View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Available online at www.sciencedirect.com

Journal of the Chinese Medical Association 75 (2012) 573e580 www.jcma-online.com Original Article Number of involved lymph nodes is important in the prediction of prognosis for primary duodenal

Tsung-Jung Liang a,b, Being-Whey Wang b,c, Shiuh-Inn Liu b,c, Nan-Hua Chou b, Cheng-Chung Tsai b, I-Shu Chen b, Ming-Hsin Yeh b, Yu-Chia Chen b, Po-Min Chang b, King-Tong Mok b,c,*

a Department of Surgery, Kaohsiung Veterans General Hospital, Pingtung Branch, Pingtung, Taiwan, ROC b Division of , Department of Surgery, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, ROC c National Yang-Ming University School of Medicine, Taipei, Taiwan, ROC

Received December 22, 2011; accepted April 16, 2012

Abstract

Background: The significance of lymph node involvement regarding the prognosis of primary duodenal adenocarcinoma remains controversial. This study aims to evaluate the prognostic accuracy of nodal metastasis using the seventh edition American Joint Committee on staging system in patients with primary duodenal adenocarcinoma. Methods: Between 1993 and 2010, 36 patients who had undergone surgical resection for primary duodenal adenocarcinoma at the Kaohsiung Veterans General Hospital were retrospectively reviewed. Results: The median disease-free survival for all patients was 19 months and the median overall survival was 21 months. Lymph node metastases were found in 26 (72%) of the patients, and 14 patients (39%) patients had in excess of three positive lymph nodes (N2). Patients with N2 disease had significantly reduced overall survival, as compared to patients with three or fewer positive lymph nodes (N1; p ¼ 0.036). In univariate analysis, factors including age >75 years, body weight loss, tumor size 4 cm, N2 disease and lymph node ratio >0.4 predicted shorter overall survival. Multivariate analysis demonstrated that N2 and lymph node ratio >0.4 are significant risk factors associated with overall survival ( p ¼ 0.026 and p ¼ 0.042 respectively). N2 is also the only independent predictive factor for disease-free survival ( p ¼ 0.023). Conclusion: Subdivision of metastatic lymph nodes into N1 and N2 improves predictive ability. The seventh edition American Joint Committee on Cancer staging system is applicable in the present study with regard to the prediction of the prognosis for primary duodenal adenocarcinoma. Copyright Ó 2012 Elsevier Taiwan LLC and the Chinese Medical Association. All rights reserved.

Keywords: American Joint Committee on Cancer; duodenal adenocarcinoma; lymph node metastasis; small bowel

1. Introduction distal duct .2e4 Studies that focus only on duodenal adenocarcinoma are scarce.5,6 Correspondingly, Primary duodenal adenocarcinoma is an uncommon tumor many questions regarding disease pathogenesis, natural that accounts for less than 0.5% of all gastrointestinal malig- history, ideal medical management, and prognostic factors are nancies.1 Because of the rarity of this type of cancer, it has still unanswered. typically been reported together with jejunal and ileal tumors Until now, the significance of lymph node involvement as a small bowel cancer, or in some cases grouped with per- regarding the prognosis of primary duodenal adenocarcinoma iampullary , along with pancreatic, ampullary, and remains controversial.7 However, some studies show a strong negative impact on survival in lymph node metastatic disease.6,8 The most common tool used to evaluate lymph * Corresponding author. Dr. King-Tong Mok, Division of General Surgery, Department of Surgery, Kaohsiung Veterans General Hospital, 386, Ta-Chung node metastasis is the American Joint Committee on Cancer First Road, Kaohsiung 813, Taiwan, ROC. (AJCC) tumor-node-metatasis (TNM) staging system. Lymph E-mail address: [email protected] (K.-T. Mok). node staging is divided into two groups: no regional lymph

