Abstract
Background
Persons newly diagnosed with pancreas cancer and who have survived a previous cancer are often excluded from clinical trials, despite limited evidence about their prognosis. We examined the association between previous cancer and overall survival.
Methods
This US population‐based cohort study included older adults (aged ≥66 years) diagnosed with pancreas cancer between 2005 and 2015 in the linked Surveillance, Epidemiology, and End Results‐Medicare data. We used Cox proportional hazards models to estimate stage‐specific effects of previous cancer on overall survival, adjusting for sociodemographic, treatment, and tumor characteristics.
Results
Of 32,783 patients, 18.7% were previously diagnosed with another cancer. The most common previous cancers included prostate (29.0%), breast (18.9%), or colorectal (9.7%) cancer. More than half of previous cancers (53.9%) were diagnosed 5 or more years prior to pancreas cancer diagnosis or at an in situ or localized stage (47.8%). The proportions of patients surviving 1, 3, and 5 years after pancreas cancer were nearly identical for those with and without previous cancer. Median survival in months was as follows for those with and without previous cancer respectively: 7 versus 8 (Stage 0/I), 10 versus 10 (Stage II), 7 versus 7 (Stage III), and 3 versus 2 (Stage IV). Cox models indicated that patients with previous cancer had very similar or statistically equivalent survival to those with no previous cancer.
Conclusions
Given nearly equivalent survival compared to those without previous cancer, cancer survivors newly diagnosed with pancreas cancer should be considered for inclusion in pancreas cancer clinical trials.
Keywords: clinical trials, multiple malignancies, pancreas cancer, survival
In this population‐based cohort of 32,783 patients diagnosed with pancreas cancer, nearly one in five were previously diagnosed with another cancer; commonly prostate, breast, or colorectal. Survival was nearly equivalent among those with and without previous cancer.
1. INTRODUCTION
In 2022, it is estimated that 62,210 people will be diagnosed with and 49,830 people will die from pancreas cancer in the US. 1 Pancreas cancer is the fourth leading cause of cancer deaths in the US; both incidence and mortality rates have increased since 2000. 2 Most patients diagnosed with pancreas cancer present with metastatic or unresectable disease, and therefore cytotoxic therapies are the mainstay of treatment. 3 , 4 These therapies provide only modest improvements in survival, often measured in weeks to months. As a result, 5‐year relative survival has remained stagnant at about 3% for patients with metastatic disease, in contrast to many other cancers, such as colorectal cancer, for which novel therapies have led to major improvements in survival. However, in recent years, there have been many advancements in adjuvant and neoadjuvant therapies resulting in a marked increase in the number of clinical trials. 3 , 4 , 5 , 6 Multiple clinical trials are currently enrolling patients to test additional targeted therapies and immunotherapies. 3 , 4
For most cancer types, clinical trials exclude cancer survivors newly diagnosed with another cancer, 7 including the majority of National Cancer Institute (NCI)‐sponsored pancreas cancer trials. 8 , 9 Exclusion of this population can hinder accrual efforts for the growing number of clinical trials. A recent systematic review indicates that only 8% of patients with cancer enroll in industry or government‐sponsored clinical trials. 10 In recent years, multiple organizations have recommended against the reflexive exclusion of cancer survivors. For example, the Food and Drug Administration (FDA) drafted new guidance in 2020 stating that patients with previous or concurrent malignancies should generally be eligible for enrollment in clinical trials. 11 The American Society of Clinical Oncology–Friends of Cancer Research Clinical Trial Eligibility Working Group similarly recommended against excluding patients solely on the basis of a previous or stable, concurrent malignancy, 12 , 13 and the NCI recently issued inclusion criteria template language relating to these recommendations. 14 , 15
Excluding patients with pancreas cancer who have survived previous cancer from clinical trials excludes a substantial number of patients, especially older adults, who may benefit from participation in clinical trials and need more effective therapies. A population‐based study from 1975 to 2017 demonstrated that the proportion of patients diagnosed with pancreas cancer and a previous cancer increased from 8.7% in 1975–1984 to 21.9% in 2015–2017. 16 As the cancer survivor population grows and ages, the number of persons diagnosed with multiple cancers will continue to increase. 17 , 18 , 19 As a consequence, by excluding cancer survivors, clinical trial results are becoming less generalizable to the overall population of people with cancer. Inclusion of patients with previous cancer may also improve the timeliness of trial accrual.
To address this gap, we used US population‐based data of older (age ≥ 65 years) patients newly diagnosed with pancreas cancer to determine whether overall survival differs for patients with previous cancer compared to those without previous cancer. Results will answer critical questions about the prevalence and impact of previous cancer and will inform evidence‐based clinical trial eligibility criteria.
2. METHODS
2.1. Data
We used linked Medicare claims and National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) program data. 20 SEER is a nationally representative collection of cancer registries in defined geographic regions and states. Linked SEER‐Medicare data provide additional clinical information on patients enrolled in Medicare. SEER‐Medicare data have been described elsewhere. 20 For this study, eligible patients included adults diagnosed with pancreas cancer at age ≥ 66 years between 2005 and 2015. We included patients diagnosed with Stage 0–IV pancreas cancer defined using the American Joint Committee on Cancer (AJCC) 6th edition; we excluded patients with the unknown stage of pancreas cancer (n = 7310). We included patients with common histologic types (see Appendix 1 for codes). To ensure complete data, we included patients with full coverage of Medicare Parts A and B, starting 1 year prior to pancreas cancer diagnosis and continuing at least 1 year after diagnosis or until death. We excluded HMO members, patients with only autopsy or death certificate records, and patients with incomplete dates of diagnosis or death, discrepancies between SEER and Medicare birthdate of 1 year or more, or discrepancies between sequence number and tumor site recode.
