Pancreaticoduodenectomy with Extended Retroperitoneal Lymphadenectomy for Periampullary Adenocarcinoma

Pancreaticoduodenectomy with Extended Retroperitoneal Lymphadenectomy for Periampullary Adenocarcinoma

Surg Oncol Clin N Am 16 (2007) 157–176 Pancreaticoduodenectomy with Extended Retroperitoneal Lymphadenectomy for Periampullary Adenocarcinoma Eugene ...

442KB Sizes 0 Downloads 84 Views

Surg Oncol Clin N Am 16 (2007) 157–176

Pancreaticoduodenectomy with Extended Retroperitoneal Lymphadenectomy for Periampullary Adenocarcinoma Eugene P. Kennedy, MD, Charles J. Yeo, MD* Department of Surgery, Thomas Jefferson University, Jefferson Medical College, 1025 Walnut Street, Suite 605 College Building, Philadelphia, PA 19107, USA

Periampullary adenocarcinoma (adenocarcinoma within the right half of the pancreas, the ampulla of Vater, distal common bile duct, or peri-Vaterian duodenum) is a common cause of cancer mortality in the United States affecting more than 30,000 individuals per year [1]. The only chance for long-term survival or cure is surgical resection of the tumor. Pancreaticoduodenectomy (the Whipple operation) is currently the standard procedure for patients with these tumors. Long-term outcomes following surgery depend on a variety of factors including primary tumor characteristics (site of origin, size, resection margin status, nodal status) [2–8], tumor ploidy level [9], molecular genetics [10], postresection CA 19-9 levels [11], and adjuvant chemotherapy and chemoradiation therapy [4,12–15]. Even with surgical resection, in the more favorable subsets of patients, 5-year survival only approaches 50%.

Historical perspective and background Surgical resection for periampullary tumors dates to the end of the nineteenth century. The first successful resection of a periampullary cancer was reported by Halsted [16] in 1899. Dr Whipple’s reports [17] of his successful resections during the 1930s and 1940s established first a two-stage and then a one-stage pancreaticoduodenectomy (with implantation of the pancreatic remnant into the jejunum) as the procedure of choice for these malignancies.

* Corresponding author. E-mail address: [email protected] (C.J. Yeo). 1055-3207/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.soc.2006.10.010 surgonc.theclinics.com

158

KENNEDY & YEO

Early experience with this procedure, however, yielded very high perioperative mortality rates. It was not until the 1960s and 1970s that reports of significant series without operative deaths began to be published [18]. As short-term results became more predictable, attention was turned to modifying the operation in ways to improve long-term oncologic outcomes. In 1973, Dr Joseph Fortner [19–22] proposed a ‘‘regional resection of the pancreas’’ involving a total pancreaticoduodenectomy with subtotal gastrectomy and resection of the transpancreatic portion of the portal vein, as well as occasional visceral vessel resection and reconstruction. Such radical or extended pancreaticoduodenal resections attempted to incorporate observations about the patterns of failure after standard resections into the operative management of the disease. Patients who develop recurrence after pancreaticoduodenectomy most often have local intraabdominal recurrences in the peripancreatic region and within the liver. Several attempts have been made at clinicopathologic analysis of lymph node involvement after pancreaticoduodenectomy as a way to better understand and predict patterns of recurrence. Cubilla and colleagues [23] studied the pattern of lymph node involvement in patients with pancreatic cancer. When 22 extended pancreaticoduodenectomy specimens were examined, it was revealed that one third of lymph node–positive patients had nodal metastases along the superior and inferior borders of the body of the pancreas. Obviously, these lymph nodes would not have been removed during standard pancreaticoduodenectomy, suggesting that a more radical operation might achieve better oncologic outcomes. Kayahara and colleagues [24] analyzed the pattern of tumor spread, with reference to nodal involvement and invasion of the extrapancreatic plexus, in 110 patients with periampullary adenocarcinoma. In patients with cancer of the pancreatic head, 76% were found to have spread to regional lymph nodes. Locations were classified anatomically as per the Japanese classification system (Fig. 1) with the posterior pancreaticoduodenal lymph nodes (stations 13a and 13b) being the most commonly involved (51%), followed by the superior mesenteric artery lymph nodes (station 14) at 37%, the anterior pancreaticoduodenal (stations 17a and 17b) at 33%, and the periaortic lymph nodes (station 16) at 16%. Again, this study identified involved regional lymph nodes not routinely removed during standard pancreaticoduodenectomy. A clinicopathologic correlation published by Ishikawa and colleagues [25] also looked at the distribution of lymph node metastases and correlated involvement with survival. Their analysis of 81 patients who underwent extended pancreaticoduodenectomy for adenocarcinoma of the head of the pancreas revealed lymph node involvement in 73% of the patients. Involved lymph nodes were grouped according to the same Japanese classification system. The group most commonly involved was the posterior pancreaticoduodenal (49%), followed by the superior mesenteric artery lymph nodes (47%) and the anterior pancreaticoduodenal lymph nodes (37%). For the other 11 lymph node groups, involvement was seen in less than 20% of

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

159

Fig. 1. Drawing showing nomenclature of the major lymph nodes involved with pancreatic carcinoma. 8: lymph nodes surrounding the common hepatic artery; 9: lymph nodes surrounding the celiac trunk; 10: lymph nodes at the hilum of the spleen; 11: lymph nodes along the splenic artery; 12: lymph nodes in the hepatoduodenal ligament; 13: posterior pancreaticoduodenal lymph nodes; 14: lymph nodes surrounding the superior mesenteric artery; 15: lymph nodes along the middle colic artery; 16: para-aortic lymph nodes; 17: anterior pancreaticoduodenal lymph nodes; and 18: inferior body lymph nodes. (From Kayahara M, Nagakawa T, Ohta T, et al. Analysis of paraaortic lymph node involvement in pancreatic carcinoma: a significant indication for surgery? Cancer 85(3):583–90; with permission. Copyright 1999, John Wiley & Sons, Inc.)

