4
British Journal of Surgery 1995,82,399-402 Lymph node metastases in carcinoma of the head of the pancreas region A. NAKAO, A. HARADA, T. NONAMI, T. KANEKO, H. MURAKAMI, S. INOUE, Y. TAKEUCHI and H. TAKAGI Department of Surgery ZZ, Nagoya lJniversi@ School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466, Japan Correspondence to: Mr A. Nakao Histopathological examination of lymph node metastatic involvement in 139 specimens obtained from patients who underwent pancreatoduodenectomy or total pancreat- ectomy combined with wide resection of lymph nodes was performed, to clarify the critical areas of lymph node dissection in patients with carcinoma of the head of the pancreas region. Perigastric lymph node involvement in patients with carcinoma of the head of the pancreas was 14 per cent, in those with carcinoma of the distal bile duct 0 per cent and in those with carcinoma of the papilla of Vater 4 per cent. Para-aortic lymph node involvement in patients with carcinoma of the head of the pancreas, the distal bile duct and the papilla of Vater was 26, 9 and 0 per cent, respectively. On the basis of these results, pylorus-preserving pancreatoduodenectomy is indicated in almost all patients with carcinoma of the distal bile duct and the papilla of Vater. In patients with carcinoma of the head of the pancreas, however, wide dissection of lymph nodes, including para-aortic lymph nodes, should be carried out because of the relatively high incidence of para-aortic lymph node involvement. Lymph node dissection is an important component of radical surgery for treating carcinoma of the head of the pancreas region. In 1973, Fortner' advocated an extended radical surgical procedure to improve the prognosis of patients with pancreatic cancer. Since 1981, the authors2 have performed extended radical surgery with wide dissection of lymph nodes, including the para-aortic lymph nodes, and portal vein resection for those with carcinoma of the head of the pancreas region. Histopathological studies of lymph node involvement have practical use and are helpful for guiding operative therapy. The present study was undertaken using resected specimens to evaluate the extent of lymph node involvement, especially involvement of the perigastric and para-aortic lymph nodes, in patients with carcinoma of the head of the pancreas, the distal bile duct and the papilla of Vater. Patients and methods Some 139 patients who underwent wide dissection of lymph nodes and an extensive histopathological examination were selected from 153 patients who had resection for cancer of the head of the pancreas region between July 1981 and March 1993. Specimens were examined histopathologically from 90 patients with duct cell carcinoma of the head of the pancreas, 22 with carcinoma of the distal bile duct and 27 with carcinoma of the papilla of Vater. Forty-eight total pancreatectomies and 42 pancreatoduodenectomies with systematic lymph node (including regional and para-aortic lymph nodes) dissection for cancer of the head of the pancreas were performed between 1981 and March 1993 (Table I). Portal vein resection was performed in all the total pancreatectomies and in 41 of 42 pancreato- duodenectomies for patients with carcinoma of the head of the pancreas using a catheter bypass procedure for the portal vein3. Pancreatoduodenectomy was usually carried out in patients with malignancy of the distal bile duct or the papilla of Vater (Table I). A total gastrectomy was performed in one patient and distal gastrectomy in the remaining 138. Pathological findings were evaluated in accordance with the General Rules for Surgical and Pathological Studies on Cancer of Paper accepted 1 August 1994 Table 1 Operative procedures for carcinoma of the head of the pancreas region No. of total No. of pancreat- pancreato- ectomies duodenectomies Carcinoma of the head 48 (48) 42 (41) of the pancreas (n = 90) Carcinoma of the distal bile duct (n = 22) of Vater (n = 27) 21 (2) Carcinoma of the papilla 10) 26 (2) Values in parentheses indicate the number of portal vein resections the Pancreas proposed by the Japan Pancreas Sociev and the General Rules for Surgical and Pathological Studies on Cancer of the Biliary Tract proposed by the Japanese Society of Biliary Surgerys. These classifications are more precise than the tumour node metastasis classification. The nomenclature of the major lymph nodes is defined in Fig. 1. The area of para-aortic lymph node dissection extended from the coeliac trunk to the inferior mesenteric trunk, and from the right margin of the inferior vena cava to the left margin of the abdominal aorta (Fig. 2). Some specimens contained only a few lymph nodes, but those with no removed nodes were treated as metastatic-negative. The relationships among the lymph node groups, especially focusing on perigastric and para-aortic lymph nodes, were studied using the xz test. P < 0.05 was considered significant. Survival, including postoperative death, was calculated by the Kaplan-Meier method. Results The number of dissected lymph nodes in each case ranged between seven and 127 (mean 48.0) in patients with carcinoma of the head of the pancreas, between four and 90 (mean 26.7) in those with carcinoma of the distal bile duct, and between ten and 54 (mean 30.4) in those with carcinoma of the papilla of Vater. The frequency of lymph node involvement is given in Table 2. Sixty-nine (77 per cent) of the 90 patients with 399