2.2. Measures
We defined previous cancer using our published approach. 8 , 21 , 22 , 23 In brief, we determined the order and timing of all primary cancer diagnoses using the SEER variable tumor site recode and associated ICD‐O‐3 values. As described elsewhere, 8 this approach identifies previous cancers reported to SEER and with a confirmed diagnosis date. We restricted our sample to two groups of patients diagnosed with pancreas cancer: (i) Patients with no previous cancer and (ii) patients with one or more previous cancer of a different type. We excluded the very small number of patients with any previous pancreas cancer (<1%).
We defined overall survival time as the interval in months between pancreas cancer diagnosis (defined as the 15th of the month because SEER provides only month and year of diagnosis) and death from any cause. Patients were followed until death or December 31, 2016.
The following patient characteristics were defined at the time of pancreas cancer diagnosis using SEER data: Age, sex, marital status, race/ethnicity, neighborhood poverty, urban/rural residence, region, and year of diagnosis. Patients with Medicaid were identified using the state buy‐in variable. 24 We measured the number of 10 frailty‐defining diagnoses (e.g., falls, dementia) and 16 non‐cancer comorbidities (e.g., diabetes, cerebrovascular disease) in the year prior to pancreas cancer diagnosis using Medicare claims (see Appendix 1). 25 , 26 , 27 We measured the following tumor characteristics of the pancreas cancer using SEER data: Tumor location, grade, histology, and stage at diagnosis defined using AJCC, 6th edition. Using both SEER and Medicare claims, we measured receipt of any chemotherapy, radiation, and surgery type. When defining covariates using Medicare claims, we searched inpatient, outpatient, and carrier claims.
2.3. Analysis
We reported the prevalence of previous cancer and compared covariates in patients with and without previous cancer using chi‐square tests. We also described the type, timing, and stage of previous cancer.
We conducted survival analyses separately by pancreas cancer stage. First, to illustrate the association of previous cancer and overall survival, we plotted Kaplan–Meier curves and used the log‐rank test to compare overall survival between patients with and without previous cancer. Second, we used Cox proportional hazard models with and without adjustment for all covariates to estimate the hazard rate for overall survival. Hazard ratios (HR) and their 95% confidence intervals (CIs) quantify the effects of previous cancer on overall survival. To assess the proportional hazard assumption, we examined log–log plots and Schoenfeld residuals; when violated, we included interaction terms of previous cancer with (log)time and qualitatively assessed results. Next, to assess whether our findings would be generalizable to a population of patients enrolled in clinical trials, we repeated Cox survival analyses in the subset of patients surviving 30 or more days after their pancreas cancer diagnosis (n = 2099, 6.4% of total population). These patients have likely survived long enough to be considered for a clinical trial.
Finally, to better understand absolute differences in survival, we estimated 1‐, 3‐ and 5‐year overall survival for patients with and without previous cancer.
We conducted analyses in SAS 9.4 (SAS Institute Inc.) and created figures in R Studio Version 1.2.5033. 28
The University of Texas Southwestern Medical Center Institutional Review Board approved this study (STU 042018–032).
3. RESULTS
3.1. Study population
Appendix 2 shows sample selection. Of 32,783 eligible patients newly diagnosed with pancreas cancer, 18.7% (n = 6139) had previous cancer. Table 1 demonstrates differences between those with and without previous cancer. For example, patients with previous cancer were older, more likely to be male and to receive chemotherapy and radiation, more likely to live in lower‐poverty census tracts, and less likely to have Medicaid; further, the proportion of White patients was higher among those with previous cancer. Characteristics of those with and without previous cancer are also shown separately by stage of pancreas cancer in Appendix 3.
TABLE 1.
Characteristics of patients diagnosed with pancreas cancer, by previous cancer history by analytic cohort
| Overall population | Clinical trial subset of patients surviving >= 30 days | |||||||
|---|---|---|---|---|---|---|---|---|
| No previous cancer(N = 26,644) | Previous cancer of other site(N = 6139) | p‐Value | Total(N = 32,783) | No previous cancer(N = 1686) | Previous cancer of other site(N = 413) | p‐Value | Total(N = 2099) | |
| Sex | ||||||||
