Curative Treatment for Pancreatic Neoplasms: Standard Resection

PANCREATIC NEOPLASMS

0039-6109/95 $0.00 + .20

CURATIVE TREATMENT FOR PANCREATIC NEOPLASMS Standard Resection Henry A. Pitt, MD

Pancreatoduodenectomy was introduced by Whipple and his associates26 60 years ago. Numerous modifications of the original two-stage operation were introduced in the 1940s and 1950s. By the 1960s, the "standard Whipple procedure" was a one-stage procedure that included a partial pancreatic resection and an antrectomy. During the 1970s, a trend toward more radical operations including total pancreatectomy, portal vein resection, and extensive retroperitoneal lymph node dissection evolved. However, concerns about increased postoperative morbidity and mortality as well as a poorer quality of life with these more radic';ll operations led Traverso and Longmire22 to introduce the pyloruspreserving pancreatoduodenectomy (PPPD) in 1978. During the 1980s, the mortality of pancreatoduodenectomy diminished markedly from more than 20% to less than 5%.7,13,23 In recent years this trend has continued, with large series now being reported with a mortality of 0 to 2%.5,6,10,11,25 Numerous factors, including improvements in anesthetic and critical care as well as increased surgical experience, concentration of patients in centers of excellence, and improved methods to diagnose and treat complications, have led to significantly better operative resultsY During the past decade, several groups doing standard resections have also reported 5-year survival rates of more than 20%.6,10,25,29 Multiple parameters are likely to be responsible for this increased survival from previous levels of 5% to

From the Department of Surgery, The Johns Hopkins University School of Medicine, The Johns Hopkins Medical Institutions, Baltimore, Maryland

SURGICAL CLINICS OF NORTH AMERICA VOLUME 75 • NUMBER 5 • OCTOBER 1995

891

892

PITT

10%, including less blood loss and fewer transfusions, lower operative mortality, and better postoperative adjuvant therapy. With these dramatic improvements following standard resection, the need for a more radical surgical approach must be seriously questioned. STANDARD RESECTIONS

Several factors must be considered when defining a "standard resection." The extent of pancreatic resection, the role of antrectomy, the degree of lymph node dissection, and the inclusion of portal vein and/ or arterial resection are all involved in the consideration of whether an individual procedure is classified as a "standard" or a "radical" resection. In general, a partial pancreatectomy, which includes resection of the head and uncinate and divides the pancreatic neck, is considered standard. Resection of a portion of the pancreatic body to achieve a negative tumor margin might also be considered standard. In comparison, total pancreatectomy and splenectomy are a more radical treatment for a tumor confined to the head and/or uncinate. The degree of gastric resection also may be a factor in differentiating a standard from a radical operation. However, this factor is probably less important than the degree of pancreatic resection or whether a radical lymph node or vascular resection is performed. In general, advocates of the PPPD (Fig. 1) do not routinely perform an extensive retroperitoneal node dissection or a segmented resection of the portal vein. Thus, the pylorus-preserving variants of the Whipple procedure are considered standard therapy. Similarly, the inclusion of an antrectomy, the "classic" Whipple procedure, in the absence of a radical lymph node or major vascular resection is also standard therapy. On the other hand, advocates of more radical surgery usually include an antrectomy in an effort to achieve a more extensive lymph node clearance. A standard pancreatoduodenectomy includes resection of the lymph nodes intimately associated with the pancreatic head as well as those along the right side of the distal bile duct. However, a standard resection usually does not include an extensive lymph node dissection, which skeletonizes the hepatic arteries and portal vein in the porta hepatis, the celiac axis, the superior mesenteric artery, or the retroperitoneal tissue surrounding the inferior vena cava and aorta. Similarly, resection of a segment of the portal vein and/ or superior mesenteric artery to remove the tumor is not standard. However, partial resection of a small area of the portal vein where a tumor is adherent might still qualify the procedure as standard. On the other hand, radical or regional pancreatectomy usually involves an extensive retroperitoneal lymph node dissection and often includes major vascular resection and reconstruction. RECONSTRUCTIVE OPTIONS

Various options exist for reconstructing the pancreatobiliary and digestive tracts after resection. To a large extent, the reconstructive

STANDARD RESECTION

893

Figure 1. Pylorus-preserving pancreatoduodenectomy. Anatomy of the resected area (A) and reconnected digestive tract with end-to-end pancreatojejunostomy (8).

