Journal of Gastrointestinal Surgery

, Volume 15, Issue 8, pp 1411–1416

The Influence of Prognostic Factors and Adjuvant Chemoradiation on Survival After Pancreaticoduodenectomy for Ampullary Carcinoma

Authors

  • Timothy N. Showalter
    • Department of Radiation OncologyJefferson Medical College of Thomas Jefferson University, Kimmel Cancer Center
  • Tingting Zhan
    • Division of Biostatistics, Department of Pharmacology and Experimental TherapeuticsThomas Jefferson University
  • P. Rani Anne
    • Department of Radiation OncologyJefferson Medical College of Thomas Jefferson University, Kimmel Cancer Center
  • Inna Chervoneva
    • Division of Biostatistics, Department of Pharmacology and Experimental TherapeuticsThomas Jefferson University
  • Edith P. Mitchell
    • Department of Medical OncologyJefferson Medical College of Thomas Jefferson University, Kimmel Cancer Center
  • Charles J. Yeo
    • Department of SurgeryJefferson Medical College of Thomas Jefferson University, and the Jefferson Pancreas, Biliary and Related Cancer Center
  • Ernest L. Rosato
    • Department of SurgeryJefferson Medical College of Thomas Jefferson University, and the Jefferson Pancreas, Biliary and Related Cancer Center
  • Eugene P. Kennedy
    • Department of SurgeryJefferson Medical College of Thomas Jefferson University, and the Jefferson Pancreas, Biliary and Related Cancer Center
    • Department of SurgeryJefferson Medical College of Thomas Jefferson University, and the Jefferson Pancreas, Biliary and Related Cancer Center
Original Article

DOI: 10.1007/s11605-011-1518-6

Cite this article as:
Showalter, T.N., Zhan, T., Anne, P.R. et al. J Gastrointest Surg (2011) 15: 1411. doi:10.1007/s11605-011-1518-6

Abstract

Introduction

The prognosis after pancreaticoduodenectomy (PD) for ampullary carcinoma (AC) is superior to that of pancreatic cancer. Decisions regarding adjuvant therapy are influenced by factors such as nodal status, stage, and grade, but the influence of these individual variables on survival is unclear.

Methods

A prospective tumor registry database was queried to identify patients who underwent PD for AC at Thomas Jefferson University between Jan 1997 and Apr 2009. The study was conducted with the approval of the institutional review board. Data were collected through review of hospital and departmental charts. Overall survival (OS) was analyzed using univariate and multivariate Cox proportional hazard models. The proportional hazard assumption was verified for the overall model and individual covariates.

Results

A total of 61 patients underwent PD for AC at our institution. There were five perioperative deaths (8.2%). Mean age was 70 years (62% male). Median survival time (MST) was 50 months for all patients. Only primary tumor stage, T1/T2 versus T3/T4 (American Joint Committee on Cancer Staging, version 6), was associated with OS in univariate analyses (p = 0.003). The association of nodal status with OS was borderline-significant (p = 0.08), with the MST being 84 months for node-negative and 17 months for node-positive patients. The remaining covariates were not predictors of OS. In the multivariate analysis, only primary tumor stage (HR, 5.1; p < 0.001) and age (HR, 1.04; p = 0.06), but not nodal status or adjuvant therapy, were associated with overall survival.

Conclusions

Advanced primary tumor stage and age were associated with inferior OS after PD for AC. Adjuvant therapy did not impact survival. Patients with advanced tumor stage should be considered for clinical trials of adjuvant therapy after PD with novel compounds and optimized radiation therapy strategies.

Keywords

Ampullary cancerAdjuvant therapyAmpulla of VaterAdenocarcinoma

Introduction

Carcinoma of the ampulla of Vater is a relatively uncommon malignancy and accounts for approximately 10% to 20% of pancreaticoduodenectomies (PDs) performed for periampullary cancers.13 Prognosis after PD for ampullary carcinoma (AC) is superior to that of pancreatic cancer (PC), with higher rates of complete surgical resection, lower recurrence rates, and longer survival.1 Published 5-year overall survival (OS) rates for ampullary carcinoma range from 38% to 68%.17 There are less published data available for AC than for PC to assist with decisions regarding adjuvant therapy. The only randomized, prospective trial of adjuvant radiotherapy and concurrent chemotherapy to include patients with AC was the one reported by the European Organization for Research and Treatment of Cancer (EORTC).8 This study randomized patients with cancers of the pancreatic head or the periampullary region, defined as the distal common bile duct, ampulla of Vater, or duodenum, to the addition of adjuvant chemoradiotherapy (CRT) after surgery versus surgery alone. The study results showed that CRT did not improve OS rates in this trial. Results of this study are controversial, but they do not lend support for adjuvant therapy for either PC or AC.8 Retrospective, single-institution studies have suggested conflicting results for the role of CRT in AC; most notably that adjuvant CRT extends survival either only in patients with involved lymph nodes,5 or only in patients with advanced tumor stage.9