| Male | 11,996 (45.0%) | 3448 (56.2%) | <0.001 | 15,444 (47.1%) | 765 (45.4%) | 226 (54.7%) | <0.001 | 991 (47.2%) |
| Female | 14,648 (55.0%) | 2691 (43.8%) | 17,339 (52.9%) | 921 (54.6%) | 187 (45.3%) | 1108 (52.8%) | ||
| Age | ||||||||
| 66–70 | 4781 (17.9%) | 837 (13.6%) | <0.001 | 5618 (17.1%) | 432 (25.6%) | 86.0 (20.8%) | 0.0755 | 518 (24.7%) |
| 70–75 | 6064 (22.8%) | 1299 (21.2%) | 7363 (22.5%) | 474 (28.1%) | 128 (31.0%) | 602 (28.7%) | ||
| 75–80 | 6002 (22.5%) | 1454 (23.7%) | 7456 (22.7%) | 430 (25.5%) | 102 (24.7%) | 532 (25.3%) | ||
| 80–85 | 5195 (19.5%) | 1366 (22.3%) | 6561 (20.0%) | 243 (14.4%) | 76.0 (18.4%) | 319 (15.2%) | ||
| >85 | 4602 (17.3%) | 1183 (19.3%) | 5785 (17.6%) | 107 (6.3%) | 21.0 (5.1%) | 128 (6.1%) | ||
| Race/ethnicity | ||||||||
| Non‐Hispanic White | 20,671 (77.6%) | 5008 (81.6%) | <0.001 | 25,679 (78.3%) | 1353 (80.2%) | 336 (81.4%) | 0.337 | 1689 (80.5%) |
| Hispanic white | 1122 (4.2%) | 183 (3.0%) | 1305 (4.0%) | 83.0 (4.9%) | 12.0 (2.9%) | 95.0 (4.5%) | ||
| Black | 2560 (9.6%) | 540 (8.8%) | 3100 (9.5%) | 101 (6.0%) | 28.0 (6.8%) | 129 (6.1%) | ||
| Other/Unknown | 2291 (8.6%) | 408 (6.6%) | 2699 (8.2%) | 149 (8.8%) | 37.0 (9.0%) | 186 (8.9%) | ||
| Marital status | ||||||||
| Married/Unmarried or domestic partner | 13,749 (51.6%) | 3433 (55.9%) | <0.001 | 17,182 (52.4%) | 1053 (62.5%) | 274 (66.3%) | 0.0031 | 1327 (63.2%) |
| Separated/Divorced/Widowed | 9763 (36.6%) | 2018 (32.9%) | 11,781 (35.9%) | 467 (27.7%) | 83.0 (20.1%) | 550 (26.2%) | ||
| Single | 2165 (8.1%) | 460 (7.5%) | 2625 (8.0%) | 114 (6.8%) | 34.0 (8.2%) | 148 (7.1%) | ||
| Unknown | 967 (3.6%) | 228 (3.7%) | 1195 (3.6%) | 52.0 (3.1%) | 22.0 (5.3%) | 74.0 (3.5%) | ||
| Medicaid | ||||||||
| Yes | 4387 (16.5%) | 739 (12.0%) | <0.001 | 5126 (15.6%) | 173 (10.3%) | 31.0 (7.5%) | 0.109 | 204 (9.7%) |
| No | 22,257 (83.5%) | 5400 (88.0%) | 27,657 (84.4%) | 1513 (89.7%) | 382 (92.5%) | 1895 (90.3%) | ||
| Poverty | ||||||||
| 0% to <10% poverty | 13,856 (52.0%) | 3397 (55.3%) | <0.001 | 17,253 (52.6%) | 1013 (60.1%) | 257 (62.2%) | 0.291 | 1270 (60.5%) |
| 10% to <20% poverty | 7529 (28.3%) | 1724 (28.1%) | 9253 (28.2%) | 431 (25.6%) | 109 (26.4%) | 540 (25.7%) | ||
| 20% to 100% poverty | 5259 (19.7%) | 1018 (16.6%) | 6277 (19.1%) | 242 (14.4%) | 47.0 (11.4%) | 289 (13.8%) | ||
| Urban–rural indicator | ||||||||
| Metro | 22,576 (84.7%) | 5253 (85.6%) | 0.103 | 27,829 (84.9%) | 1464 (86.8%) | 357 (86.4%) | 0.897 | 1821 (86.8%) |
| Non‐metropolitan | 4068 (15.3%) | 886 (14.4%) | 4954 (15.1%) | 222 (13.2%) | 56.0 (13.6%) | 278 (13.2%) | ||
| Region | ||||||||
| Northeast | 5868 (22.0%) | 1281 (20.9%) | <0.001 | 7149 (21.8%) | 430 (25.5%) | 111 (26.9%) | 0.0102 | 541 (25.8%) |
| South | 6338 (23.8%) | 1142 (18.6%) | 7480 (22.8%) | 334 (19.8%) | 65.0 (15.7%) | 399 (19.0%) | ||
| Midwest | 3265 (12.3%) | 1087 (17.7%) | 4352 (13.3%) | 178 (10.6%) | 65.0 (15.7%) | 243 (11.6%) | ||
| West | 11,173 (41.9%) | 2629 (42.8%) | 13,802 (42.1%) | 744 (44.1%) | 172 (41.6%) | 916 (43.6%) | ||
| Grade | ||||||||
| Well differentiated | 954 (3.6%) | 235 (3.8%) | 0.731 | 1189 (3.6%) | 204 (12.1%) | 53.0 (12.8%) | 0.674 | 257 (12.2%) |
| Moderately differentiated | 3338 (12.5%) | 775 (12.6%) | 4113 (12.5%) | 550 (32.6%) | 142 (34.4%) | 692 (33.0%) | ||
| Poorly differentiated/Undifferentiated | 3612 (13.6%) | 846 (13.8%) | 4458 (13.6%) | 345 (20.5%) | 74.0 (17.9%) | 419 (20.0%) | ||
| Not determined | 18,740 (70.3%) | 4283 (69.8%) | 23,023 (70.2%) | 587 (34.8%) | 144 (34.9%) | 731 (34.8%) | ||
| Surgery | ||||||||
| No surgery | 22,255 (83.5%) | 5077 (82.7%) | 0.0876 | 27,332 (83.4%) | 475 (28.2%) | 118 (28.6%) | 0.986 | 593 (28.3%) |
| Whipple | 2177 (8.2%) | 499 (8.1%) | 2676 (8.2%) | 589 (34.9%) | 143 (34.6%) | 732 (34.9%) | ||
| Non‐Whipple surgery | 2212 (8.3%) | 563 (9.2%) | 2775 (8.5%) | 622 (36.9%) | 152 (36.8%) | 774 (36.9%) | ||
| Chemotherapy | ||||||||
| Yes | 11,989 (45.0%) | 2991 (48.7%) | <0.001 | 14,980 (45.7%) | 1133 (67.2%) | 300 (72.6%) | 0.0385 | 1433 (68.3%) |
| No | 14,655 (55.0%) | 3148 (51.3%) | 17,803 (54.3%) | 553 (32.8%) | 113 (27.4%) | 666 (31.7%) | ||
| Radiation | ||||||||
| Yes | 12,337 (46.3%) | 3081 (50.2%) | <0.001 | 15,418 (47.0%) | 1180 (70.0%) | 305 (73.8%) | 0.137 | 1485 (70.7%) |
| No | 14,307 (53.7%) | 3058 (49.8%) | 17,365 (53.0%) | 506 (30.0%) | 108 (26.2%) | 614 (29.3%) | ||
| Comorbidity count | ||||||||
| 0 | 10,039 (37.7%) | 2183 (35.6%) | 0.0131 | 12,222 (37.3%) | 770 (45.7%) | 189 (45.8%) | 0.991 | 959 (45.7%) |
| 1 | 8610 (32.3%) | 2014 (32.8%) | 10,624 (32.4%) | 558 (33.1%) | 135 (32.7%) | 693 (33.0%) | ||
| 2 | 4330 (16.3%) | 1048 (17.1%) | 5378 (16.4%) | 236 (14.0%) | 60.0 (14.5%) | 296 (14.1%) | ||
| >= 3 | 3665 (13.8%) | 894 (14.6%) | 4559 (13.9%) | 122 (7.2%) | 29.0 (7.0%) | 151 (7.2%) | ||
