PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
J Am Coll Surg. Author manuscript; available in PMC Feb 21, 2013.
Published in final edited form as:
PMCID: PMC3578342
NIHMSID: NIHMS438344
Management of Patients with Pancreatic Adenocarcinoma: National Trends in Patient Selection, Operative Management, and Use of Adjuvant Therapy
Skye C Mayo, MD, MPH, Marta M Gilson, PhD, Joseph M Herman, MD, MSc, John L Cameron, MD, FACS, Hari Nathan, MD, Barish H Edil, MD, FACS, Michael A Choti, MD, FACS, Richard D Schulick, MD, FACS, Christopher L Wolfgang, MD, PhD, FACS, and Timothy M Pawlik, MD, MPH, FACS
Department of Surgery, Johns Hopkins University School of Medicine (Mayo, Gilson, Cameron, Nathan, Edil, Choti, Schulick, Wolfgang, Pawlik) and the Department of Radiation Oncology, The Johns Hopkins Hospital (Herman), Baltimore, MD
Correspondence address: Timothy M Pawlik, MD, MPH, FACS, Department of Surgery, Johns Hopkins University School of Medicine, Harvey 611, 600 N Wolfe St, Baltimore, MD 21287. tpawlik1/at/jhmi.edu
BACKGROUND
Surgical resection remains the only potentially curative option for patients with pancreatic adenocarcinoma (PAC). Advances in surgical technique and perioperative care have reduced perioperative mortality; however, temporal trends in perioperative morbidity and the use of adjuvant therapy on a population basis remain ill-defined.
STUDY DESIGN
Using Surveillance, Epidemiology, and End Results–Medicare data, 2,461 patients with resected PAC were identified from 1991 to 2005. We examined trends in preoperative comorbidity indices, adjuvant treatment, type of pancreatic resection, and changes in morbidity and mortality during 4 time intervals (ie, 1991–1996, 1997–2000, 2001–2003, and 2003–2005).
RESULTS
The majority of patients underwent pancreaticoduodenectomy (n = 1,945; 79%). There was a temporal increase in mean patient age (p < 0.05) and the number of patients with multiple preoperative comorbidities (Elixhauser comorbidities ≥3: 1991–1996, 10% vs 2003–2005, 26%; p < 0.001). Perioperative morbidity (53%) did not, however, change over time (p = 0.97) and 30-day mortality decreased by half (1991–1996: 6% vs 2003–2005: 3%; p = 0.04). Overall, 51% (n = 1,243) of patients received adjuvant therapy, with the majority receiving chemoradiation (n = 817; 33%). Among patients who received adjuvant therapy, factors associated with receipt of adjuvant chemotherapy alone relative to chemoradiation included older patient age (odds ratio = 1.75; p < 0.001) and ≥3 medical comorbidities (odds ratio = 1.57; p = 0.007). Receipt of adjuvant chemotherapy alone also increased over time (2003–2005 vs 1991–1996, odds ratio = 2.21; p < 0.001).
CONCLUSIONS
Perioperative 30-day mortality associated with resection for PAC decreased by one-half from 1991 to 2005. Although patients undergoing resection for PAC were older and had more preoperative comorbidities, the incidence of perioperative complications remained stable. The relative use of adjuvant chemotherapy alone vs chemoradiation therapy for PAC has increased in the United States during the 15 years examined.
Pancreatic adenocarcinoma (PAC) remains one of the most challenging gastrointestinal malignancies, with little improvement in long-term survival during the past 20 years. In 2010, the American Cancer Society reported 43,140 new PAC cases with 36,800 associated deaths.1 During the last several decades, advances in surgical techniques and perioperative care have substantially reduced perioperative mortality to <5% in high-volume centers.2 Perioperative morbidity and complications, including pancreatic fistula, still remain between 15% and 50%, even in institutions with high volumes of pancreatic operations.2,3 Although data on long-term outcomes after PAC surgery have been described previously,2 population-based data on temporal changes in patient selection, perioperative management, morbidity, and use of adjuvant therapy after surgery for PAC remain ill-defined.