1726-4901/$ - see front matter Copyright Ó 2012 Elsevier Taiwan LLC and the Chinese Medical Association. All rights reserved. http://dx.doi.org/10.1016/j.jcma.2012.08.002 574 T.-J. Liang et al. / Journal of the Chinese Medical Association 75 (2012) 573e580 node metastasis (N0) and positive lymph node metastasis (N1) Patients had regular postoperative follow-up examinations. according to the sixth edition of the AJCC TNM classification Tumor recurrence was identified using various diagnostic tools system for small intestinal cancer published in 2002.9 In 2010, such as sonography, computed tomography, magnetic reso- the release of the seventh edition of AJCC classification nance imaging, positron emission tomography scan, and whole included major modifications to the lymph node staging body bone scan. Survival time and pattern of recurrence were system.10 The previous N1 group is further subdivided into recorded. Distant metastasis was suspected if the distal lesion new N1 (metastasis in 1e3 nodes) and N2 (metastasis in 4 or was newly detected and did not appear in the preoperative more nodes). As a result, the Stage III disease is subclassified survey. Biopsy might be considered if the lesion was not in into Stage IIIA (TxN1M0) and Stage IIIB (TxN2M0). any of the customary metastatic sites where duodenal adeno- However, the clinical correlation of this change has yet to be carcinoma is found (e.g. liver, , lung). Informed determined. consent was obtained from all patients before operation. The Apart from the involved lymph node, studies from other study complied with institutional review board protocols. cancers have suggested that the total number of lymph nodes Statistical analyses were performed using SPSS 18.0 for harvested is also an important indicator for prognosis because Windows (SPSS, Inc., Chicago, IL, USA). Disease-free an insufficient number evaluated may not provide an accurate survival and overall survival were estimated by use of the node staging.11e13 Consequently, the idea of lymph node ratio Kaplan-Meier method and differences were compared by log- (LNR) was introduced to eliminate the variation of dissection rank test. Multivariate comparisons were conducted using the techniques by different surgeons and the number of lymph Cox proportional hazards model. Statistical significance was node harvested. This ratio is defined as the ratio of metastatic considered at p < 0.05. nodes to the total number of lymph node examined. Evidence presented in existing literature has demonstrated that the LNR 3. Results is inversely related to survival in esophageal, gastric, and colon cancers.14e16 3.1. Patient characteristics In this study, we aimed to clarify the prognostic impact of lymph node status, including AJCC node staging, the total The clinicopathological data of enrolled patients were number of lymph node examined, and the lymph node ratio in shown in Table 1. Of the 36 patients reviewed, 24 (67%) were patients with primary duodenal adenocarcinoma treated with men and 12 (33%) were women. The median age at diagnosis curative resection at a single medical center. was 66 years (range 35e88). Most of the tumors were located at the secondary portion of the (70%), followed by eight (22%) at the first portion, two (5%) at the third portion, 2. Methods and one case was found at the fourth portion. Also, 29 (80%) patients underwent classic (Whip- The medical records of all patients with primary duodenal ple procedure) and two (6%) patients received pylorus- adenocarcinoma who had undergone surgical resection at preserving pancreaticoduodenectomy (PPPD). Segmental Kaohsiung Veterans General Hospital between 1993 and 2010 duodenal resections were performed in five (14%) patients, were reviewed retrospectively. All patients enrolled in this four of whom had a tumor at the first portion, and one patient study had primary tumor located in the duodenum, and had a tumor at the fourth portion of the duodenum. All patient histologically proven adenocarcinoma. Tumors arising from surgical margins were disease-free, except for one patient in ampulla of Vater, , and were excluded. the Whipple procedure group. Grossly, this patient’s resection Patients whose pathologic reports were not available were also margins were free of tumor, but tumor cells were found excluded from the study. microscopically in the distal pancreatic cut end. The AJCC A total of 36 patients met the inclusion criteria. Data for nodal staging of this male patient was N2 and his LNR was each patient were collected, including age at diagnosis, clin- 0.686 (11/16). He suffered from local recurrence 8 months ical presentation, type of surgery, postoperative course, and the later. use of adjuvant therapy. Anemia was defined as hemoglobin The median tumor size was 4 cm in diameter and the depth level <130 g/L in males, and <110 g/L in females. Pathologic of tumor invasion was T4 in 72% of all patients. The median specimens were reviewed for tumor location, size, tumor number of lymph nodes harvested was 16 (range 5e38), and depth, histology, grade, surgical margin, pancreatic invasion, 10 (28%) patients had no metastatic lymph nodes with an LNR number of metastatic lymph node, and number of lymph nodes of zero. Among the 26 patients who had positive lymph nodes, harvested. The lymph node ratio was calculated by dividing 12 patients had one to three positive lymph nodes (N1) and the number of metastatic lymph nodes by the total number of were classified as Stage IIIA. The other 14 patients had more lymph nodes assessed. Patients were then subclassified into than three positive lymph nodes (N2) and were classified as four groups based on LNR (LNR ¼ 0; 0 < LNR 0.2; Stage IIIB. When categorized by LNR, there were 11 (30%) 0.2 < LNR 0.4; 0.4 < LNR). The cutoff value for each patients with LNR from 0 to 0.2, nine (25%) patients with group was determined by previous studies.3,17,18 The seventh LNR from 0.2 to 0.4 and six (17%) with LNR > 0.4. All of the edition of AJCC classification for malignant of the eight (22%) patients who received adjuvant had was used for staging. node-positive disease. T.-J. Liang et al. / Journal of the Chinese Medical Association 75 (2012) 573e580 575

