Current cigarette smokers have approximately a twofold increased risk of pancreatic cancer, and cessation of smoking reduces the risk of pancreatic cancer compared with current smokers, with risk returning to that of never smokers within 15 to 20 years (13,14).
Approximately, 5%-10% of pancreatic cancer patients report a close relative with pancreatic cancer (15–17). Previous studies have reported the risk of pancreatic cancer among FDRs of a patient with pancreatic cancer ranges from 1.5 to 13 (12,18–20), with risk increasing as the number of affected relatives increases (21). However, many of these studies were based on retrospective data and thereby subject to recall bias or ascertainment bias. Although risk estimates from case-control studies are in general higher than that observed in cohort studies, there was limited evidence of recall bias when family history of cancer was examined directly in several case-control studies (22). Large-scale registry studies can present an unbiased estimate of pancreatic cancer risk, but these studies have a limited number of families with multiple cases of pancreatic cancer resulting in a limited ability to quantify the impact of both family history and age of onset of pancreatic cancer in the relatives (23). Furthermore, familial aggregation studies using retrospective data can impacted by ascertainment bias resulting in an overestimation of risk (24). Prior prospective studies conducted within National Familial Pancreas Tumor Registry (NFPTR) population found an elevated risk of pancreatic cancer in members of familial pancreatic cancer kindreds (21). However, these analyses were based on a limited number of incident pancreatic cancers (n = 41). With 10 additional years of follow-up and a much larger registry population, we can now provide more accurate estimates of pancreatic cancer risk associated with family history particularly in kindreds with multiple pancreatic cancers, including age at diagnosis and the role of smoking history.
Recent studies have demonstrated that screening of individuals at high risk of pancreatic cancer based on their family history, using endoscopic ultrasound or magnetic resonance imaging, can detect pancreatic cancers at an earlier stage than those presented with symptomatic disease (25,26). Although there is some consensus on early detection screening practices, there is still considerable debate about who should be screened, particularly for those with more modest family histories, and at what age to begin screening (13). A more precise estimate of risk due to a family history can help inform and tailor pancreatic cancer interception strategies.
The aim of this study is to estimate the risk of pancreatic cancer across family history risk strata and to further understand and quantify the risk of pancreatic cancer by age at diagnosis of pancreatic cancer in relatives and personal smoking history.
Methods
Study Population
Study procedures were reviewed and approved by the institutional review board of the Johns Hopkins medical institutions, and informed consent was obtained from all study participants. The study was performed in accordance with the Declaration of Helsinki.
The NFPTR is the source of our study population. Recruitment and data collection methods have been described in detail elsewhere (21,27). Briefly, individuals, or their relatives, with pancreatic cancer are recruited via 1) self-referral through study website, 2) physician referral, or 3) patients undergoing care for pancreatic cancer at the Johns Hopkins Hospital. Patient and family history questionnaires are collected at enrollment. Pancreatic cancers are verified via pathology review or medical records, when possible. Cancers, found on pathological review not be ductal adenocarcinoma or one of its variants, were excluded. When available, germline genetic testing reports are obtained and reviewed. At least 1 family member is contacted annually to obtain updated health history and vital status information for all members of the family. The study population included individuals with at least 1 FDR with pancreatic cancer who was alive at enrollment, aged younger than 91 years, free of pancreatic cancer, and with date of birth information. Partners of probands who were alive at baseline, free of pancreatic cancer, with date of birth known were included in the analyses as a genetically unrelated comparison cohort. To minimize bias because of underlying symptoms of pancreatic cancer impacting ascertainment, follow-up time began 90 days after the family’s enrollment date. Follow-up time was calculated from enrollment plus 90 days until either the individual’s date of pancreatic cancer diagnosis, date of death, 91st birthday, last date of contact from the family, or censoring date (June 30, 2019, date at which the 2019 survey began).
Smoking history was captured at enrollment and categorized as current, former, never (<100 cigarettes in lifetime), or unknown. Youngest age of onset of pancreatic cancer was estimated by kindred. Familial pancreatic cancer kindreds are defined as kindreds with at least a pair for FDRs with pancreatic cancer, or kindreds with a pathogenic germline variant in BRCA2, BRCA1, ATM, CDKN2A, PALB2, TP53, STK11, or in one of the mismatch repair genes. The number of FDRs with pancreatic cancer was counted per individual. Family history was measured at the kindred level and then within a kindred, based on an individual’s number of FDRs (parents, siblings, children) with pancreatic cancer. The number of FDRs with pancreatic cancer and familial status were time-dependent covariates. Sporadic kindreds were all remaining families that did not meet familial criteria.