| Frailty count | ||||||||
| 0 | 16,057 (60.3%) | 3734 (60.8%) | 0.419 | 19,791 (60.4%) | 1110 (65.8%) | 298 (72.2%) | 0.0359 | 1408 (67.1%) |
| 1 | 8404 (31.5%) | 1886 (30.7%) | 10,290 (31.4%) | 510 (30.2%) | 105 (25.4%) | 615 (29.3%) | ||
| >= 2 | 2183 (8.2%) | 519 (8.5%) | 2702 (8.2%) | 66.0 (3.9%) | 10.0 (2.4%) | 76.0 (3.6%) | ||
| Tumor location | ||||||||
| Pancreas body & tail | 6721 (25.2%) | 1576 (25.7%) | 0.688 | 8297 (25.3%) | 361 (21.4%) | 115 (27.8%) | 0.00904 | 476 (22.7%) |
| Pancreas head | 13,367 (50.2%) | 3078 (50.1%) | 16,445 (50.2%) | 1075 (63.8%) | 232 (56.2%) | 1307 (62.3%) | ||
| Pancreas other | 6556 (24.6%) | 1485 (24.2%) | 8041 (24.5%) | 250 (14.8%) | 66.0 (16.0%) | 316 (15.1%) | ||
| PC stage | ||||||||
| Stage 0 & 1 | 2559 (9.6%) | 687 (11.2%) | <0.001 | 3246 (9.9%) | 451 (26.7%) | 94.0 (22.8%) | 0.299 | 545 (26.0%) |
| Stage II | 6934 (26.0%) | 1759 (28.7%) | 8693 (26.5%) | 942 (55.9%) | 247 (59.8%) | 1189 (56.6%) | ||
| Stage III | 2331 (8.7%) | 543 (8.8%) | 2874 (8.8%) | 122 (7.2%) | 26.0 (6.3%) | 148 (7.1%) | ||
| Stage IV | 14,820 (55.6%) | 3150 (51.3%) | 17,970 (54.8%) | 171 (10.1%) | 46.0 (11.1%) | 217 (10.3%) | ||
| Vital status | ||||||||
| Alive | 1326 (5.0%) | 340 (5.5%) | 0.0761 | 1666 (5.1%) | 720 (42.7%) | 166 (40.2%) | 0.384 | 886 (42.2%) |
| Dead | 25,318 (95.0%) | 5799 (94.5%) | 31,117 (94.9%) | 966 (57.3%) | 247 (59.8%) | 1213 (57.8%) | ||
Table 2 shows previous cancer type, timing, and stage. For those with more than one previous cancer, n = 945 (15.4%) we report characteristics of the most recent previous cancer (relative to the pancreas cancer). The most common previous cancers were prostate (26.5%), breast (19.7%), or colorectal (11.4%). While some previous cancers (14.4%) were diagnosed within 1 year of pancreas cancer, nearly one third (31.7%) were diagnosed within 1–5 years, and the majority (53.9%) were diagnosed more than 5 years before the pancreas cancer. Nearly half of previous cancers (47.8%) were diagnosed at an in situ or localized stage.
TABLE 2.
Characteristics of previous cancer among patients with pancreas cancer, by analytic cohort
| Overall population | Clinical trial subset of patients surviving >=30 days | |||||
|---|---|---|---|---|---|---|
| Stage 0 & I N = 687 (21.2%) | Stage II N = 1759 (20.2%) | Stage III N = 543 (18.9%) | Stage IV N = 3150 (17.5%) | All stages combined N = 6139 (18.7%) | All stages combined N = 413 | |
| Number (%) | ||||||
| Type of previous cancer | ||||||
| Breast | 135 (19.7%) | 348 (19.8%) | 123 (22.7%) | 603 (19.1%) | 1209 (19.7%) | 78 (18.9%) |
| Colon; rectum and rectosigmoid junction; Small intestine | 79 (11.5%) | 180 (10.2%) | 74 (13.6%) | 365 (11.6%) | 698 (11.4%) | 40 (9.7%) |
| Melanoma of the skin; other non‐epithelial skin | 45 (6.6%) | 148 (8.4%) | 33 (6.1%) | 279 (8.9%) | 505 (8.2%) | 29 (7.0%) |
| Prostate | 139 (20.2%) | 477 (27.1%) | 124 (22.8%) | 885 (28.1%) | 1625 (26. 5%) | 120 (29.0%) |
| Urinary bladder; ureter; other urinary organs | 40 (5.8%) | 103 (5.9%) | 35 (6.4%) | 213 (6.8%) | 391 (6.4%) | 27 (6.5%) |
| All others | 249 (36.2%) | 503 (28.6%) | 154 (28.4%) | 805 (25.6%) | 1711 (27.9%) | 119 (28.8%) |
| Previous cancer timing | ||||||
| <1 year | 134 (19.5%) | 254 (14.4%) | 73 (13.4%) | 420 (13.3%) | 881 (14.4) | 69 (16.7%) |
| 1–5 years | 205 (29.8%) | 569 (32.3%) | 189 (34.8%) | 985 (31.3%) | 1948 (31.7) | 133 (32.2%) |
| >5 years | 348 (50.7%) | 936 (53.2%) | 281 (51.7%) | 1745 (55.4%) | 3310 (53.9) | 211 (51.1%) |
| Stage of previous cancer | ||||||
| Localized/In‐situ | 319 (46.4%) | 856 (48.7%) | 273 (50.3%) | 1485 (47.1%) | 2933 (47.8) | 184 (44.6%) |
| Regional | 129 (18.8%) | 239 (13.6%) | 86 (15.8%) | 374 (11.9%) | 828 (13.5) | 59 (14.3%) |
| Distant | 41 (6%) | 77 (4.4%) | 31 (5.7%) | 183 (5.8%) | 332 (5.4) | 20 (4.8%) |
| Unknown/missing | 198 (28.8%) | 587 (33.4%) | 153 (28.2%) | 1108 (35.2%) | 2046 (33.3) | 150 (36.3%) |
3.2. Survival of patients with previous cancer
Figure 1 illustrates survival curves for patients with and without previous cancer, separately by stage of pancreas cancer. For patients with Stage 0/I pancreas cancer (n = 3246), the log‐rank test indicated a statistically significant difference (p = 0.032) in overall survival between the two groups, whereby those with previous cancer had worse survival compared to those with no previous cancer. In contrast, for patients with Stage IV pancreas cancer (n = 17,970), the survival curve demonstrated that patients with previous cancer had better survival compared to those with no previous cancer; this difference was also statistically significant (p < 0.001). There were no statistically significant differences between the two groups for patients with Stage II or III pancreas cancer.