Despite multiple randomized trials that have examined various adjuvant treatment regimens, there is still no consensus on what constitutes the most effective adjuvant treatment regimen—a fact that is reflected in the 2011 National Comprehensive Cancer Network guidelines.4 Recent reports from population-based analyses and from the most recent randomized controlled trials provide 5-year survival estimates ranging from 10% to 20% in patients resected with curative intent and treated with adjuvant chemoradiation (CRT).511 Results of recent adjuvant trials, including the European Study Group for Pancreatic Cancer-1,9 Radiation Therapy Oncology Group 97-04,11 and Charité Onkologie-00110 have called into question the value of postoperative CRT compared with chemotherapy alone. Current trends in the relative use of adjuvant CRT vs chemotherapy only for patients undergoing resection with curative intent for PAC in the United States have not been examined using the Surveillance, Epidemiology, and End Results (SEER)–Medicare database since the reporting of major clinical trials.5 The objective of the current study was to evaluate trends in the type of adjuvant treatment (ie, CRT vs chemotherapy alone) that patients received after surgical resection of PAC among Medicare beneficiaries in the United States using SEER-Medicare data. Linkage of the SEER database with Medicare billing information allows for capture of population-based data on upwards of 26% of the US population.12 Specifically, SEER-Medicare data have been evaluated for effectiveness in identifying persons receiving systemic chemotherapy13 and radiation treatment.14 In addition, use of SEER-Medicare data has been noted to be more effective at capturing perioperative comorbidities, types of procedures, and treatments, compared with the use of either database alone.15 As such, we also sought to define trends in patient comorbidities, as well as characterize the specific use of perioperative and operative procedures and treatments in patients with resected PAC on a population basis over time.
Data source
This study was a retrospective analysis of prospectively collected data from the linked SEER-Medicare database from 1991–2005. The SEER-Medicare database represents the unique linkage of 2 large population-based sources of data that provide detailed information about Medicare beneficiaries with cancer. The SEER database is derived from 18 cancer registries representing approximately 26% of the US population and is maintained by the National Cancer Institute.16 The SEER dataset includes information on patient demographics, tumor and disease characteristics, cancer-associated treatments, use of cancer-directed surgery and medical therapy, and survival for individuals with cancer. The linkage of SEER data to Medicare claims is performed by the National Cancer Institute and Center for Medicare and Medicaid Services. For Medicare-eligible individuals, the linked SEER-Medicare database also includes Medicare claims for covered health care services, including hospital, physician, outpatient, home health, and hospice bills, from the date of the individual’s enrollment until the time of death.17 The linked SEER-Medicare database is available beginning in 1991 and has successfully matched 93% of individuals aged 65 years or older at the time of primary cancer diagnosis to their Medicare enrollment file.
Study population
All Medicare-enrolled patients aged 64 years or older diagnosed with incident malignant PAC between 1991 and 2005 in a SEER area were evaluated for inclusion. Although Medicare covers patients 65 years or older, we included patients 64 years or older from SEER to capture any patients that were diagnosed at age 64 but were then treated under Medicare coverage at age 65. Patients with PAC were identified by the International Classification of Diseases for Oncology (3rd ed) topography, behavior, and histology codes.18 Histology codes (Table 1) were selected to only identify patients with adenocarcinoma; other histology codes (eg, carcinoid, intraductal mucinous neoplasm, etc) were excluded. Identification of patients with resected PAC was accomplished using the ICD-9-CM diagnosis and procedure codes and CPT codes for various pancreatic operations (Table 2, under “Pancreas”). Patients who did not undergo operative management of their PAC were excluded from analysis (Fig. 1) as were patients undergoing an open pancreatic biopsy or other procedures done without curative intent. Finally, patients with metastatic or unstaged disease as indicated by the SEER dataset were excluded from analysis. The study cohort included only patients enrolled in both Medicare Parts A and B who were not enrolled in a managed care plan during the study period.
Table 1
Table 1
International Classification of Diseases for Oncology (3rd ed) Histology Codes Specific to Pancreatic Adenocarcinoma
Table 2
Table 2
CPT/ICD-9-CM Codes Used In Identification of Medicare Claims
Figure 1
Figure 1
Flow diagram of patient selection out of the total 56,820 patients identified with pancreatic adenocarcinoma to arrive at those patients who had their cancer resected with curative intent within the Surveillance, Epidemiology, and End Results–Medicare (more ...)
Outcomes and predictor variables
Data on perioperative procedures, treatments, and complications were selected a priori based on clinical relevance and then identified from the Medicare database using both ICD-9-CM diagnosis and procedure codes, as well as CPT codes (Table 2). Information on chemotherapy was designated as preoperative (within 3 months before) and adjuvant (within 6 months after) in relation to the pancreatic operation. Medicare billing codes to assess a wide range of outcomes have previously been demonstrated to be valid.19 Information on age, sex, race, marital status, and geographic region were obtained from the SEER portion of the database. Variables were transformed into categorical and indicator variables where appropriate. The Elixhauser comorbidity index, a comprehensive set of 30 comorbidity measures, was used to identify and adjust for comorbid conditions.2024 Comorbid diagnoses related to the patient’s admission diagnosis (eg, solid tumor), as well as any comorbid condition with a frequency <5, were excluded so that a total of 20 comorbidities remained for analysis.