894

PITT

options are similar for standard and radical resections. With each approach, if a partial pancreatectomy is performed, options for reconstruction include oversewing or ductal blockage without anastomosis, pancreatojejunostomy, and pancreatogastrostomy. Pancreatojejunostomy may be performed as an end-to-end "dunking" procedure, an end-to-side anastomosis that also incorporates the pancreas into the jejunum, or an end-to-side anastomosis that sews the pancreatic duct directly into a small opening in the jejunum. Pancreatogastrostomy is normally performed in an end-to-side fashion. Technical factors such as the number of layers in the anastomosis, the type of suture, and whether to include the pancreatic duct in the anastomosis continue to be matters of debate. Another issue is whether these anastomoses are stented internally or externally. However, the texture of the pancreas is probably more important than any of these other factors in determining whether a pancreatic fistula develops postoperatively. The most commonly used reconstructive technique anastomoses the pancreas first, followed by the bile duct and then the stomach or duodenum (Fig. 2). Alternates include a separate Roux-en-Y limb of jejunum to the pancreas or anastomosis first to the stomach or duodenum followed by the biliary and then the pancreatic anastomoses. The biliary anastomosis is usually performed in an end-to-side fashion. Debate continues, however, as to the number of layers in the biliary anastomo-

EncJ.to-si(ie pancreaticojejunostomy

Hemigastrectomy with gastrojejunostomy

Figure 2. Reconstructive alternatives. A, Pylorus-preserving pancreatoduodenectomy with end-to-side pancreaticojejunostomy and end-to-side duodenojejunostomy. B, Classic Whipple procedure with end-to-end pancreaticojejunostomy and hemigastrectomy, with gastrojejunostomy. (From Cameron JL: The pancreas. In Atlas of Surgery, vol 1. Philadelphia, BC Decker, 1990, p 407; with permission; illustrations by Corinne Sandone.)

STANDARD RESECTION

895

sis, the type of suture, and whether a stent, usually aT-tube, should be used. If a gastrectomy is performed, a Hoffmeister modification with a side-to-side gastrojejunostomy is most commonly performed (Fig. 2). With the PPPD, the duodenojejunostomy is normally performed in an end-to-side fashion. However, those who prefer doing the duodenojejunostomy first usually perform an end-to-end anastomosis. Thus, many variations in reconstruction may be considered standard and may also apply to radical resections. However, regional pancreatectomy frequently involves removal of the whole pancreas, obviating pancreatic anastomosis, and usually also includes an antrectomy. MORBIDITY AND MORTALITY

Although the mortality of pancreatoduodenectomy has diminished in recent years, the morbidity remains high. In a recent series of 145 consecutive patients undergoing a standard resection at Johns Hopkins Hospital,6 the operative mortality was zero. However, in this series, in which 81% of patents underwent a PPPD, the postoperative morbidity was 52%. The most common complication was delayed gastric emptying, which occurred in 36% of the patients (Table 1). This problem was slightly more common in older patients, but this difference did not achieve statistical significance. Although some authorities have suggested that this problem occurs more commonly after PPPD, comparative data have not confirmed this contention?' 12, 17 Pancreatic fistula occurred postoperatively in 19% of the patients in the Hopkins series,6 but none of these patients died as a result of this complication, This relatively high incidence of pancreatic fistula is due, in part, to a moderately liberal definition, Pancreatic fistula can lead to intra-abdominal abscess, hemorrhage, and death. However, careful monitoring of drain amylase levels leading to early diagnosis, percutaneous drainage of intra-abdominal collections, and parenteral nutritional support have all diminished the lethality of this complication, Of course,

Table 1. MORBIDITY OF STANDARD PANCREATICODUODENECTOMY Postoperative Complication

Delayed gastric emptying Pancreatic fistula Intra-abdominal abscess Wound infection Pancreatitis Bile leak

<70 Years

2:70 Years

(n = 108)

(n=37)

Total (n = 145)

32% 19% 10% 6% 7% 6%

46% 22% 5% 11% 3% 3%

36% 19% 9% 8% 6% 6%

Adapted from Cameron JL, Pitt HA, Yeo CJ, et al: One hundred and forty-five consecutive pancreaticoduodenectomies without mortality. Ann Surg 217:430-438, 1993; with permission.

