Ann Surg Oncol DOI 10.1245/s10434-013-3473-9
ORIGINAL ARTICLE – PANCREATIC TUMORS
Revisiting the Concept of Lymph Node Metastases of Pancreatic Head Cancer: Number of Metastatic Lymph Nodes and Lymph Node Ratio According to N Stage Mee Joo Kang, MD, PhD, Jin-Young Jang, MD, PhD, Ye Rim Chang, MD, Wooil Kwon, MD, Woohyun Jung, MD, and Sun-Whe Kim, MD, PhD, FACS Department of Surgery and Cancer Research Institute, Seoul National University College of Medicine, Seoul, Korea
ABSTRACT Background. Some suggest that metastatic lymph node ratio (LNR) may be prognostic of survival in patients with pancreatic cancer. However, this phenomenon was confused by inclusion of node-negative patients in the analysis. The present study was designed to evaluate the prognostic impact of metastatic LNR and the absolute number of metastatic LNs in patients resected for pancreatic cancer. Methods. Data were collected from 398 patients who underwent curative surgery for pancreatic head cancer at Seoul National University Hospital. Long-term survival was analyzed according to LNR and absolute number of metastatic LNs. Results. Of the patients, 227 (57.0 %) had LN metastasis. The mean numbers of total retrieved and metastatic LNs were 19.5 and 1.9, respectively, and the mean LNR was 0.11. Median overall survival (OS) of patients was significantly higher in N0 than in N1 patients after curative resection (25.4 vs. 14.8 months, p \ 0.001). Median OS was significantly lower in patients with 1 than in those with 0 positive LNs (17.3 vs. 25.4 months, p = 0.001). Among N1 patients, those with 0 \ LNR B 0.2 had comparable prognosis than those with [0.2 LNR (median OS 17.2 vs. 12.8 months, p = 0.096), and the number of metastatic LNs did not correlate with median OS (p = 0.365). Conclusions. The presence of a single positive metastatic LN was associated with significantly poorer OS in patients with pancreatic cancer. When LN metastasis was present,
Ó Society of Surgical Oncology 2014 First Received: 17 September 2013 S.-W. Kim, MD, PhD, FACS e-mail: [email protected]
the number of metastatic LNs and LNR had limited prognostic relevance.
INTRODUCTION Factors prognostic of survival in patients resected for pancreatic cancer include histologic characteristics, such as tumor stage, tumor size, histologic differentiation, lymph node (LN) status, tumor grade, and resection margin status.1,2 Although many studies have shown that LN metastasis is associated with poor prognosis, other studies have found that the associations between LN metastasis and survival outcomes are not always consistent.1–6 Looking deeper, the absolute number of metastatic LNs or total number of LNs examined may be associated with patient prognosis, but incomplete lymphadenectomy or inadequate histopathologic examination can introduce bias in assessing nodal involvement or the absolute number of metastatic LNs.7–17 More extensive lymphadenectomy by the surgeon or more thorough examination by the pathologist increases the number of LNs examined. To overcome problems associated with nodal staging and to better assess the prognostic importance of nodal disease, the concept of ‘‘stage migration,’’ or the ‘‘Will Rogers Phenomenon,’’ was introduced in 1985.18 To prevent stage migration and to better stratify patients, several studies have assessed whether lymph node ratio (LNR) is a more significant and consistent prognostic indicator than absolute LN number in patients with various gastrointestinal malignancies, including pancreatic cancer.14,19–22 Although several studies have reported that LNR is a significant prognostic factor in patients with pancreatic cancer, other studies failed to show its prognostic relevance in node-positive patients.10,14,15,17,23–25 Other studies have suggested that the number of metastatic LNs is more
M. J. Kang et al.
prognostic of survival than LNR in patients with pancreatic cancer.7–9 We therefore evaluated the prognostic impact on survival of LNR and the absolute number of metastatic LNs in patients resected for pancreatic cancer.
TABLE 1 Clinicopathologic characteristics N = 398 Age (years)
60.7 ± 10.6
Sex (M:F)
1.75:1
Operation name
MATERIALS AND METHODS This study was approved by our institutional review board, which waived the requirement for informed consent. Clinicopathological data and radiological images were prospectively collected in electronic medical record form. Between January 1983 and October 2011, 717 consecutive patients underwent surgical treatment for pancreatic head adenocarcinoma at Seoul National University Hospital, Korea. Of the patients, 398 (55.5 %) underwent curative resection and they were included for further analysis. Pylorus preservation was attempted in all patients undergoing pancreatoduodenectomy, unless duodenal ischemia, a duodenal ulcer, or duodenal tumor infiltration was present. LN dissection in pancreatoduodenectomy included the removal of regional LNs to the right side of the celiac and superior mesenteric arteries and all tissues in the hepatoduodenal ligament, except for the portal vein and hepatic artery. Para-aortic LN dissection was not performed routinely. Pancreatic specimens were serially sectioned at 5- to 7-mm intervals. The slides were reviewed for pathological data including T stage, N stage, and numbers of total retrieved and metastatic LNs. Following surgical resection, all patients received chemoradiation, with or without maintenance chemotherapy, except for patients who had T1N0 lesions with no residual tumor, those with poor performance status, and those who refused adjuvant treatment. External beam radiation therapy consisted of a total dose of 45 Gy in 25 fractions delivered 5 days per week for 5 weeks, with a tumor bed boost of 5.4 Gy in 3 fractions every other day, using 5-fluorouracil (5-FU) or gemcitabine as a radiosensitizer. Maintenance chemotherapy consisted of 5-FU (375–500 mg/m2/day) or gemcitabine (1,000 mg/m2), as decided by the medical oncologists. IBM SPSS Statistics version 19.0 (IBM Corp., Somers, NY) was used for all statistical analyses. Nominal data were compared with v2 tests and continuous data with t tests and post hoc tests. Survival parameters were assessed by the Kaplan–Meier method and compared with the log-rank test. A Cox proportional hazards model was used for multivariate survival analysis. Two-sided p \ 0.05 were considered statistically significant. Variables with p \ 0.05 after univariate analysis and variables picked a priori were used in the multivariate analysis.
