DOI: https://doi.org/10.21203/rs.3.rs-1769995/v1
Background: We analyzed clinicopathologic variables and postoperative outcomes in patients with ampulla of Vater carcinoma (AVC) to identify key predictors of post-resection prognosis.
Methods: Fifty-five patients who underwent pancreaticoduodenectomy for AVC with curative intent between 2005 and 2020 were included in this study. A retrospective review of the clinical records of patients who underwent surgical exploration for AVC was performed using univariate and multivariate analyses. Clinical and pathological factors that influenced patient survival were analyzed.
Results: One-, 3-, and 5-year overall survival rates after surgery were 97.4%, 71.8%, and 63.0%, respectively. A total of 21 patients developed recurrent disease after curative resection for AVC. The most frequently observed sites of recurrence were lymph nodes in 11 patients (52%), followed by the liver in 8 (38%), lung in 6 (29%), local in 3 (14%), and peritoneal dissemination in 3 (14%). Estimated 3- and 5-year recurrence free survival rates for the entire cohort were 62.8% and 57.2%, respectively. On multivariate analysis, only the presence of lymph node metastasis extending to the pancreatic head region (hazard ratio = 5.374; 95% confidence interval: 1.279–22.58; P = 0.022) predicted inferior relapse-free survival. A significant correlation between postoperative recurrence and pathological lymph node metastasis was observed.
Conclusions: Lymph node metastasis, especially that which extends to the pancreatic head region, was clearly identified as a prognostic indicator of reduced relapse-free survival in patients who have undergone curative surgical treatment for AVC.
Ampulla of Vater carcinoma (AVC) is an uncommon malignancy, accounting for only 0.2% of gastrointestinal cancers [1]. Because AVC arises from the end of the biliary tract, typical clinical symptoms such as abdominal pain or jaundice can be detected at a relatively early stage. The radical resection rate of AVC is high, and its prognosis is better than that of other periampullary carcinomas, with a relative 5-year survival rate of 30–70%. Hence, oncologic resection of the tumor together with lymphadenectomy is currently the gold standard for AVC treatment [2, 3]. Therefore, macroscopic or microscopic tumor-free tumor resection is likely a prerequisite for favorable survival in patients with AVC [4].
In previous reports, factors predicting survival among patients with AVC included tumor node metastasis (TNM) stage, resection margin, lymph node metastasis (LNM), vascular invasion, perineural involvement, differentiation grade, and perioperative nutritional status [5–9]. Although the effects of recurrence on postoperative mortality have been well studied in patients with periampullary cancers, such as pancreatic and distal bile duct cancers, there are few analyses of its impact on AVC [10, 11]. The recurrence rate in patients with AVC has been reported to range from 23–73% and manifests in the form of metastases to the lymph node (LN), including locoregional recurrence, liver metastases, and lung metastases [8].
Whether LN resection improves survival in patients with AVC involving LNM remains controversial, although such a strategy is now well established for local infiltration, with no deleterious effects on perioperative morbidity or mortality evident as yet [12]. Thus, to prevent death from metastatic recurrent disease, this study aimed to clarify the postoperative prognosis of AVC with pathologic LNM. The resulting data provide novel insights into clinicopathological features underlying AVC progression and define a broad time window of opportunity for accurate surgical indication.
Each initial diagnosis of AVC was made following imaging, and confirmed via pathological analysis. We retrospectively reviewed the surgical pathology database of Kochi Health Sciences Center to identify patients who underwent resection of pancreatic neoplasms between March 2005 and December 2020. This study was approved by the ethics committee of the Kochi Health Sciences Center. Written informed consent was obtained from all patients.
Physical status and preoperative laboratory values were obtained within the two-week period preceding the initiation of surgery. Body mass index (BMI) was calculated by dividing body weight in kilograms by the square of the patient’s height in meters. Each surgically resected specimen was examined macroscopically to determine maximum tumor diameter. Resected specimens were evaluated by two expert pathologists according to guidelines of the eighth edition of the Union for International Cancer Control/American Joint Committee on Cancer staging system [13]. Tumor size, stage, degree of differentiation, vascular invasion, lymphatic permeation, perineural invasion, and LNM were assessed.