Table 1 (according to the current AJCC staging system) was 50%, 45% Clinicopathologic features of 36 patients with primary duodenal and 0%, respectively, ( p < 0.001; Fig. 1). Patients with N2 adenocarcinoma. disease had a significantly shorter overall survival rate Feature n (%) compared with patients with N0 and N1 disease ( p < 0.001 and Age, y, median (range) 66 (35e88) p ¼ 0.036, respectively). However, no significant difference was Male sex 24 (67) detected in the overall survival between N0 and N1 disease Tumor location ( p ¼ 0.238). Additionally, LNR is shown to be inversely First 8 (22) Second 25 (70) associated with overall survival when categorized into four Third 2 (5) groups (LNR ¼ 0; 0 < LNR 0.2; 0.2 < LNR 0.4; Fourth 1 (3) 0.4 < LNR). Their 3- and 5-year overall survival rate were 90%, Type of surgery 31%, 29%, 0%, 50%, 31%, 0%, and 0%, ( p ¼ 0.011). However, Whipple 29 (80) there is no one distinctive LNR group that has a significant PPPD 2 (6) Segmental resection 5 (14) difference in overall survival in comparison with the other three Surgical margin groups (Fig. 2). Positive 1 (3) Nineteen out of the 36 patients (53%) had recurrence of Negative 35 (97) their cancer. The median time to recurrence after operation e Tumor size, cm, median (range) 4 (1.0 9.0) was 9 months (range 3e46). Local recurrence occurred in Tumor depth (T-stage) T1 1 (3) five (14%) cases and distant metastasis in six (17%) cases. T2 4 (11) Concurrent local and distant metastases were documented in T3 5 (14) the remaining eight (22%) cases. The liver was the most T4 26 (72) a common site of distant metastasis followed by the perito- Nodal status neum, lung, and bone. Palliative chemotherapy was given to N0 10 (28) N1 12 (33) three of the 19 recurrent patients (16%). Disease-free survival N2 14 (39) rates at 3 and 5 years for all 36 patients were 39% and 35%, TNM stageb respectively. Both the AJCC node classification and LNR I 4 (11) were significantly associated with disease-free survival IIA 2 (6) ( p ¼ 0.021 and 0.040, respectively; Fig. 3 and Fig. 4). In IIB 4 (11) IIIA 12 (33) subgroup analysis, N2 disease had shorter median disease- IIIB 14 (39) free survival compared with N1 disease, but it did not IV 0 (0) reach statistical significance (8 months vs. 13 months, Lymph node harvested, median (range) 16 (5e38) p ¼ 0.174). Lymph node ratio (LNR) 0 10 (28) 0 < LNR 0.2 11 (30) 0.2 < LNR 0.4 9 (25) <0.4 6 (17) Grade Well 3 (8) Moderate 20 (56) Poor 13 (36) Adjuvant chemotherapy Yes 8 (22) No 28 (78) Recurrence No 17 (47) Local 5 (14) Distant 6 (17) Local þ Distant 8 (22) PPPD ¼ pylorus-preserving pancreaticoduodenectomy. a N0, no positive lymph node; N1, 1e3 positive lymph nodes; N2, >3 positive lymph nodes. b TNM stage: American Joint Committee on Cancer TNM staging system, seventh edition.

3.2. Survival stratified by AJCC N classification and lymph node ratio

The median follow-up duration was 41 months for patients Fig. 1. Overall survival stratified by seventh edition American Joint Committee who survived and 17 months for patients who died. The 5-year on Cancer (AJCC) N stage. ( p < 0.001). N0 ¼ node negative; N1 ¼ 1e3 overall survival rate of patients with N0, N1, and N2 disease positive nodes; N2 ¼ >3 positive nodes. 576 T.-J. Liang et al. / Journal of the Chinese Medical Association 75 (2012) 573e580

Fig. 2. Overall survival stratified by lymph node ratio ( p ¼ 0.011). Fig. 4. Disease-free survival stratified by lymph node ratio ( p ¼ 0.040). LNR ¼ lymph node ratio. LNR ¼ lymph node ratio.