Whipple’s operation
203 (51.1 %)
PPPD
179 (45.0 %)
Total pancreatectomy
16 (4.0 %)
Portal vein resection
70 (17.6 %)
R0 resection
332 (83.4 %)
Adjuvant treatment Chemotherapy only
34 (8.5 %)
Chemotherapy with radiation therapy Radiation therapy only
199 (50.0 %) 30 (7.5 %)
Follow-up (median, months)
14.7 (.2–176.4)
AJCC 7th stage IA
9 (2.3 %)
IB
8 (2.0 %)
IIA
150 (37.7 %)
IIB
231 (58.0 %)
Lymph node metastasis after curative resection
227 (57.0 %)
No. total retrieved lymph node
19.5 ± 11.9
No. metastatic lymph node
1.9 ± 2.9
Lymph node ratio (LNR)
0.11 ± 0.16
LNR = 0
171 (43.0 %)
0 \ LNR B 0.2
146 (36.7 %)
0.2 \ LNR B 0.4
56 (14.1 %)
0.4 \ LNR
23 (5.8 %)
RESULTS Patient Characteristics The clinicopathologic characteristics of the 398 patients who underwent curative surgery for pancreatic cancer are shown in Table 1. Of these, 332 (83.4 %) underwent R0 resection, and 227 (57.0 %) had LN metastasis. The mean number of retrieved LNs was 19.5 (median 17.0; range 1–82), the mean number of metastatic LNs was 1.9 (median 0; range 0–27), and the mean LNR was 0.11 (median 0.05; range 0–1.17). Pathologic Characteristics in Relation to Lymph Node Metastasis After curative resection, N0 and N1 patients had significantly different rates of R0 resection, T stage, endolymphatic invasion, microscopic venous invasion, and recurrence. None of these pathologic variables differed significantly between N1 patients with 0 \ LNR B 0.2 and LNR [ 0.2 (Table 2). When the patients were stratified
Lymph Node Ratio of Pancreatic Cancer TABLE 2 Pathologic characteristics according to lymph node metastasis after curative resection N0 (n = 171)
R0 resection
N1 (n = 227)
151 (88.3 %)
p value
181 (79.7 %)
0.023
T stage
N1 patients 0 \ LNR B 0.2 (n = 146)
LNR [ 0.2 (n = 79)
p value
117 (80.1 %)
62 (78.5 %)
0.769
4 (2.7 %)
1 (1.3 %)
0.001
T1, T2
17 (9.9 %)
T3, T4
5 (2.2 %)
0.659
154 (90.1 %)
222 (97.8 %)
142 (97.3 %)
78 (98.7 %)
108 (77.2 %)
164 (85.9 %)
0.054
104 (86.0 %)
60 (87.0 %)
0.846
Endolymphatic invasion
43 (37.4 %)
112 (73.7 %)
\0.001
69 (72.6 %)
43 (75.4 %)
0.704
Microscopic venous invasion Recurrence
15 (14.4 %) 98 (57.3 %)
50 (36.5 %) 158 (69.6 %)
\0.001 0.011
29 (35.4 %) 101 (69.2 %)
21 (38.2 %) 55 (69.6 %)
0.737 0.945
Perineural invasion
a
b P < 0.001
Survival (%) 100
c
Survival (%) 100
P < 0.001
Survival (%) 100
P < 0.001 P < 0.001
80
80
80
60
60
60
40
40
40
20
20
20
0
12
24
36
48
Follow-up (months) N0 (n = 171, median 25.4m) N1 (n = 227, median 14.8m)
60
0
12
24
36
48
60
Follow-up (months) N (−) (n = 171, median 25.4m) Positive LN = 1 (n = 73, median 17.3m) Positive LN ≥ 2 (n = 154, median 13.9m)
P = 0.096
0
12
24
36
48
60
Follow-up (months) LNR = 0 (n = 171, median 25.4m) 0 < LNR ≤ 0.2 (n = 146, median 17.2m) LNR > 0.2 (n = 79, median 12.8m)
FIG. 1 Survival after curative resection according to a lymph node metastasis, b number of metastatic lymph node, and c lymph node ratio
according to the number of positive metastatic LNs, the rate of endolymphatic invasion was significantly greater in patients with[2 than in patients with 1 metastatic LN (80.6 vs. 56.8 %, p = 0.003). Survival in Relation to Lymph Node Ratio and Number of Metastatic Lymph Nodes Median overall survival (OS) of patients who underwent curative resection was 18.4 months, and median OS was significantly longer after R0 than R1 resection (19.6 vs. 12.8 months, p \ .001). After R0 resection, median OS was significantly longer in stage I than in stage II patients (34.7 vs. 19.5 months, p \ .001). After curative resection, N0 patients had significantly longer median OS than N1 patients (25.4 vs. 14.8 months, p \ .001, Fig. 1a). Examination of C15 LNs was
associated with poorer OS than examination of \15 LNs (15.7 vs. 21.7 months, p = .029). Among N0 patients, median OS tended to be longer in patients with examination of \15 LNs (30.9 vs. 22.0 months, p = 0.077), whereas, among N1 patients, total number of LNs examined was not associated with patient prognosis (p = 0.399). Patients with 0, 1, and C2 metastatic LNs had a median OS of 25.4, 17.3, and 13.9 months, respectively (Fig. 1b). Relative to patients without LN metastases, the presence of a single positive LN metastasis significantly worsened the prognosis of patients curatively resected for pancreatic cancer (median OS, 25.4 vs. 17.3 months, p = 0.001). Among N1 patients, patients with 1 and C2 metastatic LNs had comparable median OS (17.3 vs. 13.9 months, p = 0.365), and patients with 0 \ LNR B 0.2 had comparable prognosis than those with [0.2 LNR (median OS, 17.2 vs. 12.8 months, p = 0.096, Fig. 1c).
M. J. Kang et al.
Recurrence Pattern During follow-up after curative resection, 256 patients (64.3 %) had tumor recurrences, including 71 (17.8 %) with local recurrence and 221 (55.5 %) with systemic recurrence. Local recurrence was not significantly associated with T stage (21.1 % [T1, T2] vs. 20.6 % [T3, T4], p = 1.000), N stage (23.7 % [N0] vs. 18.1 % [N1], p = 0.2062), R status (20.6 % [R0] vs. 20.4 % [R1], p = 0.974), perineural invasion (19.1 % [?] vs. 31.3 % [-], p = 0.059), endolymphatic invasion (23.3 % [?] vs. 20.0 % [-], p = 0.540), or microscopic venous invasion (14.5 % [?] vs. 22.4 % [-], p = 0.210). In contrast, systemic recurrence was significantly associated with N stage (54.9 % [N0] vs. 70.3 % [N1], p = 0.003), R status (61.2 % [R0] vs. 77.6 % [R1], p = 0.028), endolymphatic invasion (74.6 % [?] vs. 48.1 % [-], p \ 0.001), and microscopic venous invasion (74.5 % [?] vs. 57.8 % [-], p = 0.027). However, among N1 patients, systemic metastasis was not associated with the number of metastatic LNs (67.7 % [1 positive] vs. 71.5 % [C2 positive], p = 0.580) or LNR (65.9 % [0–0.2] vs. 78.1 % [[0.2], p = 0.081). Efficacy of Adjuvant Treatment According to Nodal Status After Curative Resection Median OS was significantly longer in patients who did than in those who did not receive adjuvant treatment after curative resection of pancreatic cancer (19.6 vs. 13.9 months, p = 0.006). Types of adjuvant treatment significantly affected patient prognosis, with median survival being 12.8 months in patients receiving chemotherapy alone, 21.6 months in patients receiving both chemotherapy and radiation therapy, and 22.2 months in patients who received radiation therapy alone (p \ 0.001). The systemic recurrence rate was significantly lower in patients receiving chemotherapy plus radiation therapy than in those receiving either modality alone (22.8 vs. 36.9 %, p = 0.047). Among N0 patients, median OS was significantly lower (p = 0.001) in those who received chemotherapy alone (13.7 months) than in those who received chemotherapy plus radiation therapy (32.5 months) or radiation therapy alone (31.0 months). Similarly, median OS was significantly lower in N1 patients who received chemotherapy alone (12.8 months) than in those who received chemotherapy plus radiation therapy (17.6 months) or radiation therapy alone (18.7 months, p = 0.037). Factors Prognostic of Survival in Patients Curatively Resected for Pancreatic Cancer For overall patients, univariate analysis showed that N stage, R status, endolymphatic invasion, microscopic
venous invasion, and adjuvant treatment were significant prognostic indicators in patients who underwent curative resection. Multivariate analysis revealed that R status, endolymphatic invasion, microscopic venous invasion, and adjuvant chemoradiation therapy were significant predictors for survival in patients curatively resected for pancreatic cancer (Table 3). N stage had marginal significance after multivariate analysis. Subgroup analysis was performed for N1 patients, and multivariate analysis showed that endolymphatic invasion was the most significant prognostic factor, whereas LNR was no longer statistically significant after multivariate analysis (Table 4). Adjuvant chemoradiation therapy lost its statistical significance after multivariate analysis in N1 patients.
