Abstract
To report long-term follow up of a phase II, single-center, single-arm trial of resectable pancreatic ductal adenocarcinoma (PDAC) treated with adjuvant interferon-based chemoradiation followed by gemcitabine to determine survival, recurrence, and complications.
Methods
From 2002 to 2005, 53 patients with PDAC underwent pancreaticoduodenectomy and received adjuvant interferon-based chemoradiation consisting of external-beam irradiation and simultaneous 3-drug chemotherapy of continuous daily 5-fluorouracil infusion, weekly intravenous bolus cisplatin, and subcutaneous interferon-α, followed by two 4-week courses of weekly intravenous gemcitabine.
Results
Actual overall survival for the 5- and 10-year periods were 26% and 10%, respectively, with a median overall survival of 25 months (95% CI: 16.4–38.5). Adverse prognostic factors on multivariate analysis were positive tumor margin (p<0.035), lymphovascular invasion (p<0.015), and perineural invasion (p<0.026). Median time to recurrence was 11 months. Positive tumor margin was associated with lymph node involvement (p<0.005), portal vein resection (p<0.038), and distant recurrence/metastases (p<0.018). Late complications were frequent and predominated by gastrointestinal and infectious complications.
Conclusions
Adjuvant interferon-based chemoradiation for PDAC improves long-term survival compared to standard therapy. However, recurrence rates and long-term complications remain high, thus further studies are indicated to assess patient characteristics that indicate a favorable treatment profile.
Introduction
Pancreatic adenocarcinoma (PDAC) is a devastating disease, and while surgical resection is essential for cure, adjuvant therapy has been shown to improve survival compared with resection alone. In 1985, the Gastrointestinal Tumor Study Group (GITSG) demonstrated that surgery followed by adjuvant chemoradiation with 5-fluorouracil was superior to surgery alone, improving median survival from 11 months to 20 months.1 The GITSG later enrolled 30 additional patients and confirmed the magnitude of the effects of therapy with a median survival of 18 months and 2-year actuarial survival of 46% for those treated with surgery and adjuvant therapy compared to 18% for those who only received surgery in the original cohort.2
These early results were promising and provided the impetus for further studies with adjuvant chemoradiation. In 2000, Picozzi and colleagues described their initial experience with a novel chemoradiotherapy protocol utilizing interferon-α, which was selected for its enhancement of radiation effects and tumoricidal effects of select chemotherapeutic agents.3 In their initial report, they demonstrated that the Virginia Mason protocol of external-beam irradiation and simultaneous 3-drug chemotherapy consisting of continuous daily 5-fluorouracil infusion, weekly intravenous bolus cisplatin, and subcutaneous interferon-α, followed by two 6 weeks of continuous daily 5-fluorouracil infusion was superior to a GITSG-type protocol with 2-year survival rates of 84% compared to 54%.3
In 2003, Picozzi and colleagues reported their 5-year follow-up and experience of 43 patients who received their protocol, this time demonstrating 1-, 2-, and 5-year actuarial survival of 95%, 64%, and 55%, respectively.4 However, there were significant toxicities observed, which were predominantly gastrointestinal, and 70% of patients were affected. Due to these toxicities, hospitalization was required for 42% and chemoradiation was delayed in 70% of patients. However, long-term follow-up revealed 5- and 10-year actual survival to be 42% and 28%, respectively, with a median overall survival of 42 months, indicating efficacy despite toxicity.5
Due to these findings, we were given the impetus to review our experience with a phase II, single-institution, single-arm trial of a modified version of the Virginia Mason protocol. Due to concerns for gastrointestinal toxicity, 3-dimensional conformal radiation was utilized in an attempt to limit radiation injury to the bowel as well as other surrounding strictures. Furthermore, the dose of both 5-fluorouracil and cisplatin was decreased during the chemoradiation phase, and rather than continue 5-fluorouracil as the post-chemoradiation phase, gemcitabine was selected to improve the efficacy of treatment. Our initial experience was reported in 2008, demonstrating improved survival compared to historical controls.6 However, there were significant toxicities experienced during both the chemoradiation and post-chemoradiation gemcitabine phases, and several patients developed late injuries that were presumed to be radiation-induced.
As the study is now more than 10 years out from inception, we wanted to assess whether our patients developed any further injuries that could be attributed to the treatment algorithm as well as determine outcomes and associations with recurrence and survival. Herein, we report long-term follow-up of the 53 patients with PDAC enrolled in our phase II trial of adjuvant interferon-α-based chemoradiation followed by gemcitabine.