Lymph node metastases in carcinoma of the head of the pancreas region

Embed Size (px)

Citation preview

British Journal of Surgery 1995,82,399-402

Lymph node metastases in carcinoma of the head of the pancreas region A . N A K A O , A . H A R A D A , T. N O N A M I , T. K A N E K O , H. M U R A K A M I , S . I N O U E ,

Y . T A K E U C H I and H. T A K A G I

Department of Surgery ZZ, Nagoya lJniversi@ School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466, Japan Correspondence to: Mr A. Nakao

Histopathological examination of lymph node metastatic involvement in 139 specimens obtained from patients who underwent pancreatoduodenectomy or total pancreat- ectomy combined with wide resection of lymph nodes was performed, to clarify the critical areas of lymph node dissection in patients with carcinoma of the head of the pancreas region. Perigastric lymph node involvement in patients with carcinoma of the head of the pancreas was 14 per cent, in those with carcinoma of the distal bile duct 0 per cent and in those with carcinoma of the papilla of Vater 4 per cent. Para-aortic lymph node involvement in

patients with carcinoma of the head of the pancreas, the distal bile duct and the papilla of Vater was 26, 9 and 0 per cent, respectively. On the basis of these results, pylorus-preserving pancreatoduodenectomy is indicated in almost all patients with carcinoma of the distal bile duct and the papilla of Vater. In patients with carcinoma of the head of the pancreas, however, wide dissection of lymph nodes, including para-aortic lymph nodes, should be carried out because of the relatively high incidence of para-aortic lymph node involvement.

Lymph node dissection is an important component of radical surgery for treating carcinoma of the head of the pancreas region. In 1973, Fortner' advocated an extended radical surgical procedure to improve the prognosis of patients with pancreatic cancer. Since 1981, the authors2 have performed extended radical surgery with wide dissection of lymph nodes, including the para-aortic lymph nodes, and portal vein resection for those with carcinoma of the head of the pancreas region.

Histopathological studies of lymph node involvement have practical use and are helpful for guiding operative therapy. The present study was undertaken using resected specimens to evaluate the extent of lymph node involvement, especially involvement of the perigastric and para-aortic lymph nodes, in patients with carcinoma of the head of the pancreas, the distal bile duct and the papilla of Vater.

Patients and methods Some 139 patients who underwent wide dissection of lymph nodes and an extensive histopathological examination were selected from 153 patients who had resection for cancer of the head of the pancreas region between July 1981 and March 1993.

Specimens were examined histopathologically from 90 patients with duct cell carcinoma of the head of the pancreas, 22 with carcinoma of the distal bile duct and 27 with carcinoma of the papilla of Vater. Forty-eight total pancreatectomies and 42 pancreatoduodenectomies with systematic lymph node (including regional and para-aortic lymph nodes) dissection for cancer of the head of the pancreas were performed between 1981 and March 1993 (Table I). Portal vein resection was performed in all the total pancreatectomies and in 41 of 42 pancreato- duodenectomies for patients with carcinoma of the head of the pancreas using a catheter bypass procedure for the portal vein3. Pancreatoduodenectomy was usually carried out in patients with malignancy of the distal bile duct or the papilla of Vater (Table I). A total gastrectomy was performed in one patient and distal gastrectomy in the remaining 138.

Pathological findings were evaluated in accordance with the General Rules for Surgical and Pathological Studies on Cancer of

Paper accepted 1 August 1994

Table 1 Operative procedures for carcinoma of the head of the pancreas region

No. of total No. of pancreat- pancreato- ectomies duodenectomies

Carcinoma of the head 48 (48) 42 (41) of the pancreas (n = 90)

Carcinoma of the distal bile duct (n = 22)

of Vater (n = 27)

21 (2)

Carcinoma of the papilla 1 0 ) 26 (2)

Values in parentheses indicate the number of portal vein resections

the Pancreas proposed by the Japan Pancreas Sociev and the General Rules for Surgical and Pathological Studies on Cancer of the Biliary Tract proposed by the Japanese Society of Biliary Surgerys. These classifications are more precise than the tumour node metastasis classification. The nomenclature of the major lymph nodes is defined in Fig. 1. The area of para-aortic lymph node dissection extended from the coeliac trunk to the inferior mesenteric trunk, and from the right margin of the inferior vena cava to the left margin of the abdominal aorta (Fig. 2). Some specimens contained only a few lymph nodes, but those with no removed nodes were treated as metastatic-negative.