DISCUSSION Extensive evaluation of LNs in patients with pancreatic cancer can result in more accurate staging and prognosis. The determination of LNR has gained prominence in gastrointestinal cancers, particularly in gastric cancer.19,22 The use of LNR has significantly decreased the phenomenon of ‘‘stage migration,’’ observed in 10–15 % of gastric cancer patients when using AJCC/UICC N stage based on the number of metastatic LNs.19,22 The prognostic relevance of LNR has also been shown in patients with colon, esophageal, and bladder cancers.20,21,26 LNR has also been evaluated in patients with ampulla of Vater and periampullary cancers and intraductal papillary mucinous carcinoma.8–10,27 LNR may also be prognostic in patients with pancreatic cancer.7,11–13,15,17,23–25 However, because LN metastasis has great influence in survival outcome, LNR should be stratified according to nodal status in assessing survival outcomes. Studies reporting the association of LNR with survival have included patients with an LNR of 0; thus survival differences may be due to the longer survival of the latter relative to those with LNR [ 0.10,11,14,24 These studies failed to show that LNR had prognostic significance in node-positive patients. Moreover, because N status and LNR has multicollinearity, therefore, multivariate analysis including these two variables cannot give valid results about individual predictors. As a consequence, the results of multivariate analysis of previous studies that simultaneously analyzed prognostic relevance of both N status and LNR should be reconsidered.6,10,15,23 Our findings showed that LNR was not an independent predictor of OS in patients curatively resected for pancreatic cancer. Although R0 resection rate, T stage, endolymphatic and microscopic venous invasion, and recurrence rate differed significantly in node-positive and node-negative patients, none of these factors were significantly related to LNR in node-positive patients. Survival analysis revealed that any LN
Lymph Node Ratio of Pancreatic Cancer TABLE 3 Prognostic factor analysis of the patients who underwent curative resection Variables
N
Univariate analysis Median survival
Multivariate analysis p value
Hazard ratio
95 % CI
p value
T stage (T1, T2/T3, T4)
22/376
33.1/17.8
0.259
1.869
0.616–5.675
0.270
N stage (N0/N1)
171/227
25.4/14.8
\0.001
1.383
0.958–1.996
0.084
Total examined lymph node (\ 15/C 15)
147/221
21.7/15.7
0.029
1.023
0.714–1.465
0.901
R status (R0/R1)
332/66
19.6/12.8
\0.001
1.620
1.009–2.602
0.046
Perineural invasion (±)
272/58
17.9/21.2
0.182
1.082
0.638–1.836
0.769
Endolymphatic invasion (±)
155/112
14.8/34.7
\0.001
2.019
1.377–2.961
\0.001
Microscopic venous invasion (±) Adjuvant CCRT (no/yes)
65/176 169/229
15.5/25.5 13.2/21.6
\0.001 \0.001
1.512 1.632
1.042–2.194 1.145–2.325
0.030 0.007
CCRT concurrent chemoradiation TABLE 4 Prognostic factor analysis for metastatic lymph node-positive patients after curative resection Variables
N
Univariate analysis Median survival
Multivariate analysis p value
Hazard ratio
95 % CI
p value
T stage (T1, T2/T3, T4)
5/222
13.1/14/8
0.586
0.460
0.060–3.530
0.455
R status (R0/R1)
181/46
17.3/11.4
\0.001
1.667
0.920–3.019
0.092
Perineural invasion (±)
164/27
15.5/17.2
0.683
1.323
0.614–2.853
0.475
Endolymphatic invasion (±)
112/40
14.0/25.5
0.001
2.027
1.217–3.378
0.007
Microscopic venous invasion (±)
50/87
15.5/17.8
0.034
1.386
0.873–2.201
0.166
1.154
0.733–1.815
0.536
1.407
0.890–2.226
0.144
Lymph node ratio (B0.2/[0.2)
146/79
17.2/12.8
0.096
Total examined lymph node (\15/C15)
81/144
17.8/14.4
0.432
Number of metastatic lymph node (1/C2)
73/154
17.3/13.9
0.365
Adjuvant CCRT (no/yes)
98/129
12.6/17.8
0.004
CCRT concurrent chemoradiation
metastasis significantly worsened prognosis, but among nodepositive patients, LNR did not affect prognosis. Moreover, multivariate analysis revealed LNR did not have significant prognostic significance in node-positive patients. The total number of LNs examined is dependent on how extensively the pathologist examines the surgical specimen, and the prognostic relevance of total number of LNs examined is unclear.11,12,14,15 There is no consensus on what constitutes a sufficient number of LNs examined, with cutoffs ranging from 10 to 16 LNs having been used in previous studies.9–11,13, 28,29 We found that OS was significantly lower in patients with C15 than \15 LNs examined. This is in line with experiences of stage migration effect in other cancers.30,31 However, in agreement with studies reporting that the number of LNs examined was not a prognostic indicator in node-positive pancreatic cancer, the total number of LNs examined was not statistically significant prognostic indicator in nodepositive patients in this study.17,24 In addition, multivariate analysis revealed that total number of LNs examined was not a significant prognostic factor for curatively resected pancreatic cancer. Moreover, our recent randomized
controlled study evaluating the extent of lymphadenectomy revealed that patients who underwent standard and extended LN dissection had comparable prognoses in nodepositive patients.32 These findings indicate that, because the presence of nodal metastasis is of great importance in determining the prognosis of patients with pancreatic cancer, the extent of lymphadenectomy and the number of total examined LNs are not important in node-positive patients. The number of positive LNs is another established prognostic factor in various malignancies including stomach, colon, breast, or ampullary cancer.33–35 The number of metastatic LNs may also be prognostic in patients with pancreatic cancer.7,8 Studies have reported that the prognosis of patients with 0 and 1 positive LN does not differ, suggesting that LNR may be prognostic in these patients.5,24 Other studies have found that patients with 2 positive LNs have similar prognosis as patients with 0 or 1 positive LN.11,16 In contrast, we found that the presence of only 1 positive LN was associated with a significantly poorer OS in patients curatively resected for pancreatic cancer. On the contrary, the number of metastatic LNs
M. J. Kang et al.
among N1 patients did not have prognostic difference. This observation implies that since the presence of a single positive LN has great importance for patient prognosis, patients with any type of N1 disease should be considered different from those with node-negative disease. The presence of tumor cell in regional LNs may reflect the potential of the primary tumor to metastasize. For breast cancer, even the isolated tumor cells or micrometastases are known to reduce overall and disease-free survival of the patients.36 Therefore, it can be logically inferred that presence of LN metastasis of such an aggressive tumor as pancreatic cancer will heavily impact prognosis, regardless of its number. In this study, LNR had limited prognostic value among N1 patients. Moreover, systemic metastasis rate was significantly associated with N stage and endolymphatic invasion, but it was not associated with the number of metastatic LNs or LNR among N1 patients. Therefore, it can be concluded that the presence of any nodal metastasis has a more important role in prognosis than the tumor nodal burden. In this study, types of adjuvant treatment were significantly correlated with prognosis. Chemoradiation therapy had significant survival gain over chemotherapy alone in both node-negative and node-positive patients. This result conflicts with previous studies; however, due to the retrospective nature of this study, this observation results from selection bias of the patients because those who could not endure radiation therapy received chemotherapy alone.37,38 This study has major limitation due to its retrospective study design. However, we provide one of the largest study cohorts of pancreatic head cancer who were treated with standardized treatment protocol. Multivariate analysis compensates the limitation of study design and revealed nodal status was more important than LNR, and LNR did not have statistical significance for N1 patients. In conclusion, LN metastasis per se is more important than the number of LNs examined, the number of metastatic LNs, and LNR in determining patient prognosis. ACKNOWLEDGMENT This study was supported by a Grant from the National R&D Program for Cancer Control funded by the Ministry of Health & Welfare, Republic of Korea (No. 1120310). COMPETING INTEREST
None.
REFERENCES 1. Brennan MF, Kattan MW, Klimstra D, Conlon K. Prognostic nomogram for patients undergoing resection for adenocarcinoma of the pancreas. Ann Surg. 2004;240:293–8. 2. Winter JM, Cameron JL, Campbell KA, Arnold MA, Chang DC, Coleman J, et al. 1423 pancreaticoduodenectomies for pancreatic cancer: a single-institution experience. J Gastrointest Surg. 2006;10:1199–210; discussion 1210–1.
