Abstract
Background
It is unclear whether results from recent trials of resectable pancreatic ductal adenocarcinoma (PDAC) are generalizable to older patients, who are underrepresented. We aimed to evaluate outcomes of surgery and of neoadjuvant and adjuvant therapy in older patients with resectable PDAC.
Patients and Methods
We included patients aged ≥65 years with upfront resectable PDAC from a prospectively maintained pancreatic cancer registry from 2007 to 2016. Patients were stratified into ages 65–75 and 75+ years. Overall survival (OS) was assessed in treatment comparisons: (A) surgery (n = 636) versus nonsurgical (n = 178), (B) neoadjuvant therapy (n = 139) versus upfront surgery (n = 497), and (C) adjuvant therapy (n = 379) versus surgery alone (n = 118). We compared neoadjuvant (n = 139) versus adjuvant therapy (n = 379) in an exploratory analysis.
Results
Nine hundred and three patients had a median age of 73.7 (range, 65–96.6) years. Median OS was 26.6 versus 11.9 months (adjusted hazard ratio [HRadj], 0.4; 95% confidence interval [CI], 0.31–0.52; p < .001) in Comparison A groups, 30.7 versus 25.8 months (HRadj, 0.69; 95% CI, 0.49–0.96; p = .03) in Comparison B groups, and 26.9 versus 17.4 months (HRadj, 0.62; 95% CI, 0.44–0.88; p = .008) in Comparison C groups, respectively. OS benefit in these treatment comparisons was present in age group 75+ with HRadj 0.24 (95% CI, 0.16–0.36; p < .001) in Comparison A and HRadj 0.52 (95% CI, 0.27–1; p = .049) in Comparison B, but not in Comparison C with HRadj 0.68 (95% CI, 0.43–1.08; p = .1). Statistically comparable median OS of patients who received neoadjuvant or adjuvant therapy stratified by age groups was observed.
Conclusion
Older patients with resectable PDAC who received surgery, neoadjuvant therapy, or adjuvant therapy appeared to have improved survival outcomes compared with those who did not receive such treatment.
Implications for Practice
Older patients with resectable pancreatic ductal adenocarcinoma (PDAC) in general are underrepresented in large clinical trials and less well studied in terms of the role of surgery, neoadjuvant therapy, and adjuvant therapy. This study collected data on older patients with resectable PDAC from a prospectively maintained single‐institutional pancreatic cancer registry of a tertiary referral center from 2007 to 2016. It was found that, with multidisciplinary evaluation, older patients with resectable PDAC who received surgery, neoadjuvant therapy, or adjuvant therapy appeared to have improved survival outcomes compared with those who did not receive such treatment. These results are of substantial importance to practitioners who treat older patients, who are traditionally underrepresented in most clinical trials.
Keywords: Resectable pancreatic cancer, Neoadjuvant, Adjuvant, Surgery
Short abstract
Most studies that support the benefit of perioperative therapy for resectable pancreatic ductal adenocarcinoma (PDAC) are based on younger patients. This study evaluated the role of surgery, neoadjuvant therapy, and adjuvant therapy in older patients with anatomically resectable PDAC to test the hypothesis that older patients with resectable PDAC who received surgery, neoadjuvant therapy, or adjuvant therapy could have improved survival outcomes compared with those who did not receive such treatments.
Introduction
Pancreatic ductal adenocarcinoma (PDAC) is currently the third leading cause of cancer death. PDAC is a disease of older people, with median age of 70 years at diagnosis in the U.S. [1]. Despite recent advances in the management of PDAC with the use of combination chemotherapy, patients with advanced PDAC still have poor prognosis [2]. In addition, poor performance status and other medical comorbidities in older patients bring more difficulties and complexities to their management [3].
Surgical resection remains the only potentially curative therapy for resectable PDAC [4]. For patients older than 80 years of age, the feasibility of complex pancreatic resection was demonstrated with similar overall survival (OS) benefit compared with younger patients in a retrospective study [5]. Such small retrospective studies, however, were conducted prior to the introduction of combination chemotherapy and the routine use of either neoadjuvant or adjuvant therapy [5]. In contrast, a more recent population study concluded that survival benefit associated with surgical resection was very small for patients older than 80 years; chemotherapy alone was suggested as an alternative to surgery [6]. Survival benefit of adjuvant therapy has been clearly demonstrated in a number of randomized clinical trials [7, 8, 9]. However, these trials typically enrolled a much younger patient population with median age between 61 and 64 years. As a result, patients older than 75 were underrepresented in these trials, and survival benefit of adjuvant therapy for older patients was either not reported or did not reach statistical significance in these studies [7, 8, 9, 10]. The role of neoadjuvant therapy for resectable PDAC remains controversial even in younger patients. A recent study demonstrated survival benefit of neoadjuvant gemcitabine plus S‐1 in resectable PDAC [11, 12]. This observation was not consistent with the results from previously reported small trials [13, 14]. A recent population‐based study using the National Cancer Database showed survival benefit of neoadjuvant therapy in patients with resectable PDAC. Patients older than 75 years in the matched cohort constituted only 14%, and the effect of age was not scrutinized because of age match [15].