896

PITT

total pancreatectomy avoids the problem with pancreatic fistula. However, a report from UCLA13 suggests that the overall morbidity is actually higher with total pancreatectomy than with partial pancreatoduodenectomy. Moreover, with total pancreatectomy the diabetes can be quite brittle, and late problems with hepatic fibrosis have been reported. A recent analysis by Trede 23 of more than 1500 pancreatoduodenectomies performed in the late 1980s and early 1990s also suggests that the mortality is less with a standard Whipple procedure than with total pancreatectomy (Table 2). In this collected series the hospital mortality of total pancreatectomy was 18%, compared with only 3% for the standard Whipple procedure. In a collected series of 339 PPPDs published in 1990,12 the hospital mortality was only 3.8%. In comparison, the hospital mortality in a collected series of regional pancreatoduodenectomies was 6.4%.18 Thus, in general, the mortality of regional and total pancreatoduodenectomy is greater than for the standard Whipple procedure or the pylorus-preserving variant. SURVIVAL

Although the standard Whipple and pylorus-preserving procedures have a lower hospital mortality, controversy continues as to whether they provide the best chance for long-term survival. Critics argue that the field of resection may be reduced and that PPPD should not be performed for pancreatic cancer. 16, 19, 20 On the other hand, advocates point out that the other resection margins in pancreatoduodenectomy are much closer to the tumor than the duodenum and antrum?' 4, 12, 17 Moreover, comparative survival rates after the pylorus-preserving procedure have been the same as or better than those with the classic Whipple procedure or with total pancreatectomy,3, 12, 17, 29 Type of Resection

A recent analysis of 201 patients undergoing resection for adenocarcinoma of the head of the pancreas at Johns Hopkins Hospital from 1970 to 1994 reviewed the effect of type of resection on survival,29 Partial pancreatectomy was performed in 181 patients (90%), whereas only 20 Table 2. MORTALITY OF STANDARD WHIPPLE PROCEDURE VERSUS TOTAL PANCREATECTOMY Procedure

Institutions

Patients

Mortality

Whipple Total

12

1153

6

369

3% 18%

Adapted from Trede M, Schwall G: The complications of pancreatectomy. Ann Surg 207:39-47, 1988; with permission.

STANDARD RESECTION

897

patients underwent total pancreatectomy. Median survival was 16.0 months for partial compared with 10.0 months for total pancreatectomy (P = 0.85). Of the 181 patients undergoing partial pancreatic resection, 134 (67%) underwent PPPD, and these patients had a median survival of 17.5 months, compared with 12.0 months for the classic Whipple procedure (P<0.02). Five-year survival was 24% for the PPPD group, compared with only 9% for classic pancreatoduodenectomy (Fig. 3). These data support the performance of PPPD in the management of patients with cancer of the pancreatic head. This modification of the Whipple procedure retains the entire stomach as a reservoir and maintains a more normal gastrointestinal hormone milieu. 3, 4, 15 As a result, postgastrectomy problems are minimized, the incidence of marginal ulcer is low, and enterogastric reflux is less common. 3, 4, 8, 15, 27 The pylorus-preserving procedure also has been demonstrated to be associated with normal vitamin levels and a better weight gain postoperatively.3,8 Moreover, the avoidance of the antrectomy with PPPD results in shorter operative times, less blood loss, and fewer transfusions. 2 , 17 Blood Loss and Transfusions

