Pancreatic pseudocyst

Tzu Chi Med J 2004 � 16 � No. 6 �� ! PRV

The Role of Surgery in Pancreatic Pseudocyst

Wen-Yao Yin, Hwa-Tzong Chen1, Shih-Ming Huang

1, Chih-Wen Lin

2, Shih-Pin Lin, Dah-Wen Shyu,

Chang-Kuo Wei, Ming-Che Lee1, An-Liang Chou

3, Kuo-Chih Tseng

3,Yao-Jen Chang

1

Department of Surgery, Radiology2, Gastroenterology3, Buddhist Dalin Tzu Chi General Hospital, Chiayi, Taiwan; Department

of Surgery1, Buddhist Tzu Chi General Hospital, Hualien, Taiwan

ABSTRACTObjective: Surgery was the only option available for management of pancreatic pseudocyst (PP) for many years. Recently, newmethods, such as percutaneous drainage, endoscopic transenteric drainage and transpapaillary drainage, have been used for treat-ment of a pseudocyst. However, no single technique offers the desired combination of 100% success and no complications. Wepresent our surgical experience with PP over the past 14 years. Patients and Methods: A total of 22 patients were treated for PP inour departments in Dalin and Hualien Tzu Chi General Hospitals within the last 14 years. They were retrospectively reviewed andfollowed up. Results: There were 14 (63.6%) men and 8 (36.4%) women between 15 and 79-years-old ( mean age 38.2 years).Dominating symptoms in most patients were epigastric pain, palpable mass, nausea, vomiting, fever and leukocytosis, and persis-tent elevation of serum amylase. Imaging studies, such as ultrasound, computed tomography scan, and endoscopic retrograde cholangio-pancreaticography, were helpful in establishing the diagnosis. In addition to symptomatic persistent large ( > 6 cm) pseudocysts,various complications, including infection, GI obstruction, rupture into the GI tract, peritonitis, GI bleeding, internal bleeding, andpancreatic ascites were indications for surgery in our cases. Operative procedures consisted of external drainage (ED, 9 cases),internal drainage using cystojejunostomy (CJ, 4 cases )and cystogastrostomy (CG, 8 cases), and distal pancreatectomy (1 case).There were ten complications (45.5%) including recurrence of cyst (1 patient with ED and 1 with CJ), recurrence withpancreaticopleural fistula (1 with ED), colon perforation (1 with ED), delayed massive bleeding (1 with CG), pancreatic fistula (3with ED), pancreatic abscess (1 with CJ) and persistent pain (1 with CG). Repeat surgery was needed to stop bleeding (1 patient withCG) and to construct a proximal colostomy for a colon injury (1 with ED). One patient had a CJ for recurrence of pseudocyst 9 yearsafter the first surgery. Percutaneous drainage with a wide bore tube was effective for pancreatic abscess (1 with CJ) and transpapillarydrainage with a stent was used to relieve pleural effusion with respiratory failure (1 with ED). No deaths occurred in this series.Conclusion: Although complications do occur in surgical treatment, we believe that it is still important in the management ofselected cases of pseudocyst of the pancreas. Surgical intervention, endoscopic drainage, and percutaneous drainage are comple-mentary rather than competing alternatives both for simple and complicated pseudocysts. (Tzu Chi Med J 2004; 16:359-369)

Key words: pancreatic pseudocyst, internal drainage, external drainage, percutaneous drainage, transpapillary drainage

Received: August 7, 2003, Revised: December 2, 2003, Accepted: March 26, 2004Address reprint requests and correspondence to: Dr. Wen-Yao Yin, Department of Surgery, Buddhist Dalin Tzu ChiGeneral Hospital, 2, Min Sheng Road, Dalin, Chiayi, Taiwan

ORIGINAL ARTICLE

INTRODUCTION

In 1862, Le Dentu [1] proclaimed that cysts of thepancreas "should be relegated to the list of affectationsfor which the healing aid is impotent". Fortunately, forthe patient in whom a pseudocyst develops as a compli-

cation of pancreatitis, this prediction has been provedincorrect. Despite more than 100 years of surgical expe-rience with this disease entity, controversy continues tosurround the timing of operative intervention as well asindications for specific procedures.

Surgery was the only option available for manyyears. In the opinion of most surgeons and many senior

W. Y. Yin, H. T. Chen, S. M. Huang, et al

Tzu Chi Med J 2004 � 16 � No. 6PSM �� !

gastroentrologists, surgical drainage is still the treatmentof choice. It has been a mainstay in the management ofthe pseudocyst especially in cases of recurrent or per-sistent pseudocyst, and cyst rupture, in patients withconcurrent common bile duct or duodenal stenosis, inassociation with dilated pancreatic duct, and in suspectedneoplasms. Recently, new methods, such as percutane-ous drainage, endoscopic cystoenterostomy, and endo-scopic transpapillary drainage have been developed totreat the pseudocyst. However these methods are notwidely used because of technical demands and inade-quacy of the procedures. With endoscopic internal drain-age 10%-50% of patients require surgery [2] , and withpercutaneous drainage 50%-60% of patients need fur-ther treatment [3].