Most high‐quality studies that support the benefit of perioperative therapy for resectable PDAC are based on younger patients. It is unclear whether these results are generalizable to older patients. We thus aimed to evaluate the role of surgery, neoadjuvant therapy, and adjuvant therapy in older patients (≥65 years) with anatomically resectable PDAC to test the hypothesis that older patients with resectable PDAC who received surgery, neoadjuvant therapy, or adjuvant therapy could have improved survival outcomes compared with those who did not receive such treatments.
Materials and Methods
Patients
This study was approved by the Mayo Clinic institutional review board. Patients were identified from the pancreatic cancer patient registry of the Mayo Clinic Specialized Program of Research Excellence (SPORE) in Pancreatic Cancer (P50 CA102701), a prospectively maintained patient cohort that enrolls patients with pancreatic cancer who are seen at Mayo Clinic. Patients have been consented since 2002 to participate in the SPORE Registry, usually at the time of diagnosis or first visit with long‐term follow‐up [16]. For this study, the SPORE Registry was searched to identify eligible patients; criteria included age ≥ 65 years at the time of diagnosis and pathologically confirmed anatomically upfront resectable (stage I–II) PDAC between 2007 and 2016 (to allow inclusion of the use of modern therapies and adequate follow‐up). We did not focus our analysis on patients younger than 65 because the outcome of these patients with PDAC has been extensively reported before. We felt that those findings were sufficient as a historical control. The clinical stages were defined based on radiology reports according to clinical staging system, mainly American Joint Committee on Cancer, 7th edition, for pancreatic cancer. The clinical resectability was defined by the consulting surgeons based on clinical stages defined by the radiologist and patients’ clinical characteristics at the time of diagnosis. Patients with unknown treatment information outside of Mayo Clinic were excluded. Clinical variables collected for this study included but were not limited to age, gender, comorbidities, Eastern Cooperative Oncology Group (ECOG) performance status, stage, and carbohydrate antigen 19‐9 (CA 19‐9). The median was used to categorize CA 19‐9 into low and high groups. The referral status is yes if patients were coded as being referred for diagnosis or treatment at Mayo Clinic or initial diagnosis predated first visit to Mayo Clinic.
Age and Treatment Comparison Groups
We stratified patients by age groups: 65–75 years and 75+ years. The optimal age cutoff point was determined (supplemental online Fig. 1) using the maximally selected rank statistics from the “maxstat” R package (https://cran.r-project.org/web/packages/maxstat/index.html) [17]. It was also consistent with the age cutoff in exclusion criteria of several randomized trials [18]. Patients in this study were also stratified into three comparisons based on different treatment strategies (Fig. 1): (a) patients who received curative‐intent surgical resection and those who received nonsurgical management, (b) patients who received neoadjuvant therapy followed by surgical resection and those who had upfront surgery, and (c) patients who received upfront surgical resection followed by adjuvant therapy and those who received upfront surgery alone. We also did an exploratory analysis to compare the outcome of patients who received neoadjuvant treatment versus those who received adjuvant treatment.
Figure 1.
Three treatment comparisons based on surgery and perioperative treatment modalities (CONSORT diagram). Abbreviation: PDAC, pancreatic ductal adenocarcinoma.
Statistical Analysis
Continuous variables were summarized as mean, median, standard deviation, or interquartile ranges (IQRs) and compared with t test or Wilcoxon rank‐sum test. Categorical variables were summarized as frequency counts or percentages and compared with Fisher's exact test. Comorbidity score was calculated as Charlson/Deyo Score. Age was not considered as part of this comorbidity score calculation because it was evaluated as a covariate by itself in survival analyses. Follow‐up time was defined as the time interval between the date of diagnosis and the date of last follow‐up. OS was defined as the time interval between the date of diagnosis and the date of last follow‐up (censored if alive) or death. Time‐to‐event data were summarized using Kaplan‐Meier method and compared using log‐rank test. Univariate and multivariable survival analyses for the prognostic factors were performed using standard or time‐dependent Cox proportional hazards model with surgery status as a time‐dependent variable. Covariates of multivariable Cox proportional hazards models included different treatment groups, age, comorbidity score, ECOG performance status, stage, CA 19‐9, gender, referral status, and year of diagnosis. We included common prognostic factors oncologists may use routinely for clinical decision making. Referral status and year of diagnosis were included to adjust for potential referral and selection biases. The treatment effect was also evaluated in subgroup analyses stratified by age and other covariates. We performed moderation analyses with interaction terms for age and treatment. The results were summarized as hazard ratios (HRs), 95% confidence intervals (CIs) and p values and presented in forest plots. The proportionality assumption was assessed graphically using log (−log) plots and quantitatively using the Z statistic. Our preplanned hypothesis testing includes evaluation of effects of three treatment comparisons in overall patient population and each age group. The number of hypothesis testing is preplanned and limited. We did not draw any conclusion from other subgroup analyses. For the above reasons, we did not adjust p values for multiple testing. All tests were two sided and performed in R version 3.6.0 (R Foundation for Statistical Computing, Vienna, Austria).