3.3. Prognostic factors associated with survival grade, T stage, tumor location, total lymph node examined, LNR > 2, pancreatic invasion, and the use of adjuvant Univariate analysis showed that patient factors including chemotherapy had no significant influence on overall survival. age >75 years ( p ¼ 0.020), body weight loss ( p ¼ 0.037), Subsequent multivariate analysis using 2 different models tumor size 4cm(p ¼ 0.041), AJCC N2 disease ( p < 0.001) were performed to avoid a confounding effect by putting the (Fig. 5) and LNR > 0.4 ( p ¼ 0.014) were associated with two lymph node factors (N2 disease, LNR > 0.4) together. decreased overall survival (Table 2). Gender, anemia, tumor The results demonstrated that N2 disease (hazard ratio ¼ 3.48,

Fig. 3. Disease-free survival stratified by seventh edition American Joint Committee on Cancer (AJCC) N stage ( p ¼ 0.021). N0 ¼ node negative; Fig. 5. Overall survival stratified by positive lymph node > 3 and 3 N1 ¼ 1e3 positive nodes; N2 ¼ >3 positive nodes. ( p < 0.001). T.-J. Liang et al. / Journal of the Chinese Medical Association 75 (2012) 573e580 577

Table 2 Table 3 Univariate analysis for factors associated with overall survival. Multivariate analysis for factors associated with overall survival. Factors N Median Survival p Factors Hazard ratio 95% CI p survival (%) Model 1 (mo) 3-y 5-y Age >75 y 2.79 0.89e8.77 0.079 e Age (y) Body weight loss 2.06 0.87 4.85 0.100 Tumor size 4 cm 1.48 0.56e3.90 0.432 75 29 23 50 30 0.020 > e >75 7 5 17 17 3 LN metastases (N2) 3.48 1.18 10.22 0.024 Gender Male 24 18 45 25 0.886 Model 2 Age >75 y 4.78 1.58e14.50 0.006 Female 12 21 42 31 e Anemiaa Body weight loss 2.60 1.12 6.06 0.026 e Yes 22 23 49 38 0.291 Tumor size 4 cm 2.20 0.91 5.28 0.078 LNR > 0.4 4.25 1.38e13.10 0.012 No 14 18 35 9 Body weight loss LN ¼ lymph node; LNR ¼ lymph node ratio; CI ¼ confidence interval. Yes 12 10 18 9 0.037 No 24 46 57 35 Grade remained as an independent factor predicting tumor recurrence Well þ Moderate 24 21 45 35 0.199 (hazard ratio ¼ 3.36, p ¼ 0.023; Table 5). Poor 12 18 44 0 T stage T1 þ T2 þ T3 10 56 70 35 0.296 4. Discussion T4 26 18 33 22 Tumor size (cm) 4 21 18 30 10 0.040 The principal goal of this study was to determine whether >415616655LN status, especially the new AJCC N-stage was associated Tumor location with survival. Lymph node metastasis is a well-known prog- Second portion 25 18 43 22 0.743 nostic factor for many gastrointestinal and periampullary Other portion (first, third,fourth) 11 23 47 35 .19e22 When LN assessment occurs in the context Lymph node metastasis, n 0e3 (N0, N1) 22 56 67 43 <0.001 of duodenal adenocarcinoma, however, there is some 7 >3 (N2) 14 13 8 0 disagreement in this area. Our data, in accordance with many Total lymph nodes examined other reports, have shown that patients with nodal metastases <12 9 18 50 50 0.296 have diminished survival ( p < 0.001).6e8,23e25 However, 12 27 21 43 21 other studies have found no such relationship.1,26,27 The Lymph node ratio 0.2 21 55 57 34 0.286 incidence of lymph node metastasis in patients with duodenal >0.2 15 17 24 12 adenocarcinoma has been reported to range from 22% to Lymph node ratio 0.4 30 41 51 31 0.014 >0.4 6 13 0 0 Pancreatic invasion Yes 18 16 28 17 0.224 No 18 55 64 37 Adjuvant chemotherapy Yes 8 16 25 13 0.218 No 28 46 50 32 a Anemia was defined as hemoglobin level <130 g/L in male and <110 g/L in female respectively. p ¼ 0.024) and LNR > 0.4 (hazard ratio ¼ 4.25, p ¼ 0.012) were significant prognostic predictors (Table 3). With respect to recurrence, anemia ( p ¼ 0.004), poor tumor grade ( p ¼ 0.017), tumor size 4cm(p ¼ 0.008), AJCC N2 disease ( p ¼ 0.010; Fig. 6), LNR > 0.4 ( p ¼ 0.026) and pancreatic invasion ( p ¼ 0.028) were found to have a signifi- cant negative impact on disease-free survival by univariate analysis (Table 4). Similarly, 2 separate models were con- structed using the Cox proportional hazards model to assess the influence of lymph node evaluation in disease-free survival. When multivariate analysis was employed, > LNR 0.4, along with anemia, tumor grade, and pancreatic Fig. 6. Disease-free survival stratified by positive lymph node >3 and 3 invasion lost their prognostic significance. Only N2 disease ( p ¼ 0.010). 578 T.-J. Liang et al. / Journal of the Chinese Medical Association 75 (2012) 573e580