Postoperative adjuvant therapy was not performed as a standard treatment. Patients who participated in clinical trials to evaluate the efficacy of adjuvant therapy and were assigned to treatment groups were administered S-1 adjuvant therapy postoperatively [14]. The prognostic value of LMN was evaluated based on the following three assessments; 1. total number of LNM (classified as either none, one, or more than 2); 2. metastatic LN ratio (number of metastatic LNs divided by the total number retrieved); and 3. the location of LNM (pancreatic head region [LN panc] or extended to pancreatic head [LN ex-panc]).The primary endpoint evaluated was prognostic factors of surgically resected AVCs. The secondary endpoint was recurrence patterns in patients with AVC after curative resection, particularly regarding the importance metastasis to the LN region. Furthermore, we evaluated the treatment of patients who experienced recurrence after surgical resection for AVC and the possibility of operative management in patients with recurrent AVC. Postoperative morbidities including postoperative pancreatic fistula were defined according to both the classification system of the International Study Group of Pancreatic Surgery (ISGPF) and the Clavien-Dindo (C-D) classification of surgical complications [15, 16]. Postoperative recurrence was examined via abdominal ultrasonography and/or computed tomography.
This study prospectively collected and retrospectively analyzed data from a biostatistician (TI). Overall survival (OS) and relapse-free survival (RFS) were calculated from the date of curative surgery for AVC to the date of death and the date of documentation of recurrent disease, respectively, for those with recurrent disease. For those without recurrent disease, OS was calculated from the date of curative surgery to the date of death. Patients who died of causes unrelated to the disease were censored at the last follow-up visit. Categorical variables were compared using the Fisher’s exact test. Survival was estimated using the Kaplan-Meier method. A multivariate Cox proportional hazards model was used to examine the effect of pretreatment physiobiological values as continuous variables on RFS. All statistically significant preoperative and perioperative variables were included in the model. All P-values reported are two-sided, with an alpha level of 0.05 considered statistically significant. Statistical analyses were performed using the EZR software (Saitama Medical Center, Jichi Medical University, Saitama, Japan).
In this study period, 692 patients underwent pancreaticoduodenectomy as an initial treatment for peri-ampullary diseases. Of these patients, 55 (8%) who were surgically treated for AVC were initially identified (Table 1). Clinicopathological features of included patients (40 men [73%]; median age, 70 years [range, 46–84] years) are summarized in Table 1. The median operative time was 277 min (range, 163–508 min), and the median blood loss was 350 mL (range, 25–2,115 mL). No in-hospital deaths (0%) occurred among these patients. The overall morbidity rate evaluated according to C-D classification guidelines for major complications (grade III-V) was 33%, and the incidence of postoperative pancreatic fistula (grade B/C) based on ISGPS criteria was 35% (Table 1). The median hospital stay was 25 days (range, 10–153 days). No other operative mortalities or operation-related readmissions were observed in the present study.
Pathologic database data of selected patients revealed a mean tumor size of 2.6 cm (± 1.6 cm). Among patients considered, 17 (31%), 23 (42%), and 15 (27%) were determined to possess tumors of UICC T Stage classification T1, T2, and T3, respectively. Further, 19 patients (35%) were LN-positive (Table 2). Histopathologic grade was undifferentiated carcinoma in only four (7%) patients (Table 2). Regarding the malignant pathologic behavior of tumors, lymphatic permeation was observed in 24 patients (44%), venous invasion in 12 (22%), and perineural involvement in 9 (16%) (Table 2).
The duration of patient follow-up, as of January 2021, ranged from 22 to 87 months with a median value of 48.0 months (mean: 57.2 months). One-, 3-, and 5-year OS rates after surgery were 97.4%, 71.8%, and 63.0%, respectively (Fig. 1A). Estimated 3- and 5-year RFS rates for the entire cohort were 62.8% and 57.2%, respectively (Fig. 1B). A total of 21 patients developed recurrent disease after curative resection for AVC. The major recurrence sites were the LNs in 11 patients (52%), followed by the liver in 8 (38%), lung in 6 (29%), local in 3 (14%), and peritoneal dissemination in 3 (14%). Although curative resection was the operative aim for all patients, 11 (52.4%) experienced recurrence within 24 months of curative resection for AVC. The median time to recurrence was 10.5 months (range 2–60 months). The time to recurrence seemed to be shorter when tumor recurrence was observed locally or in the liver versus other areas (Table 3).
Interestingly, there was a significant correlation between recurrence pattern and pathologic LNM (Spearman rank correlation coefficient = 0.473, P < 0.01). The most frequent sites of LNM were determined to be the pancreatic head region; LN panc (LN groups #13 and #17, as defined by the Japanese Society of Hepato-Biliary-Pancreatic Surgery staging system [JHBP system]), followed by LN ex-panc; LN-mesenteric (LN groups #14 defined by JHBP system), and the pyloric region (LN-pyloric; LN groups #5, #6, #8, and/or #12, as per the JHBP system) [17]. LN positive ratio (positive LNs/total number of retrieved LNs) was greatest for LN panc (25.3%), in contrast to LN ex-panc, including mesenteric LN (10.8%) and pyloric LN (1.7%).
