Abstract
Background
Increased risk of malignancies other than pancreatic cancer (PC) has been reported among first-degree relatives (FDRs) of PC patients; however, the roles of susceptibility gene mutations are unclear. We assessed risk for 15 cancers among FDRs of unselected PC probands.
Methods
Data on 17 162 FDRs, with more than 336 000 person-years at risk, identified through 2305 sequential PC probands enrolled at Mayo Clinic (2000–2016) were analyzed. Family history data were provided by the probands. Standardized incidence ratios (SIRs) and 95% confidence intervals (CIs) were calculated, comparing malignancies observed among the FDRs with that expected using Surveillance, Epidemiology, and End Results (SEER) data. Genetic testing was performed among a subset of probands (n = 2094), enabling stratified analyses among FDRs based on whether the related proband tested positive or negative for inherited mutation in 22 sequenced cancer susceptibility genes. All statistical tests were two-sided.
Results
Compared with SEER, PC risk was twofold higher among FDRs of PC probands (SIR = 2.04, 95% CI = 1.78 to 2.31, P < .001). Primary liver cancer risk was elevated among female FDRs (SIR = 2.10, 95% CI = 1.34 to 3.12, P < .001). PC risk was more elevated among FDRs of mutation-positive probands (SIR = 4.32, 95% CI = 3.10 to 5.86) than FDRs of mutation-negative probands (SIR = 1.77, 95% CI = 1.51 to 2.05, between-group P < .001). FDR PC risk was higher when the related proband was younger than age 60 years at diagnosis and mutation-positive (SIR = 5.24, 95% CI = 2.93 to 8.64) than when the proband was younger than age 60 years but mutation-negative (SIR = 1.76, 95% CI = 1.21 to 2.47, between-group P < .001). Breast (SIR = 1.29, 95% CI = 1.01 to 1.63) and ovarian (SIR = 2.38, 95% CI = 1.30 to 4.00) cancers were elevated among FDRs of mutation-positive probands.
Conclusions
Our study substantiates twofold risk of PC among FDRs of PC patients and suggests increased risk for primary liver cancer among female FDRs. FDRs of susceptibility mutation carriers had substantially increased risk for PC and increased risk for breast and ovarian cancers.
Pancreatic cancer (PC) is a rapidly fatal malignancy with established risk factors that include tobacco smoking, long-standing diabetes, chronic pancreatitis, obesity, and positive family history of PC (1–3). In general, family history of cancer is important for accurate cancer risk stratification toward primary prevention, early detection, and timely intervention. Studies have shown that certain cancers tend to cluster in families due in part to inherited susceptibility and shared environmental exposures (4,5). Some genetic syndromes have a pleiotropic effect and are characterized by the presence of different cancer types among mutation carriers in the family, presumably because the underlying susceptibility variants target multiple tissues (4–6). For example, inherited mutations in CDKN2A (p16) can predispose to malignant melanoma or PC, and mutations in BRCA1, BRCA2, or PALB2 predispose to breast cancer, ovarian cancer, or PC, whereas mutations in TP53 can predispose to leukemia, breast cancer, brain cancer, or soft tissue sarcoma (4,6–8). Reliable estimates for the risk of different cancers in families of PC patients who have been tested for susceptibility genes are not available.
Individuals with PC often have multiple first-degree relatives (FDRs) affected with cancers other than PC (7,9–11), but most studies have focused primarily on the PC risk (reviewed in [12]). In 2005, we examined the occurrence of 14 common malignancies among 3335 FDRs of 426 incident PC cases diagnosed at Mayo Clinic and compared the observed number of cases of each cancer type with those expected based on data from the Surveillance, Epidemiology, and End Results (SEER) Program (11). The study, which included more than 130 000 person-years at risk, found higher than expected risks of PC and liver cancer and lower than expected risks of bladder, lung, lymphoma, prostate, and breast cancers among the FDRs, compared with the SEER reference population. In 2009, Wang et al. examined risk of death from extrapancreatic malignancies among 8564 FDRs of 1328 PC probands (13). Comparing deaths from 17 cancers among FDRs with more than 200 000 person-years of follow-up with the expected number of deaths for each cancer based on SEER data, they reported an increased risk of death from breast, liver, ovarian, and bile duct cancers among the FDRs (13). When stratified by familial vs sporadic kindreds, deaths from breast, liver, ovarian, and bile duct cancers remained increased in the familial kindreds, but only bile duct cancer deaths were elevated in the sporadic kindreds (13).
Broader insight into the pattern and scope of aggregation of cancers among relatives of PC probands can be gained with large sample sizes and evaluation of the role of inherited mutations. In the present study, we extend our previous work (11) using a larger sample size, including the data from our previous work, and additionally examined the influence of germline mutation status of PC probands on FDRs’ cancer risk. Specifically, we examined the risk of 15 common malignancies among these FDRs, investigating whether risk differed by FDRs’ kinship status, the FDRs’ smoking history, or the probands’ age at diagnosis. In a subset of probands who were tested for germline mutations in 22 cancer predisposition genes, we compared cancer risk among FDRs of mutation-positive probands with FDRs of mutation-negative probands.
Methods
Study Population
The prospective Mayo Clinic Biospecimen Resource for Pancreas Research is a registry of PC probands that utilizes an ultrarapid case ascertainment process for patient recruitment, with a participation rate of approximately 70% (14,15). The main reasons for nonparticipation in the registry are the severely debilitating nature of PC and its rapidly fatal course (14). Included in the present study were data reported on 17 162 FDRs (parents, siblings, and children) of 2305 unselected, consecutively enrolled PC probands in the registry between October 2000 and June 2016. For validation analyses described below, we used self-reported personal history of cancer provided by FDRs, who themselves enrolled in a separate familial PC study (16). Informed consent was obtained from each proband and each FDR in the familial PC study. The study was approved by the Mayo Clinic Institutional Review Board.
Data Collection
The PC probands completed structured risk factor questionnaires at enrollment that sought information on each FDR, including cancer diagnosis, age at diagnosis, dates of diagnosis, birth, and death (if applicable), and smoking history. We compared reported cases of 15 common malignancies among the FDRs with the expected number of cases for each cancer using data from the SEER 9 registries (1973–2013) (17), aligning with our previous study (11). The cancers of interest were primary bladder, brain, breast, colorectal, gastric, head and neck, leukemia, liver, lung, lymphoma, melanoma, multiple myeloma, ovary, pancreas, and prostate carcinomas. Analyses of brain, liver, and lung cancers were restricted to instances where each was the only cancer reported in an FDR (hereafter referred as primary liver, brain, or lung cancer). Because these 15 malignancies rarely occur in persons younger than age 20 years, we restricted the analyses to FDRs who were age 20 years or older. Detailed pedigrees were constructed based on family history information provided by the probands.
