- Email: [email protected]
Palliative Surgery for Unresectable Pancreatic and Periampullary Cancer: A Reappraisal
Palliative Surgery for Unresectable Pancreatic and Periampullary Cancer: A Reappraisal Mickael Lesurtel, M.D., Nidal Dehni, M.D., Emmanuel Tiret, M.D., Rolland Parc, M.D., Franc¸ois Paye, M.D., Ph.D.
This study aimed to reappraise short-term and long-term results of palliative biliary and gastric bypass surgery in patients with unresectable pancreatic head carcinoma found at explorative laparotomy. We retrospectively analyzed 83 consecutive patients whose pancreatic head carcinoma appeared unresectable at laparotomy (vascular involvement [57%], liver metastases [24%], distant metastatic lymph nodes [11%], peritoneal implants [8%]) and who underwent palliative surgical concomitant biliary and gastric bypass. Postoperative mortality and morbidity rates were 4.8% and 26.5%, respectively. Postoperativedelayed gastric emptying occurred in 9 patients (10%). Antecolic (46%) and retrocolic (54%) gastrojejunostomies did not differ for the duration of nasogastric suction, the delay of oral intake, and the incidence of delayed gastric emptying. Mean hospital stay was 16 6 8 days. Median survival was 9 months (range 1–44). Late cholangitis occurred in 2 patients (2.4%) treated medically. One recurrent jaundice required transhepatic stenting 9 months from surgery. Four late gastric outlet obstructions occurred (4.8%) with a mean delay of 8 months from surgery. These data demonstrate that, in patients with unresectable pancreatic head carcinoma at laparotomy, palliative concomitant biliary and gastric bypass in a single procedure is safe and long-term efficient. This strategy remains to be compared to endoscopic palliation in this setting. ( J GASTROINTEST SURG 2006;10:286–291) Ó 2006 The Society for Surgery of the Alimentary Tract KEY
WORDS:
Pancreatic carcinoma, palliative care, gastric bypass, biliary bypass
Only 5–20% of pancreatic head carcinoma are resectable at the time of presentation.1–3 For patients with obvious nonresectable disease, endoscopic techniques have been developed as alternatives to traditional surgical management. Biliary stenting, and more recently duodenal self-expandable endoprostheses, have been promoted as the treatment of choice because of their low morbidity.4–9 However, despite improvement in imaging procedures, assessing unresectability still remains difficult in some cases, and purely nonsurgical palliation may, in these cases, overlook resectable tumors. Furthermore, pancreatic biopsies performed under radiologic or endoscopic ultrasound guidance, which have to be obtained before starting a palliative treatment by chemotherapy or radiotherapy, fail to prove the diagnosis of adenocarcinoma in 10–20% of attempted procedures. In these questionable cases where unresectability of the tumor is not proven, our policy today remains in favor of a surgical approach where
palliative surgical bypass is performed if the tumor appears unresectable at laparotomy. Historical series reported high morbidity (up to 50%) and mortality (up to 30%) rates of bypass surgery,10,11 but more recent series reported reduced morbidity and mortality rates of less than 30% and 10%, respectively.3,12–14 Whereas endoscopic palliation is widely used today, this study aimed to reappraise the short-term and long-term results of a surgical palliation policy combining biliary-enteric bypass and gastrojejunostomy in patients with unresectable pancreatic head carcinoma at laparotomy.
MATERIAL AND METHODS Patients Between June 1996 and December 2003, 340 patients without obvious contraindication to resection detected by preoperative imaging assessment
From the Department of Digestive Surgery, Saint Antoine University-Hospital, Paris, France. Reprint requests: Franc¸ois Paye, M.D., Ph.D., Department of Digestive Surgery, Hoˆpital Saint Antoine, 184, rue du Faubourg Saint Antoine 75012 Paris, France. e-mail: [email protected] Ó 2006 The Society for Surgery of the Alimentary Tract
286 Published by Elsevier Inc.
1091-255X/06/$dsee front matter doi:10.1016/j.gassur.2005.05.011
Vol. 10, No. 2 2006
(distant metastases, distant lymph nodes involvement, unresectable vascular involvement) underwent a surgical exploration for pancreatic head adenocarcinoma. If such contraindication was found intraoperatively, a biopsy was done to prove malignancy, and surgical palliation was performed in a single surgical procedure. Among the 340 patients, 257 (76%) underwent a pancreaticoduodenectomy. The remaining 83 patients (24%), who underwent surgical palliation for malignant pancreatic disease found unresectable at laparotomy, were retrospectively reviewed in this study. There were 38 women and 45 men with a mean age of 64 6 11 years. The following preoperative symptoms were recorded at first presentation: jaundice (77%), abdominal pain (54%), loss of weight more than 10% (25%), and vomiting or nausea (13%). Mean delay between the onset of symptoms and operation was 9 6 9 weeks. Resectability was routinely assessed preoperatively by chest X-ray, abdominal ultrasonography, and abdominal-computed tomography (helicoidal CT scan was used from 1998 to 2003). Echoendoscopy was not routinely performed. For 62% of the patients, the resectability of the tumor was preoperatively judged as doubtful. For the others (38%), the preoperative assessment judged the tumor as resectable. Histologic confirmation of the diagnosis of adenocarcinoma was obtained intraoperatively in 73 patients (88%). For the remaining patients (12%), either the intraoperative biopsies were not performed (n 5 4) or were negative (n 5 6), but proof of malignancy was obtained by follow-up and demonstrated disease progression in these 10 patients. In this specific setting, the sensitivity of biopsy was 92%.
Palliative Surgery for Pancreatic Carcinoma
287
Table 1. Indications for surgical palliation (n 5 83 patients) Reasons for unresectability
Vascular invasion* Liver metastases Distant metastatic lymph nodes Peritoneal implants
No. of patients (%)
47 20 9 7
(57) (24) (11) (8)
*Vascular invasion included arterial encasement or involvement of the superior mesenteric artery, the celiac axis, or the hepatic artery, and unresectable venous involvement of the mesenteric and portal veins.