Methods
Patient Selection and Study Protocol
This study was approved by the Institutional Review Board of Washington University School of Medicine and by the Human Studies Committee of the Siteman Cancer Center. Between April 2002 and September 2005, 53 patients with PDAC (pancreatic ductal adenocarcinoma) were enrolled in a single institution Phase II trial. Eligible patients underwent curative resection; all cases except one were performed by surgeons in the Section of Hepatobiliary-Pancreatic and Gastrointestinal Surgery at Washington University School of Medicine. Inclusion and exclusion criteria have been previously described.6 Enrolled patients received external-beam irradiation at a dose of 5 040 cGy (25 fractions per 5 weeks) and simultaneous 3-drug chemotherapy consisting of continuous daily infusion of 5-fluorouracil 175 mg/m2 by outpatient pump 7 days a week, weekly intravenous bolus cisplatin 25 mg/m2, and interferon-alpha 3 million units subcutaneously three times per week during the first six weeks of irradiation. This was followed by two 4-week courses of weekly intravenous infusion of gemcitabine (1 000 mg/m2, 3 of 4 weeks) one month after chemoradiation. An absolute neutrophil count ≤ 1 500/mm3 or platelet count ≤ 100 000/mm3 resulted in a treatment delay of at least 1 week. When therapy resumed, chemotherapy doses were reduced by 25%, and during consolidation chemotherapy, gemcitabine doses were reduced by 25% in the event of recovery from absolute neutrophil count < 1 500/mm3 and platelet count < 100 000/mm3. Missed gemcitabine doses were made up one week later if blood counts had recovered.
Follow-up
Patient demographics, tumor characteristics, and clinical outcome data were abstracted from electronic hospital records. Recurrent disease was determined from radiologic imaging studies and physician documentation. At the completion of adjuvant therapy, patients underwent clinical examination and laboratory studies (complete blood count, electrolytes, and liver function tests) every two months for two years, then every three months during year three, every four months during year four, every six months during year five, and then annually. Repeat CT was performed two months after completion of adjuvant therapy then every six months for two years, then annually. Additional imaging was performed if there was concern for recurrent or metastatic disease. Local recurrence was defined as tumor in the pancreatic bed and/or in the celiac, superior mesenteric, and para-aortic lymph nodes. All other recurrence sites were classified as distant. Survival was determined by review of clinical information and the Social Security Death Index for patients lost to follow-up.
Statistical Analyses
The primary outcome variable was overall survival (OS), which was defined as the time from date of surgery to date of death from any cause. Patients alive or lost to follow-up were censored at the last follow-up. Disease-free survival (DFS) wad defined as the time from date of surgery to date of death or recurrence, whichever occurred first. Patients alive without progression or lost to follow-up were censored at last follow-up. Recurrence-free survival (RFS) was defined as time from date of surgery to disease recurrence or last follow-up. Kaplan-Meier (KM) curves were generated to provide unadjusted survival estimates for all subjects and the survival probabilities at specific time points were estimated. Clinical characteristics were summarized using descriptive statistics. Categorical variables were compared by the χ2 or Fisher’s Exact test. Predicted survival was calculated utilizing the Memorial Sloan-Kettering Cancer Center Pancreatic Adenocarcinoma Nomogram as previously described.7 Briefly, points are assigned to characteristics such as age, sex, tumor location and size, margin of resection, histologic differentiation, T stage, N stage, weight loss, back pain, portal vein resection, and splenectomy and used to estimate survival. Data was complete for 51 of 53 patients for nomogram calibration. A calibration plot was formed to compare the predicted probability of overall survival with the actual survival probabilities at years 1, 2, and 3. This was performed by grouping patients according to their nomogram-predicted probabilities and then comparing the mean of the group with the observed Kaplan-Meier overall survival estimate. Univariate and multivariate binary logistic regression tests were performed to assess the role of tumor characteristics with median survival. Covariates with a p-value of 0.1 in the univariate analysis were entered into multivariate analysis using backward selection techniques. All statistical tests were two-sided using an α = 0.05 level of significance. SAS Version 9.3 (Cary, NC) was used to perform all statistical analysis.
Results
Patient Demographics and Surgical Characteristics
Patient characteristics at the time of enrollment have been previously reported.6 Briefly, of the 53 patients enrolled, 29 patients (54.7%) were men and the median age was 59 years (range: 42–78 years). All patients had an ECOG performance score of 0 (58.4%) or 1 (41.5%). Forty-three patients (81.1%) underwent pancreaticoduodenectomy, 4 (7.5%) underwent distal pancreatectomy, 4 (7.5%) underwent radical antegrade modular pancreatectomy, and 2 (3.8%) required total pancreatectomy.