patients. In only 5% of patients was lymph node involvement limited to the superior mesenteric artery lymph node group. When lymph node involvement was identified in the 11 infrequently involved groups, it was never isolated involvement, but was always associated with involvement of the posterior pancreaticoduodenal, superior mesenteric artery, or anterior pancreaticoduodenal lymph node groups. Five-year survival was found to be related to the extent of lymph node involvement. Patients without lymph node involvement had a 5-year survival of 59%. Those with involvement limited to the posterior and anterior pancreaticoduodenal lymph node groups had a 53% 5-year survival. With involvement extending to the superior mesenteric artery lymph nodes, the 5-year survival dropped to 15%. Any extension to the 11 infrequently involved lymph node groups indicated a very poor prognosis, with no 5-year survivors. Of note, while these data show dramatic differences in the 5-year survival rates for patients with secondorder lymph node involvement, the 5-year survival rates of 59% for lymph node–negative patients and 53% for patients with involved lymph nodes are remarkably high and not consistent with other reports in the literature. Following Fortner’s proposition that a more radical resection might improve the cure rate after pancreaticoduodenectomy, multiple reports were published addressing this question [24,26–28]. Most comparisons have yielded conflicting results as to any survival advantage conferred by extended or radical pancreaticoduodenectomy. Those studies that have shown a survival advantage have tended to be nonrandomized, retrospective

160

KENNEDY & YEO

studies limited by their lack of concurrent controls and lack of randomization between clearly defined surgical approaches. When evaluating these studies it is important to note that what constitutes a radical or extended pancreaticoduodenectomy (versus a standard resection) is not universally defined and has varied considerably among past trials [29].

Prospective studies Because of the lack of consensus on this issue, several prospective trials have been performed and reported in recent years (Table 1). First, HenneBruns and colleagues [30] reported a prospective nonrandomized study from Germany that evaluated 72 patients with adenocarcinoma of the pancreatic head. Twenty-six patients underwent pancreaticoduodenectomy with standard lymph node dissection including removal of all lymph nodes at the hepatoduodenal ligament, the celiac trunk including the first 3 cm of the splenic artery, the right side of the superior mesenteric artery, and the ventral surface of the vena cava and the renal veins. An additional 46 patients underwent a more extensive retroperitoneal lymphadenectomy including clearance of all lymphatic, connective, and neural tissue along the left side of the superior mesenteric artery and along the aorta between the inferior mesenteric artery and the diaphragm. Extended lymph node dissection did, as expected, yield an improved lymph node retrieval with 14 nodes in the lesser procedure versus 24 nodes in the extended dissection. There was, however, no evidence of skip metastases (positive nodes in the extended specimen with only negative nodes in the field encompassed by a standard resection) and no survival benefit associated with the extended procedure. Pedrazzoli and colleagues [31] reported a multicenter, prospective, randomized, multicenter trial of 81 patients with pancreatic adenocarcinoma treated at six Italian institutions. Patients undergoing pancreaticoduodenectomy were stratified by tumor size and then randomized to one of two Table 1 Summary of relevant clinical trials

Investigator

Year of publication

Location

Henne-Bruns Pedrazzoli Farnell

2000 1998 2005

Germany Italy US

Nimura Seiler Riall

2004 2005 2005

Japan Switzerland US

a

Indicates median survival.

Number of patients 72 81 79 101 110 280

Standard resection survival

Extended resection survival

35% 5 y 11.1 moa 16.5% 5 y 26 moa 32.4% 3 y 34 moa 25% 5 y 25 moa

17.6% 5 y 16.7 moa 16.4% 5 y 18.8 moa 16% 3 y 27 moa 31% 5 y 28 moa

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

161

different lymphadenectomy approaches. Forty patients underwent a standard dissection with removal of the anterior and posterior pancreaticoduodenal, pyloric and biliary duct, and superior and inferior pancreatic head and body lymph node stations. The other 41 patients additionally had the lymph nodes from the hepatic hilum and along the aorta from the diaphragm to the inferior mesenteric artery removed, including lateral clearance to both renal hila and circumferentially at the origin of the celiac trunk and superior mesenteric artery. No patients received postoperative adjuvant therapy, but 19 did receive intraoperative radiotherapy. Those 19 patients were evenly distributed between the two groups. Demographics and histology were comparable between the groups as were postoperative mortality (5%) and morbidity (40%). An average of 13 lymph nodes were removed in the standard dissection group and 20 in the extended group. There were no differences in survival between the two groups, with overall actual 1-, 2-, 3-, and 4-year survival rates of 51%, 22%, 9%, and 7%, respectively. The authors performed a survival analysis based on the presence or absence of metastatic lymph nodes using an a posteriori analysis that was not planned at the time of the study design. They found that lymph node–positive patients had a significantly better survival rate after undergoing the extended dissection. Several conclusions were drawn from this relatively small study. First, an extended lymphadenectomy can be done safely, without increases in morbidity or mortality. Second, an extended resection does not seem to improve overall survival. Third, there is a suggestion that node-positive patients may receive some benefit from a more extensive operation. The last point is confounded, however, by the notably low 2-year survival for patients in the standard resection group (5%), significantly lower than that reported in multiple studies previously referenced in this article. Farnell and colleagues [32] quite recently reported a randomized, prospective, single-institution study from the Mayo Clinic comparing standard pancreaticoduodenectomy to pancreaticoduodenectomy with extended lymphadenectomy, focusing on the end points of survival and quality of life. All patients underwent a classic, non–pylorus-preserving, pancreaticoduodenectomy (which included a distal gastrectomy). Their standard group included 40 patients whose resections included removal of pyloric and gastric lymph nodes, nodes to the right of the hepatoduodenal ligament, anterior and posterior pancreaticoduodenal nodes, nodes to the right of the superior mesenteric artery, and nodes anterior to the common hepatic artery. An additional 39 patients underwent the same resection with the additional removal of retroperitoneal soft tissue from the hilum of the kidneys bilaterally and the celiac axis superiorly to the inferior mesenteric artery inferiorly, as well as circumferential dissection of the hepatic artery, skeletonization of the hepatoduodenal ligament, and circumferential dissection of the superior mesenteric artery from the first jejunal branch inferiorly to its origin from the aorta