25 patients (30%) received postoperative chemotherapy, and 23 patients (28%) postoperative chemoradiation. Outcome Postoperative mortality, postoperative morbidity, resumption of oral diet, delay of flatus passage, postoperative hospital stay, and long-term survival were assessed. Postoperative mortality was defined as death within 30 days after operation. Postoperativedelayed gastric emptying was diagnosed when a normal diet was not tolerated within 10 days of surgery.12 Follow-up Follow-up information was obtained through direct patient or referring physician contacts, from hospital charts, and by contacting the French civil status registry office. Mean length of follow-up was 9 6 9 months. Statistical Analysis
Surgical Palliative Procedure Reasons for unresectability indicating surgical palliation are shown in Table 1. The palliative procedure included, routinely, both a biliary-enteric bypass and a gastrojejunostomy. Biliary bypass was a hepaticoduodenostomy in 69 patients (83%). A Roux and Y hepaticojejunostomy was performed in 14 patients (17%) because the duodenum could not be used due to tumor volume or duodenum involvement. Gastrojejunostomy was antecolic in 46 patients (antecolic group) and retrocolic in 37 patients (retrocolic group), depending on the surgeons. Both were isoperistaltic. A chemical splanchnicectomy with alcohol was performed in 5 patients (6%) because preoperative pain was not controlled by oral analgesics. Mean duration of surgical procedures was 203 6 67 minutes. Postoperatively, 35 patients (42%) had no other therapy,
Data were presented as mean 6 standard deviation or median (range). Statistical analysis was carried out using statistical software (Statview 5.0, SAS Institute Inc., Cary, NC). Differences between groups were evaluated using chi-square analysis, Fisher’s exact test, or Student’s t-test when appropriate. Survival analysis was performed with the Kaplan-Meier method. The log-rank test was used to evaluate differences in survival between groups. A P value ! 0.05 was considered as statistically significant. RESULTS Short-term Outcome Mortality rate was 4.8%, and the overall postoperative morbidity rate was 26.5% (Table 2). Postoperative deaths were due to one hepatic failure
288
Journal of Gastrointestinal Surgery
Lesurtel et al.
Table 2. Postoperative mortality and morbidity (n 5 83 patients)
Table 3. Clinical and surgical findings in the two groups of gastrojejunostomy
No. of patients (%)
Mortality* Morbidity Delayed gastric emptying Wound infection Intra-abdominal abscess Ascites Pneumonia Urinary infection Cholangitis
4 (4.8) 22 (27) 9 3 1 4 2 2 1
*Defined as death within 30 days after operation.
in an alcoholic patient, one mesenteric arterial infarction, one gastrojejunal anastomotic leak, and one pneumonia complicated by multiorgan failure. Age of the patients did not correlate with the overall postoperative complication rate (P 5 0.47). One patient (1.2%) was reoperated to drain a wound abscess. The most common complication was the postoperative-delayed gastric emptying occurring in 9 patients (10%). Seven patients (8%) received red cell units perioperatively. Overall mean hospital stay was 16 6 8 days (median 13 days, range 8–60).
Antecolic Retrocolic group (n 5 46) group (n 5 37) P
Male/Female 25/21 Age 64 6 11 Delay first symptoms 10 6 11 operation (days) Preoperative symptoms Jaundice 33 (72) Abdominal pain 25 (54) Vomiting/Nausea 7 (15) Weight loss O10% 17 (37) Reasons for unresectability Vascular invasion 29 (63) Liver metastases 10 (22) Metastatic lymph nodes 3 (7) Peritoneal implants 4 (8) Duration of surgical 207 6 72 procedure (min) Biliary bypass Hepaticoduodenostomy 40 (87) Hepaticojejunostomy 6 (13)
20/17 62 6 12 866
0.97 0.43 0.37
31 20 6 8
(84) (54) (16) (22)
0.30 0.97 0.90 0.20
18 10 6 3 197
(49) (27) (16) (8) 6 62
0.25 0.60 0.90 0.30 0.53
29 (78) 8 (22)
0.45
Values in parentheses are percentages. Continuous data are presented as mean 6 standard deviation.
Long-term Outcome Comparison Between Antecolic and Retrocolic Gastrojejunostomies No significant difference in patient demographics, preoperative symptoms, and surgical findings was observed between the two groups, as shown in Table 3. There was no statistical difference between the two groups for postoperative length of nasogastric suction, passage of flatus, resumption of oral intake, and postoperative-delayed gastric emptying (Table 4). Delayed gastric emptying was not correlated with preoperative symptoms of gastric outlet obstruction (P 5 0.92).
Comparison Between Hepaticoduodenostomy and Hepaticojejunostomy No significant difference in patient demographics, preoperative symptoms, and short-term outcomes was observed between the two groups (data not shown). The surgical time was significantly longer in patients with a hepaticojejunostomy than with a hepaticoduodenostomy (251 6 88 minutes vs. 193 6 58 minutes; P 5 0.002).