The relationships among the lymph node groups, especially focusing on perigastric and para-aortic lymph nodes, were studied using the xz test. P < 0.05 was considered significant. Survival, including postoperative death, was calculated by the Kaplan-Meier method.

Results The number of dissected lymph nodes in each case ranged between seven and 127 (mean 48.0) in patients with carcinoma of the head of the pancreas, between four and 90 (mean 26.7) in those with carcinoma of the distal bile duct, and between ten and 54 (mean 30.4) in those with carcinoma of the papilla of Vater.

The frequency of lymph node involvement is given in Table 2. Sixty-nine (77 per cent) of the 90 patients with

399

400 A . NAKAO, A . HARADA, T . NONAMIetnl.

Fig. 1 Nomenclature of a perigastric lymph nodes in patients with carcinoma of the head of the pancreas and b lymph nodes in carcinoma of the head of the pancreas region. 1, right cardiac lymph nodes; 2, left cardiac lymph nodes; 3, lesser curvature lymph nodes; 4, greater curvature lymph nodes; 5, suprapyloric lymph nodes; 6, infrapyloric lymph nodes; 7, lymph nodes around the left gastric artery; 8, lymph nodes around the common hepatic artery; 9, lymph nodes around the coeliac trunk; 10, lymph nodes at the hilus of the spleen; 11, lymph nodes along the splenic artery; 12, lymph nodes of the hepatoduodenal ligament; 13, posterior pancreaticoduodenal lymph nodes; 14, lymph nodes around the superior mesenteric artery; 15, lymph nodes along the middle colic artery; 16, para-aortic lymph nodes; 17, anterior pancreaticoduodenal lymph nodes; 18, inferior pancreatic body lymph nodes

Table 2 Lymph node involvement in patients with carcinoma of the head of the pancreas region

Fig. 2 Retroperitoneal area after total pancreatectomy combined with wide lymph node dissection and portal vein resection for carcinoma of the head of the pancreas

carcinoma of the head of the pancreas had lymph node involvement, and perigastric lymph node (especially around the right gastroepiploic artery) involvement was observed in 13 (14 per cent). A high incidence of lymph node involvement was confirmed on the posterior surface of the pancreatic head (51 per cent), anterior surface of the pancreatic head (39 per cent), para-aortic lymph nodes (26 per cent) and around the superior mesenteric artery (23 per cent). Nine of 22 patients with carcinoma of the distal bile duct had lymph node involvement: no

Carcinoma of the Carcinoma of the Carcinoma of the Lymph head of the pancreas distal bile duct nodes (n =90) (n = 22) (n = 27)

papilla of Vater

1 0 2 0 3 0 4 0 5 0 6 13 (14) 7 0 8 12 (13) 9 2 (2)

10 l ( 1 ) 11 16(18) 12 17(19) 13 46 (51) 14 21 (23) 15 0 16 23 (26) 17 35 (39) 18 3 (3)

0 0 0

0 0 0 0 0 0 0

1(4)

1 (4)

3 (11)

6 (22)

11 (41)

0 0

0

Values in parentheses represent percentages

perigastric lymph node involvement was observed, while para-aortic lymph node metastasis was noted in two. Lymph node involvement was found in the hepatoduodenal ligament in five of 22 patients. Twelve of 27 patients with carcinoma of the papilla of Vater had lymph node involvement: perigastric lymph node involvement was identified in only one but no para-aortic lymph node metastasis was recorded. A high incidence of lymph node involvement was found on the posterior (11 of 27 patients) and anterior (six) surface of the pancreatic head. The metastatic rates of lymph node groups 1, 2, 3 and 5 were zero, but one patient with carcinoma of the papilla of Vater had metastases in lymph nodes of groups

British Journal of Surgery 1995,82, 399-402

LYMPH NODE METASTASES I N C A R C I N O M A O F T H E H E A D O F T H E PANCREAS R E G I O N 401

4, 12 and 13. Metastasis in group 6 lymph nodes was observed only in those with carcinoma of the head of the pancreas and was positive in 14 per cent of such patients. Metastases in groups 7 and 15 lymph nodes were not detected in any patient.

The 13 patients with carcinoma of the head of the pancreas and involvement of group 6 lymph nodes had a much higher frequency of metastasis in node groups 9 and 12 compared with that of those without metastatic involvement of group 6 lymph nodes (P < 0.05) (Table 3).