Germline mutations in 22 cancer predisposition genes were assessed by sequencing leukocyte DNA obtained from the PC probands. Genetic testing was performed among probands who provided blood samples (2094 of 2305 probands). Among the genes tested, 144 unique mutations determined to be deleterious (excluding variants of uncertain significance) were detected in 19 of the cancer predisposition genes: APC, ATM, BARD1, BRCA1, BRCA2, BRIP1, CDKN2A, CHEK2, FANCC, MLH1, MSH2, MSH6, NBN, PALB2, PMS2, PRSS1, RAD51C, RAD51D, and TP53. None of the probands carried a mutation in CDH1, PTEN, or EPCAM (18). The FDRs were classified as belonging to mutation-positive or mutation-negative pedigrees based on whether the related proband tested positive for a mutation in any of the tested genes (n = 1465; 198 pedigrees) or tested negative for all 22 genes (n = 14 168; 1896 pedigrees). Risk analyses were performed also among full siblings of probands who were mutation-positive (n = 636; 182 pedigrees) vs mutation-negative (n = 5663; 1728 pedigrees).
Statistical Analyses
Standardized incidence ratios (SIRs) and 95% confidence intervals (CIs) were used to estimate the risk of the 15 malignancies among the FDRs. The SIRs were calculated for each cancer as the ratio of the observed to the expected number of cases. The expected number of cases was calculated from SEER data (17) based on five-year age group–specific and sex-specific incidence of each cancer multiplied by person-years at risk (17). SIRs were calculated for the overall sample as well as for males and females separately. Three subanalyses were performed based on 1) FDRs’ kinship status to the proband (mother, father, sibling, and parents plus siblings), 2) FDRs’ personal smoking history (never vs ever), and 3) age of the proband at diagnosis of PC (<60 vs ≥60 years). Additionally, SIRs were calculated among FDRs of mutation-positive vs FDRs of mutation-negative probands and stratified by age of the proband at diagnosis (<60 years vs ≥60 years) in each group, and among full siblings of mutation-positive vs mutation-negative probands. Stratified analyses by sex were performed for each of these analyses.
Because of incomplete data on some of the FDRs, date of birth was imputed for 4135 FDRs (24.1%) and date of death was imputed for 1748 (10.2%), as was done previously (Supplementary Methods, available online) (11). While PC diagnoses among all probands were pathologically (>95%) or clinically confirmed, this level of verification among the FDRs was not feasible in the present study. Therefore, we relied on probands’ report of cancer diagnosis in an FDR, which has been shown to have an accuracy rate ranging between 60% and 80% (19,20). In validation analyses, we assessed the accuracy of probands’ report of cancer diagnosis in an FDR by comparing the proband report with the specified FDR’s report of personal history of cancer. We used a subset of data extracted from the familial PC study (16), wherein 480 FDRs (both affected and unaffected) completed the same questionnaires administered to the probands; their responses were used to validate the probands’ report. All statistical tests were two-sided and performed with SAS v9.4 (SAS Institute, Cary, NC). P values for statistical difference between SIRs were calculated assuming binomial distribution, as described in detail (21). See the Supplementary Methods (available online) for more details. P values less than .05 or confidence intervals that do not include 1.00 were considered statistically significant (22,23).
Results
The analyses included 17 162 FDRs, with more than 336 000 person-years at risk, identified through 2305 incident PC probands (Supplementary Methods, available online). The FDRs consisted of 4610 parents, 6920 siblings, and 5632 offspring. There were roughly equal proportions of male (50.6%) and female (49.3%) FDRs, and 43.2% of the FDRs had ever smoked. Overall, there were 3376 reported cases of cancer among the FDRs. Carcinomas of the breast (n = 463), colon and rectum (n = 338), prostate (n = 338), lung (n = 302), and pancreas (n = 233) were the most frequently reported malignancies.
The age- and sex-adjusted SIRs showed a twofold excess risk of PC among the FDRs (SIR = 2.04, 95% CI = 1.78 to 2.31, P < .001) as compared with the SEER reference population (Table 1). Incidence ratios for bladder, breast, colorectal, gastric, head and neck, leukemia, lung, lymphoma, melanoma, multiple myeloma, and prostate carcinomas were lower than expected among the FDRs compared with the SEER population. PC risk was elevated among both male and female FDRs. Among male FDRs only, gastric, head and neck, leukemia, melanoma, and prostate carcinomas were lower than expected; among female FDRs, breast cancer was lower (SIR = 0.77, 95% CI = 0.70 to 0.84, P < .001), but primary liver cancer was higher (SIR = 2.10, 95% CI = 1.34 to 3.12, P < .001) than expected. In subgroup analyses, we observed an excess risk of PC among the FDRs regardless of their relation to the proband, their smoking status, or the proband’s age at diagnosis (Supplementary Table 1, available online). The results from the subgroup analyses are similar to those reported in Table 1, with a few notable findings. The magnitude of PC risk for FDRs of young-onset PC probands (before age 60 years) was somewhat higher (SIR = 2.39, 95% CI = 1.82 to 3.09) than that of FDRs of older-onset probands (SIR = 1.94, 95% CI = 1.66 to 2.25) (Supplementary Table 1, available online). FDRs who had ever smoked had higher PC risk (SIR = 2.40, 95% CI = 1.98 to 2.89) than nonsmoking FDRs (SIR = 1.64, 95% CI = 1.32 to 2.00). Primary liver cancer was higher among mothers of the probands (SIR = 3.38, 95% CI = 1.89 to 5.58) and among FDRs who had ever smoked (SIR = 1.74, 95% CI = 1.18 to 2.47).
Table 1.