Statistical analyses
Mean values were used to describe continuous data, with discrete variables displayed as totals and frequencies. Cells with <11 cases per variable cell were relabeled as “<11 (<%)” in compliance with the National Cancer Institute regulation for reporting of SEER-Medicare data. For purposes of analyses, the distribution of the total number of comorbid conditions per patient was divided into approximate quartiles: 0, 1, 2, and ≥3 comorbidities. When assessing temporal trends, data were similarly separated into calendar year quartiles (1991–1996, 1997–2000, 2001–2002, and 2003–2005) based on the year of operation. Trends in ordinal data were evaluated using the linear-by-linear association test25 and the differences in mean values for continuous variables across time were assessed using ANOVA test. The linear-by-linear test of trend offers a measure of significance for ordinal variables (such as calendar year quartiles ordered from lowest to highest) and is preferred when testing the significance of linear relationship between ordinal variables. Unless otherwise specified, the p values reported for an analysis of trend refers to all of the 4 calendar year quartiles.
Cumulative event rates were calculated using the method of Kaplan and Meier26 and survival curves were compared using the log-rank test. The overall survival time was calculated from the date of the pancreatic operation to the date of last follow-up. Postoperative death was defined as patient survival <30 days after the pancreatic operation. Additionally, patient survival at 60 days was also assessed.27 Univariable and multivariable logistic regression models were constructed to determine association with receipt of adjuvant therapy (ie, CRT or chemotherapy alone). To identify variables for inclusion in the multivariable model, variables were selected using a univariable significance of p < 0.25. Logistic regression models were built for patients who received adjuvant therapy to determine factors associated with receipt of CRT vs chemotherapy alone. Finally, a multivariable Cox regression model was constructed to examine the association of no adjuvant therapy, adjuvant CRT, and adjuvant chemotherapy only on survival. All reported p values were 2-tailed, and for all tests p < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS Version 18.0 (SPSS, Inc).
Patient and primary tumor characteristics
Using the SEER-Medicare database, 56,820 patients diagnosed with PAC between 1991 and 2005 were identified. After selecting patients based on a combination of CPT and ICD-9 codes, and excluding patients younger than 64 years old and those who were not surgically managed, 6,676 patients remained (Fig. 1). Of the 6,676 patients with PAC, 3,146 patients (47.1%) who underwent an open pancreatic biopsy or other procedures done without curative intent were excluded. Patients who were unstaged, had metastatic disease, and those who did not have histology codes consistent with adenocarcinoma were then excluded, leaving 2,461 patients (4.3% overall) available for analysis.
Demographic and clinical characteristics of the 2,461 patients with resected PAC are outlined in Table 3. Mean patient age was 73.2 years (SEM 0.1 years). Most patients were women (n = 1,321; 53.7%), white (n = 2,056; 83.5%), and resided in an urban setting (n = 2,304; 93.6%). The mean age of patients undergoing surgery for PAC increased over time (for 1991–1996, mean age 72.1 years vs for 2003–2005, mean age 74.0 years; p < 0.001). In addition to being older, patients in the later time periods also had more medical comorbidities as indicated by the Elixhauser comorbidity index. Specifically, only 10.4% (n = 59) of patients undergoing surgical resection for PAC from 1991 to 1996 had ≥3 preoperative medical comorbidities, compared with 26.0% (n = 222) of patients undergoing surgery from 2003 to 2005 (p < 0.001). In contrast, there was a decrease in the number of patients with no medical comorbidities from the beginning to the end of the study period (1991–1996: n = 281 [49.5%] vs 2003–2005: n = 210 [24.6%]; p < 0.001). The most common preoperative comorbidities were hypertension (n = 1,132; 46.0%), diabetes (n = 524; 21.3%), and chronic pulmonary disease (n = 322; 13.1%).
Table 3
Table 3
Trends in Baseline Demographic and Clinical Characteristics of Patients with Surgically Managed Pancreatic Adenocarcinoma, Surveillance, Epidemiology, and End Results–Medicare, 1991–2005
The most prevalent histology code (n = 1,661; 67.5%) was 8140, corresponding to “adenocarcinoma not otherwise specified” (Table 1).18 The majority of patients had regional disease (n = 2,015; 81.9%) as classified by the SEER historic stage and this proportion increased over time (P = 0.039). Most patients (n = 1,160; 47.1%) had a tumor grade classified as moderately differentiated. Overall, 54.2% (n = 1,333) of patients had lymph node metastasis on final pathologic examination, with an increase in the number of patients with nodal disease identified over time (1991–1996: n = 287 [50.5%] vs 2003–2005: n = 515 [60.3%]; p < 0.001).