Table 4 Table 5 Univariate analysis for factors associated with disease-free survival. Multivariate analysis for factors associated with disease-free survival. Factors N Median Survival p Factors Hazard ratio 95% CI p survival (mo) (%) Model 1 3-y 5-y Anemia 2.31 0.61e8.69 0.216 e Age (y) Grade, poor 1.56 0.48 5.07 0.459 Tumor size 4 cm 1.07 0.25e4.49 0.928 75 29 14 41 37 0.801 > e >75 7 19 33 33 3 LN metastasis (N2) 3.36 1.18 9.55 0.023 e Gender Pancreatic invasion 1.87 0.60 5.75 0.278 Male 24 27 43 37 0.644 Model 2 Female 12 13 33 33 e Anemia Anemia 1.70 0.42 6.88 0.461 e Yes 22 NR 58 58 0.004 Grade, poor 1.44 0.44 4.75 0.549 Tumor size 4 cm 1.88 0.43e8.27 0.406 No 14 10 15 8 LNR > 0.4 2.23 0.66e7.53 0.199 Body weight loss e Yes 12 12 33 33 0.638 Pancreatic invasion 1.47 0.47 4.65 0.508 No 24 27 42 36 LN ¼ lymph node; LNR ¼ lymph node ratio. Grade Well þ Moderate 24 34 50 50 0.017 Poor 12 9 20 0 with a single metastatic node and patients with negative lymph T stage nodes. However, they did find that patients with no more than T1 þ T2 þ T3 10 46 63 47 0.248 T4 26 13 31 31 one metastatic lymph node survived significantly longer than Tumor size (cm) patients with two or more involved lymph nodes. They 4 21 11 19 13 0.008 concluded that the presence of two or more positive lymph >415NR6868nodes was a major prognostic factor after resection of Tumor location pancreatic carcinoma. These results also implied that a small Second portion 25 19 37 31 0.661 Other portion (first, third,fourth) 11 13 44 44 number of positive lymph nodes might not affect clinical Lymph node metastases, n outcome, and that a subdivision according to the number of 0e3 (N0, N1) 22 NR 56 51 0.010 metastatic lymph nodes should be considered to provide better >3 (N2) 14 8 0 0 prognostic value. Total lymph nodes examined Regarding small bowel adenocarcinoma, Overman et al.32 <12 9 14 50 50 0.564 12 27 19 36 31 demonstrated evident improvement in prognostic efficacy by Lymph node ratio stratifying patients with nodal metastasis into those patients 0.2 21 27 47 42 0.528 with fewer than three positive lymph nodes and those with >0.2 15 14 17 17 three or more positive lymph nodes (hazard ratio 1.44; 95% Lymph node ratio confidence interval 1.12e1.86). The 5-year disease-specific 0.4 30 34 46 41 0.026 >0.4 6 5 0 0 survival rates of the two subgroups were 58% and 37% Pancreatic invasion respectively. The selected cutoff point for the number of Yes 18 11 28 21 0.028 positive lymph node in their study was slightly different No 18 NR 54 54 compared with the seventh edition of AJCC N-staging (3 vs. Adjuvant chemotherapy >3 positive lymph nodes). Because in Overman’s study, Yes 8 12 13 13 0.159 No 28 46 51 45 patients with tumors located in , ileum, and non- specified locations were all included. When all circumstances NR ¼ not reached median survival rates of 50.0%. were factored in, it could not be determined whether more than three positive lymph nodes is the cutoff value that has the 76%,6,7,28 and 72% of our patients had lymph node involve- best prognostic success in patients with duodenal adenocar- ment. Given the relatively high percentage of nodal metastasis cinoma. Based on our study, we clearly demonstrated that in our study, it is reasonable that our 5-year survival rate of having more than three lymph node metastases (N2) is the 26% appeared to be lower in compared with other series most important factor to predict survival after curative resec- (25e75%).7 tion. To our knowledge, this study is the first to validate the In this study, we found that the overall survival rate of significant association between survival and N-staging of the patients with one to three positive lymph nodes (N1) did not seventh edition AJCC classification in patients with duodenal differ significantly from that of patients