Survival
Survival status of every living patient was verified in September 2015. Kaplan-Meier overall, disease-free, and recurrence-free survival curves are depicted in the Figure, and the survival rates are listed in Table 1. Median overall survival (OS) was 25 months (95% CI: 16.4–38.5 months); median disease-free survival (DFS) was 15 months (95% CI: 10.4–21.4 months); median recurrence-free survival was 16 months (95% CI: 11.0–25.0 months).
Figure.
Kaplan-Meier survival curves of (A) overall survival, (B) disease-free survival, and (C) recurrence-free survival with number at risk for each year after surgical intervention. Number of patients at risk are displayed below the x-axis.
Table 1.
Kaplan-Meier survival statistics
| Time | Overall Survival (95% CI) | Disease-Free Survival (95% CI) | Recurrence-Free Survival (95% CI) |
|---|---|---|---|
| One year | 75% (61.5–85.0% | 57% (42.3–68.7%) | 60% (45.5–72.3%) |
| Two years | 53% (38.6–65.2%) | 36% (23.2–48.6%) | 38% (24.5–51.0% |
| Three years | 40% (26.6–52.4%) | 28% (17.0–40.7%) | 31% (18.9–44.4%) |
| Five years | 26% (15.4–38.7%) | 21% (11.1–32.5%) | 29% (16.6–41.8%) |
| Ten years | 10% (3.1–21.1%) | 9% (2.5–19.3%) | 20% (9.1–33.0%) |
The Memorial Sloan-Kettering Cancer Center Pancreatic Adenocarcinoma Nomogram was applied to study patients to assess whether our patients had improved survival compared to their predicted survival. The calibration plot compares predicted survival, based upon the nomogram, to overall survival (Supplemental Figure). Actual survival was higher than the predicted 1-, 2-, and 3-year survival of 60.2% (SE: 0.7), 30.8% (SE: 2.7), and 19.4% (SE: 21.7), respectively.
Patterns of Recurrence
Of the 53 patients, 37 (69.8%) had evidence of recurrent disease at a median of 11 months (range: 2.2–79.3 months) from the date of surgery. Recurrence occurred distally in 27 (73%) and locally in 13 (35%) patients. Three (8%) patients presented with both distal and local recurrence. Of those who suffered distant recurrence, the initial sites of disease recurrence were liver in 16, lung in 12, peritoneum in 4, and bone in 2. Eight patients presented with at least two distant sites of recurrence. Further metastases were later identified in several patients with the overall sites of metastatic disease being liver in 18, lung in 13, peritoneum in 5, and bone in 3 patients.
Evaluation of Tumor Characteristics
Tumor characteristics were unfavorable in the study population (Table 2). Seventeen patients had positive margins but the 5 patients with positive neck margins were re-resected at the time of surgery to negative margins. Forty-one patients (77.4%) had evidence of perineural invasion, 30 (56.6%) had lymphovascular invasion, and 40 (75.5%) had positive lymph nodes, and nine (16.9%) had extracapsular spread. Twelve (22.6%) required vascular resection due to tumor involvement.
Table 2.
Relation of tumor characteristics to median survival
| Covariate | N |
Median OS (year) |
Univariate Analysis
|
p |
Multivariate Analysis
|
p | ||
|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | HR | 95% CI | |||||
| Positive margin | 0.073 | 0.035 | ||||||
| Negative | 36 | 2.68 | 1.00 | 1 | ||||
| Positive | 17 | 1.22 | 1.77 | 0.95, 3.32 | 2.01 | 1.05, 3.84 | ||
| Tumor size | 0.840 | |||||||
| <2.5 cm | 14 | 2.29 | 1.00 | |||||
| ≥2.5 cm | 37 | 1.82 | 1.07 | 0.55, 2.09 | ||||
| Grade of differentiation | 0.842 | |||||||
| Well | 5 | 2.48 | 1.22 | 0.44, 3.40 | ||||
| Moderate | 30 | 2.23 | 0.92 | 0.49, 1.75 | ||||
| Poor | 18 | 1.80 | 1.00 | |||||
| Lymph node involvement | 0.051 | |||||||
| Negative | 13 | 4.29 | 1.00 | |||||
| Positive | 40 | 1.83 | 2.09 | 1.00, 4.35 | ||||
| Lymph node ratio | 0.358 | |||||||
| <50% | 47 | 2.09 | 1.00 | |||||
| ≥50% | 6 | 1.47 | 1.50 | 0.63, 3.55 | ||||
| Lymphovascular invasion | 0.018 | 0.015 | ||||||