The associations between physio-biological characteristics and RFS are detailed in Table 4. Advanced T stage (hazard ratio [HR] = 4.95; 95% CI: 1.15–21.27; P < 0.05) and pathological N stage (HR = 3.41; 95% CI: 1.43–8.14; P < 0.05), especially in LN ex-panc (HR = 5.04; 95% CI: 2.05–12.40; P < 0.05), positive lymphatic permeation (HR = 2.87; 95% CI: 1.16–7.12; P < 0.05), and high LN metastatic status (HR = 3.56; 95% CI: 1.37–9.24; P < 0.01) were associated with worsened RFS. In contrast, age, sex, BMI, ASA, pre-treatment cholangitis, serological findings such as CEA and CA19-9, histologic grade, and pathologic perineural involvement were not significantly associated with outcome (Table 4). On multivariate analysis, only the presence of LNM extended to pancreatic head (LN ex-panc) (HR = 5.374; 95% CI: 1.279–22.58; P = 0.022) predicted inferior RFS (Table 5).
During the study period, of the 21 patients who experienced disease recurrences, six (28.6%) underwent radical surgery for metastatic tumors, three hepatectomy for liver metastasis, two additional lymphadenectomy for local recurrence, and one pulmonary resection for lung metastasis. Eleven patients (42.9%) with whole-body recurrence received chemotherapy or chemoradiotherapy, and the remaining 4 patients (19.0%) received supportive care. The median overall survival rate of patients with recurrence was 14 months, and the 3-and 5-years OS rates were 29.2% and 23.4%, respectively (Fig. 2A). Median OS was 78, 11, and 4 months in the surgery, systemic therapy, and best supportive care groups, respectively (Fig. 2B). The subgroup analyses of survival according management strategy in patients with recurrence showed significant differences between surgical treatment and conservative therapy, although there were strong biases regarding patient selection.
This study showed that pathologic LNM, especially LNM extending to the pancreatic head lesion, predicts poor RFS in patients with AVC. Moreover, the current study demonstrated that the presence of LNM extending to the pancreatic head region is associated with tumor recurrence after curative surgical resection of AVC. Therefore, this is the first study to show a significant correlation between pathologic LNM and tumor recurrence after curative operative management of AVC.
In previous studies, several predictors of LNM in resected AVC have been described including number of retrieved LNs, ratio of positive LNs/retrieved LNs, number of metastatic LNs (based on UICC N stage classification), and location of regional LNs (pancreatic head or except for pancreas head) [5,6,18-20]. Matui et al. reported that regional LNM, except to the pancreatic head region, was associated with poor prognosis, comparable to that of patients with M1-stage [21]. Our results suggest that LNM extending to the pancreatic head region, especially to the superior mesenteric region, is an important unfavorable prognostic factor. We found that the pancreatic head and superior mesenteric region predicted a high LN positivity rate (positive LN/retrieved LN ratio; 25.4% and 10.8%, respectively), and six out of eight patients in the LN-mesenteric cohort had local or LN recurrence. Accurate pretreatment staging is essential for making an informed diagnosis of truly curable AVC. Unfortunately, preoperative staging methods such as endoscopic ultrasonography, computed tomography, and magnetic resonance imaging are only moderately accurate. Hence, more tools are needed to stratify patients before initiating treatment. Questions regarding whether lymphadenectomy affects RFS after curative surgical management remain, especially when considering its demonstrated impact on both local and distant metastatic recurrence. To determine the feasibility and effectiveness of LN dissection for resectable AVC, epidemiologic and pathologic data from a multi-institutional database should be considered.
Our results showed that the recurrence rate of AVC after pancreaticoduodenectomy remains high. Thus far, no AVC treatment strategies have been shown to be sufficiently effective [22]. Chemotherapy remains the principal systemic treatment for recurrent AVC after surgery or advanced-stage unresectable AVC, although optimal integration of fluorouracil-based agents remains debatable [22]. Surgical resection of the metastatic region is the only curative treatment option; however, many patients have multiple metastatic regions and are not amenable to surgery [23]. Favorable selection of patients with a single oligometastatic region might be indicative of curative surgery, and therefore promote the achievement of long-term survival. In our study cohort, 6 of 21 patients who experienced relapse (28.6%) were candidates for radical surgery and achieved long-term survival (3-year OS was 80%). The results of this research suggest that adjuvant and/or neoadjuvant strategies are required for the preoperative diagnosis of AVC with clinical LNM.