Standardized incidence ratios* for cancer risk among first-degree relatives of pancreatic cancer probands, stratified by sex of the relatives; the Mayo Clinic Biospecimen Resource for Pancreas Research, 2000–2016
| Cancer type | Overall (n = 17 162) |
Male (n = 8692) |
Female (n = 8470) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No. observed | No. expected | Person-years at risk | SIR (95% CI) | No. observed | No. expected | Person-years at risk | SIR (95% CI) | No. observed | No. expected | Person-years at risk | SIR (95% CI) | |
| Bladder | 82 | 200.7 | 685 085 | 0.41 (0.32 to 0.51)† | 58 | 176.4 | 339 152 | 0.33 (0.25 to 0.43)† | 24 | 45.0 | 345 932 | 0.53 (0.34 to 0.79)† |
| Brain‡ | 57 | 52.3 | 687 624 | 1.09 (0.83 to 1.41) | 32 | 31.7 | 340 889 | 1.01 (0.69 to 1.42) | 25 | 21.5 | 346 735 | 1.16 (0.75 to 1.72) |
| Breast | 463 | 655.4 | 680 553 | 0.71 (0.64 to 0.77)† | 1 | 5.4 | 339 754 | 0.18 (0.00 to 1.02) | 462 | 599.1 | 340 798 | 0.77 (0.70 to 0.84)† |
| Breast, Female | 462 | 599.1 | 340 798 | 0.77 (0.70 to 0.84)† | n/a | n/a | n/a | n/a | 462 | 599.1 | 340 798 | 0.77 (0.70 to 0.84)† |
| Colorectal | 338 | 516.3 | 682 944 | 0.65 (0.59 to 0.73)† | 178 | 299.9 | 338 134 | 0.59 (0.51 to 0.69)† | 160 | 227.2 | 344 810 | 0.70 (0.60 to 0.82)† |
| Gastric | 66 | 86.4 | 685 681 | 0.76 (0.59 to 0.97)† | 37 | 61.5 | 339 554 | 0.60 (0.42 to 0.83)† | 29 | 29.4 | 346 127 | 0.99 (0.66 to 1.42) |
| Head and neck | 102 | 160.7 | 686 862 | 0.63 (0.52 to 0.77)† | 71 | 125.6 | 340 451 | 0.57 (0.44 to 0.71)† | 31 | 43.3 | 346 411 | 0.72 (0.49 to 1.02) |
| Leukemia | 86 | 118.2 | 686 932 | 0.73 (0.58 to 0.90)† | 53 | 77.9 | 340 344 | 0.68 (0.51 to 0.89)† | 33 | 45.1 | 346 588 | 0.73 (0.50 to 1.03) |
| Liver‡ | 50 | 44.0 | 687 767 | 1.14 (0.84 to 1.50) | 26 | 35.1 | 340 967 | 0.74 (0.48 to 1.08) | 24 | 11.4 | 346 800 | 2.10 (1.34 to 3.12)† |
| Lung‡ | 302 | 600.1 | 686 657 | 0.50 (0.45 to 0.56)† | 189 | 406.7 | 340 329 | 0.46 (0.40 to 0.54)† | 113 | 221.0 | 346 328 | 0.51 (0.42 to 0.61)† |
| Lymphoma | 103 | 194.5 | 684 761 | 0.53 (0.43 to 0.64)† | 57 | 117.3 | 338 983 | 0.49 (0.37 to 0.63)† | 46 | 81.3 | 345 778 | 0.57 (0.41 to 0.76)† |
| Melanoma | 115 | 160.2 | 684 679 | 0.72 (0.59 to 0.86)† | 62 | 99.0 | 339 121 | 0.63 (0.48 to 0.80)† | 53 | 67.4 | 345 558 | 0.79 (0.59 to 1.03) |
| Myeloma | 21 | 56.2 | 685 946 | 0.37 (0.23 to 0.57)† | 13 | 34.7 | 339 706 | 0.38 (0.20 to 0.64)† | 8 | 23.2 | 346 240 | 0.34 (0.15 to 0.68)† |
| Ovary | 83 | 69.1 | 345 411 | 1.20 (0.96 to 1.49) | n/a | n/a | n/a | n/a | 83 | 69.1 | 345 411 | 1.20 (0.96 to 1.49) |
| Pancreas | 233 | 114.5 | 685 426 | 2.04 (1.78 to 2.31)† | 118 | 65.9 | 339 502 | 1.79 (1.48 to 2.15)† | 115 | 50.5 | 345 924 | 2.28 (1.88 to 2.73)† |
| Prostate | 338 | 697.7 | 336 398 | 0.48 (0.43 to 0.54)† | 338 | 697.7 | 336 398 | 0.48 (0.43 to 0.54)† | n/a | n/a | n/a | n/a |
Compared the observed with the expected number of cases based on data from the Surveillance, Epidemiology, and End Results program (nine registries, 1973–2013). CI = confidence interval; SIR = standardized incidence ratio.
Statistically significant association.
Where each was the only primary site reported, thereby excluding metastatic cases.
A subset of probands was tested for inherited mutations in 22 cancer predisposition genes; 9.4% (198/2094) carried a mutation in at least one of 19 genes (Supplementary Table 2, available online). Among FDRs of mutation-positive probands, we observed 4.32 (95% CI = 3.10 to 5.86) times as many cases of PC as expected in the SEER reference population; the magnitude of PC risk was lower among FDRs of mutation-negative probands (SIR = 1.77, 95% CI = 1.51 to 2.05, between-group difference, P < .001) (Table 2). Risks of breast (SIR = 1.29, 95% CI = 1.01 to 1.63) and ovarian (SIR = 2.38, 95% CI = 1.30 to 4.00) cancers were higher among FDRs of mutation-positive probands; while among FDRs of mutation negative probands, breast cancer risk was lower than expected (SIR = 0.64, 95% CI = 0.58 to 0.71), and there was no association for ovarian cancer. These results are very similar to those observed among full siblings of the probands (Table 3), such that PC risk was 4.09 (95% CI = 2.42 to 6.46) times higher than expected among siblings of mutation-positive probands, while it was 1.52 (95% CI = 1.16 to 1.94) higher than expected among siblings of mutation-negative probands (between-group difference, P = .001). Further, breast (SIR = 1.63, 95% CI = 1.17 to 2.22) and ovarian (SIR = 2.52, 95% CI = 1.01 to 5.20) cancers were elevated among siblings of mutation-positive probands. Among siblings of mutation-negative probands, breast cancer risk was lower than expected (SIR = 0.65, 95% CI = 0.55 to 0.76), and no association was found for ovarian cancer. Similar results were obtained from the sex-stratified analyses (Supplementary Tables 3 and 4, available online). Moreover, FDRs of young-onset, mutation-positive probands had much higher PC risk (SIR = 5.24, 95% CI = 2.93 to 8.64) than FDRs of young-onset probands who were mutation-negative (SIR = 1.76, 95% CI = 1.21 to 2.47, between-group difference, P < .001) (Table 4). Similarly, FDRs of mutation-positive, older-onset probands had higher PC risk (SIR = 3.93, 95% CI = 2.56 to 5.75) than FDRs of mutation-negative, older-onset probands (SIR = 1.77, 95% CI = 1.48 to 2.09, between-group difference, P = .001) (Table 5; Supplementary Table 6, available online).
