The complete removal of the pancreas has been a topic of controversial discussions ever since surgeons realized that it was feasible. The sporadic reports of total pancreatectomy (TP) in the 1940s and 1950s of the twentieth century grew to a frank enthusiasm about the potential advantages of the procedure with the accumulation of surgical experience in the 1970s but were about to vanish later due to the negative metabolic consequences of the operation.
The complete removal of the gland has already been established as a potential option in the treatment of locally advanced pancreatic cancer, multifocal or recurrent exocrine, and endocrine tumors [
Despite increasing demand for TP and increasing rate of its performance at large pancreas centres recently, reports in the literature on its current indications and results remain scarce [
The aim of this study was to analyse the current indications and the outcome between the different indications for TP at a single high-volume institution and determine the current place of TP in the spectrum of pancreatic resections.
All cases of TP were analysed within the prospectively managed pancreas data bank at the Department of General and Visceral Surgery, St. Josef Hospital Bochum, Germany. All resections were performed by three experienced pancreatic surgeons. The set of data available for every patient included detailed information on all preoperative diagnostic procedures, perioperative parameters, and postoperative complications, as well as strict follow-up documentation. The preoperative risk assessment was graded according to the American Society of Anaesthesiologists classification (ASA). Operation time, perioperative blood loss, necessity of red blood cell transfusions, and postoperative hospital stay were evaluated. Postoperative complications were determined and categorised into major and minor surgical complications and nonsurgical ones. Major complications comprised biliary leakage, postoperative bleeding, intraabdominal abscess, and burst abdomen. Minor surgical complications included delayed gastric emptying, cholangitis, chylous fistula, and wound infection. Nonsurgical complications were defined as complications within 30 days of surgery or during the hospital stay that were not directly related to the surgical procedure, for example, pneumonia, urinary tract infection, or systemic infections not related to the surgical procedure (e.g., central venous infection). Any death during the hospital stay or within the first 30 days after operation was defined as postoperative mortality.
Statistical analysis was performed using the SPSS 16 software package (SPSS Inc., Chicago, IL, USA). All data are presented as median with interquartile range and minimal and maximal values, which are shown in the figures as box-and-whisker plots, respectively. For comparison of quantitative variables were used the nonparametric Mann-Whitney and Kruskal-Wallis tests. Two-sided
During the study period of 54 months (January 2004–June 2008) 948 patients underwent surgery for pancreatic disease, of which 599 (63.2%) pancreatic resectional procedures. The total pancreatectomies were 63, that is, 6.7% of all pancreatic procedures. They were performed in 34 (54%) males and 29 (46%) females at a median age of 69 (38–87) years. In 45 cases (71.4%) a TP was performed as a primary procedure; in the rest 18 patients it was a completion pancreatectomy. A splenectomy was performed in 45 patients; in eighteen the spleen was preserved. Twenty-five patients (39.7%) were classified as ASA I -II, while 38 (60.3%) were categorized as ASA class III-IV. Substantial cardiac comorbidity was present in 13 cases (21%), pulmonary in 8 (12.7%), and renal in 4 (6.3%). Insulin-dependent diabetes mellitus was present in 11 patients (17.5%).
The indications for a TP were grouped to the
Twenty-three patients (36.5%) underwent TP for malignancy. In 22 cases it was a primary operation and in one patient with a recurrent intraductal papillary-mucinous cancer of the head after left resection a completion pancreatectomy was performed. In 13 cases the main reason for total pancreatectomy was the size of the tumor, which spread over the most of the pancreas—there were 10 T3 tumors and 3 T4 tumors. There was also one patient with a T3 carcinoma of the distal hepatic duct and a positive resection margin on the frozen section—a total pancreatectomy was necessary in that case too. In 8 cases multifocal cancer was found—3 patients had a multicentric intraductal papillary-mucionous carcinoma (all T2), 3 patients suffered a multifocal pancreatic adenocarcinoma (one of them T2 and the other two T3), one patient had simultaneously a cancer of the papilla (T3) and an adenocarcinoma of the pancreatic body (T2), and one patient had a cystadenocarcinoma of the pancreatic head (T3) and a simultaneous undifferentiated neuroendocrine cancer in the pancreatic body. There was only one small T1 cancer of the pancreatic head for which a total pancreatectomy was performed, because of the subtotal atrophy of the rest pancreas due to a severe chronic pancreatitis with preoperatively existing insulin-dependent diabetes mellitus and a severe exocrine insufficiency. There were twelve multivisceral resections (
Eighteen patients (28.6%) underwent total pancreatectomy due to technical reasons, that is, very soft and fatty pancreatic tissue in the remnant. In 13 of those patients the diagnosis was a small cancer of the pancreatic head (six T1 and seven T2), two patients had a T2 cancer of the distal common bile duct, another patient had a benign cystadenoma in processus uncinatus, and there were two cases of intraductal papillary mucinous adenoma in the pancreatic head. There were no multivisceral and no vessel resections in this group. Preoperatively 5 patients had an endocrine and 1 patients also an exocrine dysfunction of the pancreas.