162

KENNEDY & YEO

superiorly. Para-aortic lymph nodes were harvested from the celiac axis superiorly to the inferior mesenteric artery inferiorly. Patients undergoing standard resection had a median of 15 lymph nodes removed, while those undergoing the extended lymphadenectomy had 34 lymph nodes removed. Patients were offered standard adjuvant chemoradiation following resection. Demographics were similar between the two groups as were perioperative morbidity and mortality. Overall actuarial survival estimates were not different between the groups with 1-, 3-, and 5-year estimates of 82%, 41%, and 16% respectively for the standard resection group, and 71%, 25%, and 17% for the extended lymphadenectomy group. Survival analysis by lymph node status revealed no difference between the groups for either node-positive or node-negative patients. Quality of life data were also collected and analyzed by the authors. They found that the extended lymphadenectomy was associated with a significantly worse quality of life when compared with the standard resection. Specifically, patients undergoing the extended lymphadenectomy had significant difficulty with bowel control and diarrhea. Given the lack of a survival benefit and the worsening of quality of life, the authors chose to end their study after the interim analysis and report their results. Yet another study by Nimura and colleagues [33] was designed as a prospective, randomized, multicenter trial of 101 patients in Japan with pancreatic adenocarcinoma. Fifty-one patients were randomized to undergo a standard Japanese radical pancreaticoduodenectomy with regional lymph node dissection. Specimens yielded an average of 13 total lymph nodes. The other 51 patients underwent a standard Japanese radical pancreaticoduodenectomy plus an extended lymph node dissection in the para-aortic region. This procedure yielded an average of 40 total lymph nodes per specimen. Both groups were matched with respect to age, sex, and body mass index. In addition to increased lymph node retrieval, the extended procedure resulted in increased mean operating time (426 minutes for the standard regional dissection versus 547 minutes for the extended dissection) and intraoperative blood loss (1118 mL for the standard regional dissection versus 1680 for the extended dissection). There were no differences in perioperative morbidity excluding diarrhea, postoperative length of hospital stay, perioperative mortality, or histological stage of resected disease between the groups. Importantly, there were no differences in overall survival with a 3-year survival rate of 32.4% for the standard regional dissection group versus 16.0% for the extended dissection group. There was, however, a dramatic increase in the long-term complication of diarrhea in the extended lymph node dissection group (48% incidence versus 0% in the standard dissection group). Another aspect of this issue was examined by Seiler and colleagues [34] in a study that specifically looked at the issue of pylorus preservation, independent of extent of lymph node dissection. Critics of pylorus preservation have often implied that there is an oncologic disadvantage to leaving the distal stomach in situ, with its accompanying nodal stations. In a prospective, randomized trial, 110 patients with either pancreatic or periampullary cancer

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

163

were analyzed, comparing pylorus-preserving to classic (with distal gastrectomy) pancreaticoduodenectomy. The pylorus-preserving group had shorter operating times (382 minutes versus 449 minutes) and lower blood loss (1198 mL versus 1500 mL), as compared with the classic resection group. At a median follow-up for surviving patients of 63.1 months, there was no difference in long-term survival between the groups, with a median survival of 28 months after a curative (R0) classic resection versus 26 months for the pylorus-preserving group. Morbidity was similar between both groups as was quality of life.

The Johns Hopkins prospective, randomized study Because review of the available data from both observational and prospective studies failed to generate a clear consensus, we embarked on a single-institution, prospective, randomized trial to evaluate the impact of standard versus radical (or extended) pancreaticoduodenectomy with regard to postoperative morbidity, mortality, survival, and quality of life. Between April 1996 and June 2001 the study was open for accrual. Specific exclusion criteria included preoperative chemotherapy or chemoradiation, pathology revealing a process other than adenocarcinoma, or the presence of gross tumor left behind at the end of the standard pylorus-preserving pancreaticoduodenal resection. During this time period, 983 patients underwent pancreaticoduodenectomy, with 672 of these resections being for periampullary adenocarcinoma. Ultimately, 299 patients were enrolled in the study [35–37]. After obtaining informed consent, patients first underwent a standard, margin-negative resection. Pylorus preservation was preferred and the lymph node groups harvested included the anterior pancreaticoduodenal lymph nodes (station 17), the posterior pancreaticoduodenal lymph nodes (station 13), nodes in the lower hepatoduodenal ligament (stations 12b2 and 12c), and nodes along the right lateral aspect of the superior mesenteric artery and vein (some station 14b and 14v). The standard pylorus-preserving resection involved division of the duodenum 2 to 3 cm distal to the pylorus; removal of the gallbladder and common bile duct from the level of the cystic duct insertion; removal of the duodenum (beyond its proximal 2 to 3 cm) and the first 10 to 20 cm of jejunum beyond the ligament of Treitz; and resection of the head, neck, and uncinate process of the pancreas. The pancreas was divided ventral to the superior mesenteric vein–portal vein axis. In cases with an inadequate duodenal margin secondary to tumor proximity or duodenal ischemia, a distal gastrectomy removing between 10% and 40% of the stomach was performed. All patients had the uncinate process of the pancreas removed from underneath the superior mesenteric vein flush with the superior mesenteric artery. All lymphatic and neural tissues were cleared from the ventral and