Table 2.
Standardized incidence ratios* for cancer risk among first-degree relatives of pancreatic cancer probands, stratified by mutation status of the probands; the Mayo Clinic Biospecimen Resource for Pancreas Research, 2000–2016
| Cancer type | FDRs of mutation-positive probands†,‡ (n = 1465; from 198 pedigrees) |
FDRs of mutation-negative probands†,‡ (n = 14 168; from 1896 pedigrees) |
||||||
|---|---|---|---|---|---|---|---|---|
| No. observed | No. expected | Person-years at risk | SIR (95% CI) | No. observed | No. expected | Person-years at risk | SIR (95% CI) | |
| Bladder | 12 | 16.6 | 56 741 | 0.72 (0.37 to 1.26) | 65 | 165.9 | 566 204 | 0.39 (0.30 to 0.50)§ |
| Brain‖ | 4 | 4.3 | 57 092 | 0.92 (0.25 to 2.36) | 50 | 43.2 | 568 247 | 1.16 (0.86 to 1.53) |
| Breast | 70 | 54.2 | 56 268 | 1.29 (1.01 to 1.63)§ | 348 | 541.8 | 562 646 | 0.64 (0.58 to 0.71)§ |
| Breast, Female | 70 | 50.6 | 28 795 | 1.38 (1.08 to 1.75)§ | 347 | 494.9 | 281 504 | 0.70 (0.63 to 0.78)§ |
| Colorectal | 38 | 42.7 | 56 431 | 0.89 (0.63 to 1.22) | 268 | 426.8 | 564 592 | 0.63 (0.55 to 0.71)§ |
| Gastric | 5 | 7.2 | 56 875 | 0.70 (0.22 to 1.63) | 55 | 71.4 | 566 629 | 0.77 (0.58 to 1.00) |
| Head and neck | 9 | 13.3 | 56 992 | 0.67 (0.31 to 1.28) | 87 | 132.8 | 567 589 | 0.66 (0.52 to 0.81)§ |
| Leukemia | 9 | 9.8 | 57 001 | 0.92 (0.42 to 1.74) | 73 | 97.6 | 567 697 | 0.75 (0.59 to 0.94)§ |
| Liver‖ | 6 | 3.7 | 57 054 | 1.64 (0.60 to 3.58) | 42 | 36.4 | 568 390 | 1.15 (0.83 to 1.56) |
| Lung‖ | 19 | 49.8 | 56 949 | 0.38 (0.23 to 0.60)§ | 247 | 496.0 | 567 532 | 0.50 (0.44 to 0.56)§ |
| Lymphoma | 9 | 16.1 | 56 796 | 0.56 (0.25 to 1.06) | 87 | 160.7 | 565 853 | 0.54 (0.43 to 0.67)§ |
| Melanoma | 11 | 13.3 | 56 718 | 0.83 (0.41 to 1.48) | 93 | 132.4 | 565 881 | 0.70 (0.57 to 0.86)§ |
| Myeloma | 2 | 4.7 | 56 908 | 0.43 (0.05 to 1.55) | 17 | 46.5 | 566 851 | 0.37 (0.21 to 0.59)§ |
| Ovary | 14 | 5.9 | 29 389 | 2.38 (1.30 to 4.00)§ | 62 | 57.0 | 285 083 | 1.09 (0.83 to 1.39) |
| Pancreas | 41 | 9.5 | 56 807 | 4.32 (3.10 to 5.86)§ | 167 | 94.6 | 566 497 | 1.77 (1.51 to 2.05)§ |
| Prostate | 33 | 56.4 | 27 205 | 0.58 (0.40 to 0.82)§ | 278 | 577.2 | 278 322 | 0.48 (0.43 to 0.54)§ |
Compared the observed with the expected number of cases based on data from the Surveillance, Epidemiology, and End Results program (nine registries, 1973–2013). CI = confidence interval; FDR = first-degree relative; SIR = standardized incidence ratio.
Does not sum to 2305 pedigrees because not all families had an individual included in the mutation testing study.
Probands who tested positive or negative for inherited mutation in APC, ATM, BARD1, BRCA1, BRCA2, BRIP1, CDKN2A, CHEK2, FANCC, MLH1, MSH2, MSH6, NBN, PALB2, PMS2, PRSS1, RAD51C, RAD51D, and TP53.
Statistically significant association.
Where each was the only primary site reported, thereby excluding metastatic cases.
Table 3.