Trends in preoperative diagnostic imaging
In the overwhelming majority of cases, CT was the imaging modality of choice (n = 2,255; 91.6%), and MRI (n = 364; 14.8%) and PET (n = 44; 1.8%) were used in only a minority of patients (Table 4). Overall use of cross-sectional imaging increased during the time periods examined (p < 0.001), with 94.3% and 24.7% of patients staged with CT and MRI, respectively, in 2003–2005. For MRI, this corresponded to a >10-fold increase from the beginning to the end of the study period (1991–1996: n = 12 [2.1%] vs 2003–2005: n = 211 [24.7%]; p < 0.001). Similarly, use of preoperative diagnostic laparoscopy increased by >5-fold with 15.5% (n = 132) of patients having a laparoscopic staging procedure before their pancreatic resection in 2003–2005 (p < 0.001). Of note, PET scan was used in only 4.4% of patients, even in the most recent time period examined.
Table 4
Table 4
Preoperative Staging, Operative Details, and Postoperative Outcomes of Patients with Resected Pancreatic Adenocarcinoma, Surveillance, Epidemiology, and End Results–Medicare, 1991–2005
Patients from a rural area (n = 15; 9.6%) were as likely to be staged with MRI as patients from an urban setting (n = 349; 15.1%) (p = 0.06). Use of PET (p = 0.62) and diagnostic laparoscopy (p = 0.72) was not statistically different in rural vs urban areas. In contrast, patients with ≥3 medical comorbidities were more likely to have undergone an MRI (p < 0.001) or staging laparoscopy (p = 0.039) before pancreatic resection, compared with patients who had fewer comorbidities. There was no statistical difference in the use of PET imaging based up the patient’s comorbidity index (p = 0.06).
Trends in pancreatic operations and perioperative outcomes
Of the 2,461 patients who underwent pancreatic resection for PAC, 1,945 (79.0%) had a pancreaticoduodenectomy (PD), 333 (13.5%) had a distal pancreatectomy, 28 (1.1%) were managed with total pancreatectomy, and 155 (6.3%) had various other pancreatic resections, including partial and near total pancreatectomy (Table 4). There were several shifting trends noted in the type of pancreatic operations performed during the time periods examined (Table 4). Specifically, the proportion of patients managed with distal pancreatectomy increased over time (1991–1996: 13.2% vs 2003–2005: 15.9%; p = 0.035), but both PD and total pancreatectomy remained stable throughout the study period (both p > 0.05).
The overall proportion of patients who had any postoperative complication was 53.0% (n = 1,304) and did not change appreciably during the study periods examined (Table 4 and Fig. 2A). The most common complications were postoperative infection (n = 302; 12.3%), need for a percutaneous drain (n = 246; 10.0%), and postoperative hemorrhage (n = 209; 8.5%). Notably, use of percutaneous drains increased during the study period (1991–1996: n = 43; 7.6% vs 2003–2005: n = 105; 12.3%; p = 0.005). The risk of complication was associated with having ≥3 Elixhauser comorbidities (≥3 Elixhauser comorbidities: 59.8% vs <3 comorbidities: 54.2%; p = 0.025).
Figure 2
Figure 2
(A) Trends in Elixhauser medical comorbidities, postoperative complications, and 30-day postoperative mortality for patients with pancreatic adenocarcinoma (PAC) undergoing pancreatic resection. Star ([large star]) indicates the overall test for trend across (more ...)
There were 104 deaths within 30 days of surgery (4.2%) and 271 deaths within 60 days (11.0%). Both 30-day (Fig. 2A) and 60-day mortality decreased during the study period (both p < 0.05). There was no association between perioperative mortality and rural residence, race, or ≥3 Elixhauser comorbidities (all p > 0.05). Patients undergoing a distal pancreatectomy did, however, have a lower risk of 30-day mortality compared with patients undergoing other types of pancreatic operations (odds ratio [OR] = 0.25; 95% CI, 0.07–0.69; p = 0.003). Specifically, patients undergoing a PD had a nearly 4-fold increase in the risk of death at 30 days vs patients undergoing distal pancreatectomy (OR = 3.90; 95% CI, 1.42–10.69; p = 0.008).