Results
Baseline Characteristics
A total of 903 patients with upfront resectable PDAC diagnosed between 2007 and 2016 were included and further categorized into three treatment comparisons (Fig. 1). Four hundred seven patients (78.4%) in age group 65–75 died at the time of last follow‐up compared with 320 patients (83.3%) in age group 75+ (p = .08). Median follow‐up time was 17.7 (IQR, 9.7–32.4) months for the entire cohort. In contrast to age group 65–75, patients in age group 75+ had significantly higher comorbidity scores (p < .001), poor ECOG performance status (p < .001), more stage I disease (p = .001), higher CA 19‐9 level at the time of diagnosis (p = .02), and fewer being referred (p < .001). Patients in age group 75+ received less curative‐intent surgery (p < .001), neoadjuvant therapy (p = .009), or adjuvant therapy (p < .001; Table 1).
Table 1.
Baseline characteristics of patients with resectable pancreatic ductal adenocarcinoma
| Variables, levels | Entire cohort (n = 903), n (%) | Age 65–75 (n = 519), n (%) | Age 75+ (n = 384), n (%) | p value a | Missing, % |
|---|---|---|---|---|---|
| Age, median (IQR), years | 73.7 (69.4–78.8) | 69.9 (67.8–72.3) | 79.6 (77.0–82.8) | <.001 | 0 |
| Gender | |||||
| Female | 405 (44.9) | 228 (43.9) | 177 (46.1) | .563 | 0 |
| Male | 498 (55.1) | 291 (56.1) | 207 (53.9) | ||
| Race | |||||
| White | 834 (96.3) | 479 (96.0) | 355 (96.7) | .776 | 4.1 |
| Asian | 9 (1.0) | 7 (1.4) | 2 (0.5) | ||
| Hispanic | 8 (0.9) | 5 (1.0) | 3 (0.8) | ||
| Black | 7 (0.8) | 4 (0.8) | 3 (0.8) | ||
| Other | 8 (0.9) | 4 (0.8) | 4 (1.1) | ||
| Comorbidity | |||||
| 0 | 382 (42.3) | 249 (48.0) | 133 (34.6) | <.001 | 0 |
| 1 | 234 (25.9) | 139 (26.8) | 95 (24.7) | ||
| 2 | 127 (14.1) | 58 (11.2) | 69 (18.0) | ||
| 3 | 80 (8.9) | 39 (7.5) | 41 (10.7) | ||
| 4 | 43 (4.8) | 20 (3.9) | 23 (6.0) | ||
| 5 | 9 (1.0) | 1 (0.2) | 8 (2.1) | ||
| 6 | 19 (2.1) | 9 (1.7) | 10 (2.6) | ||
| 7 | 6 (0.7) | 3 (0.6) | 3 (0.8) | ||
| 9 | 2 (0.2) | 0 (0.0) | 2 (0.5) | ||
| 14 | 1 (0.1) | 1 (0.2) | 0 (0.0) | ||
| ECOG performance status | |||||
| 0 | 379 (54.2) | 252 (60.7) | 127 (44.7) | <.001 | 22.6 |
| 1 | 221 (31.6) | 118 (28.4) | 103 (36.3) | ||
| 2 | 73 (10.4) | 34 (8.2) | 39 (13.7) | ||
| 3 | 22 (3.1) | 8 (1.9) | 14 (4.9) | ||
| 4 | 4 (0.6) | 3 (0.7) | 1 (0.4) | ||
| Tumor location | |||||
| Head | 526 (69.4) | 288 (67.8) | 238 (71.5) | .308 | 16.1 |
| Other | 232 (30.6) | 137 (32.2) | 95 (28.5) | ||
| Grade | |||||
| 1 | 36 (6.3) | 22 (6.5) | 14 (6.0) | .598 | 36.8 |
| 2 | 188 (32.9) | 116 (34.2) | 72 (31.0) | ||
| 3 | 309 (54.1) | 182 (53.7) | 127 (54.7) | ||
| 4 | 38 (6.7) | 19 (5.6) | 19 (8.2) | ||
| Stage | |||||
| I | 163 (18.1) | 75 (14.5) | 88 (22.9) | .001 | 0 |
| II | 740 (81.9) | 444 (85.5) | 296 (77.1) | ||
| Surgery types | |||||
| Whipple | 444 (49.2) | 279 (53.8) | 165 (43.0) | <.001 | 0 |
| Distal pancreatectomy | 165 (18.3) | 102 (19.7) | 63 (16.4) | ||
| Total pancreatectomy | 27 (3.0) | 17 (3.3) | 10 (2.6) | ||
| Procedure aborted | 17 (1.9) | 12 (2.3) | 5 (1.3) | ||
| Palliative surgery | 3 (0.3) | 2 (0.4) | 1 (0.3) | ||
| Other surgery | 8 (0.9) | 6 (1.2) | 2 (0.5) | ||
| No | 239 (26.5) | 101 (19.5) | 138 (35.9) | ||
| Neoadjuvant therapy types | |||||
| Chemotherapy | 107 (11.8) | 70 (13.5) | 37 (9.6) | .009 | 0 |
| Radiation | 1 (0.1) | 1 (0.2) | 0 (0.0) | ||
| Chemotherapy and radiation | 139 (15.4) | 93 (17.9) | 46 (12.0) | ||
| No | 656 (72.6) | 355 (68.4) | 301 (78.4) | ||
| Adjuvant therapy types | |||||
| Chemotherapy | 235 (26.0) | 149 (28.7) | 86 (22.4) | <.001 | 0 |
| Radiation | 2 (0.2) | 1 (0.2) | 1 (0.3) | ||
| Chemoradiation | 167 (18.5) | 124 (23.9) | 43 (11.2) | ||
| No | 499 (55.3) | 245 (47.2) | 254 (66.1) | ||
| CA 19‐9, median (IQR), units/mL | 178 (44–657) | 144 (32–581) | 210 (54–756) | .019 | 20.7 |
| Referral status | |||||
| No | 495 (58.6) | 236 (48.8) | 259 (71.7) | <.001 | 6.4 |
| Yes | 350 (41.4) | 248 (51.2) | 102 (28.3) | ||
| Year of diagnosis | |||||
| 2007–2011 | 404 (44.7) | 222 (42.8) | 182 (47.4) | .189 | 0 |
| 2012–2016 | 499 (55.3) | 297 (57.2) | 202 (52.6) |
p value from Fisher's exact test for categorical variables or from Wilcoxon rank‐sum test of medians.