The 23 patients with pancreatic carcinoma and metastatic involvement of group 16 lymph nodes had a much higher frequency of metastases in node groups 12, 13, 14 (P < 0.01) and 17 (P < 0.05) than those without involvement of group 16 lymph nodes (Table 4) .

Two patients with carcinoma of the distal bile duct and involvement of group 16 lymph nodes also had metastasis in group 14 nodes.

There was no correlation between lymph node involvement and various other parameters such as tumour size and carcinoma invasion into the portal vein.

Operative mortality (within 30 days after operation) was observed in nine (10 per cent) of 90 patients with carcinoma of the head of the pancreas. There were no deaths in patients with carcinoma of the distal bile duct or the papilla of Vater.

Table 3 Frequency of lymph node metastases in patients with and without metastatic involvement of group 6 lymph nodes

Lymph Involvement of No involvement of node group 6 lymph nodes group 6 lymph nodes group (n = 13) (n = 77)

8 9

10 0 ' ' 1(1) 11 (14) 11 (14) 12 6 (46)*

13 10 (77) 36 (47) 17 (22)

15 0 17 (22) 28 (36)

11 5 (38)

14 4 (31)

16 6 (46) 17 7 (54) 18 1 (8) 2 (3)

0

Values in parentheses are percentages. *P < 0.05 (x' test)

Table 4 Frequency of lymph node metastases in patients with and without metastatic involvement of group 16 lymph nodes

Lymph Involvement of No involvement of node group 16 lymph nodes group 16 lymph nodes group (n = 23) (n = 67)

6 6 (26) 7 (10) 8 5 (22) 7 (10) 9 1(4) 1(1)

10 0 1(1) 11 7 (30) 9 (13) 12 11 (48)** 6 (9)

14 12 (52)** 9 (13)

18 1(4) 2 (3)

13 18 (78)** 28 (42)

15 0 0 17 16 (70)* 19 (28)

Values in parentheses are percentages. *P < 0.05, **P < 0.01 (x' test)

The cumulative survival rate for patients with carcinoma of the head of the pancreas was 30 per cent in the first year, 12 per cent in the third year and 5 per cent in the fifth year after operation. The 23 patients with cancer of the head of the pancreas and involvement of group 16 lymph nodes &d from reeu-rrence (Rot lyfftpk node involvement but especially liver metastasis) within 37 months after operation. The cumulative survival rate of patients with bile duct malignancy was 66 per cent in the first year, 46 per cent in the third and 31 per cent in the fifth. Two patients with involvement of group 16 lymph nodes died from liver metastasis 5 months after operation and without recurrence 54 months after surgery. The cumulative survival rate of patients with carcinoma of papilla of Vater was 89 per cent in the first year, 57 per cent in the third and 53 per cent in the fifth.

Discussion Since its description in 19356, the Whipple operation has been the standard procedure for the treatment of resectable tumours of the head of the pancreas and periampullary region. The standard Whipple procedure involves a radical pancreatoduodenectomy with a large gastric resection. In recent years, after the report of Traverso and Longmire7, numerous authors have advocated modifications of the Whipple operation with preservation of the stomach, pylorus and proximal duodenum to decrease postgastrectomy complications without compromising en bloc resection of the malignancy.

Lymph node dissection is an important component of radical surgery for those with carcinoma of the pancreatic head region, and the value of pylorus-preserving pancreatoduodenectomy is decided by the presence or absence of positive lymph node involvement in the perigastric lymph nodes. In the study by Cubilla et d8, using the specimens obtained after a Fortner's regional pancreatectomy, one of 22 patients had perigastric lymph node involvement. Perigastric lymph node involvement was also reported in one of eight patients in the study by Nagai et aL9 and in one of 44 patients in the study by Kayahara and colleagues10. In the present series, perigastric lymph node involvement was observed in 14 per cent of those with carcinoma of the head of the pancreas but only around the right gastroepiploic artery. In addition, as lymph node involvement around the left gastric artery and coeliac trunk was 0 and 2 per cent, respectively, pylorus-preserving pancreatoduodenectomy may be indicated if perigastric lymph node involvement is not observed. The incidence of other lymph node involvement, however, was very high, especially on the posterior or anterior surface of the pancreatic head, in the para-aortic region and around the superior mesenteric artery. There have been few studies concerning para- aortic lymph node involvement in patients with carcinoma of the head of the pancreas. In the study by Kayahara and colleagues" seven of 44 patients had para-aortic lymph node involvement. The incidence of para-aortic lymph node involvement in the series described here was 26 per cent and these cases had frequencies of positive lymph node involvement of 78 per cent on the posterior surface of the pancreatic head and 52 per cent around the superior mesenteric artery. Lymphatic flow from the tumour to the para-aortic lymph nodes via the posterior surface of the pancreatic head and around the superior