This work is part of a multicenter, prospective, randomized phase III trial, JCOG1920, which aims to examine the role of neoadjuvant chemotherapy and adjuvant chemotherapy using oral fluoropyrimidine (S-1) platinum-based agents and gemcitabine in patients diagnosed with potentially curative AVC via surgical resection. The final publication of this study is eagerly anticipated, particularly due to its potential relevance to AVC patients. Further, the design, setting, participants, and methodology used will potentially inform subsequent research.
Nonetheless, this study had several limitations. This was a retrospective cohort review of patients undergoing surgical resection that included postoperative patients exclusively. Therefore, the study design confers a degree of selection bias owing to the selection of patients who underwent surgical management. Among single-center databases, AVC is a rare disease and the sample size available for analysis was small. In fact, only 19 patients had LNM. Further, LNM to the pancreatic head region and LNM to an extended pancreatic head region occurred in only 6 and 13 patients, respectively. Patients with postsurgical LNMs were mainly administered S-1 adjuvant chemotherapy according to guidelines outlined by our expert team.
In conclusion, LNM was clearly identified as a prognostic indicator of reduced RFS in patients with post-surgery AVC. Locoregional recurrence was the most common type of failure, and a significant correlation between node-positive AVC and postoperative local recurrence, liver metastasis, and lung metastasis was observed.
Acknowledgment: This work was supported by the Kochi organization for medical reformation and renewal grants.
Author’s contributions: Okabayashi T contributed design of the work. Kimura J, Sui K, and Noda Y research data. Iiyama T and Iwata J analyzed data. Tabuchi M and Okabayashi T wrote the manuscript and researched data. Okabayashi T reviewed/edited the manuscript. Okabayashi T contributed to discussion and reviewed/edited the manuscript.
Data availability: The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
Conflict of interest: The authors declare that they have no competing interests.
Funding sources: This work was supported by Kochi Organization for Medical Reformation and Renewal grants.
Ethics statement: The present study complied with the standards of the Declaration of Helsinki on human research ethics and was approved by the Ethics Committee of Kochi Health Sciences Center.
Disclosure statement: no financial disclosure.
Acknowledgment: We would like to thank Editage (www.editage.com) for English language editing.
Table 1 Main clinical characteristics of patients in study registry
Demographics |
|
Number (n = 55) |
Clinical valuables |
|
|
Gender (female/male) |
|
15/40 |
Age (median [range]) |
|
70 (46-84) |
Body mass index (kg/m2, median [range]) |
|
23.1 (16.3-31.8) |
ASA (1/2/3) |
|
11/32/12 |
Total bilirubin (mg/dL, median [range]) |
|
1.1 (0.1-31.2) |
C-reative protein (mg/dL, median [range]) |
|
1.6 (0.2-36.3) |
CEA (ng/mL, median [range]) |
|
2.6 (2.0-4.1) |
CA19-9 (U/mL, median [range]) |
|
25.6 (8.4-72.7) |
Cholangitis |
|
22 |
Surgical valuables |
|
|
Operation time (min, median [range]) |
|
277 (163-508) |
Blood loss (mL, median [range]) |
|
350 (25-2115) |
C-D grade >3 |
|
18 |
POPF* (grade B/C) |
|
17/2 |
Hospital stay (days, median [range]) |
|
25 (10-153) |
ASA; American Society of Anesthesiologists Physical Status, CEA; carcinoembryonic antigen, CA19-9; carbohydrate antigen 19-9, C-D; Clavien-Dindo classification.
*POPF; postoperative pancreatic fistula according to the International Study Group on Pancreatic Surgery classification.
Table 2 Pathologic characteristics of patients in study registry
Demographics |
|
Number (n = 55) |
Tumor size (cm, + SD) |
|
2.6 + 1.6 |
T stage*, n (%) |
|
|
1 |
|
17 (31) |
2 |
|
23 (42) |
3 |
|
15 (27) |
N stage*, n (%) |
|
|
0 |
|
36 (66) |
1 |
|
10 (18) |
2 |
|
9 (16) |
Differentiation |
|
|
Differentiated |
|
51 (93) |
Undifferentiated |
|
4 (7) |
Vascular invasion |
|
13 (24) |
Lymphatic permeation |
|
2 (4) |
Perineural involvement |
|
10 (18) |
*evaluated according to the guidelines of the eighth edition of the Union for International Cancer Control/American Joint Committee on Cancer staging system.