Abbreviations: CA 19‐9, carbohydrate antigen 19‐9; ECOG, Eastern Cooperative Oncology Group; IQR, interquartile range.
As shown in Table 2, the surgery versus nonsurgical comparison group (n = 814) included 636 patients (78.1%) in the surgery group and 178 patients (21.9%) in the nonsurgical group. The use of surgical management was associated with stage II disease (85.2%; p = .01), lower CA 19‐9 level (54.7%; p < .001), and diagnosis in 2007–2011 (49.1%; p < .001). Neoadjuvant therapy versus upfront surgery comparison group (n = 636) included 139 patients (21.9%) in the neoadjuvant therapy group and 497 patients (78.1%) in the upfront surgery group. The use of neoadjuvant therapy was associated with diagnosis in 2012–2016 (82.0%; p < .001). Adjuvant therapy versus surgery comparison group (n = 497) included 379 patients (76.3%) in the adjuvant therapy group and 118 patients (23.7%) in the surgery alone group. The use of adjuvant therapy was associated with younger age (68.1%; p < .001), better ECOG performance status (91.5%; p = .01), stage II disease (88.9%; p = .02), and diagnosis in 2012–2016 (44.9%; p = .046).
Table 2.
Baseline characteristics of patients in three treatment comparisons
| Variables, levels | Surgery vs. nonsurgical (n = 814) | Neoadjuvant therapy vs. upfront surgery (n = 636) | Adjuvant therapy vs. surgery alone (n = 497) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Nonsurgical (n = 178), n (%) | Surgery (n = 636), n (%) | p value | Neoadjuvant (n = 139), n (%) | Upfront surgery (n = 497), n (%) | p value | Adjuvant (n = 379), n (%) | Surgery alone (n = 118), n (%) | p value | |
| Age | |||||||||
| 65–75 | 100 (56.2) | 398 (62.6) | .144 | 96 (69.1) | 302 (60.8) | .091 | 258 (68.1) | 44 (37.3) | <.001 |
| 75+ | 78 (43.8) | 238 (37.4) | 43 (30.9) | 195 (39.2) | 121 (31.9) | 74 (62.7) | |||
| Comorbidity | |||||||||
| 0–1 | 122 (68.5) | 459 (72.2) | .393 | 102 (73.4) | 357 (71.8) | .800 | 276 (72.8) | 81 (68.6) | .445 |
| 2+ | 56 (31.5) | 177 (27.8) | 37 (26.6) | 140 (28.2) | 103 (27.2) | 37 (31.4) | |||
| ECOG PS | |||||||||
| 0–1 | 131 (84.5) | 446 (90.5) | .055 | 101 (94.4) | 345 (89.4) | .169 | 281 (91.5) | 64 (81.0) | .012 |
| 2+ | 24 (15.5) | 47 (9.5) | 6 (5.6) | 41 (10.6) | 26 (8.5) | 15 (19.0) | |||
| Stage | |||||||||
| I | 41 (23.0) | 94 (14.8) | .012 | 28 (20.1) | 66 (13.3) | .060 | 42 (11.1) | 24 (20.3) | .015 |
| II | 137 (77.0) | 542 (85.2) | 111 (79.9) | 431 (86.7) | 337 (88.9) | 94 (79.7) | |||
| Gender | |||||||||
| Female | 82 (46.1) | 285 (44.8) | .832 | 64 (46.0) | 221 (44.5) | .815 | 170 (44.9) | 51 (43.2) | .837 |
| Male | 96 (53.9) | 351 (55.2) | 75 (54.0) | 276 (55.5) | 209 (55.1) | 67 (56.8) | |||
| CA 19‐9 | |||||||||
| Low | 54 (37.0) | 279 (54.7) | <.001 | 58 (47.5) | 221 (57.0) | .086 | 169 (58.1) | 52 (53.6) | .515 |
| High | 92 (63.0) | 231 (45.3) | 64 (52.5) | 167 (43.0) | 122 (41.9) | 45 (46.4) | |||
| Referral status | |||||||||
| No | 99 (59.6) | 318 (53.9) | .220 | 62 (47.7) | 256 (55.7) | .132 | 192 (54.1) | 64 (61.0) | .257 |
| Yes | 67 (40.4) | 272 (46.1) | 68 (52.3) | 204 (44.3) | 163 (45.9) | 41 (39.0) | |||
| Year of diagnosis | |||||||||
| 2007–2011 | 56 (31.5) | 312 (49.1) | <.001 | 25 (18.0) | 287 (57.7) | <.001 | 209 (55.1) | 78 (66.1) | .046 |
| 2012–2016 | 122 (68.5) | 324 (50.9) | 114 (82.0) | 210 (42.3) | 170 (44.9) | 40 (33.9) | |||
Abbreviations: CA 19‐9, carbohydrate antigen 19‐9; ECOG PS, Eastern Cooperative Oncology Group performance status.