| No | 23 | 4.29 | 1.00 | 1.00 | ||||
| Yes | 30 | 1.83 | 2.10 | 1.14, 3.88 | 2.17 | 1.16, 4.07 | ||
| Perineural invasion | 0.052 | 0.026 | ||||||
| No | 12 | 4.04 | 1.00 | 1.00 | ||||
| Yes | 41 | 1.84 | 2.14 | 0.99, 4.60 | 2.44 | 1.12, 5.33 | ||
| Extracapsular spread | 0.641 | |||||||
| No | 44 | 1.88 | 1.00 | |||||
| Yes | 9 | 2.32 | 0.83 | 0.39, 1.79 | ||||
| Portal vein resection | 0.038 | |||||||
| No | 41 | 2.32 | 1.00 | |||||
| Yes | 12 | 1.79 | 2.04 | 1.04, 3.99 | ||||
We assessed tumor characteristics to determine which factors were adversely associated with survival (Table 2). Patients who were identified to have a positive margin on frozen section were treated as having a positive margin for the analysis. On univariate analysis, lymphovascular invasion and portal vein resection were associated with decreased survival. Lymph node involvement and perineural invasion trended toward decreased survival but were not statistically significant. On multivariate analysis, any positive margin (hazards ratio (HR) = 2.01, 95% CI: 1.05, 3.84, p=0.035), lymphovascular invasion (HR = 2.17, 95% CI: 1.16, 4.07, p=0.115), and perineural invasion (HR = 2.44, 95% CI: 1.12, 5.33, p=0.026) were significantly associated with decreased survival. Of the 5-year and 10-year survivors, only one patient had a positive resection margin. In relation to other tumor characteristics, four of the 5-year survivors had lymphovascular invasion, eight had perineural invasion, and one underwent portal vein resection. Of the 10-year survivors, only patient had evidence of perineural invasion but none had lymphovascular invasion or underwent portal vein resection.
Relationship with Margin Status
Margin status was correlated with other tumor characteristics as well as patterns of recurrence (Table 3). Patients with positive tumor margins were more likely to have lymph node involvement (p<0.005), require portal vein resection (p<0.038), and develop distant recurrence or metastatic disease (p<0.018). Positive tumor margin was not associated with tumor size, lymphovascular invasion, perineural invasion, extracapsular spread, or other patterns of recurrence. Subset analysis of the margin sites was not associated with specific tumor characteristics or recurrence pattern.
Table 3.
Association of tumor characteristics and patterns of recurrence with margin status (n = 51*)
| Any Positive Margin | ||
|---|---|---|
|
Yes (n = 17) |
No (n = 34) |
|
| Tumor characteristic | ||
| Tumor Size (≥ 2.5 cm) | 88.2 | 64.7 |
| p value | 0.102 | |
| Lymph Node Involvement | 100.0 | 64.9 |
| p value | 0.005 | |
| Lymphovascular Invasion | 70.6 | 50.0 |
| p value | 0.158 | |
| Perineural Invasion | 82.4 | 75.0 |
| p value | 0.730 | |
| Extracapsular Spread | 17.6 | 16.7 |
| p value | 1.000 | |
| Portal Vein Resection | 41.2 | 13.9 |
| p value | 0.038 | |
| Recurrence pattern | ||
| Any | 82.3 | 66.7 |
| p value | 0.333 | |
| LR only | 5.9 | 25.0 |
| p value | 0.141 | |
| DR only | 64.7 | 36.1 |
| p value | 0.051 | |
| LR + DR | 11.8 | 5.6 |
| p value | 0.586 | |
| Any LR | 17.6 | 30.6 |
| p value | 0.506 | |
| Any DR | 76.5 | 41.7 |
| p value | 0.018 | |
Note: data was incomplete for 2 of 53 patients in the study.
Abbreviations: LR, local recurrence; DR, distant recurrence.
Any LR = LR only and LR + DR
Any DR = DR only and LR + DR
Toxicities of Therapy and Long-Term Complications
We previously described the early toxicities developed during chemoradiation and post-chemoradiation gemcitabine.6 Briefly, during chemoradiation 36 patients (68%) developed a grade 3 or 4 toxicity with the most common complications being hematologic (40%) or gastrointestinal (37%). Of the 38 patients who received post-chemoradiation gemcitabine, 26 patients (68%) developed a grade 3 or 4 toxicity; hematologic complications were predominant (63%).