Intraoperative blood loss and perioperative blood transfusions have been correlated with survival for a number of malignancies. In the data from Johns Hopkins,29 median intraoperative blood loss for the 201 patients was 800 mL, and the median number of units of red cells transfused intraoperatively was zero (mean 1.7 ± 0.2 units). Patients with less than 800 mL of estimated blood loss had significantly longer

100 90 80 70 > 60 .~ :::J 50 (J) 40 oR. 30 20 10 0

cu

P < 0.02 0

12

24

36

48

60

Months Figure 3. Survival following pylorus-preserving (PPPD; broken line, N = 134) versus classic pancreatoduodenectomy (solid line, N = 47). (From Yeo CJ, Cameron JL, Lillemoe KD, et al: Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients. Ann Surg, 1995, in press; with permission.)

898

PITT

(P<0.03) median and 5-year survival than those who lost more than 800 mL (Table 3). Similarly, patients receiving two units or less of red blood cells had significantly longer (P<0.002) median and 5-year survival than those receiving three or more transfusions (Table 3). More radical operations are likely to be associated with greater blood loss and need for transfusions. On the other hand, PPPD has been documented to be associated with less blood loss and a lower requirement for transfusions. 2, 17 Therefore, these factors must be considered when evaluating the ultimate outcome of standard and regional resections. Tumor Biology

Perhaps more important than the type of resection, the amount of blood lost or the number of transfusions is the biology of the tumor. In the data from Johns Hopkins,29 DNA content of the pancreatic cancer cells as determined by image cytometry was a highly significant determinant of survival,1,29 Of the 119 patients whose tumors were analyzed, 51 (43%) had diploid whereas 68 (57%) had aneuploid tumors (Table 4). Patients with diploid tumors had median and 5-year survivals of 24.0 months and 39%, respectively, which was significantly longer (P<0.0002) than observed in patients with aneuploid tumors (Fig. 4). As expected, the lymph node status in the resected specimen also proved to be a highly significant factor affecting survival. Of the 201 patients, 57 (28%) had negative lymph nodes (Tablf 4). The median and 5-year survivals in these node-negative patients were 28.0 months and 36%, respectively, which was significantly longer than for the nodepositive patients. The diameter of the tumor was also an important predictor of survival. Of the 198 patients for whom this measurement was available, 91 (46%) had tumors less than 3 cm in diameter (Table 4). Median and 5-year survivals in these patients with smaller tumors were 21.0 months and 28%, respectively, which was significantly longer (P<0.005) than in patients with larger tumors. The status of the resection margins proved to be another highly

Table 3. EFFECT OF BLOOD LOSS AND BLOOD TRANSFUSION ON SURVIVAL

Blood loss < 800 mL ::>: 800 mL Transfusions .,; 2 units > 2 units

Number

Median Survival (mo)

5-Year Survival

Log-rank P-value

98 96

18.0 11.5

27% 17%

0.03

146 48

18.0 10.5

26% 10%

0.002

Adapted from Yeo CJ, Cameron JL, Lillemoe KD, et al: Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients. Ann Surg, 1995, in press; with permission.

STANDARD RESECTION

899

Table 4. EFFECT OF TUMOR BIOLOGY ON SURVIVAL Number

Median Survival (mo)

5-Year Survival (%)

Log-rank Pvalue

51 68

24.0 11.5

39 8

0.0002

57 114

28.0 13.0

36 14

0.0018

91 107

21.0 11.5

28 15

0.005

143 58

18.0 10.0

26 8

0.0001

DNA ploidy Diploid Aneuploid Lymph nodes Negative Positive Diameter < 3cm 2: 3 cm Margins Negative Positive

Adapted from Yeo CJ, Cameron JL, Lillemoe KD, et al: Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients. Ann Surg, 1995, in press; with permission.

significant factor predicting survival. Specimens were routinely examined for microscopic tumor involvement at the pancreatic neck, uncinate process, retroperitoneal soft tissue, and duodenal and bile duct margins. In addition, patients with involvement of the hepatic artery and superior

0.9 0.8 at ~ 0.7

~0.6 (/)

c 0.5

o

i

Diploid

0.4

£0.3 0.2 Aneuploid

0.1

04-----~------~----~------~----~ o 2 3 4 5 Suldec18 " rilk 51

46

68

27

Years 20 9

18 5

10

2

7 1

Figure 4. Survival for patients with diploid (n = 51) and aneuploid (n = 68) tumors. (From Yeo CJ, Cameron JL, Lillemoe KD, et al: Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients. Ann Surg, 1995, in press; with permission.)