Fifteen patients (23.8%) underwent total pancreatectomy because of either early postoperative or intraoperative troubles, that is, complications. Twelve patients underwent a completion pancreatectomy because of postoperative complications after pancreatic resections of the head or tail: in 8 cases that was insufficiency of the pancreatic anastomosis/stump with additional postoperative intraabdominal bleeding in 8 patients; in 2 cases pancreatectomy was necessary because of necrotising pancreatitis of the pancreatic remnant with sepsis, and in 2 patients because of failure of the biliodigestive anastomosis with biliary peritonitis. In three patients the decision to perform a total pancreatectomy was made intraoperatively in an emergency situation due to iatrogenic perforation of the duodenum during ERCP, profuse bleeding of a duodenal carcinoma, and a bleeding of a giant pseudocyst of the pancreatic head with a coagulation disorder. There were
Seven patients (11.1%) suffered from a disabling therapy-resistant pain due to severe chronic pancreatitis with small duct disease. In two patients pancreatectomy was the primary intervention—the first one had a total atrophy of the pancreatic tail, and the second one presented intraoperatively with a large tumor mass, where a malignancy was suspected. In the other five cases a completion pancreatectomy was carried out following prior resective pancreatic surgery without adequate alleviation of pain (two cases after left resection, three cases after pancreaticoduodenectomy). All 7 patients suffered already preoperatively from severe endocrine and exocrine insufficiency and all of them were analgesic drug addicts. Multivisceral resections were necessary in two patients with inflammatory stenosis of the colon. Also there were three cases (43%) with portal vein reconstruction.
A total of 17 (27%) multivisceral resections were carried out. There was one total gastrectomy, ten portal vein resections with direct anastomosis, one resection of the common hepatic artery and reconstruction with the splenic artery, five left or right hemicolectomies, and one left adrenal resection.
Median operation time for the whole group was 420 min; the median intraoperative blood loss was 800 mL. Blood transfusions were necessary in 29 (46%) patients. The duration of the surgical procedure was significantly shorter in patients with troubles as an indication for pancreatectomy with a median time of 210 min compared to all other indications with a median of 448 min (
Operation time according to indications: (1) tumor, (2) technical, (3) trouble, (4) therapy-refractory pain,
Intraoperative blood loss according to indications: (1) tumor, (2) technical, (3) trouble, (4) therapy-refractory pain,
Intraoperative transfused RBC units according to indications: (1) tumor, (2) technical, (3) trouble, (4) therapy-refractory pain,
Major postoperative complications, either surgical or nonsurgical, occurred in 32 (50.8%) patients. A total of 23 patients (36.5%) developed one or more postoperative surgical complications. The spectrum of minor surgical complications included delayed gastric emptying, cholangitis, wound infection, and chylous fistula. The most often one was delayed gastric emptying—it occurred in nine (14.3%) of the pancreatectomised patients. Cholangitis developed postoperatively three (4.7%) patients. A wound infection was observed in 4 (6.3%) cases and a chylous fistula appeared in 3 (4.8%) patients.
The following major surgical complications were observed: five patients developed an intraabdominal abscess. In three cases the abscess was situated in the upper left abdominal quadrant after splenectomy, another two were found in the left paracolic region. All abscesses were successfully treated via CT-guided drainage. Three patients with emergency completion pancreatectomy suffered an acute postoperative erosion bleeding—in two cases a hemorrhage from the splenic artery, and in the other one from the left gastric artery. Two of these patients needed multiple reoperations. All of those three patients died due to the early or late consequences of the hemorrhagic shock. Three other patients with completion pancreatectomies were reoperated because of failure of the biliodigestive anastomosis. One of those patients developed a malignant postoperative arrhythmia, underwent a pacemaker implantation, but progressed to liver and renal failure with a fatal outcome. Two patients were treated for a wound dehiscence, whereas one of them died due to a progressive multiple organ failure.
The nonsurgical complications comprised of urinary tract infections (most common with 11.1%, 7 patients), 2 cases of pneumonia—one of those patients died due to multiple lung abscesses with respiratory failure and sepsis with multiorgan failure, a pseudomembranous colitis in one patient. One patient died of an acute myocardial infarction. Two patients developed renal failure, which was successfully treated by means of hemofiltration.
Median hospital stay of all surviving patients (
Postoperative stay in 53 patients who survived TP is according to indications: (1) tumor, (2) technical, (3) trouble, (4) therapy-refractory pain,
Ten patients (15.9%) died in the postoperative period. Five (7.9%) of them died as a consequence of major surgical complications directly related to the operation, as described above. One patient died of a severe pneumonia with lung abscesses, another one due to myocardial infarction. Another patient died because of thoracic hemorrhage after complicated chest tube insertion for pleural effusion during her stay at the intensive care unit. Two more patients died of multiple organ failure due to MRSA sepsis. Seven of these ten patients were preoperatively classified ASA IV and the other three were ASA III.