Previous retrospective studies have identified clinical and pathologic factors that predict poorer outcome after PD for AC, in an effort to identify high-risk patients who may benefit most from adjuvant therapy. Reported prognostic factors include specimen lymph node involvement, poor histologic grade, positive resection margins, and higher tumor stage (T3/T4 vs. T1/T2).1,3,5,7,10,11 This analysis of outcomes for patients treated for AC at Thomas Jefferson University Hospital was performed to evaluate the associations of reported prognostic factors with overall survival after PD, and to assess the impact of adjuvant CRT on overall survival.

Methods

A prospective tumor registry database was queried to identify patients who underwent PD for AC at Thomas Jefferson University between Jan 1997 and Apr 2009. The study was conducted with the approval of the institutional review board. Data were collected through chart review and included age, gender, perioperative complications, margin status, stage, adjuvant therapy, survival, and recurrence. Tumor staging was performed according to the 7th edition of the American Joint Committee on Cancer Staging Manual (Tables 1 and 2). During the study period, decisions regarding adjuvant therapy were made by the treating physicians without a consistent policy, but the tendency was to reserve CRT for those patients with high-risk features such as positive nodes. Based on institutional treatment policies, intraoperative radiation therapy (IORT) was also considered in addition to CRT for any patient who received PD at our institution until 1998 and was subsequently reserved for those patients thought to have higher risk of positive margins.12
Table 1

Pathologic staging system for ampullary carcinoma used in the seventh edition of the American Joint Committee on Cancer staging system

Primary tumor (T)

TX

Primary tumor cannot be assessed

T0

No evidence of primary tumor

Tis

Carcinoma in situ

T1

Tumor confined in ampulla of Vater or sphincter of Oddi

T2

Tumor involves duodenal wall

T3

Tumor invades pancreas

T4

Tumor invades peripancreatic soft tissue or other organs

Regional lymph nodes (N)

NX

Regional lymph nodes cannot be assessed

N0

No regional lymph node metastasis

N1

Regional lymph node metastasis

Distant metastasis (M)

MX

Distant metastasis cannot be assessed

M0

No distant metastasis

M1

Distant metastasis

Table 2

Stage grouping for ampullary carcinoma used in the seventh edition of the American Joint Committee on Cancer staging system

Stage grouping

Stage

T

N

M

0

Tis

N0

M0

IA

T1

N0

M0

IB

T2

N0

M0

IIA

T3

N0

M0

IIB

T1–3

N1

M0

III

T4

Any N

M0

IV

Any T

Any N

M1

Time to recurrence was estimated from the date of surgical resection, and the location of first recurrence was recorded. The recurrence rate for node-positive tumors was compared with node-negative tumors using Fisher’s exact test. Overall survival (OS) was estimated from the date of surgical resection. The overall survival was analyzed using univariate and multivariate Cox proportional hazard models and reported including 95% confidence interval (CI). For statistical analysis, tumor (T) stage was dichotomized as “low” (T1/T2) versus “high” (T3/T4), and the histological grade was dichotomized as “low” (well and moderately differentiated) versus “high” (poorly differentiated). Factors included in the multivariate analysis were age, primary tumor stage (T1/T2 vs. T3/T4), grade, nodal status (N0 vs. N1), adjuvant CRT, and adjuvant chemotherapy. The final multivariate Cox model was selected using the Akaike Information Criterion.13 The proportional hazard assumption was verified for the overall model and for the individual covariates. Data were analyzed in R 2.10 (R Foundation for Statistical Computing, http://www.R-project.org).

Results

Sixty-one patients were identified who underwent PD for AC at our institution between Jan 1997 and Apr 2009. There were five perioperative deaths (8.2%), with the causes being pulmonary embolism in one patient, respiratory failure in one patient, and cardiac events in three patients. Median follow-up of surviving patients was 22.4 months. Average age was 70 years and 62% were male. Negative microscopic margins (R0) were achieved in 58 patients (95%). Positive margins (n = 3) were identified at the uncinate (n = 1) and circumferential (n = 2) margins of resection. Sixteen patients (29%) received adjuvant RT, which was delivered in combination with concurrent 5-fluorouracil (5-FU) or capecitabine. External beam RT doses used during CRT were 45–52.2 Gy in daily fractions of 1.8 Gy, with a median dose of 50.4 Gy. Five patients treated with adjuvant CRT also received IORT boost (median dose, 15 Gy) after completion of external beam RT. A total of 25 patients received adjuvant chemotherapy alone, with or without CRT, including gemcitabine or 5-FU (Table 3).
Table 3