Standardized incidence ratios* for cancer risk among siblings of pancreatic cancer probands, stratified by mutation status of the proband; the Mayo Clinic Biospecimen Resource for Pancreas Research Registry, 2000–2016
| Cancer type | All siblings (n = 6920) |
Siblings of mutation-positive probands†,‡ (n = 636; from 182 pedigrees) |
Siblings of mutation-negative probands†,‡ (n = 5663; from 1728 pedigrees) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No. observed | No. expected | Person-years at risk | SIR (95% CI) | No. observed | No. expected | Person-years at risk | SIR (95% CI) | No. observed | No. expected | Person-years at risk | SIR (95% CI) | |
| Bladder | 35 | 88.7 | 302 654 | 0.39 (0.27 to 0.55)§ | 5 | 7.7 | 26 302 | 0.65 (0.21 to 1.51) | 28 | 72.9 | 248 785 | 0.38 (0.26 to 0.56)§ |
| Brain‖ | 24 | 23.1 | 303 711 | 1.04 (0.67 to 1.55)§ | 2 | 2.0 | 26 486 | 0.99 (0.11 to 3.59) | 19 | 19.0 | 249 598 | 1.00 (0.60 to 1.56) |
| Breast | 213 | 289.7 | 300 853 | 0.74 (0.64 to 0.84)§ | 41 | 25.1 | 26 068 | 1.63 (1.17 to 2.22)§ | 154 | 238.3 | 247 450 | 0.65 (0.55 to 0.76)§ |
| Breast, female | 213 | 258.6 | 147 081 | 0.82 (0.72 to 0.94)§ | 41 | 23.8 | 13 535 | 1.72 (1.24 to 2.34)§ | 154 | 211.8 | 120 461 | 0.73 (0.62 to 0.85)§ |
| Colorectal | 118 | 228.3 | 302 015 | 0.52 (0.43 to 0.62)§ | 15 | 19.8 | 26 187 | 0.76 (0.42 to 1.25) | 93 | 187.7 | 248 327 | 0.50 (0.40 to 0.61)§ |
| Gastric | 18 | 38.2 | 303 022 | 0.47 (0.28 to 0.75)§ | 2 | 3.3 | 26 400 | 0.60 (0.07 to 2.17) | 14 | 31.4 | 249 025 | 0.45 (0.24 to 0.75)§ |
| Head and neck | 43 | 71.0 | 303 476 | 0.61 (0.44 to 0.82)§ | 5 | 6.2 | 26 458 | 0.81 (0.26 to 1.88) | 35 | 58.4 | 249 363 | 0.60 (0.42 to 0.83)§ |
| Leukemia | 34 | 52.2 | 303 529 | 0.65 (0.45 to 0.91)§ | 2 | 4.5 | 26 438 | 0.44 (0.05 to 1.59) | 30 | 42.9 | 249 476 | 0.70 (0.47 to 1.00) |
| Liver‖ | 17 | 19.4 | 303 814 | 0.87 (0.51 to 1.40) | 4 | 1.7 | 26 448 | 2.36 (0.64 to 6.05) | 12 | 16.0 | 249 698 | 0.75 (0.39 to 1.31) |
| Lung‖ | 135 | 265.2 | 303 377 | 0.51 (0.43 to 0.60)§ | 6 | 23.1 | 26 451 | 0.26 (0.09 to 0.56)§ | 112 | 217.9 | 249 320 | 0.51 (0.42 to 0.62)§ |
| Lymphoma | 43 | 85.9 | 302 602 | 0.50 (0.36 to 0.67)§ | 3 | 7.5 | 26 369 | 0.40 (0.08 to 1.17) | 38 | 70.6 | 248 648 | 0.54 (0.38 to 0.74)§ |
| Melanoma | 59 | 70.7 | 302 337 | 0.83 (0.63 to 1.08) | 7 | 6.1 | 26 271 | 1.14 (0.46 to 2.35) | 44 | 58.2 | 248 577 | 0.76 (0.55 to 1.02) |
| Myeloma | 7 | 24.9 | 303 124 | 0.28 (0.11 to 0.58)§ | 0 | 2.2 | 26 414 | to | 6 | 20.4 | 249 112 | 0.29 (0.11 to 0.64)§ |
| Ovary | 38 | 29.8 | 149 004 | 1.28 (0.90 to 1.75) | 7 | 2.8 | 13 863 | 2.52 (1.01 to 5.20)§ | 29 | 24.4 | 121 823 | 1.19 (0.80 to 1.71) |
| Pancreas | 89 | 50.6 | 302 897 | 1.76 (1.41 to 2.17)§ | 18 | 4.4 | 26 354 | 4.09 (2.42 to 6.46) | 63 | 41.6 | 248 955 | 1.52 (1.16 to 1.94)§ |
| Prostate | 152 | 316.2 | 152 447 | 0.48 (0.41 to 0.56)§ | 12 | 25.8 | 12 449 | 0.46 (0.24 to 0.81)§ | 127 | 261.0 | 125 856 | 0.49 (0.41 to 0.58)§ |
Compared the observed with the expected number of cases based on data from the Surveillance, Epidemiology, and End Results program (nine registries, 1973–2013). CI = confidence interval; FDR = first-degree relative; SIR = standardized incidence ratio.
Does not sum to 2305 pedigrees because not all families had an individual included in the mutation testing study.
Probands who tested positive or negative for inherited mutation in APC, ATM, BARD1, BRCA1, BRCA2, BRIP1, CDKN2A, CHEK2, FANCC, MLH1, MSH2, MSH6, NBN, PALB2, PMS2, PRSS1, RAD51C, RAD51D, and TP53.
Statistically significant association.
Where each was the only primary site reported, thereby excluding metastatic cases.
Table 4.
Standardized incidence ratios* for cancer risk among first-degree relatives of pancreatic cancer probands who were younger than age 60 years at diagnosis, stratified by probands’ mutation status; the Mayo Clinic Biospecimen Resource for Pancreas Research Registry, 2000–2016
| Cancer type | FDRs of probands age < 60 y at diagnosis |
|||||||
|---|---|---|---|---|---|---|---|---|
| FDRs of mutation-positive probands†,‡ (n = 493; from 72 pedigrees) |
FDRs of mutation-negative probands†,‡ (n = 3331; from 494 pedigrees) |
|||||||
| No. observed | No. expected | Person-years at risk | SIR (95% CI) | No. observed | No. expected | Person-years at risk | SIR (95% CI) | |
| Bladder | 6 | 5.0 | 17 114 | 1.20 (0.44 to 2.60) | 11 | 32.9 | 112 158 | 0.33 (0.17 to 0.60)§ |
| Brain‖ | 2 | 1.3 | 17 281 | 1.52 (0.17 to 5.50) | 8 | 8.6 | 112 617 | 0.93 (0.40 to 1.84) |
| Breast | 22 | 16.4 | 17 026 | 1.34 (0.84 to 2.03) | 82 | 107.4 | 111 509 | 0.76 (0.61 to 0.95)§ |
| Breast, Female | 22 | 15.3 | 8703 | 1.44 (0.90 to 2.18) | 81 | 99.8 | 56 793 | 0.81 (0.64 to 1.01) |
| Colorectal | 15 | 12.9 | 17 043 | 1.16 (0.65 to 1.92) | 58 | 84.6 | 111 960 | 0.69 (0.52 to 0.89)§ |
| Gastric | 2 | 2.2 | 17 180 | 0.92 (0.10 to 3.34) | 7 | 14.2 | 112 332 | 0.49 (0.20 to 1.02) |
| Head and neck | 3 | 4.0 | 17 214 | 0.74 (0.15 to 2.18) | 6 | 26.3 | 112 444 | 0.23 (0.08 to 0.50)§ |
| Leukemia | 4 | 3.0 | 17 258 | 1.35 (0.36 to 3.45) | 19 | 19.4 | 112 583 | 0.98 (0.59 to 1.53) |
| Liver‖ | 2 | 1.1 | 17 278 | 1.81 (0.20 to 6.53) | 7 | 7.2 | 112 669 | 0.97 (0.39 to 2.00) |
| Lung‖ | 6 | 15.1 | 17 227 | 0.40 (0.15 to 0.87)§ | 49 | 98.3 | 112 519 | 0.50 (0.37 to 0.66)§ |
| Lymphoma | 2 | 4.9 | 17 176 | 0.41 (0.05 to 1.48) | 21 | 31.8 | 112 065 | 0.66 (0.41 to 1.01) |
| Melanoma | 5 | 4.0 | 17 120 | 1.25 (0.40 to 2.91) | 27 | 26.2 | 112 136 | 1.03 (0.68 to 1.50) |
| Myeloma | 1 | 1.4 | 17 191 | 0.71 (0.01 to 3.95) | 4 | 9.2 | 112 337 | 0.43 (0.12 to 1.11) |
| Ovary | 5 | 1.8 | 8866 | 2.82 (0.91 to 6.58) | 11 | 11.5 | 57 447 | 0.97 (0.58 to 1.58) |
| Pancreas | 15 | 2.9 | 17 142 | 5.24 (2.93 to 8.64)§ | 33 | 18.7 | 112 249 | 1.76 (1.21 to 2.47)§ |
| Prostate | 15 | 17.0 | 8204 | 0.88 (0.49 to 1.45) | 66 | 112.3 | 54 156 | 0.59 (0.45 to 0.75)§ |
Compared the observed with the expected number of cases based on data from the Surveillance, Epidemiology, and End Results program (nine registries, 1973–2013). CI = confidence interval; FDR = first-degree relative; SIR = standardized incidence ratio.