FIGURE 1.
Overall survival curves for pancreas cancer patients with and without previous cancer, by pancreas cancer stage.
We tested the proportional hazard assumption in the total study population; it was violated in only two of the eight Cox models (the unadjusted Stage IV model and the adjusted Stage 0/I model). Therefore, to facilitate interpretation and summarize effects, Figure 2 presents the average hazard of previous cancer over time in Cox models without time interactions. Figure 2 also illustrates both the unadjusted and adjusted association of previous cancer and overall survival.
FIGURE 2.
Impact of previous cancer on overall survival among patients with pancreas cancer for the overall population and clinical trial subset surviving ≥30 days after pancreas cancer diagnosis, by pancreas cancer stage.
Results from Cox models were consistent with the log‐rank tests. Previous cancer was associated with worse overall survival for patients with Stage 0/I pancreas cancer (adjusted hazard ratio [aHR]: 1.14;95% CI:1.04–1.26) but improved overall survival for patients with Stage IV pancreas cancer (aHR: 0.94;95% CI:0.90–0.98). Patients with Stage II or III pancreas cancer and a previous cancer had equivalent survival to patients with no previous cancer.
3.3. 1‐, 3‐, and 5‐year survival
To quantify the absolute impact of previous cancer on survival, Table 3 illustrates the proportion of patients alive at 1, 3, and 5 years after diagnosis of pancreas cancer. In general, survival between those with and without previous cancer was equivalent; differences were small, and confidence intervals overlapped in all but three scenarios. First, among patients with Stage IV pancreas cancer, at 1 year after pancreas cancer diagnosis, 11% (95% CI: 10%–12%) of those with previous cancer survived, compared to 8% (95% CI: 8%–9%) of those without previous cancer. Second, among patients with Stage 0/I pancreas cancer, at 3 years after the pancreas cancer diagnosis, 14% (95% CI: 11–17%) of patients with previous cancer were alive compared to 18% (95% CI: 17–20%) of those without. Third, in the same patients with Stage 0/I pancreas cancer, at 5 years after the pancreas cancer diagnosis, 9% (95% CI: 7%–12%) of patients with previous cancer were alive compared to 14% (95% CI: 13%–16%) of those without previous cancer.
TABLE 3.
Proportions surviving 1‐, 3‐ and 5‐ years after diagnosis of pancreas cancer and median survival time, among the overall population by stage of pancreas cancer
| Stage 0 & 1 | Stage II | Stage III | Stage IV | |
|---|---|---|---|---|
| Months (Interquartile range) | ||||
| Median survival | ||||
| No previous | 8 (3–23) | 10 (4–21) | 7 (3–14) | 2 (1–5) |
| Previous cancer | 7 (3–18) | 10 (4–21) | 7 (3–17) | 3 (1–6) |
| Proportion alive (95% confidence interval) | ||||
| 1‐year survival | ||||
| No previous | 0.38 (0.36–0.39) | 0.43 (0.42–0.44) | 0.28 (0.27–0.30) | 0.08 (0.08–0.09) |
| Previous cancer | 0.35 (0.31–0.39) | 0.42 (0.39–0.44) | 0.29 (0.26–0.33) | 0.11 (0.10–0.12) |
| 3‐year survival | ||||
| No previous | 0.18 (0.17–0.20) | 0.13 (0.12–0.13) | 0.04 (0.04–0.05) | 0.01 (0.01–0.01) |
| Previous cancer | 0.14 (0.11–0.17) | 0.13 (0.11–0.14) | 0.04 (0.02–0.06) | 0.009 (0.006–0.01) |
| 5‐year survival | ||||
| No previous | 0.14 (0.13–0.16) | 0.07 (0.07–0.08) | 0.02 (0.01–0.03) | 0.004 (0.003–0.005) |
| Previous cancer | 0.09 (0.07–0.12) | 0.07 (0.06–0.09) | 0.01 (0.01–0.03) | 0.004 (0.002–0.01) |
3.4. Subset of patients eligible for clinical trials
Less than 10% of the study population (n = 2099) survived 30 or more days after pancreas cancer diagnosis and were included in the subset of patients likely eligible for a clinical trial. Patients included in this subset differed in several ways from the total study population. They were relatively younger, had fewer comorbidities and frailties, and were diagnosed with pancreas cancer at earlier stages and with more favorable clinical characteristics such as tumor grade and location. Of these, 19.7% (n = 413) had previous cancer; a percentage similar to that seen in the total population (i.e., 18.7%). There were fewer statistically significant differences in patient characteristics by previous cancer within this subset, possibly due to the smaller sample. One of these differences indicated that, compared to those without previous cancer, those with previous cancer were less likely to have tumors in the head of the pancreas and more likely to have tumors in the body or tail of the pancreas.