164

KENNEDY & YEO

right lateral aspects of the superior mesenteric artery for a circumferential dissection between 90 and 180 . Once a margin-negative resection was confirmed, patients were randomized intraoperatively using a computer-generated random number pattern. Those patients randomized to the radical resection (Figs. 2 and 3) had their procedures extended, to include a 30% to 40% distal gastrectomy and a retroperitoneal lymph node dissection. Lymph node stations 5 and 6 as well as portions of 3 and 4 were included in the distal gastrectomy specimen, along with the portion of the greater omentum along the right gastroepiploic artery and the portion of the lesser omentum along the right gastric artery. Retroperitoneal tissue was dissected from the right renal hilum to the left lateral boarder of the aorta and from the portal vein inferiorly to a point below the third portion of the duodenum. Lymph node stations 16a2 and 16b1 were included in this dissection. A celiac lymph node (station 9) was also sampled. Partial pancreatic resection was favored and nearly all reconstructions were performed using a single retrocolic jejunal limb with a proximal pancreaticojejunostomy, downstream hepaticojejunostomy, and even farther downstream duodeno- or gastrojejunostomy. Vagotomy, tube gastrostomy, and feeding jejunostomy were not performed.

Fig. 2. Illustration depicting the components of the radical procedure. At the left is the 30% to 40% distal gastrectomy specimen, which includes the pylorus and 1- to 2-cm cuff of the duodenum. At the right is shown the retained stomach, the pancreatic body and tail, and an overview of the retroperitoneal dissection. Titanium clips have been placed to mark the extent of the retroperitoneal dissection. A celiac node is removed for histological analysis. (From Yeo CJ, Cameron JL, Sohn TA, et al. Pancreaticoduodenectomy with or without extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma: comparison of morbidity and mortality and short-term outcome. Ann Surg 1999;229(5):613–22; with permission. Copyright 1999, Lippincott Williams & Wilkins.)

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

165

Fig. 3. Illustration detailing the retroperitoneal dissection component of the radical procedure. The retroperitoneum is dissected from the hilum of the right kidney to the left lateral border of the aorta (AO) in the horizontal axis, exposing the left renal vein. In the vertical axis, the dissection extends from the level of the portal vein to below the level of the third portion of the duodenum (level of the inferior mesenteric artery [IMA] origin). Depicted here the gastric staple line and pancreatic remnant are being retracted toward the upper right. The inferior vena cava (IVC) and aorta are fully exposed, and the right gonadal vein has been preserved. A curved vascular clamp gently occludes the inferior aspect of the bile duct. The retroperitoneal fat and lymph nodes are being resected en bloc (bottom right). (From Yeo CJ, Cameron JL, Sohn TA, et al. Pancreaticoduodenectomy with or without extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma: comparison of morbidity and mortality and short-term outcome. Ann Surg 1999;229(5):613–22; with permission. Copyright 1999, Lippincott Williams & Wilkins.)

All patients were managed postoperatively using a standard critical pathway. Patients were evaluated during the immediate postoperative period by medical and radiation oncologists. Recommendations concerning adjuvant chemoradiation [12,38,39] and immunotherapy [40,41] were provided. Approximately 78% of patients in each group received postoperative chemoradiation via various regimens. A small number (2.7%) also received immunotherapy. Multiple end points were evaluated. Primary end points included postoperative survival, perioperative complications, and length of postoperative

166

KENNEDY & YEO

hospital stay. A secondary end point, postoperative quality of life, was assessed using a subgroup of the pancreaticoduodenectomy survivors. Results of primary end points Demographics and intraoperative course A total of 299 patients were initially randomized intraoperatively to either standard resection or radical pancreaticoduodenectomy (Table 2). Subsequently, five patients were excluded because the final pathology did not confirm a diagnosis of invasive adenocarcinoma. As to demographics, the groups were not significantly different with respect to age (mean 65.7 years) and sex (54% male). Distribution by race was different despite the randomization, with significantly more Caucasian patients in the radical group. Intraoperative factors such as extent of pancreatic resection, type of pancreatic anastomosis, intraoperative blood loss, and units of red blood cells transfused were comparable. As would be predicted, mean operative time was significantly longer in the radical group at 6.4 hours versus 5.9 hours for the standard procedure. Table 2 Patient demographics and intraoperative factors

Demographics Mean age, y % male % Caucasian Intraoperative factors Type of resection, n (%) Pylorus-preserving Classic (distal gastrectomy) Extent of pancreatic resection, n (%) Partial Total Type of pancreatic anastomosis Pancreaticojejunostomy, n (%) Pancreaticogastrostomy, n (%) None Vein resection, n (%) Median intraoperative blood loss Median transfusions, units PRBCs Median operative time, hr

Standard n ¼ 146

Radical n ¼ 148

66.2  0.9 58 89

65.2  0.9 51 95

.46 .19 .02

125 (86) 21 (14)

0 (0) 148 (100)

!.001 d

141 (97) 5 (3)

145 (98) 3 (2)

.46 d

140 (99) 1 (1) 5 4 (3) 600 0 5.5

141 (97) 4 (3) 3 4 (3) 700 0 6.2

.42 d d .98 .30 .96 .002

P value

Abbreviation: PRBC, packed red blood cells. Adapted from Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, Part 2. Randomized controlled trial evaluating survival, morbidity and mortality. Ann Surg 2002;236(3):355–66; with permission. Copyright 2002, Lippincott Williams & Wilkins.