Does not sum to 2305 pedigrees because not all families had an individual included in the mutation testing study.
Probands who tested positive or negative for inherited mutation in APC, ATM, BARD1, BRCA1, BRCA2, BRIP1, CDKN2A, CHEK2, FANCC, MLH1, MSH2, MSH6, NBN, PALB2, PMS2, PRSS1, RAD51C, RAD51D, and TP53.
Statistically significant association.
Where each was the only primary site reported, thereby excluding metastatic cases.
Table 5.
Standardized incidence ratios* for cancer risk among first-degree relatives of pancreatic cancer probands who were age 60 years or older at diagnosis, stratified by probands’ mutation status; the Mayo Clinic Biospecimen Resource for Pancreas Research Registry, 2000–2016
| Cancer type | FDRs of probands who were age ≥ 60 y at diagnosis |
|||||||
|---|---|---|---|---|---|---|---|---|
| FDRs of mutation-positive probands†,‡ (n = 972; from 126 pedigrees) |
FDRs of mutation-positive probands†,‡ (n = 10 837; from 1402 pedigrees) |
|||||||
| No. observed | No. expected | Person-years at risk | SIR (95% CI) | No. observed | No. expected | Person-years at risk | SIR (95% CI) | |
| Bladder | 6 | 11.6 | 39 627 | 0.52 (0.19 to 1.12) | 54 | 133.0 | 454 046 | 0.41 (0.30 to 0.53)§ |
| Brain‖ | 2 | 3.0 | 39 811 | 0.66 (0.07 to 2.39) | 42 | 34.6 | 455 630 | 1.21 (0.87 to 1.64) |
| Breast | 48 | 37.8 | 39 241 | 1.27 (0.94 to 1.68) | 266 | 434.4 | 451 137 | 0.61 (0.54 to 0.69)§ |
| Breast, Female | 48 | 35.3 | 20 092 | 1.36 (1.01 to 1.80)§ | 266 | 395.0 | 224 710 | 0.67 (0.59 to 0.76)§ |
| Colorectal | 23 | 29.8 | 39 387 | 0.77 (0.49 to 1.16) | 210 | 342.2 | 452 632 | 0.61 (0.53 to 0.70)§ |
| Gastric | 3 | 5.0 | 39 695 | 0.60 (0.12 to 1.75) | 48 | 57.2 | 454 297 | 0.84 (0.62 to 1.11) |
| Head and neck | 6 | 9.3 | 39 779 | 0.64 (0.24 to 1.40) | 81 | 106.5 | 455 144 | 0.76 (0.60 to 0.95)§ |
| Leukemia | 5 | 6.8 | 39 744 | 0.73 (0.24 to 1.71) | 54 | 78.3 | 455 114 | 0.69 (0.52 to 0.90)§ |
| Liver‖ | 4 | 2.5 | 39 776 | 1.57 (0.42 to 4.02) | 35 | 29.2 | 455 721 | 1.20 (0.84 to 1.67) |
| Lung‖ | 13 | 34.7 | 39 722 | 0.37 (0.20 to 0.64)§ | 198 | 397.7 | 455 013 | 0.50 (0.43 to 0.57)§ |
| Lymphoma | 7 | 11.3 | 39 619 | 0.62 (0.25 to 1.28) | 66 | 128.9 | 453 787 | 0.51 (0.40 to 0.65)§ |
| Melanoma | 6 | 9.3 | 39 598 | 0.65 (0.24 to 1.41) | 66 | 106.2 | 453 745 | 0.62 (0.48 to 0.79)§ |
| Myeloma | 1 | 3.3 | 39 717 | 0.31 (0.00 to 1.71) | 13 | 37.3 | 454 514 | 0.35 (0.19 to 0.60)§ |
| Ovary | 9 | 4.1 | 20 523 | 2.19 (1.03 to 4.16)§ | 43 | 45.5 | 227 635 | 0.94 (0.68 to 1.27) |
| Pancreas | 26 | 6.6 | 39 665 | 3.93 (2.56 to 5.75)§ | 134 | 75.9 | 454 247 | 1.77 (1.48 to 2.09)§ |
| Prostate | 18 | 39.4 | 19 001 | 0.46 (0.27 to 0.72)§ | 212 | 464.9 | 224 167 | 0.46 (0.40 to 0.52)§ |
Compared the observed with the expected number of cases based on data from the Surveillance, Epidemiology, and End Results program (nine registries, 1973–2013). CI = confidence interval; FDR = first-degree relative; SIR = standardized incidence ratio.
Does not sum to 2305 pedigrees because not all families had an individual included in the mutation testing study.
Probands who tested positive or negative for inherited mutation in APC, ATM, BARD1, BRCA1, BRCA2, BRIP1, CDKN2A, CHEK2, FANCC, MLH1, MSH2, MSH6, NBN, PALB2, PMS2, PRSS1, RAD51C, RAD51D, and TP53.
Statistically significant association.
‖Where each was the only primary site reported, thereby excluding metastatic cases.