Clinical Outcome
Surgery Versus Nonsurgical
In the surgery versus nonsurgical comparison group, older age, poor ECOG performance status, stage II disease, higher CA 19‐9 level, and diagnosis in 2007–2011 were associated with worse prognosis. Patients in the surgery group had significantly longer median OS than those in the nonsurgical group (26.6 vs. 11.9 months; p < .001). This finding was confirmed after adjusting for significant prognostic factors in time‐dependent multivariable survival model with an HR of 0.4 (95% CI, 0.31–0.52; p < .001) at the time and after surgery (Table 3). The results from a standard multivariable survival model were also consistent with this finding (supplemental online Fig. 2). Patients in both age groups had significant OS benefit from surgical management compared with nonsurgical management: (a) age group 65–75: HR, 0.57; 95% CI, 0.4–0.81; p = .002; (b) age group 75+: HR, 0.24; 95% CI, 0.16–0.36; p < .001 (supplemental online Fig. 3A; Fig. 2A). In addition, patients in age group 75+ had significantly more OS benefit from surgical management compared with age group 65–75 (p for age‐treatment interaction = .004). Of note, subgroup analyses demonstrated significant OS benefit of surgery compared with nonsurgical management in all subgroups (Fig. 2A).
Table 3.
Univariate and multivariable survival analyses in three treatment comparisons
| Variables, levels | Surgery vs. nonsurgical (n = 814), HR (95% CI, p) | Neoadjuvant therapy vs. upfront surgery (n = 636), HR (95% CI, p) | Adjuvant therapy vs. surgery alone (n = 497), HR (95% CI, p) | |||
|---|---|---|---|---|---|---|
| Univariate | Multivariable | Univariate | Multivariable | Univariate | Multivariable | |
| Age: 75+ vs. 65–75 | 1.16 (0.99–1.36, p = .060) | 1.28 (1.04–1.58, p = .018) | 1.08 (0.89–1.30, p = .435) | 1.10 (0.86–1.40, p = .453) | 1.13 (0.93–1.38, p = .230) | 1.07 (0.81–1.41, p = .631) |
| Comorbidity: 2+ vs. 0–1 | 0.96 (0.81–1.14, p = .641) | 0.99 (0.78–1.25, p = .920) | 0.89 (0.73–1.09, p = .264) | 1.01 (0.76–1.32, p = .970) | 0.90 (0.72–1.12, p = .341) | 1.03 (0.77–1.39, p = .839) |
| ECOG PS: 2+ vs. 0–1 | 1.63 (1.25–2.14, p < .001) | 1.98 (1.44–2.71, p < .001) | 1.55 (1.11–2.16, p = .010) | 2.07 (1.41–3.05, p < .001) | 1.59 (1.12–2.26, p = .009) | 2.02 (1.33–3.06, p = .001) |
| Stage: II vs. I | 1.61 (1.28–2.02, p < .001) | 1.62 (1.21–2.16, p = .001) | 2.14 (1.59–2.87, p < .001) | 2.03 (1.39–2.99, p < .001) | 2.17 (1.56–3.03, p < .001) | 2.21 (1.43–3.41, p < .001) |
| Gender: male vs. female | 1.06 (0.91–1.24, p = .447) | 1.15 (0.94–1.42, p = .175) | 1.12 (0.94–1.35, p = .204) | 1.16 (0.91–1.47, p = .239) | 1.11 (0.91–1.35, p = .314) | 1.27 (0.96–1.67, p = .091) |
| CA 19‐9: high vs. low | 1.56 (1.31–1.86, p < .001) | 1.38 (1.12–1.69, p = .002) | 1.47 (1.21–1.81, p < .001) | 1.48 (1.17–1.88, p = .001) | 1.62 (1.30–2.02, p < .001) | 1.56 (1.20–2.04, p = .001) |
| Referral status: yes vs. no | 1.19 (1.02–1.40, p = .032) | 1.23 (1.00–1.52, p = .048) | 1.34 (1.11–1.61, p = .002) | 1.36 (1.06–1.74, p = .014) | 1.44 (1.18–1.77, p < .001) | 1.49 (1.14–1.97, p = .004) |
| Year of diagnosis: 2012–2016 vs. 2007–2011 | 0.91 (0.78–1.07, p = .255) | 0.76 (0.61–0.94, p = .011) | 0.83 (0.69–1, p = .055) | 0.88 (0.67–1.14, p = .324) | 0.88 (0.71–1.08, p = .223) | 0.94 (0.71–1.25, p = .678) |
| Treatment | 0.41 (0.33–0.50, p < .001) | 0.40 (0.31–0.52, p < .001) | 0.78 (0.61–0.99, p = .042) | 0.69 (0.49–0.96, p = .028) | 0.76 (0.61–0.96, p = .020) | 0.62 (0.44–0.88, p = .008) |
Abbreviations: CA 19‐9, carbohydrate antigen 19‐9; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio.
Figure 2.