There was no statistically significant difference in survival by previous cancer for patients with Stages 0/I, II, or III pancreas cancer (Figure 2). Among patients with Stage IV pancreas cancer, the unadjusted model indicated no difference in survival between those with and without previous cancer; however, the adjusted model indicated that previous cancer was associated with worse overall survival (aHR: 1.60;95% CI: 1.01–2.55).
4. DISCUSSION
We observed that nearly one in five (18.7%) patients aged 66 and older who were newly diagnosed with pancreas cancer have survived a previous cancer, and these patients generally have equivalent survival compared to those without previous cancer. There were some exceptions at the extremes of staging and that manifested as small differences in survival. Notably, these small survival differences by previous cancer were largely absent in the subset of patients surviving at least 30 days. In that subset, results indicated that among patients with Stage 0–III pancreas cancer, those with and without previous cancer had equivalent survival. However, among those with Stage IV pancreas cancer, those with previous cancer had worse survival. Based on these findings, we conclude that patients with and without previous cancer generally have equivalent survival. Survival is largely driven by the incident of pancreas cancer and not the previous cancer, likely because pancreas cancer is aggressive and lethal.
One rationale for excluding patients with previous cancer is the potential for previous cancer to negatively impact trial outcomes. There is an implicit assumption that previous cancer leads to worse survival that may negatively impact trial outcomes. However, the limited research about the prognosis of patients with pancreas cancer who have survived previous cancer documents improved survival or no difference. In prior studies, there is some evidence that previous cancer is associated with improved survival among Stage IV patients with pancreatic adenocarcinoma, 29 a general pancreas cancer population, 30 and among patients with pancreatic ductal adenocarcinoma 31 ; however, another study of patients with pancreatic ductal adenocarcinoma found no difference in overall survival among those with and without previous cancer. 32 Our study is unique compared to these other studies because we focused on older patients, included patients with all stages of pancreas cancer, and examined stage‐specific patterns. Differences related to eligibility criteria or other methods may account for mixed results in the literature to date.
Several limitations of our study are worth noting. First, some pancreas cancers may represent misclassified metastases from earlier tumors. This is likely rare, however, given strict SEER rules for coding primary cancers 33 and high rates of microscopic confirmation of SEER‐reported tumors. Another limitation is our lack of data on adults younger than 66 years of age. However, because the prevalence of previous cancer is more common among older versus younger patients with pancreas cancer, 34 our study reflects the survival of the majority of patients with previous cancer. Third, there are some instances of smaller cell sizes, for example, the clinical trial subset, that may have limited our power to detect differences in survival. Finally, SEER provides only the month and year of diagnosis and therefore our subset analysis may be somewhat imprecise. Among our strengths is our adjustment for important covariates not included in previous studies, such as patient comorbidity and frailty, and our more complete ascertainment of chemotherapy and radiation therapy using data from both SEER and Medicare, 35 as opposed to previous studies relying on the use of SEER data alone. 29 , 30 , 31 , 32
We observed some noteworthy differences in our analysis of the total study population versus clinical trial subset surviving 30 days or more. First, in the total population, previous cancer was associated with improved survival for patients with Stage IV cancer; however, these results were reversed in the clinical trial subset. Lead‐time bias may be one explanation. Specifically, if Stage IV pancreas cancer patients with previous cancer experience more frequent or comprehensive medical monitoring and surveillance, their cancers may be diagnosed earlier compared to those without previous cancer. In this scenario, the lead time would result in an artificially prolonged survival (defined as months from date of diagnosis to date of death/censoring) and an apparent survival benefit for patients with previous cancer. However, if lead time is relatively short on average (e.g., a few weeks), this effect would be washed out after limiting to the subset of patients surviving 30 days. Second, in early‐stage patients, previous cancer was associated with worse survival in the total population but was not statistically significantly associated with survival in the clinical trial subset, although the effect estimate was elevated. This might be explained by the smaller size of the subset, resulting in wide confidence intervals and less power to detect modest associations. When assessing results of the subset analysis, it is important to note that the clinical trial subset represents less than 10% of the total population.
4.1. Conclusions & implications for clinical trial design
In summary, nearly one in five older patients newly diagnosed with pancreas cancer has survived a previous cancer, and these patients generally have equivalent survival compared to those without previous cancer. Given overall poor survival for patients with pancreas cancer, the clear need for new treatments to improve survival from this aggressive cancer, similar survival of patients regardless of previous cancer history, and the importance of generalizability of clinical trial participants to the larger population, pancreas cancer patients with previous cancer should not be reflexively excluded from clinical trials. Furthermore, a recent study at an academic medical center illustrated that traditionally restrictive eligibility criteria may also perpetuate racial and ethnic disparities in clinical trial enrollment. 37 The authors illustrated that more inclusive inclusion criteria could equalize trial ineligibility rates between Black and White patients with pancreatic ductal adenocarcinoma, and potentially other cancers. 37 Importantly, the NCI recently issued guidance and clinical trial template language related to the inclusion of this population in clinical trials. 14 , 15 There are recent examples of clinical trial protocols with no explicit exclusion criteria; for example, an NCI network group sponsored trial Alliance A021806 of perioperative versus adjuvant chemotherapy for resectable pancreas cancer. 36 It will be important to monitor the extent to which additional clinical trials adapt selection criteria accordingly, with the ultimate goal of ensuring the results of clinical trials are generalizable to this growing and under‐studied population.
AUTHORS' CONTRIBUTIONS
Sandi L. Pruitt designed the study and led the writing; Caitlin C. Murphy provided design and analysis guidance; Anna Tavakkoli provided code and subject matter expertise. Hong Zhu, Daniel F. Heitjan, and Bhumika Maddineni designed, critically evaluated, and/or performed data analysis. David E. Gerber, Anna Tavakkoli, Muhammad Shaalan Beg, Ethan A. Halm, Ankit J. Kansagra and Amit G. Singal provided clinical insights. All authors interpreted the results and provided critical feedback and edits to the manuscript and approved the final manuscript.
FUNDING INFORMATION
National Cancer Institute R01CA229834 (SLP) and K24CA201543 (DEG).