Trends in perioperative therapy: adjuvant chemotherapy vs chemoradiotherapy
Approximately one-half of patients (n = 1,243; 50.5%) received some form of adjuvant treatment with a temporal increase in the use of adjuvant therapy during the study period (1991–1996, 40.3%; 1997–2000, 51.8%; 2001–2002, 51.2%; 2003–2005, 56.1%; p < 0.001) (Table 4). Only a small number of patients (n = 26; 1.1%) received preoperative chemotherapy. Overall, 33.2% (n = 817) of patients received adjuvant CRT, and 10.2% (n = 251) received adjuvant chemotherapy only (Table 4). Of note, although use of adjuvant CRT decreased after 2000 (1991–1996, 26.9%; 1997–2000, 38.2%; 2001–2002, 34.4%; 2003–2005, 33.5%), use of adjuvant chemotherapy only increased during the time periods examined (1991–1996, 6.0%; 1997–2000, 5.8%; 2001–2002, 10.6%; 2003–2005, 15.6%) (p < 0.001) (Fig. 2B).
On univariable logistic regression analysis (Table 5), patient age, year of surgery, type of pancreatic operation, postoperative complications, and lymph node metastasis were each noted to be associated with receipt of adjuvant therapy (all p < 0.05). Specifically, age 72 years or younger (OR = 1.93; 95% CI, 1.65–2.27; p < 0.001), presence of lymph node metastasis (OR = 1.77; 95% CI, 1.51–2.08; p < 0.001), and those patients who had no postoperative complications (OR = 1.67; 95% CI, 1.41–1.96; p < 0.001) were the most likely to receive some type of adjuvant therapy. In addition, use of adjuvant therapy increased during the time periods examined (1991–1996, referent; 1997–2000, OR = 1.59; 2001–2002, OR = 1.55; 2003–2005, OR = 1.89; p < 0.001). On multivariable analysis, each of these factors remained independently associated with receipt of adjuvant CRT (Table 5).
Table 5
Table 5
Logistic Regression Analyses of Variables Associated with Receipt of Adjuvant Therapy in Patients with Resected Pancreatic Adenocarcinoma in Surveillance, Epidemiology, and End Results–Medicare 1991–2005
Among those patients who received some type of adjuvant therapy (n = 1,243), factors associated with receipt of adjuvant chemotherapy alone relative to CRT included older patient age (OR = 1.75; p < 0.001), ≥3 medical comorbidities (OR = 1.57; p = 0.007), and time period 2003–2005 compared with 1991–1995 (OR = 2.21; p < 0.001). On multivariable logistic regression analysis, older patient age (OR = 1.60; 95% CI, 1.20–2.13; p = 0.001) and a later time period of pancreatic resection (2003–2005, OR = 2.0; 95% CI, 1.31–3.06; p = 0.001) remained independently associated with an increased likelihood to have received adjuvant chemotherapy alone vs CRT.
Long-term outcomes
Overall median survival of patients with PAC resected with curative intent was 14.0 months (95% CI, 13.2–14.8 months) with 1-, 3-, and 5-year survival of 53.2%, 19.7%, and 12.6%, respectively. There was a modest improvement in the median overall survival during the time periods examined, with an increase in the median survival from 12.0 to 16.0 months (p = 0.005) (Table 4). On multivariable Cox regression analysis with adjustment for patient clinicodemographics, including comorbidities and tumor characteristics, receipt of adjuvant therapy was associated with improved survival (p < 0.001). Specifically, receipt of adjuvant CRT relative to no adjuvant therapy (hazard ratio = 0.70; 95% CI, 0.63–0.77; p < 0.001) or chemotherapy only vs no adjuvant therapy (hazard ratio = 0.81; 95% CI, 0.69–0.97; p = 0.018) was associated with a decreased risk of death (Fig. 3).
Figure 3
Figure 3
Multivariable Cox overall survival from the time of the pancreatic resection stratified by receipt of adjuvant therapy. Patients who received any adjuvant therapy (chemoradiation or chemotherapy only) had a lower risk of death than patients who did not (more ...)
Although surgery remains the only chance at cure, morbidity and mortality associated with pancreatic surgery have traditionally been high. In 1987, data from the Johns Hopkins Hospital noted a mortality for PD that decreased from 36% between 1969 and 1980 to 2% between 1981 and 1986.28 During the same time periods, these authors noted a decrease in morbidity from 54% to 24%. Other data, mostly from single institutional series,2,3 have similarly noted a decrease in the perioperative mortality risk of pancreatic surgery. In the current study, by using the SEER-Medicare dataset, we were able to assess the care of patients with surgically managed PAC on a population-based level. The SEER-Medicare dataset offers the ability to assess cancer surgery care in >25% of the US population and provide robust data concerning perioperative resource and procedure use.5,6 The current study is important because it examined population-based trends in perioperative management, surgical procedure use, and outcomes among patients undergoing surgical resection of PAC. We noted that over time, patients undergoing surgical resection of PAC were older and had more medical comorbidities. Despite this, the overall proportion of patients who had any postoperative complications did not change during the study periods examined and perioperative mortality actually decreased. We found that only approximately one-half of patients received any form of adjuvant treatment after surgery for PAC. In addition, although use of adjuvant CRT decreased after 2001, the use of adjuvant chemotherapy only increased during the time periods examined (Fig. 2B). These data can have important implications in helping to understand how population-level care of patients with resected PAC corresponds with recommendations (eg, National Comprehensive Cancer Network) for the treatment of PAC.