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

167

Pathology The pathology results of the resected specimens were largely comparable between the groups (Tables 3 and 4). There were no differences in site of tumor origin (57% pancreatic), differentiation (42% poorly differentiated), tumor diameter (2.6 cm), resected lymph node status (74% positive), margin status (10% microscopically positive), perineural invasion, and vascular invasion. Detailed analysis of the resected lymph nodes (see Table 4) revealed that 73% of the patients undergoing a standard resection had histologically positive lymph nodes, with a mean number of 17 nodes resected. Similarly, 74% of the patients who underwent a radical resection had histologically positive nodes. The mean number of resected nodes in the radical specimens was significantly higher (28.5 nodes) when compared with the standard group. In only one patient was there a positive node in the extended lymphadenectomy specimen without positive nodes in the portion of the specimen encompassed by a standard resection. Thus, only one patient had staging altered by the radical resection, from T3N0M0 (stage II) to T3N1M0 (stage III). Postoperative complications and hospital course Of the 294 patients with periampullary adenocarcinoma enrolled in the study, 9 died in the immediate postoperative period (Table 5). There were six deaths in the standard resection group (4%) and three in the radical group (2%) for an overall perioperative mortality rate of 3% (P ¼ NS). The overall complication rate was 36%. The complication rate in the standard group (29%) was significantly lower than that in the radical group (43%) (P ¼ .01). The most common complications were early delayed Table 3 Pathology of patients with pancreatic cancer

Mean tumor diameter, cm Tumor differentiation, % poor Resected lymph node status, % positive Resection margin status,a % positive Perineural invasion, % Vascular invasion, %

Standard (n ¼ 84)

Radical (n ¼ 83)

P value

3.0  0.1 40 82 20 84 54

2.8  0.1 51 77 5 93 55

.26 .29 .42 .003 .09 .88

a All positive margins were microscopically positive on permanent section (R1). No patient was randomized with microscopically positive margins by frozen section or grossly positive margins (R2). Adapted from Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, Part 2. Randomized controlled trial evaluating survival, morbidity and mortality. Ann Surg 2002;236(3):355–66; with permission. Copyright 2002, Lippincott Williams & Wilkins.

168

KENNEDY & YEO

Table 4 Details of resected lymph node analyses

Resected lymph node status Total lymph nodes resected Nodes in pancreaticoduodenectomy specimen

Nodes in retroperitoneal lymphadenectomy specimen

Nodes in distal gastrectomy specimen

Nodes in celiac region

Standard (n ¼ 146)

Radical (n ¼ 148)

Positive

107 (73%)

110 (74%)

.76

Negative Mean

39 (27%) 17.0  0.6

38 (26%) 28.5  0.6

d .001

Median Positive

16 107 (73%)

26 109 (74%)

d 0.83

Negative Mean number LN Mean number positive LN Positive

39 (27%) 17.0  0.6 3.3  0.3

39 (26%) 16.4  0.6 3.0  0.3

d d d

N/A

22 (15%)a

d

126 (85%) 7.0  0.3 0.4  0.1

d d d

8 (5%)b

d

140 (95%) 4.1  0.4 0.1  0.02

d d d

5 (3%)c 143 (97%) 0.8  0.1 0.04  0.02

d d d d

Negative Mean number LN Mean number positive LN Positive

Negative Mean number LN Mean number positive LN Positive Negative Mean number LN Mean number positive LN

N/A

N/A

P value

Abbreviations: LN, lymph nodes; N/A, not applicable (not resected in the standard group). a In 22 patients there were positive LN in the retroperitoneal lymphadenectomy specimen. In all 22, LN were also positive in the pancreaticoduodenectomy specimen. b In eight patients there were positive LN in the distal gastrectomy specimen. In all but one of these patients, LN were also positive in the pancreaticoduodenectomy specimen. One patient in the radical group had negative pancreaticoduodenectomy specimen LN and one positive lymph node in the distal gastrectomy specimen. c In five patients there were positive LN in the celiac region. In all five patients, LN were also positive in the pancreaticoduodenectomy specimen. From Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, Part 2. Randomized controlled trial evaluating survival, morbidity and mortality. Ann Surg 2002;236(3):355–66; with permission. Copyright 2002, Lippincott Williams & Wilkins.

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

169

Table 5 Postoperative complications and hospital course

Perioperative mortality, n (%) Reoperation, n (%) Any complication, n (%) Delayed gastric emptying, n (%) Pancreatic fistula, n (%) Wound infection, n (%) Intra-abdominal abscess, n (%) Bile leak, n (%) Cholangitis, n (%) Lymphocele, n (%) Median postoperative length of stay, d

Standard (n ¼ 146)

Radical (n ¼ 148)

P value

6 6 42 9 9 7 5 3 2 1 9

3 6 64 24 19 16 6 7 3 4 10

.30 .98 .01 .006 .05 .06 .77 .21 .66 .57 .003

(4) (4) (29) (6) (6) (5) (3) (2) (1) (1)

(2) (4) (43) (16) (13) (11) (4) (5) (2) (3)

From Riall TS, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinomadpart 3: update on 5-year survival. J Gastrointest Surg 2005;9:1191–204; with permission.

gastric emptying, pancreatic fistula, and wound infection. The length of postoperative hospital stay was also related to the type of operation, with standard patients staying significantly fewer days (11.3 days) than patients in the radical group (14.3 days) (P ¼ .003).

Survival analysis at 5-year follow-up The actuarial survival curves for the 280 patients available for survival analysis are depicted in Figs. 4 and 5. For this survival analysis, the mean length of follow-up for the entire cohort was 35.7 months, with 77 patients alive at the time of analysis. The mean live-patient follow-up was 68.6 months, with a median of 64 months. The 1-, 3-, and 5-year survival data for the entire cohort are summarized in Table 6. There were no differences between the two groups with regard to survival at any of the intervals, with a median survival of 25 months in the standard group and 28 months in the radical group (P ¼ .57). When the groups are divided into lymph node– positive and lymph node–negative subgroups, there remains no difference between the two groups. Further subgroup analysis of the 162 patients with a diagnosis of pancreatic adenocarcinoma again reveals no differences in survival between the standard and radical resection groups. The standard resection group had 1-, 3-, and 5-year actuarial survival rates of 75%, 34%, and 13% (median 20 months), compared with 73%, 38%, and 29% (median 22 months) for patients undergoing radical resection (P ¼ .13). The lack of statistical significance is even more telling in light of the fact that patients in the standard resection group had an anomalously high incidence of microscopically

170

KENNEDY & YEO

Fig. 4. The actuarial survival curves for all patients (entire cohort) who survived the immediate postoperative period, comparing the standard resection group (n ¼ 136, solid line) to the radical group (n ¼ 144, dashed line). The 1-, 3-, and 5-year survival rates are 78%, 42%, and 25% for the standard group and 76%, 44%, and 31% for the radical group (P ¼ .57). (From Riall TS, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinomadpart 3: update on 5-year survival. J Gastrointest Surg 2005;9:1191–204; with permission.)

positive margins (R1) on resection (21% versus 5% in the radical group; P ¼ .002). Again, subgroup analysis of lymph node–positive patients and lymph node–negative patients did not reveal any significant difference between standard and radical resection.