Table 3 Patterns of recurrence and time to recurrence of AVC after curative surgery
Recurrent site |
No. of Pts |
Mean time to diagnosis (month, +SD) |
Lymph node |
11 |
14 + 12 |
Liver |
8 |
8 + 4 |
Lung |
6 |
16 + 15 |
Local |
3 |
4 + 2 |
Peritoneal dissemination |
3 |
29 + 16 |
SD; standard deviation.
Table 4 Association of relapse free survival with characteristics in AVC patients
Characteristics |
N |
% |
HR |
95% CI |
P value |
Gender |
|
|
|
|
|
Female |
15 |
27 |
|
|
|
Male |
40 |
73 |
2.12 |
0.71-6.30 |
0.18 |
Age |
|
|
|
|
|
<70 |
26 |
47 |
|
|
|
>70 |
29 |
53 |
0.90 |
0.38-2.11 |
0.80 |
Body mass index (kg/m2) |
|
|
|
|
|
<25 |
27 |
49 |
|
|
|
>25 |
28 |
51 |
0.96 |
0.53-1.73 |
0.89 |
ASA |
|
|
|
|
|
1 |
11 |
20 |
|
|
|
2 |
32 |
58 |
|
|
|
3 |
12 |
22 |
0.87 |
0.45-1.67 |
0.68 |
CEA (ng/mL) |
|
|
|
|
|
<5 |
45 |
73 |
|
|
|
>5 |
10 |
27 |
1.43 |
0.47-4.31 |
0.53 |
CA19-9 (U/mL) |
|
|
|
|
|
<37 |
30 |
55 |
|
|
|
>37 |
25 |
45 |
1.94 |
0.80-4.69 |
0.14 |
Cholangitis |
|
|
|
|
|
Absent |
33 |
60 |
|
|
|
Present |
22 |
40 |
0.98 |
0.41-0.24 |
0.97 |
T stage* |
|
|
|
|
|
1 |
16 |
29 |
|
|
|
2/3 |
39 |
71 |
4.95 |
1.15-21.27 |
0.04 |
N stage |
|
|
|
|
|
Negative |
36 |
65 |
|
|
|
Positive |
19 |
35 |
3.41 |
1.43-8.14 |
0.03 |
Differentiation |
|
|
|
|
|
Differentiated |
51 |
93 |
|
|
|
Undifferentiated |
4 |
7 |
1.49 |
0.35-6.40 |
0.59 |
Vascular invasion |
|
|
|
|
|
Absent |
42 |
76 |
|
|
|
Present |
13 |
24 |
1.91 |
0.77-4.75 |
0.16 |
Lymphatic permeation |
|
|
|
|
|
Absent |
31 |
56 |
|
|
|
Present |
24 |
24 |
2.87 |
1.16-7.12 |
0.04 |
Perineural involvement |
|
|
|
|
|
Absent |
45 |
73 |
|
|
|
Present |
10 |
27 |
2.10 |
0.81-5.46 |
0.13 |
Metastatic LN ratio |
|
|
|
|
|
<0.1 |
47 |
85 |
|
|
|
>0.1 |
8 |
15 |
3.56 |
1.37-9.24 |
0.01 |
LN ex-panc** |
|
|
|
|
|
Absent |
45 |
73 |
|
|
|
Present |
10 |
27 |
5.04 |
2.05-12.40 |
0.01 |
ASA; American Society of Anesthesiologists Physical Status, CEA; carcinoembryonic antigen, CA19-9; carbohydrate antigen 19-9, LN; lymph node.
*evaluated according to the guidelines of the eighth edition of the Union for International Cancer Control/American Joint Committee on Cancer staging system.
**the location of lymph node metastasis extended to pancreatic head.
Table 5 Association of relapse free survival with clinic-pathologic characteristics in patients who underwent pancreaticoduodenectomy
Demographics |
HR |
95% CI |
P value |
Higher T stage |
2.88 |
0.62-5.81 |
0.19 |
Positive LNM |
0.40 |
0.10-1.55 |
0.18 |
Vascular invasion |
1.91 |
0.63-5.81 |
0.25 |
Lymphatic permeation |
1.33 |
0.42-4.26 |
0.62 |
Metastatic LN ratio >0.1 |
2.46 |
0.47-12.77 |
0.28 |
LNM ex-panc* |
12.51 |
1.81-86.65 |
0.01 |
LNM; lymph node metastasis.
*the location of lymph node metastasis extended to pancreatic head.