3. Yeo CJ, Cameron JL, Sohn TA, Lillemoe KD, Pitt HA, Talamini MA, et al. Six hundred fifty consecutive pancreaticoduodenectomies in the 1990s: pathology, complications, and outcomes. Ann Surg. 1997;226:248–57; discussion 257–60. 4. Lim JE, Chien MW, Earle CC. Prognostic factors following curative resection for pancreatic adenocarcinoma: a populationbased, linked database analysis of 396 patients. Ann Surg. 2003; 237:74–85. 5. Sohn TA, Yeo CJ, Cameron JL, Koniaris L, Kaushal S, Abrams RA, et al. Resected adenocarcinoma of the pancreas-616 patients: results, outcomes, and prognostic indicators. J Gastrointest Surg. 2000;4:567–79. 6. Riediger H, Makowiec F, Fischer E, Adam U, Hopt UT. Postoperative morbidity and long-term survival after pancreaticoduodenectomy with superior mesenterico-portal vein resection. J Gastrointest Surg. 2006;10:1106–15. 7. John BJ, Naik P, Ironside A, Davidson BR, Fusai G, Gillmore R, et al. Redefining the R1 resection for pancreatic ductal adenocarcinoma: tumour lymph nodal burden and lymph node ratio are the only prognostic factors associated with survival. HPB (Oxford). 2013;15:674–80. 8. Sakata J, Shirai Y, Wakai T, Ajioka Y, Akazawa K, Hatakeyama K. Assessment of the nodal status in ampullary carcinoma: the number of positive lymph nodes versus the lymph node ratio. World J Surg. 2011;35:2118–24. 9. Shamseddine AI, Mukherji D, Melki C, Elias E, Eloubeidi M, Dimassi H, et al. Lymph node ratio is an independent prognostic factor after resection of periampullary malignancies: data from a tertiary referral center in the Middle East. Am J Clin Oncol. 2012. doi:10.1097/COC.0b013e31826b9c74. 10. Falconi M, Crippa S, Dominguez I, Barugola G, Capelli P, Marcucci S, et al. Prognostic relevance of lymph node ratio and number of resected nodes after curative resection of ampulla of Vater carcinoma. Ann Surg Oncol. 2008;15:3178–86. 11. Huebner M, Kendrick M, Reid-Lombardo KM, Que F, Therneau T, Qin R, et al. Number of lymph nodes evaluated: prognostic value in pancreatic adenocarcinoma. J Gastrointest Surg. 2012;16:920–6. 12. Showalter TN, Winter KA, Berger AC, Regine WF, Abrams RA, Safran H, et al. The influence of total nodes examined, number of positive nodes, and lymph node ratio on survival after surgical resection and adjuvant chemoradiation for pancreatic cancer: a secondary analysis of RTOG 9704. Int J Radiat Oncol Biol Phys. 2011;81:1328–35. 13. Slidell MB, Chang DC, Cameron JL, Wolfgang C, Herman JM, Schulick RD, et al. Impact of total lymph node count and lymph node ratio on staging and survival after pancreatectomy for pancreatic adenocarcinoma: a large, population-based analysis. Ann Surg Oncol. 2008;15:165–74. 14. Berger AC, Watson JC, Ross EA, Hoffman JP, et al. The metastatic/examined lymph node ratio is an important prognostic factor after pancreaticoduodenectomy for pancreatic adenocarcinoma. Am Surg. 2004;70:235–40; discussion 240. 15. Bhatti I, Peacock O, Awan AK, Semeraro D, Larvin M, Hall RI. Lymph node ratio versus number of affected lymph nodes as predictors of survival for resected pancreatic adenocarcinoma. World J Surg. 2010;34:768–75. 16. House MG, Gonen M, Jarnagin WR, D’Angelica M, DeMatteo RP, Fong Y, et al. Prognostic significance of pathologic nodal status in patients with resected pancreatic cancer. J Gastrointest Surg. 2007;11:1549–55. 17. Pawlik TM, Gleisner AL, Cameron JL, Winter JM, Assumpcao L, Lillemoe KD, et al. Prognostic relevance of lymph node ratio following pancreaticoduodenectomy for pancreatic cancer. Surgery. 2007;141:610–8.
Lymph Node Ratio of Pancreatic Cancer 18. Feinstein AR, Sosin DM, Wells CK. The Will Rogers phenomenon. Stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med. 1985;312:1604–8. 19. Bando E, Yonemura Y, Taniguchi K, Fushida S, Fujimura T, Miwa K. Outcome of ratio of lymph node metastasis in gastric carcinoma. Ann Surg Oncol. 2002;9:775–84. 20. Bogoevski D, Onken F, Koenig A, Kaifi JT, Schurr P, Sauter G, et al. Is it time for a new TNM classification in esophageal carcinoma? Ann Surg. 2008;247:633–41. 21. Rosenberg R, Friederichs J, Schuster T, Gertler R, Maak M, Becker K, et al. Prognosis of patients with colorectal cancer is associated with lymph node ratio: a single-center analysis of 3,026 patients over a 25-year time period. Ann Surg. 2008;248:968–78. 22. Sierra A, Regueira FM, Hernandez-Lizoain JL, Pardo F, Martı´nez-Gonzalez MA, A-Cienfuegos J. Role of the extended lymphadenectomy in gastric cancer surgery: experience in a single institution. Ann Surg Oncol. 2003;10:219–26. 23. La Torre M, Cavallini M, Ramacciato G, Cosenza G, Rossi Del Monte S, Nigri G, et al. Role of the lymph node ratio in pancreatic ductal adenocarcinoma. Impact on patient stratification and prognosis. J Surg Oncol. 2011;104:629–33. 24. Riediger H, Keck T, Wellner U, zur Hausen A, Adam U, Hopt UT, et al. The lymph node ratio is the strongest prognostic factor after resection of pancreatic cancer. J Gastrointest Surg. 2009;13:1337–44. 25. Sierzega M, Popiela T, Kulig J, Nowak K. The ratio of metastatic/resected lymph nodes is an independent prognostic factor in patients with node-positive pancreatic head cancer. Pancreas. 2006;33:240–5. 26. Herr HW. Superiority of ratio based lymph node staging for bladder cancer. J Urol. 2003;169:943–5. 27. Partelli S, Fernandez-Del Castillo C, Bassi C, Mantovani W, Thayer SP, Crippa S, et al. Invasive intraductal papillary mucinous carcinomas of the pancreas: predictors of survival and the role of lymph node ratio. Ann Surg. 2010;251:477–82. 28. Schwarz RE, Smith DD. Extent of lymph node retrieval and pancreatic cancer survival: information from a large US population database. Ann Surg Oncol. 2006;13:1189–200.
29. Tomlinson JS, Jain S, Bentrem DJ, Sekeris EG, Maggard MA, Hines OJ, et al. Accuracy of staging node-negative pancreas cancer: a potential quality measure. Arch Surg. 2007;142:767–73; discussion 773–4. 30. Kong SH, Lee HJ, Ahn HS, Kim JW, Kim WH, Lee KU, et al. Stage migration effect on survival in gastric cancer surgery with extended lymphadenectomy: the reappraisal of positive lymph node ratio as a proper N-staging. Ann Surg. 2012;255:50–8. 31. Yoshikawa T, Sasako M, Sano T, Nashimoto A, Kurita A, Tsujinaka T, et al. Stage migration caused by D2 dissection with para-aortic lymphadenectomy for gastric cancer from the results of a prospective randomized controlled trial. Br J Surg. 2006;93:1526–9. 32. Jang JY, Kang MJ, Heo JS, Choi SH, Choi DW, Park SJ, et al. A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer. Ann Surg. 2013. doi:10.1097/SLA. 0000000000000384. 33. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti III A. AJCC Cancer Staging Manual. 7th ed. New York: Springer, 2010. 34. Carter CL, Allen C, Henson DE. Relation of tumor size, lymph node status, and survival in 24,740 breast cancer cases. Cancer 1989;63:181–7. 35. Sakata J, Shirai Y, Wakai T, Yokoyama N, Sakata E, Akazawa K, et al. Number of positive lymph nodes independently affects long-term survival after resection in patients with ampullary carcinoma. Eur J Surg Oncol. 2007;33:346–51. 36. de Boer M, van Deurzen CH, van Dijck JA, Borm GF, van Diest PJ, Adang EM, et al. Micrometastases or isolated tumor cells and the outcome of breast cancer. N Engl J Med. 2009;361:653–63. 37. Oettle H, Post S, Neuhaus P, Gellert K, Langrehr J, Ridwelski K, et al. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer: a randomized controlled trial. JAMA. 2007;297:267–77. 38. Neoptolemos JP, Dunn JA, Stocken DD, Almond J, Link K, Beger H, et al. Adjuvant chemoradiotherapy and chemotherapy in resectable pancreatic cancer: a randomised controlled trial. Lancet. 2001;358:1576–85.