Age, tumor, and adjuvant therapy information for 61 ampullary carcinoma patients who were treated with pancreaticoduodenectomy

Factor

Value

Age

 Mean (year (interquartile range))

70 (54–76)

Grade (N (%))

 Well differentiated

6 (10%)

 Moderately differentiated

36 (59%)

 Poorly differentiated

16 (26%)

 Unknown

3 (5%)

Stage (N (%))

 I

19 (31%)

 II

29 (48%)

 III

12 (20%)

 IV

1 (2%)

Nodal status (N (%))

 N0

20 (33%)

 N1

41 (67%)

Surgical margin status (N (%))

 R0 (negative)

58 (95%)

 R1 (microscopic positive)

3 (5%)

 R2 (macroscopic positive)

0 (0%)

Adjuvant chemoradiotherapy (N (%))

 Yes

16 (26%)

 No

36 (59%)

 Unknown

9 (15%)

Adjuvant chemotherapy (N (%))

 Yes

25 (41%)

 No

28 (46%)

 Unknown

8 (13%)

Recurrence

Tumor recurrence was observed in 20 patients (36%) at a median of 10.5 months after surgery. Distant metastases were present at the time of first recurrence in 13 patients, including liver in 11 patients and lungs in three patients. Regional lymph nodes were involved at time of first recurrence for six patients. Initial recurrence of AC was noted to be locoregional only, without any distant dissemination, for four patients. Three of 20 (15%) node-negative patients developed recurrence, which occurred in regional lymph nodes (n = 1) and liver (n = 2). Seventeen of 41 (41%) node-positive patients developed recurrence, which occurred in regional lymph nodes (n = 5), liver (n = 9), and lungs (n = 3). The recurrence rate for node-negative patients was significantly lower than for node-positive patients (p = 0.03).

Survival

The median survival time (MST) was 50 months for all patients, with 2- and 5-year OS rates of 64% (95% CI, 52–80%) and 44% (95% CI, 30–65%), respectively (Fig. 1). The 2- and 5-year OS rates were 85% (95% CI, 73–100%) and 65% (95% CI, 46–92%), respectively, for T1/T2 tumors (n = 33) versus 39% (95% CI, 23–67%) and 20% (95% CI, 6–60%) for T3/T4 tumors (n = 28) (p = 0.003). The MST was not reached for the T1/T2 group, and was 16 months for the T3/T4 group (Fig. 2). The MST, 2- and 5-year OS rates were 84 months, 85% (95% CI, 71–100%) and 60% (95% CI, 38–94%), respectively, for node-negative tumors, versus 17 months, 50% (95% CI, 34–72%) and 33% (95% CI, 17–65%) for node-positive tumors (Fig. 3) (p = 0.081). As concerns OS, by univariate analyses, neither tumor grade, age, gender, adjuvant chemotherapy, nor adjuvant CRT were associated with overall survival. However, tumor stage was associated with OS (p = 0.003), and nodal status demonstrated a borderline-significant association (p = 0.08) (Table 4). In multivariate analysis, only tumor stage (T1/2 vs. T3/4) was significantly associated with OS, with HR 5.08 (95% CI, 1.95–13.25, p < 0.001), and age had borderline-significant association, with HR 1.04 (95% CI 1.00–1.08, p = 0.06). Tumor grade, nodal status, adjuvant CRT, and adjuvant chemotherapy were not significant independent predictors of survival.
https://static-content.springer.com/image/art%3A10.1007%2Fs11605-011-1518-6/MediaObjects/11605_2011_1518_Fig1_HTML.gif
Fig. 1

Kaplan–Meier plot of overall survival (solid line) for all ampullary carcinoma patients (n = 61) after pancreaticoduodenectomy, with 95% confidence interval (dashed line)

https://static-content.springer.com/image/art%3A10.1007%2Fs11605-011-1518-6/MediaObjects/11605_2011_1518_Fig2_HTML.gif
Fig. 2

Kaplan–Meier plot of overall survival for ampullary carcinoma patients after pancreaticoduodenectomy by primary tumor stage, comparing T1/T2 tumors (solid line, n = 33) to T3/T4 tumors (dashed line, n = 28; p = 0.003)

https://static-content.springer.com/image/art%3A10.1007%2Fs11605-011-1518-6/MediaObjects/11605_2011_1518_Fig3_HTML.gif
Fig. 3