900

PITT

mesenteric or portal veins were considered to have positive margins even if the vessel was resected. Of the 201 patients, 143 (71%) had negative margins (Table 4). Median and 5-year survivals in these patients with negative margins were 18.0 months and 26%, respectively, signifiqmtly longer (P
In the 201 patients resected at Johns Hopkins from 1970 to 1994, median survival was 15.1 months and 5-year actuarial survival was 21 %.29 An initially surprising observation was that survival increased significantly (P<0.002) by decade (Fig. 5). Median survivals in the 1970s, 1980s, and 1990s were 7.5, 14.0, and 17.5 months, respectively. Similarly, I-year survivals were 32%, 53%, and 64%, respectively, and 3-year survivals were 14%, 21%, and 36% by decade. Five-year survivals for the 1970s and 1980s were 9% and 15%, respectively, whereas these data are not yet available for the 1990s. An analysis of factors that may have caused this recent improvement in survival did not demonstrate any significant changes in tumor biology (Table 5). Tumor size did not differ by decade. A slightly higher percentage of patients had negative nodes and negative margins in the 1980s and 1990s, but these trends did not achieve statistical significance. However, three factors have changed dramatically over the 25 years of this analysis (Table 5). First, the number of units of blood transfused has decreased from 5.6 units in the 1970s to 2.0 units in the 1980s to 0.9 unit in the 1990s. Hospital mortality was 30% among the 23 patients undergoing surgery in the 1970s but dropped to only 3% in 63 patients in the 1980s and to 1% in 115 patients undergoing resection in the 1990s. The third factor that may explain the improving survival is the increased use of postoperative adjuvant radiation and chemotherapy. Adjuvant Therapy

Adjuvant therapy was not recommended at Johns Hopkins for any patient prior to the Gastrointestinal Tumor Study Group's reports of efficacy in 19859 and 1987. 14 As a result, less than 25% of the patients resected in the 1980s received adjuvant therapy.27 However, since Octo-

STANDARD RESECTION

901

0.9 '90s: 68/115 alive, Median flu 11 mos. '80s: 4/63 alive, Median flu 76 mos. '70s: 1/23 alive, 181 mos.

0.8 CIt

~0.7

~0.6 rn

c::: 0.5

o

~ 0.4

'90s

Q,

20.3 Q, 0.2

'80s

0.1 70s

O;------;r------.------.------.------,

o

2

SUl!.lecta at riak 115

53

63 23

38 7

3

4

5

12 14 3

5 9 3

0 9 2

Years 19 20 4

'90s '80s

'70s

Figure 5. Survival for patients managed in the 1970s (n = 23), 1980s (n = 63), and 1990s (n = 115). (From Yeo CJ, Cameron JL, Lillembe KD, et al: Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients. Ann Surg, 1995, in press; with permission.)

ber, 1991, adjuvant therapy has been routinely recommended, and 56 of 78 recent patients (72%) have received external beam radiation and intravenous 5-fluorouracil chemotherapy.29 Median and 2-year survival in the 56 treated patients have been 20 months and 35%, respectively, which is significantly longer (P
Negative lymph nodes Mean diameter (cm) Negative margins Transfusions (units) Hospital mortality Adjuvant therapy Median survival (mo)

19705 (n=23)

19805 (n=63)

19905 (n=115)

17% 3.1 57% 5.6 30% 0% 7.5

30% 3.0 71% 2.0 3% <25% 14.0

30% 3.2 74% 0.9 1% >65% 17.5

Adapted from Yeo CJ, Cameron JL, Lillemoe KD, et al: Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients. Ann Surg, 1995, in press; with permission