ORIGINAL ARTICLE – PANCREATIC TUMORS
Revisiting the Concept of Lymph Node Metastases of Pancreatic Head Cancer: Number of Metastatic Lymph Nodes and Lymph Node Ratio According to N Stage Mee Joo Kang, MD, PhD, Jin-Young Jang, MD, PhD, Ye Rim Chang, MD, Wooil Kwon, MD, Woohyun Jung, MD, and Sun-Whe Kim, MD, PhD, FACS Department of Surgery and Cancer Research Institute, Seoul National University College of Medicine, Seoul, Korea
ABSTRACT Background. Some suggest that metastatic lymph node ratio (LNR) may be prognostic of survival in patients with pancreatic cancer. However, this phenomenon was confused by inclusion of node-negative patients in the analysis. The present study was designed to evaluate the prognostic impact of metastatic LNR and the absolute number of metastatic LNs in patients resected for pancreatic cancer. Methods. Data were collected from 398 patients who underwent curative surgery for pancreatic head cancer at Seoul National University Hospital. Long-term survival was analyzed according to LNR and absolute number of metastatic LNs. Results. Of the patients, 227 (57.0 %) had LN metastasis. The mean numbers of total retrieved and metastatic LNs were 19.5 and 1.9, respectively, and the mean LNR was 0.11. Median overall survival (OS) of patients was significantly higher in N0 than in N1 patients after curative resection (25.4 vs. 14.8 months, p \ 0.001). Median OS was significantly lower in patients with 1 than in those with 0 positive LNs (17.3 vs. 25.4 months, p = 0.001). Among N1 patients, those with 0 \ LNR B 0.2 had comparable prognosis than those with [0.2 LNR (median OS 17.2 vs. 12.8 months, p = 0.096), and the number of metastatic LNs did not correlate with median OS (p = 0.365). Conclusions. The presence of a single positive metastatic LN was associated with significantly poorer OS in patients with pancreatic cancer. When LN metastasis was present,
Ó Society of Surgical Oncology 2014 First Received: 17 September 2013 S.-W. Kim, MD, PhD, FACS e-mail: [email protected]
the number of metastatic LNs and LNR had limited prognostic relevance.
INTRODUCTION Factors prognostic of survival in patients resected for pancreatic cancer include histologic characteristics, such as tumor stage, tumor size, histologic differentiation, lymph node (LN) status, tumor grade, and resection margin status.1,2 Although many studies have shown that LN metastasis is associated with poor prognosis, other studies have found that the associations between LN metastasis and survival outcomes are not always consistent.1–6 Looking deeper, the absolute number of metastatic LNs or total number of LNs examined may be associated with patient prognosis, but incomplete lymphadenectomy or inadequate histopathologic examination can introduce bias in assessing nodal involvement or the absolute number of metastatic LNs.7–17 More extensive lymphadenectomy by the surgeon or more thorough examination by the pathologist increases the number of LNs examined. To overcome problems associated with nodal staging and to better assess the prognostic importance of nodal disease, the concept of ‘‘stage migration,’’ or the ‘‘Will Rogers Phenomenon,’’ was introduced in 1985.18 To prevent stage migration and to better stratify patients, several studies have assessed whether lymph node ratio (LNR) is a more significant and consistent prognostic indicator than absolute LN number in patients with various gastrointestinal malignancies, including pancreatic cancer.14,19–22 Although several studies have reported that LNR is a significant prognostic factor in patients with pancreatic cancer, other studies failed to show its prognostic relevance in node-positive patients.10,14,15,17,23–25 Other studies have suggested that the number of metastatic LNs is more
M. J. Kang et al.
prognostic of survival than LNR in patients with pancreatic cancer.7–9 We therefore evaluated the prognostic impact on survival of LNR and the absolute number of metastatic LNs in patients resected for pancreatic cancer.
TABLE 1 Clinicopathologic characteristics N = 398 Age (years)
60.7 ± 10.6
Sex (M:F)
1.75:1
Operation name
MATERIALS AND METHODS This study was approved by our institutional review board, which waived the requirement for informed consent. Clinicopathological data and radiological images were prospectively collected in electronic medical record form. Between January 1983 and October 2011, 717 consecutive patients underwent surgical treatment for pancreatic head adenocarcinoma at Seoul National University Hospital, Korea. Of the patients, 398 (55.5 %) underwent curative resection and they were included for further analysis. Pylorus preservation was attempted in all patients undergoing pancreatoduodenectomy, unless duodenal ischemia, a duodenal ulcer, or duodenal tumor infiltration was present. LN dissection in pancreatoduodenectomy included the removal of regional LNs to the right side of the celiac and superior mesenteric arteries and all tissues in the hepatoduodenal ligament, except for the portal vein and hepatic artery. Para-aortic LN dissection was not performed routinely. Pancreatic specimens were serially sectioned at 5- to 7-mm intervals. The slides were reviewed for pathological data including T stage, N stage, and numbers of total retrieved and metastatic LNs. Following surgical resection, all patients received chemoradiation, with or without maintenance chemotherapy, except for patients who had T1N0 lesions with no residual tumor, those with poor performance status, and those who refused adjuvant treatment. External beam radiation therapy consisted of a total dose of 45 Gy in 25 fractions delivered 5 days per week for 5 weeks, with a tumor bed boost of 5.4 Gy in 3 fractions every other day, using 5-fluorouracil (5-FU) or gemcitabine as a radiosensitizer. Maintenance chemotherapy consisted of 5-FU (375–500 mg/m2/day) or gemcitabine (1,000 mg/m2), as decided by the medical oncologists. IBM SPSS Statistics version 19.0 (IBM Corp., Somers, NY) was used for all statistical analyses. Nominal data were compared with v2 tests and continuous data with t tests and post hoc tests. Survival parameters were assessed by the Kaplan–Meier method and compared with the log-rank test. A Cox proportional hazards model was used for multivariate survival analysis. Two-sided p \ 0.05 were considered statistically significant. Variables with p \ 0.05 after univariate analysis and variables picked a priori were used in the multivariate analysis.