Kaplan–Meier plot of overall survival for ampullary carcinoma patients after pancreaticoduodenectomy by nodal status, comparing lymph node negative (N0, solid line, n = 20) to lymph node positive (N1, dashed line, n = 41; p = 0.081)

Table 4

Univariate Cox proportional models of overall survival

Variable

Hazard ratio (95% confidence interval)

p value

Tumor stage (T1/2 vs. T3/4)

4.05 (1.6–10.1)

0.003

Grade (well/moderate vs. poor)

1.8 (0.8–4.5)

0.183

Age

1.0 (1.0–1.1)

0.225

Nodal status (negative vs. positive)

2.2 (0.9–5.4)

0.081

Adjuvant chemoradiotherapy (yes vs. no)

1.2 (0.5–3.0)

0.713

Adjuvant chemotherapy alone (yes vs. no)

0.9 (0.4–2.0)

0.741

Gender (male vs. female)

0.6 (0.2–1.4)

0.235

Discussion

In the current study, tumor stage and age were associated with OS, and the MST was not reached for T1/T2 tumors. A similar association between tumor stage and OS has been reported in previous series, including recent reports from MD Anderson Cancer Center9 and Johns Hopkins Hospital14 that showed inferior OS for T3 and T4 AC tumors compared with T1 and T2 tumors. The median survival seen in the current series compares favorably with that found at Hopkins (50 vs. 36 months). Additionally, 5-year survival rates were similar for lymph node-negative (60%) and T1/T2 tumors (65%) to those seen in the group reported by M.D. Anderson. The significant association between age and OS differs from the findings of other investigators, including Krishnan et al. and Yeh et al., who found that advanced age was not associated with inferior survival after PD for AC.9,15 Conclusions regarding the influence of age on OS for AC are limited due to the relatively small numbers of patients in these series. Lymph node involvement, which has been shown to be associated with OS in multiple series,1,3,5,16,17 was a borderline-significant predictor of OS in the current series, with median survival time of 84 months for N0 patients versus 17 months for N1 patients. In our experience, lymph node-negative AC patients had lower rates of recurrence and better overall survival than patients with nodal involvement. The current study is limited by its retrospective design, with non-randomized treatment allocation, as well as the relatively small number of patients treated for this uncommon cancer diagnosis. It is possible that a significant association between nodal status and OS would be revealed with a larger population of AC patients.

The role of adjuvant CRT in the management of AC is not clear. In the present study, adjuvant CRT did not influence OS. Interpretation of the observed lack of effect of CRT on OS is limited by the small number of patients in this study; less than 50% of patients in our series received any adjuvant therapy. The EORTC prospective trial of CRT for periampullary cancers included 92 patients with AC. In this EORTC trial, which comprised the only prospective, randomized data for CRT in AC, CRT did not improve OS, with a 2-year survival rate of 70% versus 64% with and without CRT, respectively.8 Similarly, Zhou and colleagues, in their 2009 report of 111 AC patients treated at the Johns Hopkins Hospital, also observed no clinical benefit from CRT.14 A recent report by Krishnan et al., from MD Anderson, observed that CRT improved survival for AC patients with T3/T4 tumors, but not for patients with involved lymph nodes.9 On the other hand, adjuvant CRT has been reported by other authors to improve survival for AC patients with positive lymph nodes.5,18 Future studies that include a larger number of patients may provide further insight into which subsets of AC patients will benefit from adjuvant CRT.

Previous reports have focused on the identification of high-risk factors that predict for higher rates of recurrence, using either models of multiple pathologic factors or lymph-node-based parameters, such as number of positive nodes and lymph node ratio.10,16,17 Given the influence of lymph node status on rates of recurrence and survival after PD for AC, it seems reasonable to consider closely these lymph node parameters when evaluating prognosis. Recent data have demonstrated that primary tumor stage and nodal status are correlated in AC, with increased risk of nodal metastasis for T3/4 (71.1–77.3%) versus T1/2 (28.0–50.9%) tumors.19 It is unclear whether these prognostic approaches may be used as the basis for decisions regarding adjuvant CRT for AC patients. Although the current analysis suggests that tumor stage, age, and lymph node involvement are associated with overall survival, the addition of adjuvant CRT after PD did not improve survival. Since they are associated with inferior survival after PD, these adverse factors should be considered when developing novel strategies for adjuvant therapy for AC patients.

In conclusion, only primary tumor stage, and not nodal status or the use of adjuvant CRT, was associated with OS after PC for AC. Patients with advanced primary tumor stage should be considered for clinical trials of adjuvant therapy with novel compounds and optimized RT strategies, while observation following surgical resection may be considered for patients with T1/2 and node-negative AC.

Copyright information

© The Society for Surgery of the Alimentary Tract 2011