without lymph node adenocarcinoma. metastasis (N0; p ¼ 0.238). On the contrary, more than three There were five (14%) patients in our series who received lymph nodes involved (N2) was an independent predictor of segmental resection of the duodenum. Some investigators have overall and disease-free survival ( p ¼ 0.024 and p ¼ 0.023, proposed a pancreaticoduodenectomy procedure for all respectively). Similar results had been found in patients with duodenal adenocarcinoma regardless of tumor location.1 pancreatic carcinoma.22,29e31 Zacharias et al31 reported no Others suggest duodenal segmentectomy as a less morbid statistically significant difference in survival between patients choice for tumors located in the first, third, or fourth portions T.-J. Liang et al. / Journal of the Chinese Medical Association 75 (2012) 573e580 579 of the duodenum.8,25,33 Inadequate surgical margins and analysis (overall survival, p ¼ 0.040; disease-free survival, lymph node clearance are the two major concerns for limited p ¼ 0.008). This seemingly counterintuitive result has also resection.34 Kaklamanos et al23 justified the use of segmental been observed by Hurtuk et al,40 who found that larger tumors resection by showing a comparable number of lymph nodes were less likely to behave aggressively. On the contrary, removed, whether a pancreaticoduodenectomy or a duodenal smaller tumors had a predilection to invade periduodenal fat or segmentectomy was performed. A similar result was also nearby structures. Furthermore, when they focused on tumors found in our study ( p ¼ 0.436). Furthermore, no patient in our that invaded nearby structures, patients with larger tumors still limited resection group had surgical margins involved by showed better prognosis than those with smaller tumors. The tumor. In all, our data support the use of the segmental authors explained those results by hypothesizing that there resection procedure in selected patients with duodenal were differences in biological behaviors between duodenal adenocarcinoma. adenocarcionomas and some tumors that tend to be less The total number of lymph nodes examined did not affect aggressive. Further study is required to clarify the biologic survival in our study (overall survival, p ¼ 0.296; disease-free nature of this tumor. However, in the meantime, the authors survival, p ¼ 0.564) and in a small series reported by Struck suggested patients undergo aggressive surgical resection et al.35 Nevertheless, several investigators proposed that an regardless of the tumor size.40 inadequate number of lymph nodes removed may impair In conclusion, our data validate the predictive power of prognostic discrimination due to inaccurate assessment of nodal metastasis using N classification of the seventh edition lymph node metastasis, which might understage cancer AJCC staging system. More than three positive lymph nodes patients.3,24,32 The correlation between survival and the (N2 disease) is the most important prognostic factor in patients number of total lymph node evaluated have also been shown in with primary duodenal adenocarcinoma after curative resec- esophageal, gastric, pancreatic, and colorectal cancers.11e13,36 tion. The prognostic efficacy of LNR is promising and worthy However, the exact number of lymph nodes that needed to be of further larger scale research. removed varied from study to study. For example, in pancre- atic adenocarcinoma, the proposed number of lymph nodes References that should be collected was reported to be 10, 12, or 15 13,17,37 nodes. 