902

PITT

ever, 59% of the treated patients had aneuploid tumors, compared with only 22% of the untreated patients. This factor should have adversely affected survival in the treated patients. Quality

Another importaht issue for patients with pancreatic cancer undergoing resection is the quality of postoperative survival. McLeod and colleagues 15 from the University of Toronto have recently reported a careful study of this issue. They analyzed the quality of life, nutritional status, and gastrointestinal symptomatology in 13 patients at least 6 months following a Whipple procedure and in 13 control postcholecystectomy patients. Their Whipple patients had either a standard or pylorus-preserving procedure for benign or malignant neoplasms. No significant differences were observed between post-Whipple and control patients on five quality-of-life assessments. Nutritionally, the Whipple group was at 102% of preillness weight and 113% of ideal body weight and did not differ from controls on multiple other nutritional parameters. Similarly, gastrointestinal symptomatology and mean Visick scores did not differ between post-Whipple and postcholecystectomy patients. Yasada and his associates 28 from Teikyo University in Tokyo have al~o studied the quality of survival after pancreatoduodenectomy. They analyzed 33 patients with pancreatic cancer who had survived at least 1 year after surgery. Eight underwent a PPPD, 11 patients had undergone a standard Whipple procedure, and 14 had an "extended" pancreatoduodenectomy. These authors used Mitsuno's system28 for classifying postoperative social function, which also can be used to describe functional normalcy and return to the premorbid state. All eight patients undergoing PPPD had normal social function, functional normalcy, and return to the premorbid state after surgery. Of the 11 standard Whipple patients, 100% had normal social function, but 91% achieved functional normalcy and only 36% returned to their premorbid state. In comparison, only 50% of the 14 extended pancreatoduodenectomy patients were socially functional, only 21% achieved functional normalcy, and none returned to their premorbid state. These authors concluded that PPPD is superior to the Whipple procedure and the extended pancreatoduodenectomy with respect to postoperative quality of life in patients with pancreatic cancer. SUMMARY

In the mid-1990s pancreatoduodenectomy can be performed at regional referral centers with an operative mortality of 1% to 2%. In addition, a number of centers around the world are now reporting 5year survivals between 20% and 30% for patients with resected pancreatic cancer. In recent years a debate has continued as to the proper

STANDARD RESECTION

903

extent of resection for these patients, in part because numerous resective and reconstructive options are available. In broad terms, operations can be categorized as "standard" or "radical." In general, standard operations tend to have lower postoperative morbidity and mortality. Multiple factors influence postoperative survival, including intraoperative blood loss, perioperative transfusions, type of operation, hospital mortality, tumor biology, and adjuvant therapy. Among these parameters, tumor biology is the most important by multivariate analysis. 29 Recent data suggest that the quality of survival is better with standard operations and best with the pylorus-preserving pancreatoduodenectomy. Nevertheless, debate will continue about the proper extent of resection until a well-controlled randomized trial has addressed this issue.