Although decreased mortality associated with pancreatic surgery has been reported,2 heterogeneity in data reporting (eg, in-hospital, 30 days, 60 days), as well as the fact that data have largely come from large, single-center academic centers, making interpretation of the data difficult. In the current study, we report an overall population-based 30-day mortality of 4.2% after pancreatic surgery. As expected, mortality was associated with the type of surgical procedure, as PD had a 4-fold increased risk of 30-day mortality (OR = 3.90; p = 0.008) compared with distal pancreatectomy. Although the 30-day mortality we report was consistent with the mortality reported at most centers (eg, <5%),2,3 one important finding of the current study was the 11.0% incidence of postoperative death at the 60-day time period. Carroll and colleagues29 had previously noted the importance of reporting mortality after pancreatic surgery beyond the traditional 30-day time period. The authors29 noted a 60-day mortality of 11.9% similar to our report. We did, however, note a decrease in both 30- and 60-day perioperative mortality during the time periods examined (Fig. 2A). Importantly, this reduction in mortality was accompanied by an increase in patient age and the number of preoperative medical comorbidities (Fig. 2A). Although the morbidity associated with pancreatic surgery did not decrease over time, it did remain stable, albeit high at 53.0%. In aggregate, these data suggest that additional progress is needed to decrease the perioperative morbidity associated with pancreatic resection similar to the manner in which mortality has been reduced.
Another important finding of the current study was the overall and relative use of adjuvant therapy. We found that only one-half of patients (50.5%) received any form of adjuvant treatment. The reason for this is likely multifactorial, and might be related, in part, to the fact that the dataset included only patients aged 65 years or older. Several studies have reported that complications and prolonged recovery after PD for PAC can prevent delivery of postoperative therapy in at least 25% to 30% of patients.2,3032 In the current study, we noted that PD and total pancreatectomy, as well as the occurrence of postoperative complications, were associated with a decreased likelihood of receiving adjuvant therapy (Table 5). Among those patients who did receive adjuvant therapy, we also noted some interesting trends with regard to the type of postoperative therapy delivered. Specifically, factors associated with receipt of adjuvant chemotherapy alone relative to CRT included older patient age, ≥3 medical comorbidities, and the time period treated. In fact, patients undergoing pancreatic resection from 2003–2005 had a 2-fold increased likelihood of receiving chemotherapy alone rather than CRT therapy. Conflicting results from several adjuvant studies have lead to differences in how patients with PAC are treated in the adjuvant setting. Although both the Gastrointestinal Tumor Study Group33 and Radiation Therapy Oncology Group 97-0411 trials reported a benefit for CRT therapy, the European Study Group for Pancreatic Cancer-1 trial9 found a benefit for adjuvant chemotherapy alone and a deleterious effect for combined CRT. As such, although combined CRT remains the de facto treatment modality for resected PAC among many centers in the United States, adjuvant chemotherapy alone has become the standard of care in many European centers. Data from the current study suggest, however, that use of chemotherapy alone as adjuvant therapy in the United States has increased during the last decade.
Although the aim of the study was not to assess the survival benefit conferred by adjuvant therapy, we did note a survival benefit of adjuvant therapy for patients with resected PAC, even after controlling for competing risk factors using a multivariable Cox regression model (Fig. 3). Using a propensity-adjusted model, Simons and colleagues27 similarly demonstrated a survival benefit for adjuvant therapy (chemotherapy and CRT) among a cohort of patients with resected PAC derived from SEER. Mayo and colleagues8 noted an improvement in survival associated with adjuvant therapy when using population-based data from the Oregon state cancer registry. In a large, single institutional database analysis of 908 patients, Herman and colleagues32 reported that adjuvant fluorouracil-based CRT therapy improved survival compared with patients not receiving CRT. Although assessing the therapeutic effect of any treatment modality using retrospective data can be problematic, data from the current study as well as others5,6 suggest a beneficial role of adjuvant therapy after surgical resection of PAC.