Median postoperative survival was 9.2 months (range 1–44; Fig. 1). There was no statistical difference in survival between patients without postoperative treatment (median 7.1 months, range 1–44), patients treated by postoperative chemotherapy (median 10.6 months, range 1–35), or postoperative chemoradiation (median 10.3 months, range 1–30). Two patients (2.4%) developed cholangitis 8 months and 17 months, respectively, after initial palliative surgery and were efficiently treated by antibiotics. These two patients had undergone a hepaticoduodenostomy. One patient (1.2%) developed recurrent jaundice 9 months after initial surgery with a hepaticojejunostomy. He was treated by biliary transhepatic stenting. There was no significant difference in terms of late biliary complications between patients with hepaticoduodenostomy and patients with hepaticojejunostomy (3% vs. 7%, NS). Four patients (4.8%) had late gastric outlet obstruction with a mean delay of 8 months from the initial palliative surgery. Three of them had an antecolic gastrojejunostomy and one had a retrocolic gastrojejunostomy (NS). One of them underwent a subsequent second gastroenterostomy, whereas the others had medical support only. Thus, jaundice and gastric outlet obstruction were efficiently prevented
Vol. 10, No. 2 2006
Palliative Surgery for Pancreatic Carcinoma
Table 4. Postoperative course following antecolic and retrocolic gastrojejunostomies Antecolic group n 5 46
Retrocolic group n 5 37
Nasogastric 765 764 suction (days) Passage of 562 561 flatus (days) Oral intake (days) 763 763 Postoperative delayed 6 (13%) 3 (8%) gastric emptying Morbidity 11 (24%) 11 (30%) Mortality* 4 (8.6%) 0 (0%) Hospital stay (days) 16 6 8 16 6 9 Late outlet 3 (7 6 3 months) 1 (10 months) obstruction
antecolic and retrocolic gastrojejunostomy yield similar morbidity, including postoperative delayed gastric emptying. Finally, the low rates of late cholangitis (1.2%), recurrent jaundice (2.4%), and late gastric outlet obstruction (4.8%) confirm the long-term efficacy of the palliative procedure. Procedure-related morbidity and mortality rates in our study (26% and 4.8%, respectively) compare favorably with those in the three prospective randomized studies comparing biliary bypass surgery to endoscopic biliary stenting; surgical morbidity rates were reported ranging from 26% to 40% and high mortality rates from 15% to 31%.15–17 It is noteworthy that significantly lower morbidity (11% vs. 29%; P 5 0.02) and lower procedure-related mortality rates (3% vs. 14%; P 5 0.006) in patients treated endoscopically were only demonstrated in Smith’s study.17 Mortality did not differ between patients with endoscopic or surgical bypass in the meta-analysis of these trials by Taylor et al.18 Patients included in these studies had a clear contraindication to resection that was preoperatively established, whereas these patients were excluded from our study. Our improved results may be due to improvement of preoperative imaging assessment and the use of endoscopic palliative procedures for patients whose preoperative workup detects a clear contraindication to surgical resection. Thus, the subgroup of patients who finally undergo surgery with a curative intent may have here a less extended disease than in these older studies. The resulting better general conditions may contribute to the better results of palliative surgery observed in our series. More recently, other groups reported reduced
P
0.96 0.64 0.52 0.71 0.17 0.12 0.90 0.76
Continuous data are presented as mean 6 standard deviation. *Was defined as death within 30 days after operation.
by this surgical palliative procedure, until death, in 78 patients (94%). DISCUSSION Despite improvements in preoperative diagnosis and staging, pancreatic head carcinoma remains frequently found to be unresectable at the time of the laparotomy, leading to an intraoperative shift from an initial curative intent toward a palliative procedure. This study confirms, first, that both biliary and gastric bypasses can be performed in this setting, with low morbidity (26.5%) and low mortality (4.8%) rates. Second, it shows that 1
Survival rate
.8
.6
.4
.2
0 0
5
10
15
20
25
30
35
40
45
3
1
1
0
Months No. at risk 83
51
28
12
289
7
4
Fig. 1. Survival in unresectable patients. Kaplan-Meier survival curve (n 5 83).
290
Journal of Gastrointestinal Surgery
Lesurtel et al.
morbidity and mortality rates less than 30% and 5%, respectively, in patients with tumors appearing unresectable at laparotomy.3,12–14 In our series, the mean hospital stay was 16 days (median 13 days, range 8–60). This long hospital stay is partially explained by a frequent difficulty in our country, the lack of beds in rehabilitation centers that offer medical support to patients before they return home. When the pancreatic tumor seems unresectable at laparotomy, the alternative would be to stop the surgical procedure and to consider endoscopic palliation. We do not use this strategy. Its first disadvantage is exposure of patients to several anesthesiology procedures, each with their own risk. Secondly, meta-analyses of randomized trials comparing biliary bypass surgery to endoscopic biliary stenting clearly demonstrated that more treatment sessions are required for recurrent jaundice after stent placement than after surgery, thus impairing the benefit of the endoscopic procedure.18,19 The use of metallic expendable stents may reduce this difference with surgical bypass, but that has not been tested yet in a randomized trial.20 Third, the systematic addition of a gastrojejunostomy to surgical biliary bypass efficiently prevents gastric outlet obstruction symptoms.3,21–25 In our series only 4 patients (4.8%) developed late gastric obstruction. Two recent prospective randomized trials comparing concomitant biliary and gastric bypass with biliary bypass alone, in patients found at exploratory laparotomy to have unresectable periampullary carcinoma, did not show any added morbidity in the group with prophylactic gastric bypass combined with the biliary bypass (30%).2,26 Follow-up demonstrated in the double bypass groups a dramatic drop of the incidence of late outlet obstruction (19–41% in patients without gastrojejunostomy vs. 0–5% in patients undergoing a gastrojejunostomy, P ! 0.001 in both trials). Furthermore, morbidity rates are high, approaching 25%, in patients who require a second surgical procedure in this setting.27 The incidence of postoperative-delayed gastric emptying after prophylactic gastrojejunostomy combined with biliary bypass for palliative treatment has been reported to vary from 8–15%.3,12,13,25 In the study by Lillemoe et al.,12 there was no significant difference of incidence of delayed gastric emptying between retrocolic and antecolic gastrojejunostomy (6% vs. 17%, respectively; P 5 0.08). Recently, endoscopic or radiologic palliations of duodenal obstruction using large-caliber metallic stents have been considered as an alternative to surgical gastric bypass.7–9,28–30 In the largest pub-
lished series by Nassif et al.9 including 63 patients, stent placement and resumption of oral diet were achieved in 95% and 92% of the patients, respectively. However, the median patency time of these expensive stents was only 5.5 weeks, and 20% of the patients experience recurrent gastric outlet obstruction despite a short median survival of 7 weeks. The here-adopted surgical palliation strategy, simultaneously combining biliary and gastric bypasses in jaundiced patients with preserved general condition and nonobvious diffuse metastatic disease, yielded low procedure-related morbidity and mortality rates. It allows us to identify resectable tumors and reduces the need for expensive sophisticated preoperative assessment such as systematic echoendoscopy or laparoscopic exploration. In a single procedure, it obtains proof of malignancy by a conclusive surgical biopsy in 92% of the cases and provides a long-acting efficient palliation in patients whose survival may increase with new palliative chemotherapies. The cost-effectiveness of this strategy should be compared with more recent and frequently proposed complex procedures, which combine extensive imaging assessment and expensive repeated endoscopic procedures.