Statistical Analysis
Standardized incidence ratios (SIR) were calculated by comparing the number of incident pancreatic cancers observed with the expected number based on Surveillance Epidemiology and End Results (SEER) data. SEER incidence rates for pancreatic ductal adenocarcinoma from 1994 to 2018 were obtained using SEER*Stat software (International Classification of Diseases–0-3/World Health Organization 2008 cancer site codes C25.0-C25.3, C25.5-C25.9, excluding C25.4 and International Classification of Diseases–0-3 morphology codes 8470, 8452, 8441) (28). Expected numbers of cancers were calculated by multiplying age-, race-, sex-, and calendar year–specific SEER incidence rates by total person-time in each stratum. Standardized incidence ratios compare the relative incidence in our cohort with that of the general US population but do not provide a direct estimate of cumulative incidence. Therefore, we estimated cause-specific cumulative incidence functions using semiparametric Fine-Gray competing risk models with pancreatic cancer as the outcome, death from other causes as the competing event, and age as the timescale and thus left truncation order to allow individuals to enter the risk set at their age at time of enrollment (29). The subdistribution hazards and hazard ratios were used to combine estimates together into the cumulative incidence functions. Predictors with a P value less than .10 were retained in the final model. Candidate predictors included family history, smoking, age of onset, and sex. Robust variance estimates were used to allow for correlation among family members. Analysis was conducted using SAS 9.4 (SIR), STATA 16.1 (time-to-event) and R 4.0.3 (30–32).
Results
Of the 21 141 individuals from 4432 families enrolled in the NFPTR and included in this study, 7942 individuals from 1673 families entered the risk set as familial kindreds, and 10 673 individuals from 2632 families entered the risk set as sporadic kindred members. The comparison cohort was composed of 2526 individuals. Overall, 91% of the families reported White ethnicity, and 9% non-White. The average follow-up was 7.6 years, and the median age at enrollment was 53 years. The majority of individuals were never smokers (n = 11 349, 53.7%), about one-third were former smokers (n = 7155, 33.8%), and only 1.9% (n = 406) were current smokers. Smoking history was unknown or not reported for 10.6%. Familial pancreatic cancer kindred members had slightly higher rates of current smoking (n = 187, 2.3%) than did sporadic kindred members (n = 191, 1.8%) and comparison cohort members (n = 28, 1.1%).
From 1994 to 2019, 167 incident cases of pancreatic cancers developed in 21 141 individuals. Of these incident cases, 103 (61.7%) were diagnosed in familial kindred members, 53 were sporadic kindred members, and 11 in the comparison cohort members. The average age of diagnosis of pancreatic cancer was 69.6 (SD = 11) years, which was consistent across all groups (Table 1). We observed higher incidence rates for pancreatic cancer in the 21 141 NFPTR participants compared with the general population (SEER), after controlling for age, sex, race and ethnicity, and calendar year (Table 2). Risk in familial and sporadic kindred members was statisitically significantly greater than the risk in the general population (familial SIR = 4.86, 95% confidence interval [CI] = 4.01 to 5.90; sporadic SIR = 2.55, 95% CI = 1.95 to 3.34). Risk of pancreatic cancer in the comparison cohort was comparable to the SEER population (SIR = 1.13, 95% CI = 0.63 to 2.04; P = .65).
Table 1.
Demographics of study populationa
| Participants group | Family members at risk |
Incident pancreatic cancers among family members |
||
|---|---|---|---|---|
| No. of individuals | Mean age at baseline (SD), y | No. of individuals | Mean age at diagnosis of pancreatic cancer (SD), y | |
| FPC kindred members | ||||
| Women | 4233 | 53.16 (17.20) | 64 | 70.14 (12.32) |
| Men | 3744 | 51.83 (16.99) | 39 | 67.56 (9.87) |
| SPC kindred members | ||||
| Women | 5784 | 52.26 (17.56) | 27 | 69.63 (10.38) |
| Men | 4854 | 50.74 (17.27) | 26 | 70.81 (11.18) |
| Genetically unrelated members | ||||
| Women | 1489 | 63.53 (11.21) | 3 | 59.67 (6.03) |
| Men | 1037 | 66.73 (11.28) | 8 | 75.13 (3.68) |
| Total | 21 146 | 53.51 (17.22) | 167 | 69.61 (11.01) |
FPC = familial pancreatic cancer; SPC = sporadic pancreatic cancer.
Table 2.