CONFLICT OF INTEREST
Daniel F. Heitjan reports consulting (Bluejay Diagnostics, Braintree labs, Creatics LLC, Abbott Labs, Macrogenics, Inc.,) and legal consulting (Noven, Pfizer, Women's Talc Project); Caitlin C. Murphy reports consulting (Freenome); Amit G. Singal reports consulting/advisory boards (Bayer, Eisai, Genentech, BMS, Exelixis, Exact Sciences, GRAIL); David E. Gerber reports research funding (Astra‐Zeneca, BerGenBio, Karyopharm), stock ownership (Gilead) and consulting (Samsung Bioepis, Jansen, Catalyst Pharmaceuticals). Ankit J. Kansagra reports advisory board (Alynylam, BMS/Celgene, Cota Health, GSK, Janssen, Oncopeptide).
Supporting information
Appendix S1‐S3
ACKNOWLEDGMENTS
The authors acknowledge Himani Radadiya's analysis assistance and efforts of the Applied Research Program, NCI; the Office of Research, Development and Information, CMS; Information Management Services (IMS), Inc.; and Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of SEER‐Medicare. Contents of this paper are solely the responsibility of the authors and do not necessarily represent the official view of the NIH.
Pruitt SL, Tavakkoli A, Zhu H, et al. Survival of cancer survivors with a new pancreatic cancer diagnosis. Cancer Med. 2023;12:200‐212. doi: 10.1002/cam4.4903
Muhammad Shaalan Beg is currently employed by Science37; this research was completed while he was employed by UT Southwestern.
Ankit J. Kansagra is currently employed by Johnson and Johnson; this research was completed while he was employed by UT Southwestern.
DATA AVAILABILITY STATEMENT
This study used SEER‐MEDICARE data. The Centers for Medicare and Medicaid Services do not allow the redistribution of their data by researchers. SEER‐MEDICARE data are distinct from the publicly available SEER database, and can be obtained by researchers, by following the process described on https://healthcaredelivery.cancer.gov/seermedicare/obtain/requests.html (access requirements include Institutional Review Board approval and the completion of a Data Use Agreement). Code available upon request.
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. American Cancer Society . Cancer Facts & Figures 2021. Atlanta: American Cancer Society; 2021. Accessed July 29, 2021. Available at: https://www.cancer.org/research/cancer‐facts‐statistics/all‐cancer‐facts‐figures/cancer‐facts‐figures‐2021.html
- 3. Mizrahi JD, Surana R, Valle JW, Shroff RT. Pancreatic cancer. Lancet. 2020;395(10242):2008‐2020. [DOI] [PubMed] [Google Scholar]
- 4. Nevala‐Plagemann C, Hidalgo M, Garrido‐Laguna I. From state‐of‐the‐art treatments to novel therapies for advanced‐stage pancreatic cancer. Nat Rev Clin Oncol. 2020;17(2):108‐123. [DOI] [PubMed] [Google Scholar]
- 5. Janssen QP, Buettner S, Suker M, et al. Neoadjuvant FOLFIRINOX in patients with borderline resectable pancreatic cancer: a systematic review and patient‐level meta‐analysis. J Natl Cancer Inst. 2019;111(8):782‐794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Conroy T, Hammel P, Hebbar M, et al. FOLFIRINOX or gemcitabine as adjuvant therapy for pancreatic cancer. N Engl J Med. 2018;379(25):2395‐2406. [DOI] [PubMed] [Google Scholar]
- 7. Jin S, Pazdur R, Sridhara R. Re‐evaluating eligibility criteria for oncology clinical trials: analysis of investigational new drug applications in 2015. J Clin Oncol. 2017;35(33):3745‐3752. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Gerber DE, Laccetti AL, Xuan L, Halm EA, Pruitt SL. Impact of prior cancer on eligibility for lung cancer clinical trials. J Natl Cancer Inst. 2014;106(11):dju302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Murphy CC, Garcia S, Perez M, et al. Impact of prior cancer on enrollment in clinical trials of gastrointestinal cancer. In Preparation.
- 10. Unger JM, Vaidya R, Hershman DL, Minasian LM, Fleury ME. Systematic review and meta‐analysis of the magnitude of structural, clinical, and physician and patient barriers to cancer clinical trial participation. J Natl Cancer Inst. 2019;111(3):245‐255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Cancer clinical trial eligibility criteria: Patients with organ dysfunction or prior or concurrent malignancies . Washington D.C: US Department of Health and Human Services, Food and Drug Administration, Oncology Center of Excellence, Center for Drug Evaluation and Research and Center for Biologics Evaluation and Research. Draft guidance released March 2019. Accessed May 28, 2020. Available from: https://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/default.htm
- 12. Lichtman SM, Harvey RD, Damiette Smit MA, et al. Modernizing clinical trial eligibility criteria: recommendations of the American Society of Clinical Oncology‐friends of cancer research organ dysfunction, prior or concurrent malignancy, and comorbidities working group. J Clin Oncol. 2017;35:3753‐3759. [DOI] [PubMed] [Google Scholar]
- 13. Kim ES, Bruinooge SS, Roberts S, et al. Broadening eligibility criteria to make clinical trials more representative: American Society of Clinical Oncology and friends of cancer research joint research statement. J Clin Oncol. 2017;35(33):3737‐3744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Mooney M. Modernizing and broadening eligibility criteria initiative chrome‐extension://efaidnbmnnnibpcajpcglclefindmkaj/viewer.Html?Pdfurl=https%3A%2F%2Fctep.cancer.gov%2FprotocolDevelopment%2Fdocs%2FCTEP_Memorandum_Broaden_Eligibility_Criteria_Initiative.Pdf&clen=100012&chunk=true. Published 2021.