Whipple’s operation
203 (51.1 %)
PPPD
179 (45.0 %)
Total pancreatectomy
16 (4.0 %)
Portal vein resection
70 (17.6 %)
R0 resection
332 (83.4 %)
Adjuvant treatment Chemotherapy only
34 (8.5 %)
Chemotherapy with radiation therapy Radiation therapy only
199 (50.0 %) 30 (7.5 %)
Follow-up (median, months)
14.7 (.2–176.4)
AJCC 7th stage IA
9 (2.3 %)
IB
8 (2.0 %)
IIA
150 (37.7 %)
IIB
231 (58.0 %)
Lymph node metastasis after curative resection
227 (57.0 %)
No. total retrieved lymph node
19.5 ± 11.9
No. metastatic lymph node
1.9 ± 2.9
Lymph node ratio (LNR)
0.11 ± 0.16
LNR = 0
171 (43.0 %)
0 \ LNR B 0.2
146 (36.7 %)
0.2 \ LNR B 0.4
56 (14.1 %)
0.4 \ LNR
23 (5.8 %)
RESULTS Patient Characteristics The clinicopathologic characteristics of the 398 patients who underwent curative surgery for pancreatic cancer are shown in Table 1. Of these, 332 (83.4 %) underwent R0 resection, and 227 (57.0 %) had LN metastasis. The mean number of retrieved LNs was 19.5 (median 17.0; range 1–82), the mean number of metastatic LNs was 1.9 (median 0; range 0–27), and the mean LNR was 0.11 (median 0.05; range 0–1.17). Pathologic Characteristics in Relation to Lymph Node Metastasis After curative resection, N0 and N1 patients had significantly different rates of R0 resection, T stage, endolymphatic invasion, microscopic venous invasion, and recurrence. None of these pathologic variables differed significantly between N1 patients with 0 \ LNR B 0.2 and LNR [ 0.2 (Table 2). When the patients were stratified
Lymph Node Ratio of Pancreatic Cancer TABLE 2 Pathologic characteristics according to lymph node metastasis after curative resection N0 (n = 171)
R0 resection
N1 (n = 227)
151 (88.3 %)
p value
181 (79.7 %)
0.023
T stage
N1 patients 0 \ LNR B 0.2 (n = 146)
LNR [ 0.2 (n = 79)
p value
117 (80.1 %)
62 (78.5 %)
0.769
4 (2.7 %)
1 (1.3 %)
0.001
T1, T2
17 (9.9 %)
T3, T4
5 (2.2 %)
0.659
154 (90.1 %)
222 (97.8 %)
142 (97.3 %)
78 (98.7 %)
108 (77.2 %)
164 (85.9 %)
0.054
104 (86.0 %)
60 (87.0 %)
0.846
Endolymphatic invasion
43 (37.4 %)
112 (73.7 %)
\0.001
69 (72.6 %)
43 (75.4 %)
0.704
Microscopic venous invasion Recurrence
15 (14.4 %) 98 (57.3 %)
50 (36.5 %) 158 (69.6 %)
\0.001 0.011
29 (35.4 %) 101 (69.2 %)
21 (38.2 %) 55 (69.6 %)
0.737 0.945
Perineural invasion
a
b P < 0.001
Survival (%) 100
c
Survival (%) 100
P < 0.001
Survival (%) 100
P < 0.001 P < 0.001
80
80
80
60
60
60
40
40
40
20
20
20
0
12
24
36
48
Follow-up (months) N0 (n = 171, median 25.4m) N1 (n = 227, median 14.8m)
60
0
12
24
36
48
60
Follow-up (months) N (−) (n = 171, median 25.4m) Positive LN = 1 (n = 73, median 17.3m) Positive LN ≥ 2 (n = 154, median 13.9m)
P = 0.096
0
12
24
36
48
60
Follow-up (months) LNR = 0 (n = 171, median 25.4m) 0 < LNR ≤ 0.2 (n = 146, median 17.2m) LNR > 0.2 (n = 79, median 12.8m)
FIG. 1 Survival after curative resection according to a lymph node metastasis, b number of metastatic lymph node, and c lymph node ratio
according to the number of positive metastatic LNs, the rate of endolymphatic invasion was significantly greater in patients with[2 than in patients with 1 metastatic LN (80.6 vs. 56.8 %, p = 0.003). Survival in Relation to Lymph Node Ratio and Number of Metastatic Lymph Nodes Median overall survival (OS) of patients who underwent curative resection was 18.4 months, and median OS was significantly longer after R0 than R1 resection (19.6 vs. 12.8 months, p \ .001). After R0 resection, median OS was significantly longer in stage I than in stage II patients (34.7 vs. 19.5 months, p \ .001). After curative resection, N0 patients had significantly longer median OS than N1 patients (25.4 vs. 14.8 months, p \ .001, Fig. 1a). Examination of C15 LNs was
associated with poorer OS than examination of \15 LNs (15.7 vs. 21.7 months, p = .029). Among N0 patients, median OS tended to be longer in patients with examination of \15 LNs (30.9 vs. 22.0 months, p = 0.077), whereas, among N1 patients, total number of LNs examined was not associated with patient prognosis (p = 0.399). Patients with 0, 1, and C2 metastatic LNs had a median OS of 25.4, 17.3, and 13.9 months, respectively (Fig. 1b). Relative to patients without LN metastases, the presence of a single positive LN metastasis significantly worsened the prognosis of patients curatively resected for pancreatic cancer (median OS, 25.4 vs. 17.3 months, p = 0.001). Among N1 patients, patients with 1 and C2 metastatic LNs had comparable median OS (17.3 vs. 13.9 months, p = 0.365), and patients with 0 \ LNR B 0.2 had comparable prognosis than those with [0.2 LNR (median OS, 17.2 vs. 12.8 months, p = 0.096, Fig. 1c).
M. J. Kang et al.
Recurrence Pattern During follow-up after curative resection, 256 patients (64.3 %) had tumor recurrences, including 71 (17.8 %) with local recurrence and 221 (55.5 %) with systemic recurrence. Local recurrence was not significantly associated with T stage (21.1 % [T1, T2] vs. 20.6 % [T3, T4], p = 1.000), N stage (23.7 % [N0] vs. 18.1 % [N1], p = 0.2062), R status (20.6 % [R0] vs. 20.4 % [R1], p = 0.974), perineural invasion (19.1 % [?] vs. 31.3 % [-], p = 0.059), endolymphatic invasion (23.3 % [?] vs. 20.0 % [-], p = 0.540), or microscopic venous invasion (14.5 % [?] vs. 22.4 % [-], p = 0.210). In contrast, systemic recurrence was significantly associated with N stage (54.9 % [N0] vs. 70.3 % [N1], p = 0.003), R status (61.2 % [R0] vs. 77.6 % [R1], p = 0.028), endolymphatic invasion (74.6 % [?] vs. 48.1 % [-], p \ 0.001), and microscopic venous invasion (74.5 % [?] vs. 57.8 % [-], p = 0.027). However, among N1 patients, systemic metastasis was not associated with the number of metastatic LNs (67.7 % [1 positive] vs. 71.5 % [C2 positive], p = 0.580) or LNR (65.9 % [0–0.2] vs. 78.1 % [[0.2], p = 0.081). Efficacy of Adjuvant Treatment According to Nodal Status After Curative Resection Median OS was significantly longer in patients who did than in those who did not receive adjuvant treatment after curative resection of pancreatic cancer (19.6 vs. 13.9 months, p = 0.006). Types of adjuvant treatment significantly affected patient prognosis, with median survival being 12.8 months in patients receiving chemotherapy alone, 21.6 months in patients receiving both chemotherapy and radiation therapy, and 22.2 months in patients who received radiation therapy alone (p \ 0.001). The systemic recurrence rate was significantly lower in patients receiving chemotherapy plus radiation therapy than in those receiving either modality alone (22.8 vs. 36.9 %, p = 0.047). Among N0 patients, median OS was significantly lower (p = 0.001) in those who received chemotherapy alone (13.7 months) than in those who received chemotherapy plus radiation therapy (32.5 months) or radiation therapy alone (31.0 months). Similarly, median OS was significantly lower in N1 patients who received chemotherapy alone (12.8 months) than in those who received chemotherapy plus radiation therapy (17.6 months) or radiation therapy alone (18.7 months, p = 0.037). Factors Prognostic of Survival in Patients Curatively Resected for Pancreatic Cancer For overall patients, univariate analysis showed that N stage, R status, endolymphatic invasion, microscopic
venous invasion, and adjuvant treatment were significant prognostic indicators in patients who underwent curative resection. Multivariate analysis revealed that R status, endolymphatic invasion, microscopic venous invasion, and adjuvant chemoradiation therapy were significant predictors for survival in patients curatively resected for pancreatic cancer (Table 3). N stage had marginal significance after multivariate analysis. Subgroup analysis was performed for N1 patients, and multivariate analysis showed that endolymphatic invasion was the most significant prognostic factor, whereas LNR was no longer statistically significant after multivariate analysis (Table 4). Adjuvant chemoradiation therapy lost its statistical significance after multivariate analysis in N1 patients.