Standardized incidence ratios (SIRs) of pancreatic cancer among family members at risk: stratified by family history adjusted for race, sex, and calendar yeara
| Family history | No. of individuals | Person-years of follow-up | Observed cases | Expected cases | SIR (95% CI) |
|---|---|---|---|---|---|
| Familial | |||||
| Overall | 8225 | 71 070.29 | 103 | 21.20 | 4.86 (4.01 to 5.90) |
| Number of first-degree relatives with pancreatic cancer | |||||
| 1 first-degree relative | 5756 | 50 573.97 | 38 | 10.97 | 3.46 (2.52 to 4.76) |
| 2 first-degree relatives | 2276 | 17 685.93 | 46 | 8.46 | 5.44 (4.07 to 7.26) |
| ≥3 first-degree relatives | 407 | 2810.39 | 19 | 1.76 | 10.78 (6.87 to 16.89) |
| Kindred youngest age of onset | |||||
| Classic onset (50 years and older) | 6610 | 56 361.09 | 76 | 17.29 | 4.40 (3.51 to 5.50) |
| Early onset (younger than 50 years) | 1621 | 14 709.20 | 27 | 3.91 | 6.91 (4.74 to 10.08) |
| Smoking status | |||||
| Never smoker | 4531 | 39 343.54 | 55 | 10.35 | 5.32 (4.08 to 6.92) |
| Current smoker | 194 | 2711.99 | 3 | 0.66 | 4.55 (1.47 to 14.10) |
| Former smoker | 2669 | 23 111.15 | 32 | 8.01 | 3.99 (2.83 to 5.65) |
| Unknown smoking history | 831 | 5903.61 | 13 | 2.18 | 5.97 (3.47 to 10.28) |
| Sporadic | |||||
| Overall | 10 638 | 72 272.24 | 53 | 20.77 | 2.55 (1.95 to 3.34) |
| Family history | |||||
| 1 PC in family | 8515 | 56 008.75 | 38 | 15.74 | 2.42 (1.76 to 3.32) |
| ≥2 PC in family | 2123 | 16 263.49 | 15 | 5.04 | 2.98 (1.80 to 4.94) |
| Parent with PC | |||||
| No PC in parents | 4584 | 27 527.57 | 35 | 14.30 | 2.45 (1.76 to 3.41) |
| 1 parent with PC | 4889 | 34 895.32 | 15 | 4.82 | 3.12 (1.88 to 5.17) |
| 2 parents with PC | 213 | 1983.70 | 1 | 0.49 | 2.04 (0.29 to 14.58) |
| Kindred youngest age of onset | |||||
| Classic onset (50 years and older) | 9287 | 63 517.16 | 46 | 18.62 | 2.47 (1.85 to 3.30) |
| Early onset (younger than 50 years) | 1351 | 8755.08 | 7 | 2.15 | 3.25 (1.55 to 6.82) |
| Smoking status | |||||
| Never smoker | 5788 | 40 676.87 | 22 | 9.78 | 2.25 (1.48 to 3.42) |
| Current smoker | 191 | 1900.04 | 3 | 0.35 | 8.61 (2.78 to 26.69) |
| Former smoker | 3509 | 23 242.87 | 24 | 8.29 | 2.90 (1.94 to 4.32) |
| Unknown smoking history | 1150 | 6452.45 | 4 | 2.36 | 1.69 (0.64 to 4.51) |
| Unrelated | |||||
| Overall | 2526 | 17 798.54 | 11 | 9.71 | 1.13 (0.63 to 2.04) |
Stratification variables differ between sporadic and familial to capture variation in family history within these 2 categories. CI = confidence interval; PC = pancreatic cancer.
In the familial pancreatic cancer kindreds, an individual’s number of FDRs with pancreatic cancer was directly related to their magnitude of risk, with a risk of 3.46 (95% CI = 2.52 to 4.76) in individuals with 1 FDR, a risk of 5.44 (95% CI = 4.07 to 7.26) in individuals with 2 FDR, and a risk of 10.78 (95% CI = 6.87 to 16.89) in individuals with 3 or more FDRs. In addition to family history, having a family member develop pancreatic cancer at a young age (aged younger than 50 years) was also associated with an increased risk of pancreatic cancer in familial kindreds (SIR = 6.91, 95% CI = 4.74 to 10.08) compared with those without a young-onset relative (SIR = 4.40, 95% CI = 3.51 to 5.50) (Table 2). In the familial cohort, current smoking (SIR = 4.55, 95% CI = 1.47 to 14.10; P = .03) was not associated with a risk of pancreatic cancer beyond that of having a strong family history as risk in never smokers (SIR = 5.32, 95% CI = 4.08 to 6.92; P < .001) (Table 2).
In the sporadic pancreatic cancer kindreds, the risk of pancreatic cancer among individuals where there was more than 1 pancreatic cancer in more distant relatives (SIR = 2.98, 95% CI = 1.80 to 4.94) was only moderately higher than the risk observed in individuals with a family history of a single pancreatic cancer (SIR = 2.42, 95% CI = 1.76 to 3.32). Similar to the familial pancreatic cancer kindred members, pancreatic cancer risk in in the sporadic pancreatic cancer kindred members was higher for individuals with a family history of early onset pancreatic cancer (aged younger than 50 years), with a standardized incidence ratio of 3.25 (95% CI = 1.55 to 6.82) compared with kindreds where the pancreatic cancer occurred in those aged 50 years and older (SIR = 2.47, 95% CI = 1.85 to 3.30). When examining the impact of current smoking, a larger effect (SIR = 8.61, 95% CI = 2.78 to 26.69) was observed in the sporadic pancreatic cancer kindred members who currently smoked vs those who never smoked (SIR = 2.25, 95% CI = 1.48 to 3.42). However, our ability to accurately estimate the impact of smoking was limited given the overall low smoking prevalence (1.9%) and the number of pancreatic cancers (n = 6, 3 familial and 3 sporadic) that developed in smokers.