Secondary end points Before this study, there had been only one large study [42] and several small studies [43–47] that looked at health-related quality of life (QOL) in postoperative pancreaticoduodenectomy patients. The earlier studies indicated that pancreaticoduodenectomy survivors that were without evidence of recurrence had a near normal QOL. Patients who had been enrolled in the Johns Hopkins randomized study of standard versus radical pancreaticoduodenectomy were further assessed as to health-related QOL [48]. A standard Functional Assessment of Cancer Therapy-Hepatobiliary (FACT-Hep) QOL survey designed for hepatobiliary cancer was sent to 150 surviving patients who underwent pancreaticoduodenectomy. A total

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

171

Fig. 5. The actuarial survival curves for all patients with pancreatic cancer who survived the immediate postoperative period, comparing the standard resection group (n ¼ 80, solid line) to the radical group (n ¼ 82, dashed line). The 1-, 3-, and 5-year survival rates are 75%, 34%, and 13% for the standard group and 73%, 38%, and 29% for the radical group (P ¼ .13). (From Riall TS, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinomadpart 3: update on 5-year survival. J Gastrointest Surg 2005;9:1191–204; with permission.)

of 105 QOL surveys were returned and analyzed, with 55 of the respondents originally belonging to the standard group and 50 to the radical group. The patients were evaluated at a mean of 2.2 years after pancreaticoduodenectomy. QOL and functional status were assessed via a series of subscale scores for physical, social, emotional, and functional well-being. The two groups were statistically similar with regard to parameters such as age, race, intraoperative blood transfusions, pathology and staging, and perioperative complications. Patients who had undergone the radical operation did have a longer postoperative hospital stay, as was noted during analysis of the entire study cohort. Analysis of the data revealed no difference in any of the measured parameters between those patients undergoing a standard resection and those who underwent a radical resection. When certain postoperative parameters were looked at closely (such as pancreatic enzyme use and number of bowel motions per day), there was clearly a significant increase compared with preoperative baselines; however, there was no difference between the two study groups in the degree of increase. In summary, this careful QOL analysis in a subgroup of the Johns Hopkins patients found no differences

172

KENNEDY & YEO

Table 6 Survival rates following standard versus radical pancreaticoduodenectomy 1-year, %

3-year, %

Entire cohort (all pathologic diagnoses) Standard (n ¼ 136) 78 42 Radical (n ¼ 144) 76 44 Node positive patients (all pathologic diagnoses) Standard (n ¼ 99) 76 31 Radical (n ¼ 108) 76 39 Node negative patients (all pathologic diagnoses) Standard (n ¼ 37) 84 73 Radical (n ¼ 36) 94 61 Pancreatic cancer Standard (n ¼ 80) 75 34 Radical (n ¼ 82) 73 38 Pancreatic cancer, node positive Standard (n ¼ 66) 73 26 Radical (n ¼ 64) 69 33 Pancreatic cancer, node negative Standard (n ¼ 14) 86 71 Radical (n ¼ 18) 89 56 Ampullary cancer Standard (n ¼ 32) 81 59 Radical (n ¼ 27) 81 52 Distal bile duct cancer Standard (n ¼ 21) 81 38 Radical (n ¼ 27) 78 44

5-year, %

Median, mo

P value

25 31

25 28

.57 d

15 28

19 20

.23 d

50 41

59 45

.51 d

13 29

20 22

.13 d

10 24

19 18

.36 d

29 44

43 44

.21 d

46 43

48 40

.67 d

28 15

23 25

.51 d

Excludes perioperative mortalities and patients lost to follow-up. Mean live patients followup ¼ 68.6 months. From Riall TS, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinomadpart 3: update on 5-year survival. J Gastrointest Surg 2005;9:1191–204; with permission.

in health-related QOL between the standard and the radical resection groups.

Discussion Conflicting reports initially created significant discussion about the role of extended versus radical pancreaticoduodenectomy in patients with periampullary adenocarcinoma. The large Johns Hopkins trial discussed at some length in this review was designed to provide a definitive answer to this question. As previously noted, radical resection did not yield any survival benefit, even when subgroups of patients with pancreatic cancer were stratified by lymph node status. Radical resection did have a cost however. Patients undergoing radical resection had higher rates of delayed gastric emptying (16% versus 6% for standard), pancreatic fistula (13% versus