REFERENCES 1. Singh SM, Reber HA. Surgical palliation for pancreatic cancer. Surg Clin North Am 1989;69(3):599–611. 2. Lillemoe KD, Cameron JL, Hardacre JM, et al. Is prophylactic gastrojejunostomy indicated for unresectable periampullary cancer? A prospective randomized trial. Ann Surg 1999;230(3):322–328. 3. Sohn TA, Lillemoe KD, Cameron JL, Huang JJ, Pitt HA, Yeo CJ. Surgical palliation of unresectable periampullary adenocarcinoma in the 1990s. J Am Coll Surg 1999;188(6): 658–666. 4. Gilbert DA, DiMarino AJ, Jr., Jensen DM, et al. Status evaluation: biliary stents. Gastrointest Endosc 1992;38(6): 750–752. 5. Truong S, Bohndorf V, Geller H, Schumpelick V, Gunther RW. Self-expanding metal stents for palliation of malignant gastric outlet obstruction. Endoscopy 1992;24(5): 433–435. 6. de Baere T, Harry G, Ducreux M, et al. Self-expanding metallic stents as palliative treatment of malignant gastroduodenal stenosis. AJR Am J Roentgenol 1997;169(4):1079– 1083. 7. Feretis C, Benakis P, Dimopoulos C, Manouras A, Tsimbloulis B, Apostolidis N. Duodenal obstruction caused by pancreatic head carcinoma: palliation with self-expandable endoprostheses. Gastrointest Endosc 1997;46(2):161–165. 8. Soetikno RM, Lichtenstein DR, Vandervoort J, et al. Palliation of malignant gastric outlet obstruction using an endoscopically placed Wallstent. Gastrointest Endosc 1998; 47(3):267–270. 9. Nassif T, Prat F, Meduri B, et al. Endoscopic palliation of malignant gastric outlet obstruction using self-expandable
Vol. 10, No. 2 2006
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
metallic stents: Results of a multicenter study. Endoscopy 2003;35(6):483–489. Feduska NJ, Dent TL, Lindenauer SM. Results of palliative operations for carcinoma of the pancreas. Arch Surg 1971; 103(2):330–334. Pretre R, Huber O, Robert J, Soravia C, Egeli RA, Rohner A. Results of surgical palliation for cancer of the head of the pancreas and periampullary region. Br J Surg 1992;79(8):795–798. Lillemoe KD, Sauter PK, Pitt HA, Yeo CJ, Cameron JL. Current status of surgical palliation of periampullary carcinoma. Surg Gynecol Obstet 1993;176(1):1–10. van Wagensveld BA, Coene PP, van Gulik TM, Rauws EA, Obertop H, Gouma DJ. Outcome of palliative biliary and gastric bypass surgery for pancreatic head carcinoma in 126 patients. Br J Surg 1997;84(10):1402–1406. Borie F, Rodier JG, Guillon F, Millat B. Palliative surgery of pancreatic adenocarcinoma [in French]. Gastroenterol Clin Biol 2001;2 Pt 2:C7–C14. Shepherd HA, Royle G, Ross AP, Diba A, Arthur M, Colin-Jones D. Endoscopic biliary endoprosthesis in the palliation of malignant obstruction of the distal common bile duct: a randomized trial. Br J Surg 1988;75(12): 1166–1168. Andersen JR, Sorensen SM, Kruse A, Rokkjaer M, Matzen P. Randomised trial of endoscopic endoprosthesis versus operative bypass in malignant obstructive jaundice. Gut 1989;30(8):1132–1135. Smith AC, Dowsett JF, Russell RC, Hatfield AR, Cotton PB. Randomised trial of endoscopic stenting versus surgical bypass in malignant low bileduct obstruction. Lancet 1994; 344(8938):1655–1660. Taylor MC, McLeod RS, Langer B. Biliary stenting versus bypass surgery for the palliation of malignant distal bile duct obstruction: a meta-analysis. Liver Transpl 2000;6(3): 302–308. Schwarz A, Beger HG. Biliary and gastric bypass or stenting in nonresectable periampullary cancer: analysis on the basis of controlled trials. Int J Pancreatol 2000;27(1):51–58.
Palliative Surgery for Pancreatic Carcinoma
291
20. Maosheng D, Ohtsuka T, Ohuchida J, et al. Surgical bypass versus metallic stent for unresectable pancreatic cancer. J Hepatobiliary Pancreat Surg 2001;8(4):367–373. 21. Watanapa P, Williamson RC. Surgical palliation for pancreatic cancer: developments during the past two decades. Br J Surg 1992;79(1):8–20. 22. Huguier M, Baumel H, Manderscheid JC, Houry S, Fabre JM. Surgical palliation for unresected cancer of the exocrine pancreas. Eur J Surg Oncol 1993;19(4):342–347. 23. Neuberger TJ, Wade TP, Swope TJ, Virgo KS, Johnson FE. Palliative operations for pancreatic cancer in the hospitals of the U.S. Department of Veterans Affairs from 1987 to 1991. Am J Surg 1993;166(6):632–636. 24. Bakkevold KE, Kambestad B. Morbidity and mortality after radical and palliative pancreatic cancer surgery. Risk factors influencing the short-term results. Ann Surg 1993;217(4): 356–368. 25. Shyr YM, Su CH, Wu CW, Lui WY. Prospective study of gastric outlet obstruction in unresectable periampullary adenocarcinoma. World J Surg 2000;24(1):60–64. 26. Van Heek NT, De Castro SM, van Eijck CH, et al. The need for a prophylactic gastrojejunostomy for unresectable periampullary cancer: a prospective randomized multicenter trial with special focus on assessment of quality of life. Ann Surg 2003;238(6):894–902. 27. Watanapa P, Williamson RC. Single-loop biliary and gastric bypass for irresectable pancreatic carcinoma. Br J Surg 1993; 80(2):237–239. 28. Dumas R, Demarquay JF, Caroli-Bosc FX, et al. Palliative endoscopic treatment of malignant duodenal stenosis by metal prosthesis [in French]. Gastroenterol Clin Biol 2000; 24(8–9):714–718. 29. Yim HB, Jacobson BC, Saltzman JR, et al. Clinical outcome of the use of enteral stents for palliation of patients with malignant upper GI obstruction. Gastrointest Endosc 2001; 53(3):329–332. 30. Wong YT, Brams DM, Munson L, et al. Gastric outlet obstruction secondary to pancreatic cancer: surgical vs endoscopic palliation. Surg Endosc 2002;16(2):310–312.