Multivariable survival analyses of subgroups in three treatment comparisons. (A): Surgery versus nonsurgical management. (B): Neoadjuvant therapy versus upfront surgery. (C): Adjuvant therapy versus surgery alone. Abbreviations: CA 19‐9, carbohydrate antigen 19‐9; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; OS, overall survival.
Neoadjuvant Therapy Versus Upfront Surgery
In neoadjuvant therapy versus upfront surgery comparison group, poor ECOG performance status, stage II disease, higher CA 19‐9 level, and being referred were associated with worse prognosis. Patients in the neoadjuvant therapy group had significantly longer median OS than those in the upfront surgery group (30.7 vs. 25.8 months; p = .04). This finding was confirmed after adjusting for significant prognostic factors in multivariable survival analysis with an HR of 0.69 (95% CI, 0.49–0.96; p = .03; Table 3). Interestingly, patients in age group 75+ had significant OS benefit from neoadjuvant therapy with an HR of 0.52 (95% CI, 0.27–1.0; p = .049) in contrast to patients in age group 65–75, who did not reach statistical significance with an HR of 0.77 (95% CI, 0.52–1.14; p = .2; supplemental online Fig. 3B; Fig. 2B). However, the extent of OS benefit from neoadjuvant therapy was not significantly different between age group 65–75 and age group 75+ (p for age‐treatment interaction = .2). Of note, subgroup analyses demonstrated significant OS benefit of neoadjuvant therapy compared with upfront surgery in patients with stage II disease, higher CA 19‐9 level, and diagnosis in 2012–2016 (Fig. 2B).
Adjuvant Therapy Versus Surgery Alone
In adjuvant therapy versus surgery alone comparison group, poor ECOG performance status, stage II disease, higher CA 19‐9 level, and being referred were associated with worse prognosis. Patients in the adjuvant therapy group had significantly longer median OS than those in the surgery alone group (26.9 vs. 17.4 months; p = .02). This finding was confirmed after adjusting for significant prognostic factors in multivariable survival analysis with an HR of 0.62 (95% CI, 0.44–0.88; p = .008; Table 3). Patients in age group 65–75 had significant OS benefit from adjuvant therapy with an HR of 0.4 (95% CI, 0.23–0.69; p = .001). Patients in age group 75+ appeared to have OS benefit from adjuvant therapy with an HR of 0.68 (95% CI, 0.43–1.08; p = .1; supplemental online Fig. 3C; Fig. 2C), but this did not reach statistical significance. However, the extent of OS benefit from adjuvant therapy was not significantly different between age group 65–75 and 75+ (p for age‐treatment interaction = .2). Of note, subgroup analyses demonstrated significant OS benefit of adjuvant therapy compared with surgery alone in patients with less comorbidity, better ECOG performance status, stage II disease, higher CA 19‐9 level, male patients, and referred patients (Fig. 2C).
Neoadjuvant Versus Adjuvant Therapy
As shown in Figure 3, in age group 75+, neoadjuvant therapy appeared to be associated with longer median OS compared with adjuvant therapy (39.1 vs. 27.7 months, p = .08), but this did not reach statistical significance. In contrast, for age group 65–75, neoadjuvant therapy was associated with similar median OS to adjuvant therapy (27.8 vs. 26.8 months, p = .5).
Figure 3.
Overall survival of patients who received neoadjuvant or adjuvant therapy stratified by age groups. (A): Neoadjuvant versus adjuvant therapy in patients of age group 75+ (n = 164). (B): Neoadjuvant versus adjuvant therapy in patients of age group 65–75 (n = 354).
Discussion
In contrast to previous small retrospective studies [19], this large cohort study on older patients with anatomically resectable PDAC from a prospectively maintained cancer registry helps address the literature gap of the roles of surgery and perioperative therapy in this understudied patient population [20]. Our study specifically focused on the patient population traditionally defined as older (age ≥ 65 years). In particular, this study aimed to analyze the outcomes of patients older than 75 years who were either explicitly excluded (e.g., PRODIGE/ACCORD 4) or underrepresented in PDAC clinical trials [21, 22]. As such, our findings here are particularly clinically relevant, pertinent, and valuable to guide the clinical management of these patients with increasing incidence of resectable PDAC [3]. In addition, our patient population was representative as common prognostic factors found in our study including age, performance status, stage, and tumor burden reflected by CA 19‐9 were all consistent with our clinical experience.
Our study provided evidence that carefully selected patients older than 75 years with resectable PDAC benefited from surgery. Immortal time bias inherent in this treatment comparison group was accounted for with time‐dependent survival model. The findings reinforced the previously reported observations in both retrospective and population‐based studies in terms of outcome of surgical resection in this patient population [5, 6, 19]. Of note, the median OS of both surgery and nonsurgical groups in our study appeared to be longer than those in previous population studies as 15 versus 10 months, respectively [6]. This discrepancy is possibly due to the differences in surgical management at a tertiary referral center compared with the community, selection bias, and/or recent advances of perioperative chemotherapy regimens.