Despite making significant changes to our interferon-based chemoradiation protocol, early complications and toxicities related to adjuvant therapy were quite common as were the related long-term complications (Table 4). Of those who survived to at least one year postoperatively (N = 40), grade 3 or greater toxicities were identified in 15 patients (37.5%). Infectious complications developed in 7 patients (17.5%), with sepsis being the most severe and frequent. Gastrointestinal toxicities occurred in 6 patients (15.0%); gastrointestinal bleeding was most common. Hepatobiliary complications occurred in 5 patients (12.5%).
Table 4.
Late (more than 1 year out from surgery) complications
| Grade 3 or higher | Grade 3 | Grade 4 | Grade 5 | Grade 6 | |
|---|---|---|---|---|---|
| Gastrointestinal | 6 (15.0%) | 4 (10.0%) | 2 (5.0%) | ||
| Hemorrhage | 4 (10.0%) | 4 (10.0%) | 1 (2.5%) | ||
| Ischemia | 1 (2.5%) | 1 (2.5%) | |||
| Obstruction | 1 (2.5%) | 1 (2.5%) | |||
| Fistula | 1 (2.5%) | 1 (2.5%) | |||
| Delayed empty | 1 (2.5%) | 1 (2.5%) | |||
| Infection/Sepsis | 7 (17.5%) | 3 (7.5%) | 2 (5.0%) | 2 (5.0%) | 1 (2.5%) |
| Intraabdominal infection/Sepsis | 5 (12.5%) | 2 (5.0%) | 2 (5.0%) | 1 (2.5%) | |
| Hepatic abscess | 3 (7.5%) | 3 (7.5%) | |||
| Cholangitis | 1 (2.5%) | 1 (2.5%) | |||
| Hepatobiliary | 5 (12.5%) | 5 (12.5%) | |||
| Stricture | 5 (12.5%) | 5 (12.5%) | |||
| Hepatic failure | 4 (10.0%) | 4 (10.0%) | |||
| Fibrosis | 1 (2.5%) | 1 (2.5%) | |||
| Ventral Hernia | 2 (5.0%) | 2 (5.0%) | |||
| Respiratory | 3 (7.5%) | 2 (5.0%) | 1 (2.5%) | ||
| Neurologic | 2 (5.0%) | 1 (2.5%) | 1 (2.5%) | ||
| Renal | 2 (5.0%) | 1 (2.5%) | 1 (2.5%) |
Note: 40 of 53 patients were alive at least one year after surgery and available for late-term complication analysis. Adverse effects were graded according to Extended Accordion Severity Grading System.19
We previously reported specific concern for treatment-related injuries within the field of radiation in three patients despite utilizing 3-dimensional conformal radiation to limit off-site radiation, and we have identified additional patients that may have specific radiation-associated adverse events. Patient 11, who remained recurrence-free until death, developed multiple sequelae concerning for radiation-induced injuries. This patient developed multiple bowel obstructions, an enteroenteric fistula, and an incisional hernia with recurrent ventral hernias requiring multiple laparotomies and repair. The patient also developed sepsis from an intraabdominal abscess and subsequently underwent laparotomy with washout and lysis of adhesions but ultimately developed endocarditis and became neurologically devastated from multiple infarcts prior to his death. Patient 35 developed multiple episodes of erosive gastropathy leading to gastrointestinal hemorrhage in addition to multiple incisional and ventral hernias requiring repair.
Discussion
Pancreatic adenocarcinoma (PDAC) is a devastating malignancy, and despite decades of research and clinical trials, the 5-year survival rate remains dismal at only 7%.8 High rates of local recurrence and metastatic disease necessitate the role of adjuvant therapy, but optimal therapy remains controversial. Decades ago, the Gastrointestinal Tumor Study Group (GITSG) demonstrated a survival advantage to adjuvant combined radiation and chemotherapy compared to surgery alone.1 Leukopenia affected a minority of patients but none had life-threatening reactions or deaths. While the study was terminated early due to poor patient accrual, it has become a landmark study and advanced pancreatic cancer research. However, more recently, the role and necessity of radiation has been questioned with numerous conflicting studies among the literature. In a meta-analysis, adjuvant chemotherapy alone was found to be effective against pancreatic cancer, but upon subgroup analysis, chemoradiation was more effective in patients with positive resection margins.9 The European Organization for Research and Treatment of Cancer (EORTC) Gastrointestinal Tract Cancer Cooperative Group’s phase III trial of adjuvant chemoradiotherapy demonstrated an insignificant improvement in survival.10 Conversely, the European Study Group for Pancreatic Cancer 1 (ESPAC-1) trial found that adjuvant chemotherapy alone was associated with improved survival and that adjuvant chemoradiotherapy was associated with decreased survival.11
Picozzi and colleagues published their experience with their Virginia Mason protocol, a novel, interferon-α-based adjuvant chemoradiation protocol consisting of external-beam radiation therapy, continuous 5-fluorouracil infusion, weekly intravenous bolus cisplatin, and subcutaneous interferon-α3–5. Most recently, they have found favorable survival rates at the 5- and 10-year mark, at 42% and 28% respectively. But despite these long-term survival rates, the protocol was associated with a high rate of complications, causing 70% of patients to delay or interrupt therapy and 42% to be re-hospitalized. Adjuvant therapy was marred by a high frequency of gastrointestinal complications, with 26 of 43 (60%) patients suffering a grade 3 or higher gastrointestinal toxicity.5 During post-chemoradiation chemotherapy, 11 of 43 (26%) patients developed a grade 3 or higher gastrointestinal toxicity.5 Thus, despite this long-term survival, complications remained high.