This study aimed to reappraise short-term and long-term results of palliative biliary and gastric bypass surgery in patients with unresectable pancreatic head carcinoma found at explorative laparotomy. We retrospectively analyzed 83 consecutive patients whose pancreatic head carcinoma appeared unresectable at laparotomy (vascular involvement [57%], liver metastases [24%], distant metastatic lymph nodes [11%], peritoneal implants [8%]) and who underwent palliative surgical concomitant biliary and gastric bypass. Postoperative mortality and morbidity rates were 4.8% and 26.5%, respectively. Postoperativedelayed gastric emptying occurred in 9 patients (10%). Antecolic (46%) and retrocolic (54%) gastrojejunostomies did not differ for the duration of nasogastric suction, the delay of oral intake, and the incidence of delayed gastric emptying. Mean hospital stay was 16 6 8 days. Median survival was 9 months (range 1–44). Late cholangitis occurred in 2 patients (2.4%) treated medically. One recurrent jaundice required transhepatic stenting 9 months from surgery. Four late gastric outlet obstructions occurred (4.8%) with a mean delay of 8 months from surgery. These data demonstrate that, in patients with unresectable pancreatic head carcinoma at laparotomy, palliative concomitant biliary and gastric bypass in a single procedure is safe and long-term efficient. This strategy remains to be compared to endoscopic palliation in this setting. ( J GASTROINTEST SURG 2006;10:286–291) Ó 2006 The Society for Surgery of the Alimentary Tract KEY
WORDS:
Pancreatic carcinoma, palliative care, gastric bypass, biliary bypass
Only 5–20% of pancreatic head carcinoma are resectable at the time of presentation.1–3 For patients with obvious nonresectable disease, endoscopic techniques have been developed as alternatives to traditional surgical management. Biliary stenting, and more recently duodenal self-expandable endoprostheses, have been promoted as the treatment of choice because of their low morbidity.4–9 However, despite improvement in imaging procedures, assessing unresectability still remains difficult in some cases, and purely nonsurgical palliation may, in these cases, overlook resectable tumors. Furthermore, pancreatic biopsies performed under radiologic or endoscopic ultrasound guidance, which have to be obtained before starting a palliative treatment by chemotherapy or radiotherapy, fail to prove the diagnosis of adenocarcinoma in 10–20% of attempted procedures. In these questionable cases where unresectability of the tumor is not proven, our policy today remains in favor of a surgical approach where
palliative surgical bypass is performed if the tumor appears unresectable at laparotomy. Historical series reported high morbidity (up to 50%) and mortality (up to 30%) rates of bypass surgery,10,11 but more recent series reported reduced morbidity and mortality rates of less than 30% and 10%, respectively.3,12–14 Whereas endoscopic palliation is widely used today, this study aimed to reappraise the short-term and long-term results of a surgical palliation policy combining biliary-enteric bypass and gastrojejunostomy in patients with unresectable pancreatic head carcinoma at laparotomy.
MATERIAL AND METHODS Patients Between June 1996 and December 2003, 340 patients without obvious contraindication to resection detected by preoperative imaging assessment
From the Department of Digestive Surgery, Saint Antoine University-Hospital, Paris, France. Reprint requests: Franc¸ois Paye, M.D., Ph.D., Department of Digestive Surgery, Hoˆpital Saint Antoine, 184, rue du Faubourg Saint Antoine 75012 Paris, France. e-mail: [email protected] Ó 2006 The Society for Surgery of the Alimentary Tract
286 Published by Elsevier Inc.
1091-255X/06/$dsee front matter doi:10.1016/j.gassur.2005.05.011
Vol. 10, No. 2 2006
(distant metastases, distant lymph nodes involvement, unresectable vascular involvement) underwent a surgical exploration for pancreatic head adenocarcinoma. If such contraindication was found intraoperatively, a biopsy was done to prove malignancy, and surgical palliation was performed in a single surgical procedure. Among the 340 patients, 257 (76%) underwent a pancreaticoduodenectomy. The remaining 83 patients (24%), who underwent surgical palliation for malignant pancreatic disease found unresectable at laparotomy, were retrospectively reviewed in this study. There were 38 women and 45 men with a mean age of 64 6 11 years. The following preoperative symptoms were recorded at first presentation: jaundice (77%), abdominal pain (54%), loss of weight more than 10% (25%), and vomiting or nausea (13%). Mean delay between the onset of symptoms and operation was 9 6 9 weeks. Resectability was routinely assessed preoperatively by chest X-ray, abdominal ultrasonography, and abdominal-computed tomography (helicoidal CT scan was used from 1998 to 2003). Echoendoscopy was not routinely performed. For 62% of the patients, the resectability of the tumor was preoperatively judged as doubtful. For the others (38%), the preoperative assessment judged the tumor as resectable. Histologic confirmation of the diagnosis of adenocarcinoma was obtained intraoperatively in 73 patients (88%). For the remaining patients (12%), either the intraoperative biopsies were not performed (n 5 4) or were negative (n 5 6), but proof of malignancy was obtained by follow-up and demonstrated disease progression in these 10 patients. In this specific setting, the sensitivity of biopsy was 92%.
Palliative Surgery for Pancreatic Carcinoma
287
Table 1. Indications for surgical palliation (n 5 83 patients) Reasons for unresectability
Vascular invasion* Liver metastases Distant metastatic lymph nodes Peritoneal implants
No. of patients (%)
47 20 9 7
(57) (24) (11) (8)
*Vascular invasion included arterial encasement or involvement of the superior mesenteric artery, the celiac axis, or the hepatic artery, and unresectable venous involvement of the mesenteric and portal veins.