In the sporadic pancreatic cancer kindred members, we also explored the risk of pancreatic cancer based on having a single parent with pancreatic cancer compared with individuals who had 2 parents diagnosed with pancreatic cancer. To examine this effect without the confounding impact of additional family history of pancreatic cancer, we examined risk comparing individuals where family history was limited to 1 parent with pancreatic cancer with individuals where family history was limited to 2 parents with pancreatic cancer (ie, where there was no additional history of pancreatic cancer beyond that of the parents). One pancreatic cancer developed in the 213 individuals where both parents had been diagnosed with pancreatic cancer (SIR = 2.04, 95% CI = 0.29 to 14.58) compared with 15 pancreatic cancers in 4889 individuals where only 1 parent had a pancreatic cancer diagnosis (SIR = 3.12, 95% CI = 1.88 to 5.17).
In addition to comparing risk to SEER, which provides the relative increase in pancreatic cancer in our cohort compared with SEER, we also wanted to estimate the age-specific cumulative risk of pancreatic cancer to better quantify risk. Therefore, we also used competing risk regression methods (Table3). We retained predictors with a P value less than .10. For the familial kindreds, risk increased with the number of FDR with pancreatic cancer. In addition, having a family member develop pancreatic cancer at a young age (aged younger than 50 years) was also independently associated with increased risk (subdistribution ratio [SHR] = 1.43, 95% CI = 0.98 to 2.09; P = .06) compared with individuals from families where all pancreatic cancers developed in those aged older than 50 years (Table 3 and Table 4). Thereby, the cumulative risk of pancreatic cancer by age 85 years for familial kindred members in kindreds with a young-onset case (aged younger than 50 years) was 7.02% (95% CI = 3.33% to 10.72%), 11.97% (95% CI = 6.04% to 17.89%), and 24.06% (95% CI = 10.76% to 37.36%) for familial kindred members with 1, 2, and 3 or more FDR with pancreatic cancer, respectively, and 5.74% (95% CI = 3.62% to 7.86%) for sporadic kindred members. In contrast, cumulative risk by age 85 years was 4.96% (95% CI = 2.95% to 6.98%), 8.53% (95% CI = 5.23% to 11.82%), and 17.51% (95% CI = 8.97% to 26.04%) in familial kindreds and 4.05% (95% CI = 1.92% to 6.18%) in sporadic kindreds where all pancreatic cancers occurred after the age of 50 years (Table 4 and Figure 1).
Table 3.
Competing risk regression estimatesa
| Predictor | Death due to causes other than pancreatic cancer SHR (95% CI) | Pancreatic cancer SHR (95% CI) |
|---|---|---|
| Age at pancreatic cancer diagnosis in family | ||
| Classic onset (50 years and older) (referent) | — | — |
| Early-onset (younger than 50 years) | — | 1.43 (0.98 to 2.09) |
| No. of FDRs with pancreatic cancer | ||
| Sporadic (referent) | — | — |
| FPC, 1 FDR | 0.99 (0.83 to 1.19) | 1.23 (0.81 to 1.87) |
| FPC, 2 FDRs | 1.21 (1.00 to 1.47) | 2.16 (1.43 to 3.23) |
| FPC, 3 or more FDRs | 1.38 (0.99 to 1.91) | 4.66 (2.72 to 7.97) |
| Smoking history | ||
| Never/former/unknown (referent) | ||
| Current smoker | 1.62 (1.11 to 2.37) | — |
| Sex | ||
| Female (referent) | ||
| Male | 1.64 (1.42 to 1.91) | — |
“—” denotes variables not included in the estimation the specific outcome (ie, pancreatic cancer or death due to other causes.). CI = confidence interval; FDR = first-degree relative; FPC = familial pancreatic cancer; SHR = subdistribution hazard ratio.
Table 4.