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

173

6% for standard), and wound infection (11% versus 5% for standard). This resulted in an increase in the length of postoperative hospital stay by more than 3 days in the radical group. It should not be surprising that a radical lymphadenectomy failed to provide a survival advantage, as this result mirrors data and approaches being established in other malignancies. For many years now, breast cancer has been moving away from the radical mastectomy as a mainstay of surgical therapy to progressively more limited interventions. Even in the disease process (breast cancer) where Halsted’s concept that a wider resection would result in a greater cure was introduced, the clinical advances have been toward limited surgery and more reliance on nonsurgical adjuvant therapy. In other disease processes such as gastric cancer, most of the data including randomized prospective trials have failed to show a survival advantage for extended lymphadenectomies. Another approach to the question has been taken by Pisters and colleagues [49] by means of a mathematical model. The model predicts a lack of a survival benefit for extended resections in pancreatic adenocarcinoma based on three assumptions: complete gross and microscopic tumor resection (R0) is required for extended resection to confer a survival benefit; only patients with second-order lymph node involvement will benefit when extended resections remove these nodes; and only node-positive patients without metastatic disease outside the resection zone will derive a benefit from extended resections. If one assumes an R0 resection rate of 80%, a second-order lymph node involvement in pancreatic cancer of 10%, and a node-positive but distant metastatic disease negative rate of 5%, the authors calculated a theoretical survival benefit of only 0.4% (¼ 0.8  0.1  0.05  100) for extended resection (meaning that only 4 per 1000 patients would theoretically benefit). It is admitted that the author’s starting assumptions and percentages could be challenged, but their argument is generally sound and minor adjustments to their percentages would not yield much difference in the end calculation. In the final analysis, the failure to impact overall survival by increasing the extent of surgical resection may be best explained by the evolving understanding of cancer as a disease process. We currently recommend a pylorus-preserving pancreaticoduodenectomy with standard (en bloc) lymphadenectomy as the resection of choice for patients with periampullary adenocarcinoma. Pancreatic and periampullary cancer, along with many other cancers, may be best thought of as a systemic disease for most, if not all patients, even at the time of diagnosis of apparently resectable disease. Thus, improvements in local therapy for this systemic process might be expected to yield only small benefits. It has been clearly demonstrated that local resectional therapy for pancreatic cancer can be performed safely, with measurable but often disappointing efficacy. The next leap forward in this disease process awaits the results of the ongoing research in the areas of chemotherapy, immunotherapy, and early detection. Significant improvements in long-term survival await these coming advances.

174

KENNEDY & YEO

References [1] U.S. Cancer Statistics Working Group. United States cancer statistics: 1999–2002 incidence and mortality web-based report. Available at: www.cdc.gov/cancer/npcr/uscs. Accessed February 2, 2006. [2] Geer RJ, Brennan MF. Prognostic indicators for survival after resection of pancreatic adenocarcinoma. Am J Surg 1993;165:68–72 [discussion: 72–3]. [3] Conlon KC, Klimstra DS, Brennan MF. Long-term survival after curative resection for pancreatic ductal adenocarcinoma. Clinicopathologic analysis of 5-year survivors. Ann Surg 1996;223:273–9. [4] Sohn TA, Yeo CJ, Cameron JL, et al. Resected adenocarcinoma of the pancreasd616 patients: results, outcomes, and prognostic indicators. J Gastrointest Surg 2000;4:567–79. [5] Yeo CJ, Sohn TA, Cameron JL, et al. Periampullary adenocarcinoma: analysis of 5-year survivors. Ann Surg 1998;227:821–31. [6] Yeo CJ, Cameron JL. Prognostic factors in ductal pancreatic cancer. Langenbecks Arch Surg 1998;383:129–33. [7] Delcore R, Rodriguez FJ, Forster J, et al. Significance of lymph node metastases in patients with pancreatic cancer undergoing curative resection. Am J Surg 1996;172:463–8 [discussion: 468–9]. [8] Nakeeb A, Pitt HA, Sohn TA, et al. Cholangiocarcinoma. A spectrum of intrahepatic, perihilar, and distal tumors. Ann Surg 1996;224:463–73 [discussion: 473–5]. [9] Allison DC, Piantadosi S, Hruban RH, et al. DNA content and other factors associated with ten-year survival after resection of pancreatic carcinoma. J Surg Oncol 1998;67: 151–9. [10] Tascilar M, Skinner HG, Rosty C, et al. The SMAD4 protein and prognosis of pancreatic ductal adenocarcinoma. Clin Cancer Res 2001;7:4115–21. [11] Montgomery RC, Hoffman JP, Riley LB, et al. Prediction of recurrence and survival by postresection CA 19-9 values in patients with adenocarcinoma of the pancreas. Ann Surg Oncol 1997;4:551–6. [12] Yeo CJ, Abrams RA, Grochow LB, et al. Pancreaticoduodenectomy for pancreatic adenocarcinoma: postoperative adjuvant chemoradiation improves survival. A prospective, singleinstitution experience. Ann Surg 1997;225:621–33 [discussion: 633–6]. [13] Neoptolemos JP, Dunn JA, Stocken DD, et al. Adjuvant chemoradiotherapy and chemotherapy in resectable pancreatic cancer: a randomised controlled trial. Lancet 2001;358: 1576–85. [14] Nukui Y, Picozzi VJ, Traverso LW. Interferon-based adjuvant chemoradiation therapy improves survival after pancreaticoduodenectomy for pancreatic adenocarcinoma. Am J Surg 2000;179:367–71. [15] Klinkenbijl JH, Jeekel J, Sahmoud T, et al. Adjuvant radiotherapy and 5-fluorouracil after curative resection of cancer of the pancreas and periampullary region: phase III trial of the EORTC gastrointestinal tract cancer cooperative group. Ann Surg 1999;230:776–82 [discussion: 782–4]. [16] Halsted WS. Contribution to the surgery of the bile passages, especially of the common bileduct. Boston Medical and Surgical Journal 1899;141:645–54. [17] Whipple AO. Observations on radical surgery for lesions of the pancreas. Surg Gyn Obst 1946;82:62. [18] Howard JM. Pancreatico-duodenectomy: forty-one consecutive Whipple resections without an operative mortality. Ann Surg 1968;168:629–40. [19] Fortner JG. Regional resection of cancer of the pancreas: a new surgical approach. Surgery 1973;73:307–20. [20] Fortner JG. Recent advances in pancreatic cancer. Surg Clin North Am 1974;54:859–63. [21] Fortner JG, Kim DK, Cubilla A, et al. Regional pancreatectomy: en bloc pancreatic, portal vein and lymph node resection. Ann Surg 1977;186:42–50.