25 patients (30%) received postoperative chemotherapy, and 23 patients (28%) postoperative chemoradiation. Outcome Postoperative mortality, postoperative morbidity, resumption of oral diet, delay of flatus passage, postoperative hospital stay, and long-term survival were assessed. Postoperative mortality was defined as death within 30 days after operation. Postoperativedelayed gastric emptying was diagnosed when a normal diet was not tolerated within 10 days of surgery.12 Follow-up Follow-up information was obtained through direct patient or referring physician contacts, from hospital charts, and by contacting the French civil status registry office. Mean length of follow-up was 9 6 9 months. Statistical Analysis
Surgical Palliative Procedure Reasons for unresectability indicating surgical palliation are shown in Table 1. The palliative procedure included, routinely, both a biliary-enteric bypass and a gastrojejunostomy. Biliary bypass was a hepaticoduodenostomy in 69 patients (83%). A Roux and Y hepaticojejunostomy was performed in 14 patients (17%) because the duodenum could not be used due to tumor volume or duodenum involvement. Gastrojejunostomy was antecolic in 46 patients (antecolic group) and retrocolic in 37 patients (retrocolic group), depending on the surgeons. Both were isoperistaltic. A chemical splanchnicectomy with alcohol was performed in 5 patients (6%) because preoperative pain was not controlled by oral analgesics. Mean duration of surgical procedures was 203 6 67 minutes. Postoperatively, 35 patients (42%) had no other therapy,
Data were presented as mean 6 standard deviation or median (range). Statistical analysis was carried out using statistical software (Statview 5.0, SAS Institute Inc., Cary, NC). Differences between groups were evaluated using chi-square analysis, Fisher’s exact test, or Student’s t-test when appropriate. Survival analysis was performed with the Kaplan-Meier method. The log-rank test was used to evaluate differences in survival between groups. A P value ! 0.05 was considered as statistically significant. RESULTS Short-term Outcome Mortality rate was 4.8%, and the overall postoperative morbidity rate was 26.5% (Table 2). Postoperative deaths were due to one hepatic failure
288
Journal of Gastrointestinal Surgery
Lesurtel et al.
Table 2. Postoperative mortality and morbidity (n 5 83 patients)
Table 3. Clinical and surgical findings in the two groups of gastrojejunostomy
No. of patients (%)
Mortality* Morbidity Delayed gastric emptying Wound infection Intra-abdominal abscess Ascites Pneumonia Urinary infection Cholangitis
4 (4.8) 22 (27) 9 3 1 4 2 2 1
*Defined as death within 30 days after operation.
in an alcoholic patient, one mesenteric arterial infarction, one gastrojejunal anastomotic leak, and one pneumonia complicated by multiorgan failure. Age of the patients did not correlate with the overall postoperative complication rate (P 5 0.47). One patient (1.2%) was reoperated to drain a wound abscess. The most common complication was the postoperative-delayed gastric emptying occurring in 9 patients (10%). Seven patients (8%) received red cell units perioperatively. Overall mean hospital stay was 16 6 8 days (median 13 days, range 8–60).
Antecolic Retrocolic group (n 5 46) group (n 5 37) P
Male/Female 25/21 Age 64 6 11 Delay first symptoms 10 6 11 operation (days) Preoperative symptoms Jaundice 33 (72) Abdominal pain 25 (54) Vomiting/Nausea 7 (15) Weight loss O10% 17 (37) Reasons for unresectability Vascular invasion 29 (63) Liver metastases 10 (22) Metastatic lymph nodes 3 (7) Peritoneal implants 4 (8) Duration of surgical 207 6 72 procedure (min) Biliary bypass Hepaticoduodenostomy 40 (87) Hepaticojejunostomy 6 (13)
20/17 62 6 12 866
0.97 0.43 0.37
31 20 6 8
(84) (54) (16) (22)
0.30 0.97 0.90 0.20
18 10 6 3 197
(49) (27) (16) (8) 6 62
0.25 0.60 0.90 0.30 0.53
29 (78) 8 (22)
0.45
Values in parentheses are percentages. Continuous data are presented as mean 6 standard deviation.
Long-term Outcome Comparison Between Antecolic and Retrocolic Gastrojejunostomies No significant difference in patient demographics, preoperative symptoms, and surgical findings was observed between the two groups, as shown in Table 3. There was no statistical difference between the two groups for postoperative length of nasogastric suction, passage of flatus, resumption of oral intake, and postoperative-delayed gastric emptying (Table 4). Delayed gastric emptying was not correlated with preoperative symptoms of gastric outlet obstruction (P 5 0.92).
Comparison Between Hepaticoduodenostomy and Hepaticojejunostomy No significant difference in patient demographics, preoperative symptoms, and short-term outcomes was observed between the two groups (data not shown). The surgical time was significantly longer in patients with a hepaticojejunostomy than with a hepaticoduodenostomy (251 6 88 minutes vs. 193 6 58 minutes; P 5 0.002).