The highest mortality (7/15, 47%) was found in the group of patients who underwent TP because of troubles. Three patients died in the “tumor” group, 3/23, 13%. There was no perioperative or late mortality in patients with therapy-refractory chronic pancreatitis (0/7) and in the group with TP for technical or soft tissue-related reasons (0/18). In summery the mortality in elective TP was 6.25% (3/48). A statistically significant difference was found only between the “troubles” and “technical” groups,
A summary of the most important characteristics of the four groups with total pancreatectomy is given in Tables
Summary of data related to the four groups of patients who underwent total pancreatectomy (TP).
Group | Number [%] | Operative time [min] | Blood loss [mL] | RBC units | Stay [days] | Mortality [%] |
(1) Tumors | 23 (36.5%) | 450 (360–750) | 800 (300–2500) | 0 (0–18) | 20 (12–89) | 3/23 (13%) |
(2) Technical reasons | 18 (28.6%) | 445 (320–535) | 600 (300–2200) | 0 (0–4) | 21 (10–108) | 0/18 (0%) |
(3) Troubles | 15 (23.8%) | 210 (120–525) | 1500 (500–3000) | 15 (0–34) | 48 (22–90) | 7/15 (47%) |
(4) Therapy-refractory pancreatitis | 7 (11.1%) | 430 (350–480) | 600 (300–2000) | 0 (0–2) | 15 (7–30) | 0/7 (0%) |
Median (min–max range) for all 63 patients: | 420 (120–750) | 800 (300–3000) | 0 (0–34) | 21 (7–108) | 10/63 (15.9%) | |
Median (min–max range) for | 440 (320–750) | 800 (300–2500) | 0 (0–18) | 24 (11–109) | 3/48 (6,25%) |
Characteristics of 63 patients with total pancreatectomy (TP).
Variable | Tumors | Technical reasons | Troubles | Therapyi-resistant pain |
Number | 23 | 18 | 15 | 7 |
Age (median with interquartile range) | 66 (62–72) | 72.5 (65.5–77.25) | 71 (64–75) | 50 (46–58) |
Male | 10 | 13 | 6 | 5 |
Female | 13 | 5 | 9 | 2 |
Cancer | 20 | 15 | 5 | 0 |
Benign tumors | 3 | 3 | 3 | 0 |
Chronic pancreatitis | 0 | 0 | 6 | 7 |
others | 0 | 0 | 1 | 0 |
Primary TP | 22 | 18 | 3 | 2 |
Completion TP | 1 | 0 | 12 | 5 |
Portal vein/SMV resection | 9 | 0 | 1 | 2 |
Splenectomy | 20 | 6 | 14 | 5 |
Pylorus-preserving TP | 6 | 17 | 9 | 5 |
Multivisceral resection | 4 | 0 | 1 | 3 |
Enzyme replacement | 2 | 1 | 1 | 7 |
Endocrine insufficiency | 3 | 5 | 3 | 6 |
I | 1 | 0 | 0 | 0 |
II | 13 | 5 | 2 | 4 |
III | 8 | 12 | 13 | 2 |
IV | 1 | 1 | 0 | 1 |
TP indicates total pancreatectomy; SMV: superior mesenteric vein; ASA: American Society of Anesthesiologists.
Comparison of total pancreatectomy to other pancreatic resections.
Parameter | Total pancreatectomy, | other pancreatic resections, | Statistical difference |
Morbidity | |||
erosion bleeding | 3 (4.8%) | 16 (3%) | |
biliary leakage | 3 (4.8%) | 8 (1.5%) | |
pancreatic leakage | n.a. | 25 (4.7%) | n.a. |
intraabdominal abscess | 5 (7.9%) | 26 (4.9%) | |
chylous fistula | 3 (4.8%) | 21 (3.9%) | |
delayed gastric emptying | 9 (14.3%) | 41 (7.7%) | |
wound infection | 4 (6.3%) | 22 (4.2%) | |
Mortality in all 63 patients with TP | 10 (15.9%) | 6 (1.1%) | |
Mortality in elective TP (48 patients) | 3 (6.25%) | 6 (1.1%) | |
Postoperative hospital stay [days] | 21 | 14 |
In 1943/1944 Rocky [
In 2003 Büechler et al. [
In agree with this recent data of Büechler et al. our study has shown that TP is a demanding surgical procedure. Nevertheless there exist a broad range of indications. We present the data of 63 patients who underwent a TP during the period of 54 months. Indications for TP were analysed and classified in the “
Size or localisation of pancreatic
In all 15 patients of the
The majority of patients died due to medical complications (e.g., pneumonia, myocardial infection, arrhythmia).
In the reported series 18 TPs were performed because of “
Another indication for TP is
Several improvements in postoperative management including treatment of diabetes mellitus and substitution of pancreatic enzymes and fat-soluble vitamins have significantly reduced postoperative morbidity and improved quality of life after TP. Recent studies described that there was no significant difference in quality of life in patients with elective TP and partial pancreatectomy [
In conclusion the reported literature and our data exposed a satisfactory outcome for patients when the indication for TP was well reflected. We subdivided the indication of TP into the
M. S. Janot and O. Belyaev contributed equally to this paper.