1. Sohn TA, Lillemoe KD, Cameron JL, Pitt HA, Kaufman HS, Hruban RH, In a retrospective review of 522 patients with resected et al. Adenocarcinoma of the duodenum: factors influencing long-term periampullary malignancies, Hurtuk et al3 found that LNR is survival. J Gastrointest Surg 1998;2:79e87. inversely associated with survival in pancreatic and ampullary 2. Wu TJ, Yeh CN, Chao TC, Jan YY, Chen MF. Prognostic factors of primary small bowel adenocarcinoma: univariate and multivariate anal- cancer, but not in bile duct and duodenal cancer. In contrast, ysis. World J Surg 2006;30:391e8. our results showed that LNR is predictive of disease-free 3. Hurtuk MG, Hughes C, Shoup M, Aranha GV. Does lymph node ratio survival and overall survival in patients with primary impact survival in resected periampullary malignancies? Am J Surg duodenal adenocarcinoma by categorizing patients into four 2009;197:348e52. groups using the same LNR cutoff values of 0, 0.2, and 0.4 4. Dabaja BS, Suki D, Pro B, Bonnen M, Ajani J. Adenocarcinoma of the ¼ ¼ > small bowel: presentation, prognostic factors, and outcome of 217 ( p 0.011 and p 0.040 respectively). Besides, LNR 0.4 patients. Cancer 2004;101:518e26. was identified as an independent prognostic factor for overall 5. Lee CC, Ng WK, Lin KW, Lai TW, Li SM. Adenocarcinoma of the survival ( p ¼ 0.012). Two studies of small bowel adenocar- duodenum. Hong Kong Med J 2008;14:67e9. cinoma also reported significantly reduced survival rates in 6. Bakaeen FG, Murr MM, Sarr MG, Thompson GB, Farnell MB, patients with LNR 0.5 and LNR 0.75, respectively.4,38 Nagorney DM, et al. What prognostic factors are important in duodenal adenocarcinoma? Arch Surg 2000;135:635e41. Theoretically, LNR is more discriminative than the number 7. Solej M, D’Amico S, Brondino G, Ferronato M, Nano M. Primary of positive lymph nodes (N-stage) in prognostic prediction duodenal adenocarcinoma. Tumori 2008;94:779e86. because it incorporates the extent of lymphadenectomy. 8. Lee HG, You DD, Paik KY, Heo JS, Choi SH, Choi DW. Prognostic Furthermore, LNR may also eliminate the variation on spec- factors for primary duodenal adenocarcinoma. World J Surg 2008;32: e imen examination by different pathologists (e.g., the number 2246 52. 17 9. Greene FL, Page DL, Fleming ID, Fritz A, Balch CM, Haller DG. AJCC of lymph nodes harvested). Certain reports of cases of cancer staging manual: TNM classification of malignant tumors. 6th ed. gastric and breast cancer have already demonstrated the New York: Springer-Verlag; 2002. prognostic efficacy of LNR and its superiority over current 10. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. AJCC AJCC nodal staging categories.15,39 In the current study, more cancer staging handbook. 7th ed. New York: Springer-Verlag; 2010. than three positive lymph nodes (N2 disease) exhibited prog- 11. Altorki NK, Zhou XK, Stiles B, Port JL, Paul S, Lee PC, et al. Total number of resected lymph nodes predicts survival in . nostic value in both disease-free and overall survival rates, Ann Surg 2008;248:221e6. while LNR >0.4 failed to influence disease-free survival using 12. Smith DD, Schwarz RR, Schwarz RE. Impact of total lymph node count multivariate analysis. Based on our findings, the number of on staging and survival after gastrectomy for gastric cancer: data from metastatic lymph nodes was the more potent prognostic factor a large US-population database. J Clin Oncol 2005;23:7114e24. after resection of primary duodenal adenocarcinoma. 13. Hellan M, Sun CL, Artinyan A, Mojica-Manosa P, Bhatia S, Ellenhorn JD, et al. The impact of lymph node number on survival in patients with Another interesting finding of the present study was that lymph node-negative . Pancreas 2008;37:19e24. smaller tumors (4 cm) were associated with a significantly 14. Tachibana M, Dhar DK, Kinugasa S, Kotoh T, Shibakita M, Ohno S, et al. poorer outcome than were larger ones (>4 cm) in univariate Esophageal cancer with distant lymph node metastasis: prognostic 580 T.-J. Liang et al. / Journal of the Chinese Medical Association 75 (2012) 573e580