Cumulative risk of pancreatic cancer by family historya
| Family history | Percent cumulative risk of pancreatic cancer (95% CI) |
|||
|---|---|---|---|---|
| By age 55 | By age 65 | By age 75 | By age 85 | |
| Familial | ||||
| Early onset (younger than 50 years) | ||||
| ≥3 FDRs | 2.83 (1.06 to 4.60) | 7.97 (3.11 to 12.83) | 16.90 (7.11 to 26.69) | 24.06 (10.76 to 37.36) |
| 2 FDRs | 1.32 (0.63 to 2.01) | 3.77 (1.82 to 5.73) | 8.22 (4.06 to 12.37) | 11.97 (6.04 to 17.89) |
| 1 FDR | 0.76 (0.34 to 1.17) | 2.17 (1.00 to 3.35) | 4.78 (2.23 to 7.33) | 7.02 (3.33 to 10.72) |
| Sporadic | 0.61 (0.38 to 0.85) | 1.77 (1.10 to 2.43) | 3.90 (2.44 to 5.35) | 5.74 (3.62 to 7.86) |
| Late onset (50 years and older) | ||||
| ≥3 FDRs | 1.99 (0.93 to 3.04) | 5.64 (2.70 to 8.59) | 12.14 (6.03 to 18.26) | 17.51 (8.97 to 26.04) |
| 2 FDRs | 0.93 (0.55 to 1.30) | 2.65 (1.59 to 3.71) | 5.82 (3.53 to 8.10) | 8.53 (5.23 to 11.82) |
| 1 FDR | 0.53 (0.31 to 0.75) | 1.53 (0.90 to 2.15) | 3.37 (1.99 to 4.74) | 4.96 (2.95 to 6.98) |
| Sporadic | 0.43 (0.20 to 0.66) | 1.23 (0.58 to 1.90) | 2.74 (1.29 to 4.20) | 4.05 (1.92 to 6.18) |
CI = confidence interval; FDR = first-degree relatives.
Figure 1.
Lifetime risk estimates. A) Predicted cumulative risk of pancreatic cancer by family history where at least 1 family member was diagnosed with pancreatic cancer aged younger than 50 years. B) Predicted cumulative risk of pancreatic cancer by family history where all family members with pancreatic cancer were diagnosed at age 50 years or older. FDR = first-degree relatives; FPC = familial pancreatic cancer; SPC = sporadic pancreatic cancer.
Discussion
Our results demonstrate the importance of family history of pancreatic cancer risk and provides age-specific cumulative probability of pancreatic cancer by family history to guide decision making. In familial kindreds where all pancreatic cancers developed after age 50 years, individuals with 3 or more FDR with pancreatic cancer had a lifetime risk by age 85 years of 18% (95% CI = 9% to 26%), individuals with 2 FDRs with pancreatic cancer, lifetime risk was 9% (95% CI = 5% to 12%), and individuals with 1 FDR with pancreatic cancer, lifetime risk was 5% (95% CI = 3% to 7%). Lifetime risk was even higher at 24% (95% CI = 11% to 37%), 12% (95% CI = 6% to 18%), and 7% (95% CI = 3% to 11%), respectively, if there was a young-onset (aged younger than 50 years) pancreatic cancer in the family.
A more precise understanding of the risk of pancreatic cancer among individuals with a family history of pancreatic cancer is critical to early detection screening efforts. The US Preventative Task Force does not support pancreatic cancer screening for the general population (grade of D, recommend against screening) (33) and states that screening tests with high sensitivity and specificity are needed. The report states the need for continued studies into the risks and benefits of pancreatic cancer screening in high-risk populations, specifically in those with a family history of pancreatic cancer and/or genetic predisposition. Recent studies continue to support the likely benefit of annual imaging-based surveillance in high-risk populations (25,34), and efforts to identify less invasive blood-based detection methods continue (35). As randomized trials of screening interventions are developed, an accurate quantification of pancreatic cancer risk at an individual level, as we present in this study, is critical to ensure samples size estimation and assessment of risk-benefit ratios. Our study supports tailoring the age at which screening starts to the youngest age of pancreatic cancer in the family, as risk was higher in families where a pancreatic cancer occurred in those aged younger than 50 years. Although the number of family members with pancreatic cancer provides the most risk discrimination (ie, highest SHR), the age at which the relatives develop pancreatic cancer is an important indicator of future risk. This work is an important step forward in developing screening guidelines in that we provide estimates of pancreatic cancer risk in high-risk populations. However, screening guidelines must weigh the benefits of a screening modality against the false-positive rate and consequences of false-positive findings (ie, additional testing or surgical resection). Therefore, we agree with the current recommendations that screening, even in high-risk groups, should be conducted as part of a clinical trial.
Clinical management of individuals with 2 parents with pancreatic cancer who lack additional family members with pancreatic cancer remains unclear. Our results do not support a substantively increased risk of pancreatic cancer among individuals with 2 parents with pancreatic cancer compared with 1 parent with pancreatic cancer, when there is not additional family history of pancreatic cancer. Recent retrospective studies have suggested a very high risk of pancreatic cancer to siblings of probands when both parents had pancreatic cancer (36), however, this study did not examine full pedigree data. In contrast, we limited our analysis to individuals without additional family history of pancreatic cancer beyond the parents, which may confound the impact of having affected parents alone. Furthermore, our analysis was based on prospective data. However, larger sample sizes are needed to more precisely estimate these risks.