The current study had several limitations. Although the data derived from the linked SEER-Medicare dataset offer a perspective that is generalizable to many practice patterns, the data do not include detailed data on the characteristics of the adjuvant treatment regimen. As such, although the dataset allowed for accurate identification with regard to whether a patient received no adjuvant therapy, chemotherapy alone, or CRT therapy, it did not contain data on the type of chemotherapy or dose of radiation delivered. Data from the SEER-Medicare dataset are also limited to Medicare beneficiaries (aged 65 years and older) and might not be appropriate to extrapolate to younger populations. Finally, data in the current study include patient data only through 2005 and do not include the time period in which 2 of the most recent randomized clinical trials for adjuvant treatment for PAC (ie, Charité Onkologie-01 and Radiation Therapy Oncology Group 97-04) were reported.
Mortality associated with pancreatic resection for PAC has decreased by one-half during the 15 years examined in the study. Surgical resection for PAC is being offered to older patients with more preoperative comorbidities without an increase in perioperative mortality or morbidity. Although perioperative mortality has decreased, the incidence of perioperative complications has remained unchanged. Only about 50% of patients received adjuvant therapy after resection of PAC and failure to receive adjuvant therapy was associated with advanced age and a history of postoperative complications. In addition, the specific use of adjuvant chemotherapy alone vs CRT therapy has increased in the United States during the 15 years examined. These data should help to inform our collective understanding of how patients with PAC are being managed at the population-based level and serve as a reference point when trying to interpret and extrapolate published data from high-volume tertiary academic centers.
Abbreviations and Acronyms
CRTchemoradiation/chemoradiotherapy
ORodds ratio
PACpancreatic adenocarcinoma
PDpancreaticoduodenectomy
SEERSurveillance, Epidemiology, and End Results

Footnotes
Disclosure Information: Authors have nothing to disclose. Timothy J Eberlein, Editor-in-Chief, has nothing to disclose.
Presented at the Society for Surgery of the Alimentary Tract 52nd Annual Meeting and 26th Annual Residents and Fellows Research Conference, Chicago, IL, May 2010.
Author Contributions
Study conception and design: Mayo, Herman, Cameron, Nathan, Edil, Choti, Schulick, Pawlik
Acquisition of data: Mayo, Gilson, Nathan, Pawlik
Analysis and interpretation of data: Mayo, Gilson, Pawlik
Drafting of manuscript: Mayo, Gilson, Pawlik
Critical revision: Mayo, Gilson, Herman, Cameron, Nathan, Edil, Choti, Schulick, Pawlik
1. Jemal A, Siegel R, Xu J, et al. Cancer statistics, 2010. CA Cancer J Clin. 2010;60:277–300. [PubMed]
2. Sohn TA, Yeo CJ, Cameron JL, et al. Resected adenocarcinoma of the pancreas-616 patients: results, outcomes, and prognostic indicators. J Gastrointest Surg. 2000;4:567–579. [PubMed]
3. Fischer M, Matsuo K, Gonen M, et al. Relationship between intraoperative fluid administration and perioperative outcome after pancreaticoduodenectomy: results of a prospective randomized trial of acute normovolemic hemodilution compared with standard intraoperative management. Ann Surg. 2010;252:952–958. [PubMed]
4. National Comprehensive Cancer Network. The NCCN Clinical Practice Guidelines in Oncology Pancreatic Adenocarcinoma. [Accessed May 21, 2011];Version 2. 2011 Available at: www.NCCN.org.
5. Davila JA, Chiao EY, Hasche JC, et al. Utilization and determinants of adjuvant therapy among older patients who receive curative surgery for pancreatic cancer. Pancreas. 2009;38:e18–e25. [PubMed]
6. Hazard L, Tward JD, Szabo A, et al. Radiation therapy is associated with improved survival in patients with pancreatic adenocarcinoma: results of a study from the Surveillance, Epidemiology, and End Results (SEER) registry data. Cancer. 2007;110:2191–2201. [PubMed]
7. Slidell MB, Chang DC, Cameron JL, et al. Impact of total lymph node count and lymph node ratio on staging and survival after pancreatectomy for pancreatic adenocarcinoma: a large, population-based analysis. Ann Surg Oncol. 2008;15:165–174. [PubMed]
8. Mayo SC, Austin DF, Sheppard BC, et al. Adjuvant therapy and survival after resection of pancreatic adenocarcinoma: a population-based analysis. Cancer. 2010;116:2932–2940. [PubMed]
9. 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. [PubMed]
10. Oettle H, Post S, Neuhaus P, et al. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer: a randomized controlled trial. JAMA. 2007;297:267–277. [PubMed]
11. Regine WF, Winter KA, Abrams RA, et al. Fluorouracil vs gemcitabine chemotherapy before and after fluorouracil-based chemoradiation following resection of pancreatic adenocarcinoma: a randomized controlled trial. JAMA. 2008;299:1019–1026. [PubMed]
12. Horner MJ, Ries LAG, Krapcho M, et al., editors. Bethesda, MD: National Cancer Institute; [Accessed November 2010]. SEER Cancer Statistics Review, 1975–2006. Available at: http://seer.cancer.gov/csr/1975_2006/, based on November 2008 SEER data submission, posted to the SEER Web site, 2009.