British Journal of Surgery 1995,82, 399-402

402 A. N A K A O , A . H A R A D A , T. N O N A M I et al.

mesenteric artery has been suspected and supported by other datalo. Thus, wide dissection of lymph nodes, including the para-aortic lymph nodes, and retroperitoneal connective tissue may be necessary''. The postoperative survival rate, however, is still low in patients with cancer of the head of the pancreas. The high incidence of liver metastasis after operation is a prime cause of the poor outcome, and effective therapy for postoperative liver recurrence requires evaluation.

Perigastric lymph node involvement was not observed in patients with carcinoma of the distal bile duct, but two patients had para-aortic lymph node involvement. These two patients also had lymph node involvement around the superior mesenteric artery. Pylorus-preserving pancreato- duodenectomy may be indicated in almost all cases of cancer of the distal bile duct, but lymph nodes of the hepatoduodenal ligament should be dissected and para- aortic lymph node dissection combined with lymph node dissection around the superior mesenteric artery may be required. Intraoperative pathological diagnosis of lymph node involvement using frozen sections provides important information as to whether more extensive dissection is necessary.

One patient with carcinoma of the papilla of Vater had perigastric lymph node involvement around the peripheral right gastroepiploic artery. This patient also had lymph node involvement on the posterior surface of the pancreatic head and in the hepatoduodenal ligament. Lymph node involvement around the superior mesenteric artery was identified in three of 27 patients, but no para- aortic lymph node involvement was observed. Thus, pylorus-preserving pancreatoduodenectomy may be indicated in almost all cases of cancer of the papilla of Vater, but lymph node dissection around the superior mesenteric artery and in the hepatoduodenal ligament may be necessary. Intraoperative pathological diagnosis of lymph node involvement using frozen sections is important to determine the extent of dissection.

On the basis of these results, pylorus-preserving pancreatoduodenectomy may be indicated in almost all

patients with carcinoma of the distal bile duct or the papilla of Vater. Because patients with carcinoma of the head of the pancreas often have more extensive lymph node involvement, however, wide dissection of lymph nodes including the para-aortic lymph nodes may be required.

References 1 Fortner JG. Regional resection of cancer of the pancreas: a

new surgical approach. Surgey 1973; 73: 307-20. 2 Nakao A, Horisawa M, Kondo T et al. Total pancreatectomy

accompanied by portal vein resection using catheter-bypass of the portal vein. Shujutsu (Operation) 1983; 37: 1-6 (in Japanese).

3 Nakao A, Nonami T, Harada A, Kasuga T, Takagi H. Portal vein resection with a new antithrombogenic catheter. Surgey

4 Japan Pancreas Society. General Rules for Surgical and Pathological Studies on Cancer of the Pancreas. 3rd ed. Tokyo: Kanehara Publishing, 1987 (in Japanese).

5 Japanese Society of Biliary Surgery. General Rules for Surgical and Pathological Studies on Cancer of the Biliay Cact. 3rd ed. Tokyo: Kanehara Publishing, 1993 (in Japacese).

6 Whipple AO, Parsons WB, Mullins CR. Treatment of carcinoma of the ampulla of Vater. Ann Surg 1935; 102:

7 Traverso LW, Longmire WP Jr. Preservation of the pylorus in pancreaticoduodenectomy. Surg Gynecol Obstet 1978; 146:

8 Cubilla AL, Fortner J, Fizgerald PJ. Lymph node involvement in carcinoma of the head of the pancreas area. Cancer 1978; 41: 880-7.

9 Nagai H, Kuroda A, Morioka Y. Lymphatic and local spread of TI and T2 pancreatic cancer. A study of autopsy material. Ann Surg 1986; 204: 65-71.

10 Kayahara M, Nagakawa T, Kobayashi H et al. Lymphatic flow in carcinoma of the head of the pancreas. Cancer 1992; 70:

11 Ishikawa 0, Ohhigashi H, Sasaki Y et al. Practical usefulness of lymphatic and connective tissue clearance for the carcinoma of the pancreas head. Ann Surg 1988; 208: 215-20.

1990; 108: 913-18.

763-79.

959-62.

2061-6.

British Journal of Surgey 1995,82,399-402