It is well established that smoking cessation lowers the risk of pancreatic cancer and other types of cancer, and an individual’s risk of pancreatic cancer decreases to that of a never smoker after 20 years quitting (13). Smoking history was not observed to be a statistically significant predictor for pancreatic cancer in our analysis; this may in part be because of the low prevalence of current smoking (<2%). However, current smoking was associated with an increased risk of death because of causes other than pancreatic cancer (SHR = 1.62; P = .01) (Table 3).
The NFPTR is the largest single registry of pancreatic cancer families. This registry enables us to examine the risk of pancreatic cancer prospectively, independent of the ascertainment events motivating the families to enroll in the registry. Furthermore, the long duration of follow-up of the families in our registry (an average of more than 8 years) is a strength. Other strengths include the ability to estimate risk based on family history as well as other risk factors such as sex and smoking history. Our analysis did not take into account individuals who may have chosen to undergo asymptomatic screening for pancreatic cancer. However, although promising, early detection screening in the high-risk setting is not widely available. Furthermore, our endpoint was the development of pancreatic cancer. Screening has been shown to increase survival after cancer detection in high-risk patients because of detection of cancer at an earlier stage, however, it has not been shown to prevent the development of pancreatic cancer. Any reduction in the risk of pancreatic cancer because of screening would result in our risk estimates being overly conservative (underestimation vs overestimation of risk). Enrollment into the NFPTR is based on self or physician referral or direct recruitment of patients seeking care for their pancreatic cancer at Johns Hopkins, a tertiary referral center drawing patients from across the United States. The NFPTR population is likely representative of individuals seeking out pancreatic cancer screening and/or those who seek out cancer genetics services, however, further studies are needed to assess the generalizability of these findings to other populations. We confirmed 33% of reported pancreatic cancers, and data were not obtainable on the remaining 67%, but the positive predictive value for reported family history of pancreatic cancer has been shown to be high (>75%) (37). The availability of germline genetic testing data in this cohort was limited and highly selected as guidelines recommending universal genetic testing were in place after our study period. The cost, insurance coverage, and availability of testing varied widely. Furthermore, it has been shown that family report of genetic testing results is differential as relatives are less likely to share noninformative findings (38).
Our study demonstrates the high pancreatic cancer risk in members of familial pancreatic cancer kindreds and a moderate pancreatic cancer risk in sporadic kindred members. In familial and sporadic kindreds, having a young-onset pancreatic cancer in the family was associated with a higher risk of pancreatic cancer. We provide some evidence to support that risk among individuals with 2 parents with pancreatic cancer appears to be of a magnitude similar to that observed in other sporadic pancreatic cancer kindreds. This data can help guide risk management of individuals with a family history of pancreatic cancer.
Funding
This work was supported by National Cancer Institute P50 CA62924 (to APK), Lustgarten Foundation, Rolfe Pancreatic Cancer Foundation, Dennis Troper and Susan Wojcicki, and the Sol Goldman Pancreatic Cancer Research Center.
Notes
Role of the funder: The funder of this work was not involved with nor influenced the design, conduct, or interpretation of this study. The funder was also not involved in the writing of this manuscript or the decision to submit it for publication.
Disclosures: Drs Hruban and Lennon reported royalty distributions from a patent that is licensed under an agreement between Thrive Earlier Detection (a subsidiary of Exact Sciences Corporation) and Johns Hopkins University.
Author contributions: NP: Conceptualization, data curation, formal analysis, investigation, methodology, writing—original draft, writing—review & editing; DL data curation, investigation, resources, writing—review & editing; BL: data curation, methodology, supervision, resources, writing—review & editing; JH: data curation, investigation, resources, writing—review & editing; LZ: data curation, investigation, resources, writing—review & editing; AN: data curation, investigation, resources, writing—review & editing; NJR: data curation, investigation, resources, writing—review & editing; MIC: data curation, investigation, resources, writing—review & editing; AML: data curation, investigation, resources, writing—review & editing; MGG: Conceptualization, investigation, resources, writing—review & editing; RHH: Conceptualization, Funding Acquisition, investigation, resources, writing—review & editing; APK: Conceptualization, data curation, Formal Analysis, Funding Acquisition, investigation, methodology, project administration, resources, supervision, writing—original draft, writing—review & editing.
Acknowledgements: Deepest gratitude to the patients and families who have enrolled and actively participated in the National Familial Pancreas Tumor Registry (NFPTR). The research team of the NFPTR.
Contributor Information
Nancy Porter, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Daniel Laheru, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Bryan Lau, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA; Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA.
Jin He, Department of Surgery, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Lei Zheng, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Amol Narang, Division of Radiation Oncology, Department of Medicine, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Nicholas J Roberts, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA; The Sol Goldman Pancreatic Cancer Research Center, Department of Pathology, Johns Hopkins University, Baltimore, MD, USA.