13. Warren JL, Harlan LC, Fahey A, et al. Utility of the SEER-Medicare data to identify chemotherapy use. Med Care. 2002;40(Suppl):IV-55–IV-61. [PubMed]
14. Virnig BA, Warren JL, Cooper GS, et al. Studying radiation therapy using SEER-Medicare-linked data. Med Care. 2002;40(Suppl):IV-49–IV-54. [PubMed]
15. Cooper GS, Virnig B, Klabunde CN, et al. Use of SEER-Medicare data for measuring cancer surgery. Med Care. 2002;40(Suppl):IV-43–IV-48. [PubMed]
16. National Cancer Institute Data, Surveillance Research Program, Cancer Statistics Branch. Surveillance, Epidemiology, and End Results (SEER) program. [Accessed November 2010]; Available at: http://www.seer.cancer.gov. SEER* Stat database: Incidence—SEER 17 regs public-use, Nov 2005 sub (1973–2003 varying), linked to county attributes, total US 1969–2003 counties. Released April 2006, based on the November 2005 submission.
17. Warren JL, Klabunde CN, Schrag D, et al. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care. 2002;40(Suppl):IV-3–IV-18. [PubMed]
18. Fritz AG. 3rd ed. Geneva: World Health Organization; 2000. International classification of diseases for oncology.
19. Potosky AL, Warren JL, Riedel ER, et al. Measuring complications of cancer treatment using the SEER-Medicare data. Med Care. 2002;40(Suppl):IV-62–IV-68. [PubMed]
20. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care. 1998;36:8–27. [PubMed]
21. Baldwin LM, Klabunde CN, Green P, et al. In search of the perfect comorbidity measure for use with administrative claims data: does it exist? Med Care. 2006;44:745–753. [PMC free article] [PubMed]
22. Li B, Evans D, Faris P, et al. Risk adjustment performance of Charlson and Elixhauser comorbidities in ICD-9 and ICD-10 administrative databases. BMC Health Serv Res. 2008;8:12. [PMC free article] [PubMed]
23. Quan H, Parsons GA, Ghali WA. Validity of information on comorbidity derived from ICD-9-CCM administrative data. Med Care. 2002;40:675–685. [PubMed]
24. Southern DA, Quan H, Ghali WA. Comparison of the Elixhauser and Charlson/Deyo methods of comorbidity measurement in administrative data. Med Care. 2004;42:355–360. [PubMed]
25. Agresti A. Introduction to categorical data analysis. New York: Wiley; 1996. pp. 182–185. 187–189, 201, 217, 279.
26. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–481.
27. Simons JP, Ng SC, McDade TP, et al. Progress for resectable pancreatic [corrected] cancer? A population-based assessment of US practices. Cancer. 2010;116:1681–1690. [PubMed]
28. Crist DW, Sitzmann JV, Cameron JL. Improved hospital morbidity, mortality, and survival after the Whipple procedure. Ann Surg. 1987;206:358–365. [PubMed]
29. Carroll JE, Smith JK, Simons JP, et al. Redefining mortality after pancreatic cancer resection. J Gastrointest Surg. 2010;14:1701–1708. [PubMed]
30. Spitz FR, Abbruzzese JL, Lee JE, et al. Preoperative and postoperative chemoradiation strategies in patients treated with pancreaticoduodenectomy for adenocarcinoma of the pancreas. J Clin Oncol. 1997;15:928–937. [PubMed]
31. Wayne JD, Abdalla EK, Wolff RA, et al. Localized adenocarcinoma of the pancreas: the rationale for preoperative chemoradiation. Oncologist. 2002;7:34–45. [PubMed]
32. Herman JM, Swartz MJ, Hsu CC, et al. Analysis of fluorouracil-based adjuvant chemotherapy and radiation after pancreaticoduodenectomy for ductal adenocarcinoma of the pancreas: results of a large, prospectively collected database at the Johns Hopkins Hospital. J Clin Oncol. 2008;26:3503–3510. [PMC free article] [PubMed]
33. Gastrointestinal Tumor Study Group. Further evidence of effective adjuvant combined radiation and chemotherapy following curative resection of pancreatic cancer. Gastrointestinal Tumor Study Group. Cancer. 1987;59:2006–2010. [PubMed]