Although neoadjuvant therapy is increasingly used, the exact role, sequence, and composition remain to be determined. Previous prospective trials suffered from the use of nonstandard chemotherapy regimens [11, 12] and inadequate statistical power because of early termination [13, 14]. Our study demonstrates a statistically significant OS benefit of neoadjuvant therapy compared with upfront surgery in patients aged ≥65 years, specifically patients aged >75 years after adjustment for other common prognostic factors. This was with the caveat that patients in the neoadjuvant therapy group were those who received both neoadjuvant therapy and surgery instead of those who received therapy with neoadjuvant intent. Older patients may still benefit from neoadjuvant therapy as with younger patients, in terms of becoming candidate for surgery with improved R0 resection rate [23].
We demonstrated that older patients had significantly longer OS from adjuvant therapy compared with surgery alone, consistent with that reported in the CONKO‐001 trial [8]. However, our study appeared to show survival benefit from adjuvant therapy in patients older than 75 years, but this did not reach statistical significance. Potential reasons include complications and longer recovery from the surgery that may delay or stop the older patients from receiving adjuvant therapy. In addition, older patients are less likely to receive more aggressive and dose‐intensive regimens compared with younger patients [24]. In contrast, patients older than 75 years seemed to benefit more from neoadjuvant therapy compared with adjuvant therapy. This finding is provocative, suggesting that neoadjuvant therapy may be preferred in patients older than 75 years with resectable PDAC, but is likely due to selection bias, in which oncologists choose the most “fit” patients to receive neoadjuvant therapy. It should be a topic of future clinical trials.
Limitations
This study was limited by the pitfalls such as selection bias typical of a single‐institutional study and the heterogeneity of treatment modalities used precludes meaningful analysis of the impact of specific chemotherapy regimens and radiation. Our patient population comprised cases encountered in the multidisciplinary clinic of a large tertiary referral center with the existence of referral bias. For example, the referred patients who travelled from a distance may have less comorbidity than local patients. To address this, referral status information was collected to the best of our ability and was accounted for in multivariable survival analyses. Almost all of the patients included in our study were white without available information on socioeconomic status such as education status, income, marital status, etc. This limited our ability to evaluate the effect of these factors. In addition, we did not collect information on different chemotherapy regimens and lacked data about the use of palliative care and patient‐reported outcomes including disease‐related symptoms and quality of life, which could be important for geriatric oncology population.
Conclusion
This article reports a real‐world experience on the use of contemporary treatment strategies in older patients with resectable PDAC who are underrepresented in standard‐setting randomized trials. In addition, the recent 10 years analyzed here encompass the availability of highly active systemic chemotherapy regimens, increasing use of neoadjuvant therapy, and improving supportive care. With the limitations discussed, our study is meant to be hypothesis generating. Ongoing and further prospective and, if possible, randomized trials may provide more definitive answers. Despite the lack of some granularity, our study showed that certain older patients with resectable PDAC can benefit from surgery and from neoadjuvant and adjuvant therapy.
Author Contributions
Conception/design: Hao Xie, Jun Yin, Mark J. Truty, Gloria M. Petersen, Aminah Jatoi, Joleen M. Hubbard, Wen Wee Ma
Provision of study material or patients: Hao Xie, Amit Mahipal, Robert R. McWilliams, Mark J. Truty, Tanios S. Bekaii‐Saab, Gloria M. Petersen, Aminah Jatoi, Joleen M. Hubbard, Wen Wee Ma
Collection and/or assembly of data: Hao Xie, Junjia Liu, Jun Yin, John R. Ogden, Gloria M. Petersen, Wen Wee Ma
Data analysis and interpretation: Hao Xie, Junjia Liu, Jun Yin, Wen Wee Ma
Manuscript writing: Hao Xie, Junjia Liu, Jun Yin, Wen Wee Ma
Final approval of manuscript: Hao Xie, Junjia Liu, Jun Yin, John R. Ogden, Amit Mahipal, Robert R. McWilliams, Mark J. Truty, Tanios S. Bekaii‐Saab, Gloria M. Petersen, Aminah Jatoi, Joleen M. Hubbard, Wen Wee Ma
Disclosures
Robert R. McWilliams: Bristol‐Meyer Squib, Merck (RF). The other authors indicated no financial relationships.
(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board
Supporting information
See http://www.TheOncologist.com for supplemental material available online.
Appendix S1. Figures.
Acknowledgments
This research was supported by National Cancer Institute grant P50CA102701, Mayo Clinic Specialized Program of Research Excellence in Pancreatic Cancer.
Disclosures of potential conflicts of interest may be found at the end of this article.
No part of this article may be reproduced, stored, or transmitted in any form or for any means without the prior permission in writing from the copyright holder. For information on purchasing reprints contact Commercialreprints@wiley.com. For permission information contact permissions@wiley.com.