Marcia I Canto, Division of Gastroenterology, Department of Medicine, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Anne Marie Lennon, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA; Department of Surgery, Johns Hopkins School of Medicine, Baltimore, MD, USA; Division of Radiation Oncology, Department of Medicine, Johns Hopkins School of Medicine, Baltimore, MD, USA; The Sol Goldman Pancreatic Cancer Research Center, Department of Pathology, Johns Hopkins University, Baltimore, MD, USA.
Michael G Goggins, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA; The Sol Goldman Pancreatic Cancer Research Center, Department of Pathology, Johns Hopkins University, Baltimore, MD, USA; Division of Gastroenterology, Department of Medicine, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Ralph H Hruban, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA; The Sol Goldman Pancreatic Cancer Research Center, Department of Pathology, Johns Hopkins University, Baltimore, MD, USA.
Alison P Klein, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD, USA; Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA; The Sol Goldman Pancreatic Cancer Research Center, Department of Pathology, Johns Hopkins University, Baltimore, MD, USA; Division of Gastroenterology, Department of Medicine, Johns Hopkins School of Medicine, Baltimore, MD, USA.
Data Availability
Individual participant data cannot be made publicly available as it contains dates and family structure. Requests for data sharing can be made to corresponding author Alison Klein and will be shared pending approval of the Johns Hopkins institutional review board and appropriate human subjects protections.
References
- 1. Siegel RL, Miller KD, Fuchs HE, Jemal A.. Cancer statistics, 2022. CA Cancer J Clin. 2022;72(1):7-33. [DOI] [PubMed] [Google Scholar]
- 2. Chen F, Childs EJ, Mocci E, et al. Analysis of heritability and genetic architecture of pancreatic cancer: a PanC4 study. Cancer Epidemiol Biomarkers Prev. 2019;28(7):1238-1245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Lichtenstein P, Holm NV, Verkasalo PK, et al. Environmental and heritable factors in the causation of cancer–analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med. 2000;343(2):78-85. [DOI] [PubMed] [Google Scholar]
- 4. Roberts NJ, Norris AL, Petersen GM, et al. Whole genome sequencing defines the genetic heterogeneity of familial pancreatic cancer. Cancer Discov. 2016;6(2):166-175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Roberts NJ, Jiao Y, Yu J, et al. ATM mutations in patients with hereditary pancreatic cancer. Cancer Discov. 2012;2(1):41-46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Goggins M, Schutte M, Lu J, et al. Germline BRCA2 gene mutations in patients with apparently sporadic pancreatic carcinomas. Cancer Res. 1996;56(23):5360-5364. [PubMed] [Google Scholar]
- 7. Thompson D, Easton DF.. Cancer Incidence in BRCA1 mutation carriers. J Natl Cancer Inst. 2002;94(18):1358-1365. [DOI] [PubMed] [Google Scholar]
- 8. Jones S, Hruban RH, Kamiyama M, et al. Exomic sequencing identifies PALB2 as a pancreatic cancer susceptibility gene. Science. 2009;324(5924):217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kastrinos F, Mukherjee B, Tayob N, et al. Risk of pancreatic cancer in families with Lynch syndrome. JAMA. 2009;302(16):1790-1795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Goldstein AM, Fraser MC, Struewing JP, et al. Increased risk of pancreatic cancer in melanoma-prone kindreds with p16INK4 mutations. N Engl J Med. 1995;333(15):970-974. [DOI] [PubMed] [Google Scholar]
- 11. Klein AP, Wolpin BM, Risch HA, et al. Genome-wide meta-analysis identifies five new susceptibility loci for pancreatic cancer. Nat Commun. 2018;9(1):556. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Klein AP. Pancreatic cancer epidemiology: understanding the role of lifestyle and inherited risk factors. Nat Rev Gastroenterol Hepatol. 2021;18(7):493-502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Bosetti C, Lucenteforte E, Silverman DT, et al. Cigarette smoking and pancreatic cancer: an analysis from the International Pancreatic Cancer Case-Control Consortium (Panc4). Ann Oncol. 2012;23(7):1880-1888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Lynch SM, Vrieling A, Lubin JH, et al. Cigarette smoking and pancreatic cancer: a pooled analysis from the pancreatic cancer cohort consortium. Am J Epidemiol. 2009;170(4):403-413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hu C, Hart SN, Polley EC, et al. Association between inherited germline mutations in cancer predisposition genes and risk of pancreatic cancer. JAMA. 2018;319(23):2401-2409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Shindo K, Yu J, Suenaga M, et al. Deleterious germline mutations in patients with apparently sporadic pancreatic adenocarcinoma. J Clin Oncol. 2017;35(30):3382-3390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Yurgelun MB, Chittenden AB, Morales-Oyarvide V, et al. Germline cancer susceptibility gene variants, somatic second hits, and survival outcomes in patients with resected pancreatic cancer. Genet Med. 2019;21(1):213-223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Ghadirian P, Boyle P, Simard A, Baillargeon J, Maisonneuve P, Perret C.. Reported family aggregation of pancreatic cancer within a population-based case-control study in the Francophone community in Montreal, Canada. Int J Pancreatol. 1991;10(3-4):183-196. [DOI] [PubMed] [Google Scholar]