Median postoperative survival was 9.2 months (range 1–44; Fig. 1). There was no statistical difference in survival between patients without postoperative treatment (median 7.1 months, range 1–44), patients treated by postoperative chemotherapy (median 10.6 months, range 1–35), or postoperative chemoradiation (median 10.3 months, range 1–30). Two patients (2.4%) developed cholangitis 8 months and 17 months, respectively, after initial palliative surgery and were efficiently treated by antibiotics. These two patients had undergone a hepaticoduodenostomy. One patient (1.2%) developed recurrent jaundice 9 months after initial surgery with a hepaticojejunostomy. He was treated by biliary transhepatic stenting. There was no significant difference in terms of late biliary complications between patients with hepaticoduodenostomy and patients with hepaticojejunostomy (3% vs. 7%, NS). Four patients (4.8%) had late gastric outlet obstruction with a mean delay of 8 months from the initial palliative surgery. Three of them had an antecolic gastrojejunostomy and one had a retrocolic gastrojejunostomy (NS). One of them underwent a subsequent second gastroenterostomy, whereas the others had medical support only. Thus, jaundice and gastric outlet obstruction were efficiently prevented
Vol. 10, No. 2 2006
Palliative Surgery for Pancreatic Carcinoma
Table 4. Postoperative course following antecolic and retrocolic gastrojejunostomies Antecolic group n 5 46
Retrocolic group n 5 37
Nasogastric 765 764 suction (days) Passage of 562 561 flatus (days) Oral intake (days) 763 763 Postoperative delayed 6 (13%) 3 (8%) gastric emptying Morbidity 11 (24%) 11 (30%) Mortality* 4 (8.6%) 0 (0%) Hospital stay (days) 16 6 8 16 6 9 Late outlet 3 (7 6 3 months) 1 (10 months) obstruction
antecolic and retrocolic gastrojejunostomy yield similar morbidity, including postoperative delayed gastric emptying. Finally, the low rates of late cholangitis (1.2%), recurrent jaundice (2.4%), and late gastric outlet obstruction (4.8%) confirm the long-term efficacy of the palliative procedure. Procedure-related morbidity and mortality rates in our study (26% and 4.8%, respectively) compare favorably with those in the three prospective randomized studies comparing biliary bypass surgery to endoscopic biliary stenting; surgical morbidity rates were reported ranging from 26% to 40% and high mortality rates from 15% to 31%.15–17 It is noteworthy that significantly lower morbidity (11% vs. 29%; P 5 0.02) and lower procedure-related mortality rates (3% vs. 14%; P 5 0.006) in patients treated endoscopically were only demonstrated in Smith’s study.17 Mortality did not differ between patients with endoscopic or surgical bypass in the meta-analysis of these trials by Taylor et al.18 Patients included in these studies had a clear contraindication to resection that was preoperatively established, whereas these patients were excluded from our study. Our improved results may be due to improvement of preoperative imaging assessment and the use of endoscopic palliative procedures for patients whose preoperative workup detects a clear contraindication to surgical resection. Thus, the subgroup of patients who finally undergo surgery with a curative intent may have here a less extended disease than in these older studies. The resulting better general conditions may contribute to the better results of palliative surgery observed in our series. More recently, other groups reported reduced
P
0.96 0.64 0.52 0.71 0.17 0.12 0.90 0.76
Continuous data are presented as mean 6 standard deviation. *Was defined as death within 30 days after operation.
by this surgical palliative procedure, until death, in 78 patients (94%). DISCUSSION Despite improvements in preoperative diagnosis and staging, pancreatic head carcinoma remains frequently found to be unresectable at the time of the laparotomy, leading to an intraoperative shift from an initial curative intent toward a palliative procedure. This study confirms, first, that both biliary and gastric bypasses can be performed in this setting, with low morbidity (26.5%) and low mortality (4.8%) rates. Second, it shows that 1
Survival rate
.8
.6
.4
.2
0 0
5
10
15
20
25
30
35
40
45
3
1
1
0
Months No. at risk 83
51
28
12
289
7
4
Fig. 1. Survival in unresectable patients. Kaplan-Meier survival curve (n 5 83).
290
Journal of Gastrointestinal Surgery
Lesurtel et al.
morbidity and mortality rates less than 30% and 5%, respectively, in patients with tumors appearing unresectable at laparotomy.3,12–14 In our series, the mean hospital stay was 16 days (median 13 days, range 8–60). This long hospital stay is partially explained by a frequent difficulty in our country, the lack of beds in rehabilitation centers that offer medical support to patients before they return home. When the pancreatic tumor seems unresectable at laparotomy, the alternative would be to stop the surgical procedure and to consider endoscopic palliation. We do not use this strategy. Its first disadvantage is exposure of patients to several anesthesiology procedures, each with their own risk. Secondly, meta-analyses of randomized trials comparing biliary bypass surgery to endoscopic biliary stenting clearly demonstrated that more treatment sessions are required for recurrent jaundice after stent placement than after surgery, thus impairing the benefit of the endoscopic procedure.18,19 The use of metallic expendable stents may reduce this difference with surgical bypass, but that has not been tested yet in a randomized trial.20 Third, the systematic addition of a gastrojejunostomy to surgical biliary bypass efficiently prevents gastric outlet obstruction symptoms.3,21–25 In our series only 4 patients (4.8%) developed late gastric obstruction. Two recent prospective randomized trials comparing concomitant biliary and gastric bypass with biliary bypass alone, in patients found at exploratory laparotomy to have unresectable periampullary carcinoma, did not show any added morbidity in the group with prophylactic gastric bypass combined with the biliary bypass (30%).2,26 Follow-up demonstrated in the double bypass groups a dramatic drop of the incidence of late outlet obstruction (19–41% in patients without gastrojejunostomy vs. 0–5% in patients undergoing a gastrojejunostomy, P ! 0.001 in both trials). Furthermore, morbidity rates are high, approaching 25%, in patients who require a second surgical procedure in this setting.27 The incidence of postoperative-delayed gastric emptying after prophylactic gastrojejunostomy combined with biliary bypass for palliative treatment has been reported to vary from 8–15%.3,12,13,25 In the study by Lillemoe et al.,12 there was no significant difference of incidence of delayed gastric emptying between retrocolic and antecolic gastrojejunostomy (6% vs. 17%, respectively; P 5 0.08). Recently, endoscopic or radiologic palliations of duodenal obstruction using large-caliber metallic stents have been considered as an alternative to surgical gastric bypass.7–9,28–30 In the largest pub-
lished series by Nassif et al.9 including 63 patients, stent placement and resumption of oral diet were achieved in 95% and 92% of the patients, respectively. However, the median patency time of these expensive stents was only 5.5 weeks, and 20% of the patients experience recurrent gastric outlet obstruction despite a short median survival of 7 weeks. The here-adopted surgical palliation strategy, simultaneously combining biliary and gastric bypasses in jaundiced patients with preserved general condition and nonobvious diffuse metastatic disease, yielded low procedure-related morbidity and mortality rates. It allows us to identify resectable tumors and reduces the need for expensive sophisticated preoperative assessment such as systematic echoendoscopy or laparoscopic exploration. In a single procedure, it obtains proof of malignancy by a conclusive surgical biopsy in 92% of the cases and provides a long-acting efficient palliation in patients whose survival may increase with new palliative chemotherapies. The cost-effectiveness of this strategy should be compared with more recent and frequently proposed complex procedures, which combine extensive imaging assessment and expensive repeated endoscopic procedures.