- 15. (CTEP) NCINCTEP . Broadening/Modernizing Eligiblity Criteria for National Cancer Institute (NCI) Sponsored Clinical Trials. Accessed January 13, 2022. chrome‐extension://efaidnbmnnnibpcajpcglclefindmkaj/viewer.html?pdfurl=https%3A%2F%2Fctep.cancer.gov%2FprotocolDevelopment%2Fdocs%2FCTEP_Broadened_Eligibility_Criteria_Guidance.pdf&clen=431762&chunk=true
- 16. Anderson C, Mayer DK, Nichols HB. Trends in the proportion of second or later primaries among all newly diagnosed malignant cancers. Cancer. 2021;127(15):2736‐2742. [DOI] [PubMed] [Google Scholar]
- 17. Glicksman AS, Fulton JP. Metachronous cancer. R I Med J (2013). 2013;96(4):41–44. [PubMed] [Google Scholar]
- 18. Weir HK, Johnson CJ, Thompson TD. The effect of multiple primary rules on population‐based cancer survival. Cancer Causes Control. 2013;24(6):1231‐1242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Ye Y, Neil AL, Wills KE, Venn AJ. Temporal trends in the risk of developing multiple primary cancers: a systematic review. BMC Cancer. 2016;16(1):849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER‐Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care. 2002;40(8 Suppl):IV‐3‐IV‐18. [DOI] [PubMed] [Google Scholar]
- 21. Laccetti AL, Pruitt SL, Xuan L, Halm EA, Gerber DE. Effect of prior cancer on outcomes in advanced lung cancer: implications for clinical trial eligibility and accrual. J Natl Cancer Inst. 2015;107(4):djv002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Pruitt SL, Zhu H, Heitjan DF, et al. Survival of women diagnosed with breast cancer and who have survived a previous cancer. Breast Cancer Res Treat. 2021;187:853‐865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Pruitt SL, Gerber DE, Zhu H, et al. Survival of patients newly diagnosed with colorectal cancer and with a history of previous cancer. Online ahead of print. Cancer Med. 2021;10:4752‐4767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Koroukian SM, Dahman B, Copeland G, Bradley CJ. The utility of the state buy‐in variable in the Medicare denominator file to identify dually eligible Medicare‐Medicaid beneficiaries: a validation study. Health Serv Res. 2010;45(1):265‐282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Nieman CL, Pitman KT, Tufaro AP, Eisele DW, Frick KD, Gourin CG. The effect of frailty on short‐term outcomes after head and neck cancer surgery. Laryngoscope. 2018;128(1):102‐110. [DOI] [PubMed] [Google Scholar]
- 26. Nass Sharyl J MHLaMJ. National Cancer Institute, division of Cancer Control & Population Sciences, healthcare delivery Reserach program. Comorbidity SAS macro (2014 version). Accessed on March 5, 2019. 2010. Available Here: https://healthcaredelivery.cancer.gov/seermedicare/considerations/macro‐2014.html [Google Scholar]
- 27. Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD‐9‐CM and ICD‐10 administrative data. Med Care. 2005;43(11):1130‐1139. [DOI] [PubMed] [Google Scholar]
- 28. R Core Team (2017) . R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R‐project.org/. Published 2017. Accessed 11/17/2020, 2020.
- 29. Al‐Husseini MJ, Saad AM, Turk T, Tabash MA, Abdel‐Rahman O. Impact of prior malignancy on survival outcomes of stage IV pancreatic adenocarcinoma: SEER‐based cohort. J Gastrointest Cancer. 2019;50(4):794‐800. [DOI] [PubMed] [Google Scholar]
- 30. Zhou H, Huang Y, Qiu Z, et al. Impact of prior cancer history on the overall survival of patients newly diagnosed with cancer: a pan‐cancer analysis of the SEER database. Int J Cancer. 2018;143(7):1569‐1577. [DOI] [PubMed] [Google Scholar]
- 31. Xea H. The impact of a history of cancer on pancreatic ductal adenocarcinoma survival. United European Gastroenterol J. 2018;6(6):888‐894. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. He C, Zhang Y, Cai Z, Lin X. Effect of prior cancer on survival outcomes for patients with pancreatic adenocarcinoma: a propensity score analysis. BMC Cancer. 2019;19(1):509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Adamo MB, Dickie L, Ruhl J. SEER program coding and staging manual 2018. National Cancer Institute, Bethesda, MD 20892. US Department of Health and Human Services, National Institutes of Health, National Cancer Institute. Accessed on March 9, 2020. 11/17/2020 2018. Available at: https://seer.cancer.gov/manuals/2018/SPCSM_2018_maindoc.pdf [Google Scholar]
- 34. Murphy CC, Gerber DE, Pruitt SL. Prior cancer among persons newly diagnosed with cancer: an initial report from the surveillance, Epidemiology, and End Results Program. JAMA Oncology. 2018;4(6):832‐836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Noone AM, Lund JL, Mariotto A, et al. Comparison of SEER treatment data with Medicare claims. Med Care. 2016;54(9):e55‐e64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Alliance for Clinical Trials in Oncology (Alliance A021806): A phase III trial of perioperative versus adjuvant chemotherapy for resectable pancreatic cancer. ClinicalTrials.gov Identifier: NCT04340141. Study start date: July 1, 2020. Accessed on April 6, 2021. Available at: https://clinicaltrials.gov/ct2/show/NCT04340141?titles=A+phase+III+trial+of+perioperative+versus+adjuvant+chemotherapy+for+resectable+pancreatic+cancer&draw=2&rank=1
- 37. Riner AN, Girma S, Vudatha V, et al. Eligibility criteria perpetuate disparities in enrollment and participation of black patients in pancreatic cancer clinical trials. J Clin Oncol. 2022;JCO2102492. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1‐S3
Data Availability Statement
This study used SEER‐MEDICARE data. The Centers for Medicare and Medicaid Services do not allow the redistribution of their data by researchers. SEER‐MEDICARE data are distinct from the publicly available SEER database, and can be obtained by researchers, by following the process described on https://healthcaredelivery.cancer.gov/seermedicare/obtain/requests.html (access requirements include Institutional Review Board approval and the completion of a Data Use Agreement). Code available upon request.