To assess the accuracy of reporting by the probands, we compared probands’ report of a history of cancer in an FDR with the FDR’s self-report among those with available data (n = 480). For PC, 98.5% (473/480) of reports were concordant. For the other cancers, 94% (434/462) of reports were concordant (Supplementary Methods, available online).
Discussion
We examined the risk of 15 common malignancies among FDRs of PC probands by comparing numbers of each cancer type observed among the FDRs with those expected using data from SEER. We observed a twofold higher than expected risk of PC among FDRs of the PC probands. Risk estimates for PC were consistently increased among parents and siblings of the probands, and suggest a potential aggregation of PC with primary liver cancer among female FDRs. Further analyses showed that the increased PC risk was higher among FDRs who had ever smoked than among nonsmoking FDRs, and the magnitude of the SIR for PC was higher among FDRs of young-onset probands than among FDRs of older-onset probands. We also observed lower than expected risk of bladder, breast, colorectal, gastric, head and neck, leukemia, lung, lymphoma, melanoma, multiple myeloma, and prostate carcinomas among the FDRs compared with the SEER reference population.
Familial aggregation of certain cancers may be due to shared genetic susceptibility and/or shared etiologic risk factors, or chance alone (24). For PC, inherited genetic factors account for 5% to 10% of cases, while tobacco smoking accounts for 20% to 25% of cases (2,25). A nested case–control study found that shared tobacco smoking behavior is an independent risk for familial aggregation of PC (26). Our results are consistent with independent reports (9,12,26–30) and with our previous smaller study (11), in which we observed a nearly twofold increase in PC risk among FDRs of PC probands (SIR = 1.88, 95% CI = 1.51 to 4.46) (11). Ghadirian et al. reported a fivefold increased risk among FDRs of PC probands compared with FDRs of noncancer controls and a twofold higher PC risk among FDRs of young-onset probands (10). Klein and colleagues reported a ninefold excess risk of PC among FDRs of PC probands compared with SEER; and consistent with our findings, they found higher PC risk among FDRs who had ever smoked (SIR = 19.2, 95% CI = 7.7 to 39.5) than among nonsmoking FDRs (SIR = 6.25, 95% CI = 1.70 to 16.0) (31).
Compared with the FDRs of 1896 PC probands who did not carry a mutation in any of the cancer genes tested, PC risk was statistically significantly higher among FDRs of probands who tested positive for a mutation; this would be expected based on Mendelian law probability that 50% of the full siblings likely carry the mutation(s) detected in their related mutation-positive proband. Ovarian and breast cancers were elevated among FDRs, particularly siblings, of PC probands who were mutation-positive. Moreover, SIRs for PC were much higher among FDRs of young-onset probands who carry a susceptibility gene mutation than among FDRs of young-onset probands who were mutation-negative, or FDRs of older-onset probands regardless of proband's mutation status.
Because the liver, brain, and lung are common metastatic sites, we restricted reported cancer cases for each of these sites to instances where they were reported as the only primary site (ie, liver only, lung only, or brain only). Thus, the increased risk of liver cancer observed among female FDRs, which concurs with our earlier findings (11), is potentially less prone to confounding by metastatic cases. Wang et al. (13) reported excess risk of death from liver cancer among FDRs of PC probands (standardized mortality ratio [SMR] = 1.91, 95% CI = 1.07 to 3.14), with a stronger trend toward increased risk among female FDRs (SMR = 2.38, 95% CI = 0.94 to 4.98) than among male FDRs (SMR = 1.65, 95% CI = 0.74 to 53.18) (13). Ghadirian et al. (10) also observed a statistically nonsignificant increased risk of liver cancer among FDRs of PC probands (relative risk = 1.89). While the sample sizes were small in these reports, these findings together suggest aggregation of PC with liver cancer that may be due to suggest the existence of a yet-to-be-identified pleiotropic genetic variant associated with both conditions, shared environmental risk factors among family members, or a combination of these factors. Further research involving pathologically confirmed primary liver cancers is needed before firm conclusions can be drawn about the familial aggregation of PC with liver cancer.
Strengths of the study include the use of an extensive risk factor questionnaire that allowed for uniform collection of data on multiple malignancies, the very large sample size, and the availability of cancer susceptibility mutation status on probands. A major limitation of this study is the reliance on probands’ reports of cancer diagnosis in an FDR as pathologic or clinical verification of cancer in the FDRs was not feasible in this study. In our subanalysis of FDRs from whom we had collected self-reported data, we found 98% agreement in the reporting of family history of PC between the probands and the FDRs and 94% agreement for reporting of other malignancies. This suggests that the potential impact of misclassification of cancer in the FDRs is likely minimal. However, there is the possibility that the PC probands may have had a greater awareness of a family history of PC than family history of other cancers, which may explain the observed lower than expected risk of some cancers among the FDRs as compared with the SEER population. Thus, underreporting of cancers other than PC has potential impact on the lower than expected risk observed for some cancers. Our study is limited also in its generalizability by race; 97% of the probands were White. We imputed missing data on dates of birth and death for some FDRs to ensure that the analysis reflects the population of FDRs in our registry to the fullest extent possible. Imputations were done a priori while blinded to the cancer status of the FDRs; thus, potential misclassification would be nondifferential by cancer status and would tend to attenuate effect estimates toward the null (19,32).
Our study substantiates a twofold risk of PC among FDRs of PC patients. It further suggests aggregation of PC with liver cancer among female FDRs. We found fourfold increased risk of PC and elevated risk of breast and ovarian cancer in FDRs of susceptibility gene mutation carriers, suggesting a potential role of inherited mutations in familial cancer risk. Younger age at the proband’s PC diagnosis was also associated with increased PC risk in the FDRs. Our findings enable more refined cancer risk estimation for genetic counseling and inform cancer screening recommendations for FDRs of PC probands.
Funding
This research was funded in part by a National Institutes of Health (NIH) Specialized Program of Research Excellence (SPORE) in Pancreatic Cancer grant P50 CA102701, NIH grants R01 CA97075, R01 CA192393, and R25 CA92049, Rolfe Pancreatic Cancer Foundation, and the Mayo Clinic Center for Individualized Medicine.
Notes
We thank the study participants and the dedicated staff of the Mayo Clinic Biospecimen Resource for Pancreas Research for their invaluable contributions. The study sponsors did not play any role in the design of the study, data collection, analyses, interpretation of the data, or the decision to submit the manuscript for publication.