References 1. Allison DC, Bose KK, Hruban RH, et al: Pancreatic cancer cell DNA content correlates with long-term survival after pancreaticoduodenectomy. Ann Surg 214:648--656, 1991 2. Belli L, Riolo F, Romani F, et al: Pylorus-preserving pancreatoduodenectomy versus Whipple procedure for adenocarcinoma of the head of the pancreas. HPB Surgery 1:195-200, 1980 3. Braasch JV: Pancreaticoduodenal resection. Curr Probl Surg 25:323-363, 1988 4. Braasch JV: Pylorus preserving pancreatoduodenectomy. In Braasch JV, Tompkins RK (eds): Surgical Disease of the Biliary Tract and Pancreas. St. Louis, CV Mosby, 1994, P 565 5. Cameron JL, Crist OW, Sitzmann JV, et al: Factors influencing survival after pancreaticoduodenectomy for pancreatic cancer. Am J Surg 161:120-125, 1991 6. Cameron JL, Pitt HA, Yeo CJ, et al: One hundred and forty-five consecutive pancreaticoduodenectomies without mortality. Ann Surg 217:430--438, 1993 7. Crist OW, Sitzmann JV, Cameron JL: Improved hospital morbidity, mortality and survival after the Whipple procedure. Ann Surg 206:358-365, 1987 8. Fink AS, DeSouza LR, Mayer EA, et al: Long-term evaluation of pylorus preservation during pancreatoduodenectomy. World J Surg 12:663--670, 1988 9. Gastrointestinal Tumor Study Group: Further evidence of effective adjuvant combined radiation and chemotherapy following curative resection of pancreatic cancer. Cancer 59:2006-2010, 1987 10. Geer RJ, Brennan MF: Prognostic indicators for survival after resection of pancreatic adenocarcinoma. Am J Surg 165:68-73, 1993 11. Gordon TA, Burleyson GP, Tielsch JM, et al: The effects of regionalization on cost and outcome for one high-risk surgical procedure. Ann Surg 221:43--49, 1995 12. Grace P A, Pitt HA, Longmire WP: Pylorus preserving pancreatoduodenectomy: An overview. Br J Surg 77:968-974, 1980 13. Grace PA, Pitt HA, Tompkins RK, et al: Decreased morbidity and mortality after pancreatoduodenectomy. Am J Surg 151:141-149, 1986 14. KaIser MH, Ellenberg SS: Pancreatic cancer. Adjuvant combined radiation and chemotherapy following curative resection. Arch Surg 120:899-903, 1985 15. 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 169:179185,1995 16. Moosa AR: Pancreatic cancer: Approach to diagnosis, selection for surgery and choice of operation. Cancer 50:2689-2698, 1982 17. Pitt HA, Grace PA: Pylorus-preserving resection of the pancreas. Bailliere's Clin GastroenteroI4:917-930, 1990

904

PITT

18. Reber HA, Ashley SW, McFadden D: Curative treatment for pancreatic neoplasms: Radical resection. Surg Clin North Am 75:905-912, 1995 19. Roder JD, Stein HJ, Huttl W, et al: Pylorus-preserving versus standard pancreaticoduodenectomy: An analysis of 110 pancreatic and periampullary carcinomas. Br J Surg 79:152-155, 1992 20. Sharp KW, Ross CB, Halter SA, et al: Pancreatoduodenectomy with pylorus preservation for carcinoma of the pancreas: A cautionary note. Surgery 105:645--653, 1989 21. Takada T, Yasuda H, Shikata JI, et al: Postprandial plasma gastrin and secretin concentrations after a pancreatoduodenectomy. Ann Surg 210:47-51, 1989 22. Traverso LW, Longmire WP Jr: Preservation of the pylorus in pancreaticoduodenectomy. Surg Gynecol Obstet 146:959-962, 1978 23. Trede M: The surgical treatment of pancreatic carcinoma. Surgery 94:28--35, 1985 24. Trede M, Schwall G: The complications of pancreatectomy. Ann Surg 207:39-47, 1988 25. Trede M, Schwall G, Saeger HD: Survival after pancreatoduodenectomy: 118 consecutive resections without an operative mortality. Ann Surg 211:447-458, 1990 26. Whipple AO, Parson WB, Mullens CR: Treatment of carcinoma of the ampulla of Vater. Ann Surg 102:763--779, 1935 27. Williamson RCN, Bliouras N, Cooper MJ, et al: Gastric emptying and enterogastric reflux after conservative and conventional pancreatoduodenectomy. Surgery 114:8286,1993 28. Yasuda H, Takada T, Uchiyama K, et al: Social function following pylorus-preserving pancreatico-duodenectomy for cancer of the head of the pancreas. Asian J Surg 16:228-231, 1993 29. Yeo CJ, Cameron JL, Lillemoe KD, et al: Pancreaticoduodenectomy for cancer of the head of the pancreas: 201 patients. Ann Surg, 1995, in press Address reprint requests to Henry A. Pitt, MD The Johns Hopkins Hospital 600 North Wolfe Street Blalock 679 Baltimore, MD 21287