REFERENCES 1. Singh SM, Reber HA. Surgical palliation for pancreatic cancer. Surg Clin North Am 1989;69(3):599–611. 2. Lillemoe KD, Cameron JL, Hardacre JM, et al. Is prophylactic gastrojejunostomy indicated for unresectable periampullary cancer? A prospective randomized trial. Ann Surg 1999;230(3):322–328. 3. Sohn TA, Lillemoe KD, Cameron JL, Huang JJ, Pitt HA, Yeo CJ. Surgical palliation of unresectable periampullary adenocarcinoma in the 1990s. J Am Coll Surg 1999;188(6): 658–666. 4. Gilbert DA, DiMarino AJ, Jr., Jensen DM, et al. Status evaluation: biliary stents. Gastrointest Endosc 1992;38(6): 750–752. 5. Truong S, Bohndorf V, Geller H, Schumpelick V, Gunther RW. Self-expanding metal stents for palliation of malignant gastric outlet obstruction. Endoscopy 1992;24(5): 433–435. 6. de Baere T, Harry G, Ducreux M, et al. Self-expanding metallic stents as palliative treatment of malignant gastroduodenal stenosis. AJR Am J Roentgenol 1997;169(4):1079– 1083. 7. Feretis C, Benakis P, Dimopoulos C, Manouras A, Tsimbloulis B, Apostolidis N. Duodenal obstruction caused by pancreatic head carcinoma: palliation with self-expandable endoprostheses. Gastrointest Endosc 1997;46(2):161–165. 8. Soetikno RM, Lichtenstein DR, Vandervoort J, et al. Palliation of malignant gastric outlet obstruction using an endoscopically placed Wallstent. Gastrointest Endosc 1998; 47(3):267–270. 9. Nassif T, Prat F, Meduri B, et al. Endoscopic palliation of malignant gastric outlet obstruction using self-expandable
Vol. 10, No. 2 2006
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
metallic stents: Results of a multicenter study. Endoscopy 2003;35(6):483–489. Feduska NJ, Dent TL, Lindenauer SM. Results of palliative operations for carcinoma of the pancreas. Arch Surg 1971; 103(2):330–334. Pretre R, Huber O, Robert J, Soravia C, Egeli RA, Rohner A. Results of surgical palliation for cancer of the head of the pancreas and periampullary region. Br J Surg 1992;79(8):795–798. Lillemoe KD, Sauter PK, Pitt HA, Yeo CJ, Cameron JL. Current status of surgical palliation of periampullary carcinoma. Surg Gynecol Obstet 1993;176(1):1–10. van Wagensveld BA, Coene PP, van Gulik TM, Rauws EA, Obertop H, Gouma DJ. Outcome of palliative biliary and gastric bypass surgery for pancreatic head carcinoma in 126 patients. Br J Surg 1997;84(10):1402–1406. Borie F, Rodier JG, Guillon F, Millat B. Palliative surgery of pancreatic adenocarcinoma [in French]. Gastroenterol Clin Biol 2001;2 Pt 2:C7–C14. Shepherd HA, Royle G, Ross AP, Diba A, Arthur M, Colin-Jones D. Endoscopic biliary endoprosthesis in the palliation of malignant obstruction of the distal common bile duct: a randomized trial. Br J Surg 1988;75(12): 1166–1168. Andersen JR, Sorensen SM, Kruse A, Rokkjaer M, Matzen P. Randomised trial of endoscopic endoprosthesis versus operative bypass in malignant obstructive jaundice. Gut 1989;30(8):1132–1135. Smith AC, Dowsett JF, Russell RC, Hatfield AR, Cotton PB. Randomised trial of endoscopic stenting versus surgical bypass in malignant low bileduct obstruction. Lancet 1994; 344(8938):1655–1660. Taylor MC, McLeod RS, Langer B. Biliary stenting versus bypass surgery for the palliation of malignant distal bile duct obstruction: a meta-analysis. Liver Transpl 2000;6(3): 302–308. Schwarz A, Beger HG. Biliary and gastric bypass or stenting in nonresectable periampullary cancer: analysis on the basis of controlled trials. Int J Pancreatol 2000;27(1):51–58.
Palliative Surgery for Pancreatic Carcinoma
291
20. Maosheng D, Ohtsuka T, Ohuchida J, et al. Surgical bypass versus metallic stent for unresectable pancreatic cancer. J Hepatobiliary Pancreat Surg 2001;8(4):367–373. 21. Watanapa P, Williamson RC. Surgical palliation for pancreatic cancer: developments during the past two decades. Br J Surg 1992;79(1):8–20. 22. Huguier M, Baumel H, Manderscheid JC, Houry S, Fabre JM. Surgical palliation for unresected cancer of the exocrine pancreas. Eur J Surg Oncol 1993;19(4):342–347. 23. Neuberger TJ, Wade TP, Swope TJ, Virgo KS, Johnson FE. Palliative operations for pancreatic cancer in the hospitals of the U.S. Department of Veterans Affairs from 1987 to 1991. Am J Surg 1993;166(6):632–636. 24. Bakkevold KE, Kambestad B. Morbidity and mortality after radical and palliative pancreatic cancer surgery. Risk factors influencing the short-term results. Ann Surg 1993;217(4): 356–368. 25. Shyr YM, Su CH, Wu CW, Lui WY. Prospective study of gastric outlet obstruction in unresectable periampullary adenocarcinoma. World J Surg 2000;24(1):60–64. 26. Van Heek NT, De Castro SM, van Eijck CH, et al. The need for a prophylactic gastrojejunostomy for unresectable periampullary cancer: a prospective randomized multicenter trial with special focus on assessment of quality of life. Ann Surg 2003;238(6):894–902. 27. Watanapa P, Williamson RC. Single-loop biliary and gastric bypass for irresectable pancreatic carcinoma. Br J Surg 1993; 80(2):237–239. 28. Dumas R, Demarquay JF, Caroli-Bosc FX, et al. Palliative endoscopic treatment of malignant duodenal stenosis by metal prosthesis [in French]. Gastroenterol Clin Biol 2000; 24(8–9):714–718. 29. Yim HB, Jacobson BC, Saltzman JR, et al. Clinical outcome of the use of enteral stents for palliation of patients with malignant upper GI obstruction. Gastrointest Endosc 2001; 53(3):329–332. 30. Wong YT, Brams DM, Munson L, et al. Gastric outlet obstruction secondary to pancreatic cancer: surgical vs endoscopic palliation. Surg Endosc 2002;16(2):310–312.