DISCUSSION Extensive evaluation of LNs in patients with pancreatic cancer can result in more accurate staging and prognosis. The determination of LNR has gained prominence in gastrointestinal cancers, particularly in gastric cancer.19,22 The use of LNR has significantly decreased the phenomenon of ‘‘stage migration,’’ observed in 10–15 % of gastric cancer patients when using AJCC/UICC N stage based on the number of metastatic LNs.19,22 The prognostic relevance of LNR has also been shown in patients with colon, esophageal, and bladder cancers.20,21,26 LNR has also been evaluated in patients with ampulla of Vater and periampullary cancers and intraductal papillary mucinous carcinoma.8–10,27 LNR may also be prognostic in patients with pancreatic cancer.7,11–13,15,17,23–25 However, because LN metastasis has great influence in survival outcome, LNR should be stratified according to nodal status in assessing survival outcomes. Studies reporting the association of LNR with survival have included patients with an LNR of 0; thus survival differences may be due to the longer survival of the latter relative to those with LNR [ 0.10,11,14,24 These studies failed to show that LNR had prognostic significance in node-positive patients. Moreover, because N status and LNR has multicollinearity, therefore, multivariate analysis including these two variables cannot give valid results about individual predictors. As a consequence, the results of multivariate analysis of previous studies that simultaneously analyzed prognostic relevance of both N status and LNR should be reconsidered.6,10,15,23 Our findings showed that LNR was not an independent predictor of OS in patients curatively resected for pancreatic cancer. Although R0 resection rate, T stage, endolymphatic and microscopic venous invasion, and recurrence rate differed significantly in node-positive and node-negative patients, none of these factors were significantly related to LNR in node-positive patients. Survival analysis revealed that any LN
Lymph Node Ratio of Pancreatic Cancer TABLE 3 Prognostic factor analysis of the patients who underwent curative resection Variables
N
Univariate analysis Median survival
Multivariate analysis p value
Hazard ratio
95 % CI
p value
T stage (T1, T2/T3, T4)
22/376
33.1/17.8
0.259
1.869
0.616–5.675
0.270
N stage (N0/N1)
171/227
25.4/14.8
\0.001
1.383
0.958–1.996
0.084
Total examined lymph node (\ 15/C 15)
147/221
21.7/15.7
0.029
1.023
0.714–1.465
0.901
R status (R0/R1)
332/66
19.6/12.8
\0.001
1.620
1.009–2.602
0.046
Perineural invasion (±)
272/58
17.9/21.2
0.182
1.082
0.638–1.836
0.769
Endolymphatic invasion (±)
155/112
14.8/34.7
\0.001
2.019
1.377–2.961
\0.001
Microscopic venous invasion (±) Adjuvant CCRT (no/yes)
65/176 169/229
15.5/25.5 13.2/21.6
\0.001 \0.001
1.512 1.632
1.042–2.194 1.145–2.325
0.030 0.007
CCRT concurrent chemoradiation TABLE 4 Prognostic factor analysis for metastatic lymph node-positive patients after curative resection Variables
N
Univariate analysis Median survival
Multivariate analysis p value
Hazard ratio
95 % CI
p value
T stage (T1, T2/T3, T4)
5/222
13.1/14/8
0.586
0.460
0.060–3.530
0.455
R status (R0/R1)
181/46
17.3/11.4
\0.001
1.667
0.920–3.019
0.092
Perineural invasion (±)
164/27
15.5/17.2
0.683
1.323
0.614–2.853
0.475
Endolymphatic invasion (±)
112/40
14.0/25.5
0.001
2.027
1.217–3.378
0.007
Microscopic venous invasion (±)
50/87
15.5/17.8
0.034
1.386
0.873–2.201
0.166
1.154
0.733–1.815
0.536
1.407
0.890–2.226
0.144
Lymph node ratio (B0.2/[0.2)
146/79
17.2/12.8
0.096
Total examined lymph node (\15/C15)
81/144
17.8/14.4
0.432
Number of metastatic lymph node (1/C2)
73/154
17.3/13.9
0.365
Adjuvant CCRT (no/yes)
98/129
12.6/17.8
0.004
CCRT concurrent chemoradiation
metastasis significantly worsened prognosis, but among nodepositive patients, LNR did not affect prognosis. Moreover, multivariate analysis revealed LNR did not have significant prognostic significance in node-positive patients. The total number of LNs examined is dependent on how extensively the pathologist examines the surgical specimen, and the prognostic relevance of total number of LNs examined is unclear.11,12,14,15 There is no consensus on what constitutes a sufficient number of LNs examined, with cutoffs ranging from 10 to 16 LNs having been used in previous studies.9–11,13, 28,29 We found that OS was significantly lower in patients with C15 than \15 LNs examined. This is in line with experiences of stage migration effect in other cancers.30,31 However, in agreement with studies reporting that the number of LNs examined was not a prognostic indicator in node-positive pancreatic cancer, the total number of LNs examined was not statistically significant prognostic indicator in nodepositive patients in this study.17,24 In addition, multivariate analysis revealed that total number of LNs examined was not a significant prognostic factor for curatively resected pancreatic cancer. Moreover, our recent randomized
controlled study evaluating the extent of lymphadenectomy revealed that patients who underwent standard and extended LN dissection had comparable prognoses in nodepositive patients.32 These findings indicate that, because the presence of nodal metastasis is of great importance in determining the prognosis of patients with pancreatic cancer, the extent of lymphadenectomy and the number of total examined LNs are not important in node-positive patients. The number of positive LNs is another established prognostic factor in various malignancies including stomach, colon, breast, or ampullary cancer.33–35 The number of metastatic LNs may also be prognostic in patients with pancreatic cancer.7,8 Studies have reported that the prognosis of patients with 0 and 1 positive LN does not differ, suggesting that LNR may be prognostic in these patients.5,24 Other studies have found that patients with 2 positive LNs have similar prognosis as patients with 0 or 1 positive LN.11,16 In contrast, we found that the presence of only 1 positive LN was associated with a significantly poorer OS in patients curatively resected for pancreatic cancer. On the contrary, the number of metastatic LNs
M. J. Kang et al.
among N1 patients did not have prognostic difference. This observation implies that since the presence of a single positive LN has great importance for patient prognosis, patients with any type of N1 disease should be considered different from those with node-negative disease. The presence of tumor cell in regional LNs may reflect the potential of the primary tumor to metastasize. For breast cancer, even the isolated tumor cells or micrometastases are known to reduce overall and disease-free survival of the patients.36 Therefore, it can be logically inferred that presence of LN metastasis of such an aggressive tumor as pancreatic cancer will heavily impact prognosis, regardless of its number. In this study, LNR had limited prognostic value among N1 patients. Moreover, systemic metastasis rate was significantly associated with N stage and endolymphatic invasion, but it was not associated with the number of metastatic LNs or LNR among N1 patients. Therefore, it can be concluded that the presence of any nodal metastasis has a more important role in prognosis than the tumor nodal burden. In this study, types of adjuvant treatment were significantly correlated with prognosis. Chemoradiation therapy had significant survival gain over chemotherapy alone in both node-negative and node-positive patients. This result conflicts with previous studies; however, due to the retrospective nature of this study, this observation results from selection bias of the patients because those who could not endure radiation therapy received chemotherapy alone.37,38 This study has major limitation due to its retrospective study design. However, we provide one of the largest study cohorts of pancreatic head cancer who were treated with standardized treatment protocol. Multivariate analysis compensates the limitation of study design and revealed nodal status was more important than LNR, and LNR did not have statistical significance for N1 patients. In conclusion, LN metastasis per se is more important than the number of LNs examined, the number of metastatic LNs, and LNR in determining patient prognosis. ACKNOWLEDGMENT This study was supported by a Grant from the National R&D Program for Cancer Control funded by the Ministry of Health & Welfare, Republic of Korea (No. 1120310). COMPETING INTEREST
None.
REFERENCES 1. Brennan MF, Kattan MW, Klimstra D, Conlon K. Prognostic nomogram for patients undergoing resection for adenocarcinoma of the pancreas. Ann Surg. 2004;240:293–8. 2. Winter JM, Cameron JL, Campbell KA, Arnold MA, Chang DC, Coleman J, et al. 1423 pancreaticoduodenectomies for pancreatic cancer: a single-institution experience. J Gastrointest Surg. 2006;10:1199–210; discussion 1210–1.