References
- 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2019. CA Cancer J Clin 2019;69:7–34. [DOI] [PubMed] [Google Scholar]
- 2. Hall BR, Cannon A, Atri P et al. Advanced pancreatic cancer: A meta‐analysis of clinical trials over thirty years. Oncotarget 2018;9:19396–19405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Higuera O, Ghanem I, Nasimi R et al. Management of pancreatic cancer in the elderly. World J Gastroenterol 2016;22:764–775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Ferrone CR, Brennan MF, Gonen M et al. Pancreatic adenocarcinoma: The actual 5‐year survivors. J Gastrointest Surg 2008;12:701–706. [DOI] [PubMed] [Google Scholar]
- 5. Khan S, Sclabas G, Lombardo KR et al. Pancreatoduodenectomy for ductal adenocarcinoma in the very elderly; is it safe and justified? J Gastrointest Surg 2010;14(11):1826–1831. [DOI] [PubMed] [Google Scholar]
- 6. Marmor S, Burke EE, Virnig BA et al. A comparative analysis of survival outcomes between pancreatectomy and chemotherapy for elderly patients with adenocarcinoma of the pancreas. Cancer 2016;122:3378–3385. [DOI] [PubMed] [Google Scholar]
- 7. Neoptolemos JP, Stocken DD, Friess H et al. A randomized trial of chemoradiotherapy and chemotherapy after resection of pancreatic cancer. N Engl J Med 2004;350:1200–1210. [DOI] [PubMed] [Google Scholar]
- 8. Oettle H, Neuhaus P, Hochhaus A et al. Adjuvant chemotherapy with gemcitabine and long‐term outcomes among patients with resected pancreatic cancer: The CONKO‐001 randomized trial. JAMA 2013;310:1473–1481. [DOI] [PubMed] [Google Scholar]
- 9. Conroy T, Hammel P, Hebbar M et al. FOLFIRINOX or gemcitabine as adjuvant therapy for pancreatic cancer. N Engl J Med 2018;379:2395–2406. [DOI] [PubMed] [Google Scholar]
- 10. Neoptolemos JP, Palmer DH, Ghaneh P et al. Comparison of adjuvant gemcitabine and capecitabine with gemcitabine monotherapy in patients with resected pancreatic cancer (ESPAC‐4): A multicentre, open‐label, randomised, phase 3 trial. Lancet 2017;389:1011–1024. [DOI] [PubMed] [Google Scholar]
- 11. Satoi S, Unno M, Motoi F et al. The effect of neoadjuvant chemotherapy with gemcitabine and S‐1 for resectable pancreatic cancer (randomized phase II/III trial; Prep‐02/JSAP‐05). J Clin Oncol 2019;37(suppl 15):4126s. [DOI] [PubMed] [Google Scholar]
- 12. Motoi F, Kosuge T, Ueno H et al. Randomized phase II/III trial of neoadjuvant chemotherapy with gemcitabine and S‐1 versus upfront surgery for resectable pancreatic cancer (Prep‐02/JSAP05). Jpn J Clin Oncol 2019;49:190–194. [DOI] [PubMed] [Google Scholar]
- 13. Golcher H, Brunner TB, Witzigmann H et al. Neoadjuvant chemoradiation therapy with gemcitabine/cisplatin and surgery versus immediate surgery in resectable pancreatic cancer: Results of the first prospective randomized phase II trial. Strahlenther Onkol 2015;191:7–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Casadei R, Di Marco M, Ricci C et al. Neoadjuvant chemoradiotherapy and surgery versus surgery alone in resectable pancreatic cancer: A single‐center prospective, randomized, controlled trial which failed to achieve accrual targets. J Gastrointest Surg 2015;19:1802–1812. [DOI] [PubMed] [Google Scholar]
- 15. Mokdad AA, Minter RM, Zhu H et al. Neoadjuvant therapy followed by resection versus upfront resection for resectable pancreatic cancer: A propensity score matched analysis. J Clin Oncol 2017;35:515–522. [DOI] [PubMed] [Google Scholar]
- 16. Hu C, Hart SN, Polley EC et al. Association between inherited germline mutations in cancer predisposition genes and risk of pancreatic cancer. JAMA 2018;319:2401–2409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Lausen B, Schumacher M. Maximally selected rank statistics. Biometrics 1992;48:73. [Google Scholar]
- 18. White MN, Dotan E, Catalano PJ et al. Advanced pancreatic cancer clinical trials: The continued underrepresentation of older patients. J Geriatr Oncol 2019;10:540–546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Barbas AS, Turley RS, Ceppa EP et al. Comparison of outcomes and the use of multimodality therapy in young and elderly people undergoing surgical resection of pancreatic cancer. J Am Geriatr Soc 2012;60:344–350. [DOI] [PubMed] [Google Scholar]
- 20. Scher KS, Hurria A. Under‐representation of older adults in cancer registration trials: Known problem, little progress. J Clin Oncol 2012;30:2036–2038. [DOI] [PubMed] [Google Scholar]
- 21. Lara PN, Higdon R, Lim N et al. Prospective evaluation of cancer clinical trial accrual patterns: Identifying potential barriers to enrollment. J Clin Oncol 2001;19:1728–1733. [DOI] [PubMed] [Google Scholar]
- 22. Talarico L, Chen G, Pazdur R. Enrollment of elderly patients in clinical trials for cancer drug registration: A 7‐year experience by the US Food and Drug Administration. J Clin Oncol 2004;22:4626–4631. [DOI] [PubMed] [Google Scholar]
- 23. Truty MJ, Kendrick ML, Nagorney DM et al. Factors predicting response, perioperative outcomes, and survival following total neoadjuvant therapy for borderline/locally advanced pancreatic cancer. Ann Surg 2019. [Epub ahead of print]. [DOI] [PubMed] [Google Scholar]
- 24. Parmar AD, Vargas GM, Tamirisa NP et al. Trajectory of care and use of multimodality therapy in older patients with pancreatic adenocarcinoma. Surgery 2014;156:280–289. [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
See http://www.TheOncologist.com for supplemental material available online.
Appendix S1. Figures.