Several multicenter trials were performed utilizing adjuvant chemoradiation with interferon-α. Results from the American College of Surgeons Oncology Group (ACOSOG) Trial Z05031 demonstrated a median overall survival of 25.4 months, which is congruent with our experience, but 95% of the study patients experienced a grade 3 or higher complication, and 44% were not able to complete all phases of therapy.12 The ChemoRadioImmunotherapy of pancreatic cancer (CapRI) trial of adjuvant chemoradiation and interferon-α versus fluorouracil and folinic acid demonstrated no difference in survival and also found that 85% of patients in the chemoradiation arm suffered a grade 3 or 4 toxicity.13 Furthermore, a single-center trial of adjuvant interferon-based chemoradiation reported by Katz et al. found the median overall survival to be 42.3 months but 89% of patients experienced a grade 3 toxicity and 11% of patients were hospitalized.14 Overall, a high rate of complications were consistent among all studies.
We sought to report our experience with a modified version of the Virginia Mason Protocol, utilizing 3-dimensional conformal radiation, decreasing doses of chemotherapeutics, and utilizing gemcitabine rather than 5-fluorouracil in the post-chemoradiation phase. The 5- and 10-year overall survival rates in our study were lower than those reported at Virginia Mason at 26% and 10%, respectively, with a median overall survival of 25 months. However, our patients had greater survival rates than predicted when risk adjusted using a pancreatic nomogram7 and were greater than historical controls. Interestingly, when we compared our results to those published by others we found different factors were associated with overall survival. For example, Virginia Mason investigators found that lymph node ratio < 50%, ECOG performance status of 0, and treatment continuance without delay were associated with improved survival while these were not significant in our study. In contrast we found that having a positive margin, or evidence of lymphovascular or perineural invasion were the strongest adverse prognostic factors. Some differences might be explained by differences in practice. For example, all of our patients required treatments to be held or reduced. There were two important similarities in these two experiences. Pancreatic cancer was not solved by this regimen as recurrence rates remained high in both studies, with 58% having disease recurrence in the Virginia Mason protocol compared to 73% of patients in our study. Systemic recurrence (metastases) as opposed to local recurrence was the site of first recurrence in both studies.
Tumor characteristics were assessed by univariate and multivariate analysis, demonstrating positive tumor margin and lymphovascular and perineural invasion were associated with a poor outcome. The role of positive margin status with tumor characteristics and recurrence was previously evaluated by our group. After careful pathologic review, Gnerlich et al. observed a high rate (34.0%) of positive microscopic margins and determined that positive posterior margin and lymph node involvement were associated with increased risk of local recurrence.15 Similarly, we found that 32.1% of patients in this study had a positive microscopic margin and determined that positive margin status was associated with unfavorable factors such as lymph node involvement, requirement for portal vein resection, and distal recurrence.
Despite an improvement in overall survival, there was a high rate of complications among our study patients. We previously reported 36 (67.9%) patients experienced a grade 3 or greater toxicity during adjuvant therapy, with 21 (40.4%) experiencing a hematologic toxicity such as leukopenia and neutropenia and 19 (36.5%) experiencing a gastrointestinal toxicity such as diarrhea or mucositis.6 The Virginia Mason protocol demonstrated a similar toxicity profile during interferon-based chemoradiation, with 26 (60%) suffering a grade 3 or 4 gastrointestinal toxicity and 24 (56%) suffering a grade 3 or 4 hematologic toxicity.