3. Yeo CJ, Cameron JL, Sohn TA, Lillemoe KD, Pitt HA, Talamini MA, et al. Six hundred fifty consecutive pancreaticoduodenectomies in the 1990s: pathology, complications, and outcomes. Ann Surg. 1997;226:248–57; discussion 257–60. 4. Lim JE, Chien MW, Earle CC. Prognostic factors following curative resection for pancreatic adenocarcinoma: a populationbased, linked database analysis of 396 patients. Ann Surg. 2003; 237:74–85. 5. Sohn TA, Yeo CJ, Cameron JL, Koniaris L, Kaushal S, Abrams RA, et al. Resected adenocarcinoma of the pancreas-616 patients: results, outcomes, and prognostic indicators. J Gastrointest Surg. 2000;4:567–79. 6. Riediger H, Makowiec F, Fischer E, Adam U, Hopt UT. Postoperative morbidity and long-term survival after pancreaticoduodenectomy with superior mesenterico-portal vein resection. J Gastrointest Surg. 2006;10:1106–15. 7. John BJ, Naik P, Ironside A, Davidson BR, Fusai G, Gillmore R, et al. Redefining the R1 resection for pancreatic ductal adenocarcinoma: tumour lymph nodal burden and lymph node ratio are the only prognostic factors associated with survival. HPB (Oxford). 2013;15:674–80. 8. Sakata J, Shirai Y, Wakai T, Ajioka Y, Akazawa K, Hatakeyama K. Assessment of the nodal status in ampullary carcinoma: the number of positive lymph nodes versus the lymph node ratio. World J Surg. 2011;35:2118–24. 9. Shamseddine AI, Mukherji D, Melki C, Elias E, Eloubeidi M, Dimassi H, et al. Lymph node ratio is an independent prognostic factor after resection of periampullary malignancies: data from a tertiary referral center in the Middle East. Am J Clin Oncol. 2012. doi:10.1097/COC.0b013e31826b9c74. 10. Falconi M, Crippa S, Dominguez I, Barugola G, Capelli P, Marcucci S, et al. Prognostic relevance of lymph node ratio and number of resected nodes after curative resection of ampulla of Vater carcinoma. Ann Surg Oncol. 2008;15:3178–86. 11. Huebner M, Kendrick M, Reid-Lombardo KM, Que F, Therneau T, Qin R, et al. Number of lymph nodes evaluated: prognostic value in pancreatic adenocarcinoma. J Gastrointest Surg. 2012;16:920–6. 12. Showalter TN, Winter KA, Berger AC, Regine WF, Abrams RA, Safran H, et al. The influence of total nodes examined, number of positive nodes, and lymph node ratio on survival after surgical resection and adjuvant chemoradiation for pancreatic cancer: a secondary analysis of RTOG 9704. Int J Radiat Oncol Biol Phys. 2011;81:1328–35. 13. Slidell MB, Chang DC, Cameron JL, Wolfgang C, Herman JM, Schulick RD, et al. Impact of total lymph node count and lymph node ratio on staging and survival after pancreatectomy for pancreatic adenocarcinoma: a large, population-based analysis. Ann Surg Oncol. 2008;15:165–74. 14. Berger AC, Watson JC, Ross EA, Hoffman JP, et al. The metastatic/examined lymph node ratio is an important prognostic factor after pancreaticoduodenectomy for pancreatic adenocarcinoma. Am Surg. 2004;70:235–40; discussion 240. 15. Bhatti I, Peacock O, Awan AK, Semeraro D, Larvin M, Hall RI. Lymph node ratio versus number of affected lymph nodes as predictors of survival for resected pancreatic adenocarcinoma. World J Surg. 2010;34:768–75. 16. House MG, Gonen M, Jarnagin WR, D’Angelica M, DeMatteo RP, Fong Y, et al. Prognostic significance of pathologic nodal status in patients with resected pancreatic cancer. J Gastrointest Surg. 2007;11:1549–55. 17. Pawlik TM, Gleisner AL, Cameron JL, Winter JM, Assumpcao L, Lillemoe KD, et al. Prognostic relevance of lymph node ratio following pancreaticoduodenectomy for pancreatic cancer. Surgery. 2007;141:610–8.
Lymph Node Ratio of Pancreatic Cancer 18. Feinstein AR, Sosin DM, Wells CK. The Will Rogers phenomenon. Stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med. 1985;312:1604–8. 19. Bando E, Yonemura Y, Taniguchi K, Fushida S, Fujimura T, Miwa K. Outcome of ratio of lymph node metastasis in gastric carcinoma. Ann Surg Oncol. 2002;9:775–84. 20. Bogoevski D, Onken F, Koenig A, Kaifi JT, Schurr P, Sauter G, et al. Is it time for a new TNM classification in esophageal carcinoma? Ann Surg. 2008;247:633–41. 21. Rosenberg R, Friederichs J, Schuster T, Gertler R, Maak M, Becker K, et al. Prognosis of patients with colorectal cancer is associated with lymph node ratio: a single-center analysis of 3,026 patients over a 25-year time period. Ann Surg. 2008;248:968–78. 22. Sierra A, Regueira FM, Hernandez-Lizoain JL, Pardo F, Martı´nez-Gonzalez MA, A-Cienfuegos J. Role of the extended lymphadenectomy in gastric cancer surgery: experience in a single institution. Ann Surg Oncol. 2003;10:219–26. 23. La Torre M, Cavallini M, Ramacciato G, Cosenza G, Rossi Del Monte S, Nigri G, et al. Role of the lymph node ratio in pancreatic ductal adenocarcinoma. Impact on patient stratification and prognosis. J Surg Oncol. 2011;104:629–33. 24. Riediger H, Keck T, Wellner U, zur Hausen A, Adam U, Hopt UT, et al. The lymph node ratio is the strongest prognostic factor after resection of pancreatic cancer. J Gastrointest Surg. 2009;13:1337–44. 25. Sierzega M, Popiela T, Kulig J, Nowak K. The ratio of metastatic/resected lymph nodes is an independent prognostic factor in patients with node-positive pancreatic head cancer. Pancreas. 2006;33:240–5. 26. Herr HW. Superiority of ratio based lymph node staging for bladder cancer. J Urol. 2003;169:943–5. 27. Partelli S, Fernandez-Del Castillo C, Bassi C, Mantovani W, Thayer SP, Crippa S, et al. Invasive intraductal papillary mucinous carcinomas of the pancreas: predictors of survival and the role of lymph node ratio. Ann Surg. 2010;251:477–82. 28. Schwarz RE, Smith DD. Extent of lymph node retrieval and pancreatic cancer survival: information from a large US population database. Ann Surg Oncol. 2006;13:1189–200.
29. Tomlinson JS, Jain S, Bentrem DJ, Sekeris EG, Maggard MA, Hines OJ, et al. Accuracy of staging node-negative pancreas cancer: a potential quality measure. Arch Surg. 2007;142:767–73; discussion 773–4. 30. Kong SH, Lee HJ, Ahn HS, Kim JW, Kim WH, Lee KU, et al. Stage migration effect on survival in gastric cancer surgery with extended lymphadenectomy: the reappraisal of positive lymph node ratio as a proper N-staging. Ann Surg. 2012;255:50–8. 31. Yoshikawa T, Sasako M, Sano T, Nashimoto A, Kurita A, Tsujinaka T, et al. Stage migration caused by D2 dissection with para-aortic lymphadenectomy for gastric cancer from the results of a prospective randomized controlled trial. Br J Surg. 2006;93:1526–9. 32. Jang JY, Kang MJ, Heo JS, Choi SH, Choi DW, Park SJ, et al. A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer. Ann Surg. 2013. doi:10.1097/SLA. 0000000000000384. 33. Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti III A. AJCC Cancer Staging Manual. 7th ed. New York: Springer, 2010. 34. Carter CL, Allen C, Henson DE. Relation of tumor size, lymph node status, and survival in 24,740 breast cancer cases. Cancer 1989;63:181–7. 35. Sakata J, Shirai Y, Wakai T, Yokoyama N, Sakata E, Akazawa K, et al. Number of positive lymph nodes independently affects long-term survival after resection in patients with ampullary carcinoma. Eur J Surg Oncol. 2007;33:346–51. 36. de Boer M, van Deurzen CH, van Dijck JA, Borm GF, van Diest PJ, Adang EM, et al. Micrometastases or isolated tumor cells and the outcome of breast cancer. N Engl J Med. 2009;361:653–63. 37. Oettle H, Post S, Neuhaus P, Gellert K, Langrehr J, Ridwelski K, et al. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer: a randomized controlled trial. JAMA. 2007;297:267–77. 38. Neoptolemos JP, Dunn JA, Stocken DD, Almond J, Link K, Beger H, et al. Adjuvant chemoradiotherapy and chemotherapy in resectable pancreatic cancer: a randomised controlled trial. Lancet. 2001;358:1576–85.