significance of metastatic lymph node ratio. J Clin Gastroenterol 27. Rose DM, Hochwald SN, Klimstra DS, Brennan MF. Primary duodenal 2000;31:318e22. adenocarcinoma: a ten-year experience with 79 patients. J Am Coll Surg 15. Bando E, Yonemura Y, Taniguchi K, Fushida S, Fujimura T, Miwa K. 1996;183:89e96. Outcome of ratio of lymph node metastasis in gastric carcinoma. Ann Surg 28. Gibbs JF. Duodenal adenocarcinoma: is total lymph node sampling Oncol 2002;9:775e84. predictive of outcome? Ann Surg Oncol 2004;11:354e5. 16. De Ridder M, Vinh-Hung V, Van Nieuwenhove Y, Hoorens A, Sermeus A, 29. Massucco P, Ribero D, Sgotto E, Mellano A, Muratore A, Capussotti L. Storme G. Prognostic value of the lymph node ratio in node positive colon Prognostic significance of lymph node metastases in pancreatic head cancer. Gut 2006;55:1681. cancer treated with extended lymphadenectomy: not just a matter of 17. Pawlik TM, Gleisner AL, Cameron JL, Winter JM, Assumpcao L, numbers. Ann Surg Oncol 2009;16:3323e32. Lillemoe KD, et al. Prognostic relevance of lymph node ratio following 30. Riediger H, Keck T, Wellner U, zur Hausen A, Adam U, Hopt UT, et al. pancreaticoduodenectomy for pancreatic cancer. Surgery 2007;141: The lymph node ratio is the strongest prognostic factor after resection of 610e8. pancreatic cancer. J Gastrointest Surg 2009;13:1337e44. 18. Kawai M, Tani M, Kobayashi Y, Tsuji T, Tabuse K, Horiuchi T, et al. The 31. Zacharias T, Jaeck D, Oussoultzoglou E, Neuville A, Bachellier P. Impact ratio between metastatic and examined lymph nodes is an independent of lymph node involvement on long-term survival after R0 pan- prognostic factor for patients with resectable middle and distal bile duct creaticoduodenectomy for ductal adenocarcinoma of the pancreas. J carcinoma. Am J Surg 2010;199:447e52. Gastrointest Surg 2007;11:350e6. 19. Hornick JR, Johnston FM, Simon PO, Younkin M, Chamberlin M, 32. Overman MJ, Hu CY,Wolff RA, Chang GJ. Prognostic value of lymph node Mitchem JB, et al. A single-institution review of 157 patients presenting evaluation in small bowel adenocarcinoma: analysis of the surveillance, with benign and malignant tumors of the ampulla of Vater: management epidemiology, and end results database. Cancer 2010;116:5374e82. and outcomes. Surgery 2011;150:169e76. 33. Tocchi A, Mazzoni G, Puma F, Miccini M, Cassini D, Bettelli E, et al. 20. Lee JH, Lee KG, Ha TK, Jun YJ, Paik SS, Park HK, et al. Pattern Adenocarcinoma of the third and fourth portions of the duodenum: results analysis of lymph node metastasis and the prognostic importance of of surgical treatment. Arch Surg 2003;138:80e5. number of metastatic nodes in ampullary adenocarcinoma. Am Surg 34. Lai EC, Doty JE, Irving C, Tompkins RK. Primary adenocarcinoma of the 2011;77:322e9. duodenum: analysis of survival. World J Surg 1988;12:695e9. 21. Murakami Y, Uemura K, Hayashidani Y, Sudo T, Ohge H, Sueda T. 35. Struck A, Howard T, Chiorean EG, Clarke JM, Riffenburgh R, Pancreatoduodenectomy for distal : prognostic impact Cardenes HR. Non-ampullary duodenal adenocarcinoma: factors impor- of lymph node metastasis. World J Surg 2007;31:337e42. tant for relapse and survival. J Surg Oncol 2009;100:144e8. 22. Murakami Y, Uemura K, Sudo T, Hayashidani Y, Hashimoto Y, 36. Le Voyer TE, Sigurdson ER, Hanlon AL, Mayer RJ, Macdonald JS, Nakashima A, et al. Number of metastatic lymph nodes, but not lymph Catalano PJ, et al. Colon cancer survival is associated with increasing node ratio, is an independent prognostic factor after resection of pancre- number of lymph nodes analyzed: a secondary survey of intergroup trial atic carcinoma. J Am Coll Surg 2010;211:196e204. INT-0089. J Clin Oncol 2003;21:2912e9. 23. Kaklamanos IG, Bathe OF, Franceschi D, Camarda C, Levi J, 37. Pawlik TM, Abdalla EK, Barnett CC, Ahmad SA, Cleary KR, Vauthey JN, Livingstone AS. Extent of resection in the management of duodenal et al. Feasibility of a randomized trial of extended lymphadenectomy for adenocarcinoma. Am J Surg 2000;179:37e41. pancreatic cancer. Arch Surg 2005;140:584e9. 24. Sarela AI, Brennan MF, Karpeh MS, Klimstra D, Conlon KC. Adeno- 38. Halfdanarson TR, McWilliams RR, Donohue JH, Quevedo JF. A single- carcinoma of the duodenum: importance of accurate lymph node staging institution experience with 491 cases of small bowel adenocarcinoma. and similarity in outcome to gastric cancer. Ann Surg Oncol 2004;11: Am J Surg 2010;199:797e803. 380e6. 39. Schiffman SC, McMasters KM, Scoggins CR, Martin RC, Chagpar AB. 25. Hu JX, Miao XY, Zhong DW, Dai WD, Liu W, Hu W. Surgical treatment Lymph node ratio: a proposed refinement of current axillary staging in of primary duodenal adenocarcinoma. Hepatogastroenterology 2006;53: breast cancer patients. J Am Coll Surg 2011;213:45e52. 858e62. 40. Hurtuk MG, Devata S, Brown KM, Oshima K, Aranha GV, Pickleman J, 26. Ryder NM, Ko CY, Hines OJ, Gloor B, Reber HA. Primary duodenal et al. Should all patients with duodenal adenocarcinoma be considered for adenocarcinoma: a 40-year experience. Arch Surg 2000;135:1070e4. aggressive surgical resection? Am J Surg 2007;193:319e24.