LYMPHADENECTOMY FOR PERIAMPULLARY ADENOCARCINOMA

175

[22] Fortner JG, Klimstra DS, Senie RT, et al. Tumor size is the primary prognosticator for pancreatic cancer after regional pancreatectomy. Ann Surg 1996;223:147–53. [23] Cubilla AL, Fortner J, Fitzgerald PJ. Lymph node involvement in carcinoma of the head of the pancreas area. Cancer 1978;41:880–7. [24] Kayahara M, Nagakawa T, Ueno K, et al. Surgical strategy for carcinoma of the pancreas head area based on clinicopathologic analysis of nodal involvement and plexus invasion. Surgery 1995;117:616–23. [25] Ishikawa O, Ohigashi H, Sasaki Y, et al. Practical grouping of positive lymph nodes in pancreatic head cancer treated by an extended pancreatectomy. Surgery 1997;121: 244–9. [26] Satake K, Nishiwaki H, Yokomatsu H, et al. Surgical curability and prognosis for standard versus extended resection for T1 carcinoma of the pancreas. Surg Gynecol Obstet 1992;175: 259–65. [27] Manabe T, Ohshio G, Baba N, et al. Radical pancreatectomy for ductal cell carcinoma of the head of the pancreas. Cancer 1989;64:1132–7. [28] McFadden DW, Reber HA. Cancer of the pancreas: radical resection–supporting view. Adv Surg 1994;27:257–72. [29] Pancreas Club Inc. Extended lymphadenectomy during pancreaticoduodenectomy for cancer of the pancreas: summary of ‘‘how I do it’’ session. J Gastrointest Surg 2000;4: 225–32. [30] Henne-Bruns D, Vogel I, Luttges J, et al. Surgery for ductal adenocarcinoma of the pancreatic head: staging, complications, and survival after regional versus extended lymphadenectomy. World J Surg 2000;24:595–601 [discussion: 601–2]. [31] Pedrazzoli S, DiCarlo V, Dionigi R, et al. Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy study group. Ann Surg 1998;228:508–17. [32] Farnell MB, Pearson RK, Sarr MG, et al. A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma. Surgery 2005;138:618–28 [discussion: 628–30]. [33] Nimura Y, Nagino M, Kato H, et al. Regional versus extended lymph node dissection in radical pancreaticoduodenectomy for pancreatic cancer: a multicenter, randomized controlled trial. HPB Surg 2004;6(Suppl 1):2. [34] Seiler CA, Wagner M, Bachmann T, et al. Randomized clinical trial of pylorus-preserving duodenopancreatectomy versus classical Whipple resection-long term results. Br J Surg 2005;92:547–56. [35] Yeo CJ, Cameron JL, Sohn TA, et al. Pancreaticoduodenectomy with or without extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma: comparison of morbidity and mortality and short-term outcome. Ann Surg 1999;229:613–22 [discussion: 622–4]. [36] Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, part 2: randomized controlled trial evaluating survival, morbidity, and mortality. Ann Surg 2002;236:355–66 [discussion: 366–8]. [37] Riall TS, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma–part 3: update on 5-year survival. J Gastrointest Surg 2005;9:1191–204 [discussion: 1204–6]. [38] Abrams RA, Grochow LB, Chakravarthy A, et al. Intensified adjuvant therapy for pancreatic and periampullary adenocarcinoma: survival results and observations regarding patterns of failure, radiotherapy dose and CA19-9 levels. Int J Radiat Oncol Biol Phys 1999; 44:1039–46.

176

KENNEDY & YEO

[39] Chakravarthy A, Abrams RA, Yeo CJ, et al. Intensified adjuvant combined modality therapy for resected periampullary adenocarcinoma: acceptable toxicity and suggestion of improved 1-year disease-free survival. Int J Radiat Oncol Biol Phys 2000;48:1089–96. [40] Jaffee EM, Abrams R, Cameron J, et al. A phase I clinical trial of lethally irradiated allogeneic pancreatic tumor cells transfected with the GM-CSF gene for the treatment of pancreatic adenocarcinoma. Hum Gene Ther 1998;9:1951–71. [41] Jaffee EM, Hruban RH, Biedrzycki B, et al. Novel allogeneic granulocyte-macrophage colony-stimulating factor-secreting tumor vaccine for pancreatic cancer: a phase I trial of safety and immune activation. J Clin Oncol 2001;19:145–56. [42] Huang JJ, Yeo CJ, Sohn TA, et al. Quality of life and outcomes after pancreaticoduodenectomy. Ann Surg 2000;231:890–8. [43] Wenger FA, Jacobi CA, Haubold K, et al. Gastrointestinal quality of life after duodenopancreatectomy in pancreatic carcinoma. Preliminary results of a prospective randomized study: pancreatoduodenectomy or pylorus-preserving pancreatoduodenectomy. Chirurg 1999;70: 1454–9. [44] Kokoska ER, Stapleton DR, Virgo KS, et al. Quality of life measurements do not support palliative pancreatic cancer treatments. Int J Oncol 1998;13:1323–9. [45] Melvin WS, Buekers KS, Muscarella P, et al. Outcome analysis of long-term survivors following pancreaticoduodenectomy. J Gastrointest Surg 1998;2:72–8. [46] Patel AG, Toyama MT, Kusske AM, et al. Pylorus-preserving Whipple resection for pancreatic cancer. Is it any better? Arch Surg 1995;130:838–42 [discussion: 842–3]. [47] McLeod RS, Taylor BR, O’Connor BI, et al. Quality of life, nutritional status, and gastrointestinal hormone profile following the Whipple procedure. Am J Surg 1995;169:179–85. [48] Nguyen TC, Sohn TA, Cameron JL, et al. Standard vs. radical pancreaticoduodenectomy for periampullary adenocarcinoma: a prospective, randomized trial evaluating quality of life in pancreaticoduodenectomy survivors. J Gastrointest Surg 2003;7:1–9 [discussion: 9–11]. [49] Pisters PW, Evans DB, Leung DH, et al. Re: surgery for ductal adenocarcinoma of the pancreatic head. World J Surg 2001;25:533–4.