Given specific concerns for radiation-induced injuries addressed in our initial report, we sought to perform long-term follow-up of this cohort to determine if patients suffered any long-term effects that were possibly attributable to the combination of chemo-immuno-radiation therapy. We previously identified a high rate of toxicities during chemoradiation and post-chemoradiation gemcitabine which was consequential as it limited therapy, as all patients required a dose reduction or delay in therapy or developed late-term sequela. Late complications and related poor functional status are likely to affect tolerance of further therapy and enrollment in clinical trials, which is the preferred pathway for patients with recurrent disease after resection.16 These late effects of radiation-related problems may include gastrointestinal tract injuries such as gastritis or enteritis further complicated by perforation, fistula, stricture, or hemorrhage.17, 18 In our series, the most common late overall complications were infectious (17.5%) and gastrointestinal (15.0%). Furthermore, there was specific concern for radiation-induced injuries in several patients. However, as to our knowledge this is the first report to assess these severe late-term complications with relation to this protocol, we are not able to compare our findings. Our long-term follow-up did identify several additional patients who have suffered complications that may be related to their treatments. In this paper, we report several patients who developed such severe late sequelae necessitating additional surgical procedures for treatment of sepsis despite remaining recurrence-free.
Despite its relative effectiveness, most centers have abandoned this regimen because of early treatment-related toxicities. The presence of severe late complications further highlights the relative cost at which the small survival gains have been achieved, and our assessment has failed to identify points for modification in order to improve tolerance and decrease complications. All patients required dose reduction or delay in adjuvant therapy and the majority suffered a grade 3 or higher complication. While survival gains are real, this protocol and variations thereof will need to be carefully reformulated in order to minimize both early and late toxicities.
Supplementary Material
Supplemental Figure. Calibration of the Memorial Sloan-Kettering Cancer Center Pancreatic Adenocarcinoma Nomogram to the study patients. The x axis is the observed overall survival, and the y axis is the predicted calculated survival based on the nomogram. The diagonal line represents the performance of an ideal nomogram, and the means with standard error bars at 1 year, 2 years, and 3 years were plotted. Actual overall survival was obtained from Kaplan-Meier estimates and was higher than that predicted by the pancreatic nomogram.
Acknowledgments
Funding: Salary support for KAO was provided by the Washington University Surgical Oncology Training Grant (T32 CA009621).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Presented at the 50th Annual Pancreas Club Meeting, San Diego, California, May 20–21, 2016.
References
- 1.Kalser MH, Ellenberg SS. Pancreatic cancer. Adjuvant combined radiation and chemotherapy following curative resection. Arch Surg. 1985;120:899–903. doi: 10.1001/archsurg.1985.01390320023003. [DOI] [PubMed] [Google Scholar]
- 2.Gastrointestinal Tumor Study Group. Further evidence of effective adjuvant combined radiation and chemotherapy following curative resection of pancreatic cancer. Cancer. 1987;59:2006–2010. doi: 10.1002/1097-0142(19870615)59:12<2006::aid-cncr2820591206>3.0.co;2-b. [DOI] [PubMed] [Google Scholar]
- 3.Nukui Y, Picozzi VJ, Traverso LW. Interferon-based adjuvant chemoradiation therapy improves survival after pancreaticoduodenectomy for pancreatic adenocarcinoma. Am J Surg. 2000;179:367–371. doi: 10.1016/s0002-9610(00)00369-x. [DOI] [PubMed] [Google Scholar]
- 4.Picozzi VJ, Kozarek RA, Traverso LW. Interferon-based adjuvant chemoradiation therapy after pancreaticoduodenectomy for pancreatic adenocarcinoma. Am J Surg. 2003;185:476–480. doi: 10.1016/s0002-9610(03)00051-5. [DOI] [PubMed] [Google Scholar]
- 5.Rocha FG, Hashimoto Y, Traverso LW, Dorer R, Kozarek R, Helton WS, et al. Interferon-based Adjuvant Chemoradiation for Resected Pancreatic Head Cancer: Long-term Follow-up of the Virginia Mason Protocol. Ann Surg. 2016;263:376–384. doi: 10.1097/SLA.0000000000001190. [DOI] [PubMed] [Google Scholar]