- 19. Silverman DT. Risk factors for pancreatic cancer: a case-control study based on direct interviews. Teratog Carcinog Mutagen. 2001;21(1):7-25. [DOI] [PubMed] [Google Scholar]
- 20. Jacobs EJ, Chanock SJ, Fuchs CS, et al. Family history of cancer and risk of pancreatic cancer: a pooled analysis from the Pancreatic Cancer Cohort Consortium (PanScan). Int J Cancer. 2010;127(6):1421-1428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Brune KA, Lau B, Palmisano E, et al. Importance of age of onset in pancreatic cancer kindreds. J Natl Cancer Inst. 2010;102(2):119-126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Permuth-Wey J, Egan KM.. Family history is a significant risk factor for pancreatic cancer: results from a systematic review and meta-analysis. Fam Cancer. 2009;8(2):109-117. [DOI] [PubMed] [Google Scholar]
- 23. Amundadottir LT, Thorvaldsson S, Gudbjartsson DF, et al. Cancer as a complex phenotype: pattern of cancer distribution within and beyond the nuclear family. PLoS Med. 2004;1(3):e65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Petersen GM, de Andrade M, Goggins M, et al. Pancreatic cancer genetic epidemiology consortium. Cancer Epidemiol Biomarkers Prev. 2006;15(4):704-710. [DOI] [PubMed] [Google Scholar]
- 25. Goggins M, Overbeek KA, Brand R, et al. ; for the International Cancer of the Pancreas Screening (CAPS) consortium. Management of patients with increased risk for familial pancreatic cancer: updated recommendations from the International Cancer of the Pancreas Screening (CAPS) Consortium. Gut. 2020;69(1):7-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Canto MI, Kerdsirichairat T, Yeo CJ, et al. Surgical outcomes after pancreatic resection of screening-detected lesions in individuals at high risk for developing pancreatic cancer. J Gastrointest Surg. 2020;24(5):1101-1110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Klein AP, Brune KA, Petersen GM, et al. Prospective risk of pancreatic cancer in familial pancreatic cancer kindreds. Cancer Res. 2004;64(7):2634-2638. [DOI] [PubMed] [Google Scholar]
- 28. National Cancer Institute. Surveillance Research Program. SEERStat software (seer.cancer.gov/seerstat) version 8.3.9.
- 29. Fine JP, Gray RJ.. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94(446):496-509. [Google Scholar]
- 30. SAS Institute Inc. SAS® 9.4. [computer program].
- 31. R Core Team. R: A language and environment for statistical computing. 2020.
- 32. StataCorp. Stata Statistical Software: Release 16. 2019. 2019.
- 33. Owens DK, Davidson KW, Krist AH, et al. ; for the US Preventive Services Task Force. Screening for pancreatic cancer: US Preventive Services Task Force reaffirmation recommendation statement. JAMA. 2019;322(5):438-444. [DOI] [PubMed] [Google Scholar]
- 34. Vasen H, Ibrahim I, Ponce CG, et al. Benefit of surveillance for pancreatic cancer in high-risk individuals: outcome of long-term prospective follow-up studies from three European expert centers. J Clin Oncol. 2016;34(17):2010-2019. [DOI] [PubMed] [Google Scholar]
- 35. Cohen JD, Javed AA, Thoburn C, et al. Combined circulating tumor DNA and protein biomarker-based liquid biopsy for the earlier detection of pancreatic cancers. Proc Natl Acad Sci USA. 2017;114(38):10202-10207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Rabe KG, Stevens MA, Hernández AT, et al. Pancreatic cancer risk to siblings of probands in bilineal cancer settings. Genet Med. 2022;24(5):1008-1016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Fiederling J, Shams AZ, Haug U.. Validity of self-reported family history of cancer: a systematic literature review on selected cancers. Int J Cancer. 2016;139(7):1449-1460. [DOI] [PubMed] [Google Scholar]
- 38. Hughes C, Lerman C, Schwartz M, et al. All in the family: evaluation of the process and content of sisters’ communication about BRCA1 and BRCA2 genetic test results. Am J Med Genet. 2002;107(2):143-150. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Individual participant data cannot be made publicly available as it contains dates and family structure. Requests for data sharing can be made to corresponding author Alison Klein and will be shared pending approval of the Johns Hopkins institutional review board and appropriate human subjects protections.