Supplementary Material
References
- 1. Li D, Xie K, Wolff R, et al. Pancreatic cancer. Lancet. 2004;363(9414):1049–1057. [DOI] [PubMed] [Google Scholar]
- 2. Lowenfels AB, Maisonneuve P.. Epidemiology and risk factors for pancreatic cancer. Best Pract Res Clin Gastroenterol. 2006;20(2):197–209. [DOI] [PubMed] [Google Scholar]
- 3. Michaud D. Epidemiology of pancreatic cancer. Minerva Chir. 2004;59(2):99–111. [PubMed] [Google Scholar]
- 4. Marsh D, Zori R.. Genetic insights into familial cancers—update and recent discoveries. Cancer Lett. 2002;181(2):125–164. [DOI] [PubMed] [Google Scholar]
- 5. Garber JE, Offit K.. Hereditary cancer predisposition syndromes. J Clin Oncol. 2005;23(2):276–292. [DOI] [PubMed] [Google Scholar]
- 6. Lynch HT, Brand RE, Hogg D, et al. Phenotypic variation in eight extended CDKN2A germline mutation familial atypical multiple mole melanoma-pancreatic carcinoma-prone families: The familial atypical mole melanoma-pancreatic carcinoma syndrome. Cancer. 2002;94(1):84–96. [DOI] [PubMed] [Google Scholar]
- 7. 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]
- 8. Lynch HT, Smyrk T, Kern SE, et al. Familial pancreatic cancer: A review. Semin Oncol. 1996;23(2):251–275. [PubMed] [Google Scholar]
- 9. Silverman D, Schiffman M, Everhart J, et al. Diabetes mellitus, other medical conditions and familial history of cancer as risk factors for pancreatic cancer. Br J Cancer. 1999;80(11):1830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ghadirian P, Liu G, Gallinger S, et al. Risk of pancreatic cancer among individuals with a family history of cancer of the pancreas. Int J Cancer. 2002;97(6):807–810. [DOI] [PubMed] [Google Scholar]
- 11. McWilliams RR, Rabe KG, Olswold C, et al. Risk of malignancy in first-degree relatives of patients with pancreatic carcinoma. Cancer. 2005;104(2):388–394. [DOI] [PubMed] [Google Scholar]
- 12. 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]
- 13. Wang L, Brune KA, Visvanathan K, et al. Elevated cancer mortality in the relatives of patients with pancreatic cancer. Cancer Epidemiol Biomarkers Prev. 2009;18(11):2829–2834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Antwi SO, Oberg AL, Shivappa N, et al. Pancreatic cancer: Associations of inflammatory potential of diet, cigarette smoking and long-standing diabetes. Carcinogenesis. 2016;37(5):481–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Antwi SO, Eckert EC, Sabaque CV, et al. Exposure to environmental chemicals and heavy metals, and risk of pancreatic cancer. Cancer Causes Control. 2015;26(11):1583–1591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Zhen DB, Rabe KG, Gallinger S, et al. BRCA1, BRCA2, PALB2, and CDKN2A Mutations in Familial Pancreatic Cancer (FPC): A PACGENE study. Genet Med. 2015;17(7):569–577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.National Cancer Institute Database, Surveillance Research Program, Cancer Statistics Branch. Surveillance, Epidemiology, and End Results (SEER) Program SEER*Stat Database: Incidence -SEER 9 Regs Public-Use. www.seer.cancer.gov. Accessed July 2016.
- 18. 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]
- 19. Kerber RA, Slattery ML.. Comparison of self-reported and database-linked family history of cancer data in a case-control study. Am J Epidemiol. 1997;146(3):244–248. [DOI] [PubMed] [Google Scholar]
- 20. Murff HJ, Spigel DR, Syngal S.. Does this patient have a family history of cancer? An evidence-based analysis of the accuracy of family cancer history. JAMA. 2004;292(12):1480–1489. [DOI] [PubMed] [Google Scholar]
- 21. Altman D, Machin D, Bryant T, et al. Statistics with Confidence: Confidence Intervals and Statistical Guidelines. 2nd ed. London: BMJ Books; 2000;171–190. [Google Scholar]
- 22. Breslow NE, Day NE, eds. Statistical Methods in Cancer Research, Vol. 2 The Design and Analysis of Cohort Studies. IARC Scientific Publication No. 82. Lyon: International Agency for Research on Cancer; 1987. [PubMed] [Google Scholar]
- 23. Rothman KJ, Boice JD.. Epidemiologic Analysis with a Programmable Calculator, New Edition Boston, MA: Epidemiology Resources, Inc; 1982. [Google Scholar]
- 24. Hruban RH, Petersen GM, Goggins M, et al. Familial pancreatic cancer. Ann Oncol. 1999;10(suppl 4):69–73. [PubMed] [Google Scholar]
- 25. Raimondi S, Maisonneuve P, Lowenfels AB.. Epidemiology of pancreatic cancer: An overview. Nat Rev Gastroenterol Hepatol. 2009;6(12):699–708. [DOI] [PubMed] [Google Scholar]
- 26. Rulyak SJ, Lowenfels AB, Maisonneuve P, et al. Risk factors for the development of pancreatic cancer in familial pancreatic cancer kindreds. Gastroenterology. 2003;124(5):1292–1299. [DOI] [PubMed] [Google Scholar]
- 27. Schenk M, Schwartz AG, O'Neal E, et al. Familial risk of pancreatic cancer. J Natl Cancer Inst. 2001;93(8):640–644. [DOI] [PubMed] [Google Scholar]
- 28. Hemminki K, Li X.. Familial and second primary pancreatic cancers: A nationwide epidemiologic study from Sweden. Int J Cancer. 2003;103(4):525–530. [DOI] [PubMed] [Google Scholar]
- 29. Fernandez E, La Vecchia C, D'Avanzo B, et al. Family history and the risk of liver, gallbladder, and pancreatic cancer. Cancer Epidemiol Biomarkers Prev. 1994;3(3):209–212. [PubMed] [Google Scholar]
- 30. Tersmette AC, Petersen GM, Offerhaus GJ, et al. Increased risk of incident pancreatic cancer among first-degree relatives of patients with familial pancreatic cancer. Clin Cancer Res. 2001;7(3):738–744. [PubMed] [Google Scholar]
- 31. 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]
- 32. Glanz K, Grove J, Le Marchand L, et al. Underreporting of family history of colon cancer: Correlates and implications. Cancer Epidemiol Biomarkers Prev. 1999;8(7):635–639. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.