- 6.Linehan DC, Tan MC, Strasberg SM, Drebin JA, Hawkins WG, Picus J, et al. Adjuvant interferon-based chemoradiation followed by gemcitabine for resected pancreatic adenocarcinoma: a single-institution phase II study. Ann Surg. 2008;248:145–151. doi: 10.1097/SLA.0b013e318181e4e9. [DOI] [PubMed] [Google Scholar]
- 7.Ferrone CR, Kattan MW, Tomlinson JS, Thayer SP, Brennan MF, Warshaw AL. Validation of a postresection pancreatic adenocarcinoma nomogram for disease-specific survival. J Clin Oncol. 2005;23:7529–7535. doi: 10.1200/JCO.2005.01.8101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Siegel RL, Miller KD, Jemal A. Cancer statistics, 2015. CA Cancer J Clin. 2015;65:5–29. doi: 10.3322/caac.21254. [DOI] [PubMed] [Google Scholar]
- 9.Stocken DD, Buchler MW, Dervenis C, Bassi C, Jeekel H, Klinkenbijl JH, et al. Meta-analysis of randomised adjuvant therapy trials for pancreatic cancer. Br J Cancer. 2005;92:1372–1381. doi: 10.1038/sj.bjc.6602513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Klinkenbijl JH, Jeekel J, Sahmoud T, van Pel R, Couvreur ML, Veenhof CH, et al. Adjuvant radiotherapy and 5-fluorouracil after curative resection of cancer of the pancreas and periampullary region: phase III trial of the EORTC gastrointestinal tract cancer cooperative group. Ann Surg. 1999;230:776–784. doi: 10.1097/00000658-199912000-00006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Neoptolemos JP, Stocken DD, Friess H, Bassi C, Dunn JA, Hickey H, et al. A randomized trial of chemoradiotherapy and chemotherapy after resection of pancreatic cancer. N Engl J Med. 2004;350:1200–1210. doi: 10.1056/NEJMoa032295. [DOI] [PubMed] [Google Scholar]
- 12.Picozzi VJ, Abrams RA, Decker PA, Traverso W, O’Reilly EM, Greeno E, et al. Multicenter phase II trial of adjuvant therapy for resected pancreatic cancer using cisplatin, 5-fluorouracil, and interferon-alfa-2b-based chemoradiation: ACOSOG Trial Z05031. Ann Oncol. 2011;22:348–354. doi: 10.1093/annonc/mdq384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Schmidt J, Abel U, Debus J, Harig S, Hoffmann K, Herrmann T, et al. Open-label, multicenter, randomized phase III trial of adjuvant chemoradiation plus interferon Alfa-2b versus fluorouracil and folinic acid for patients with resected pancreatic adenocarcinoma. J Clin Oncol. 2012;30:4077–4083. doi: 10.1200/JCO.2011.38.2960. [DOI] [PubMed] [Google Scholar]
- 14.Katz MH, Wolff R, Crane CH, Varadhachary G, Javle M, Lin E, et al. Survival and quality of life of patients with resected pancreatic adenocarcinoma treated with adjuvant interferon-based chemoradiation: a phase II trial. Ann Surg Oncol. 2011;18:3615–3622. doi: 10.1245/s10434-011-1847-4. [DOI] [PubMed] [Google Scholar]
- 15.Gnerlich JL, Luka SR, Deshpande AD, Dubray BJ, Weir JS, Carpenter DH, et al. Microscopic margins and patterns of treatment failure in resected pancreatic adenocarcinoma. Arch Surg. 2012;147:753–760. doi: 10.1001/archsurg.2012.1126. [DOI] [PubMed] [Google Scholar]
- 16.National Comprehensive Cancer Network. Pancreatic Adenocarcinoma (Version 1.2016) doi: 10.6004/jnccn.2019.0014. https://www.nccn.org/professionals/physician_gls/pdf/pancreatic.pdf. Accessed June 6, 2016. [DOI] [PubMed]
- 17.Galland RB, Spencer J. The natural history of clinically established radiation enteritis. Lancet. 1985;1:1257–1258. doi: 10.1016/s0140-6736(85)92322-0. [DOI] [PubMed] [Google Scholar]
- 18.Coia LR, Myerson RJ, Tepper JE. Late effects of radiation therapy on the gastrointestinal tract. Int J Radiat Oncol Biol Phys. 1995;31:1213–1236. doi: 10.1016/0360-3016(94)00419-L. [DOI] [PubMed] [Google Scholar]
- 19.Strasberg SM, Linehan DC, Hawkins WG. The accordion severity grading system of surgical complications. Ann Surg. 2009;250:177–186. doi: 10.1097/SLA.0b013e3181afde41. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Figure. Calibration of the Memorial Sloan-Kettering Cancer Center Pancreatic Adenocarcinoma Nomogram to the study patients. The x axis is the observed overall survival, and the y axis is the predicted calculated survival based on the nomogram. The diagonal line represents the performance of an ideal nomogram, and the means with standard error bars at 1 year, 2 years, and 3 years were plotted. Actual overall survival was obtained from Kaplan-Meier estimates and was higher than that predicted by the pancreatic nomogram.