No single technique offers the desired combinationof a 100% success and no complications. We reviewedthe results of surgical therapy by the same surgical teamfor pancreatic pseudocysts including various complica-tions at our institutions over the past 14 consecutiveyears. Only a few papers in this field has been publishedduring last two decades [4-7]. One of the most impor-tant reasons is that these cases were not common in oursurgical practice. The aim of this paper is to share oursurgical experience with this disease entity with ourcolleagues.

PATIENTS AND METHODS

A total of 22 patients treated for pancreatic pseudo-cyst presenting with different manifestations at the sur-gical departments of both Hualien and Dalin Tzu ChiGeneral Hospitals from 1989 through 2003 were col-lected retrospectively. Pancreatic pseudocysts were con-firmed by their lack of epithelial lining on cyst wallbiopsy, a high concentration of pancreatic enzymes, andtheir formation at least 4 weeks after an episode of acutepancreatitis or pancreatic trauma. A small specimen wasexcised in 11 elective cases and 5 out of 6 infected cases.Only the patients who needed surgery due to acute com-plications such as infection, peritonitis, ascites, or bleed-ing or who had a symptomatic PP, were included in thisstudy. The sizes of the PPs were recorded from opera-tive findings or measurement of the PPs on CT images.Pertinent data including age, sex, associated diseases,clinical findings, operative procedures performed, andpostoperative course were analyzed. Duration of follow- up was 2 months to 14 years with an average of 5.91years.

RESULTS

From 1989 through 2000, 17 patients with pancre-atic pseudocysts were treated at Hualien Tzu Chi Gen-eral Hospital. From 2001 through 2003, 5 patients weretreated at Dalin Tzu Chi Genreal Hospital by the samegeneral surgery team. The male to female ratio was 14:8. The mean age was 38.2 years (range 15-79 years).Etiologies are listed in Table 1. Alcoholism was the mostcommon cause of pseudocyst. Abdominal pain or full-ness was the most common clinical manifestation andwas noted in 95.5% (n=21) of patients. Fever and leu-

Fig. 1. Both sonography and computed tomography areequally good for diagnosis of PP. (A) Sonographicfindings in a huge pseudocyst. (B) A big cyst at thetail of the pancreas on CT scan.

1A

1B

Pancreatic pseudocyst

Tzu Chi Med J 2004 � 16 � No. 6 �� ! PSN

cocytosis were detected in most of the complicated casesand a few uncomplicated cysts (n=10, 45.5%). Sevenpatients presented with a palpable mass in the abdomen(31.8%). There were also some nonspecific gastrointes-tinal (GI) symptoms such as nausea and vomiting innearly half of the cases.

Objective documentation of a pancreatic pseudo-cyst was obtained by ultrasonography (n=17, 77.2%)(Fig.1A), computed axial tomography (n=18, 81.8%)(Fig.1B), or endoscopic retrograde pancreatography(n=5, 22.7%). Most of the PPs resolved and were fol-lowed by sonography (Fig. 2A) and CT (Fig. 2B). Theindications for surgery and the procedure performed aresummarized in Tables 2 and 3, respectively. The sizesof the PP could be recorded for 13 elective cases and 7emergency cases. All of the PP measured more than 6cm and the average size was 8.2 cm. Eleven cysts werelarger than 10 cm. The largest one measured 40 × 30 ×15 cm and occupied more than half of the abdomen in-cluding the pelvic cavity. ERCP was performed in 5 casesand four were done preoperatively. One was done for apatient with a recurrent cyst complicated by respiratoryfailure from massive pleural effusion secondary to apancreaticopleural fistula. The problem was solved bytranspapillary drainage by an endoscopist. The cyst wasinfected in a case following ERCP and urgent laparot-omy with external drainage was performed. Percutane-ous drainage under sonography or CT was done in 4cases, 2 with postoperative pancreatic abscesses and 2

2A

Fig. 2. Sonography is as good as CT in postoperative fol-low-up. (A) Shrinkage of the cyst seen on sonog-raphy. (B) No definite cyst is seen behind thestomach.

2B

Table 2. Indications for Surgery

Type of Presentation Number

Symptomatic, persistent large pseudocyst ( > 6 cm) 11Complications

Infected cyst 6Free rupture with peritonitis 1Rupture into stomach with peritonitis 1Ruptured pseudoaneurysm (gastroduodenal) into duodenum or peritoneal cavity 2Pancreatic ascites 1

Total 22

Table1. Presumed Causes of Pancreatic Pseudocyst

Cause Number (%)

Alcohol 10 (45.5%)Biliary 2 ( 9.1%)Alcoholic and trauma 2 ( 9.1%)Trauma 3 (13.6%)Hyperlipoidemia 1 ( 4.5%)Idiopathic 4 (18.2%)Total 22 (100%)

W. Y. Yin, H. T. Chen, S. M. Huang, et al

Tzu Chi Med J 2004 � 16 � No. 6PSO �� !

with recurrent pseudocysts. Preoperative usage of drain-age catheter in one case was unfortunately complicatedwith leakage of pancreatic juice around the catheter lead-ing to diffuse peritonitis.

Operative procedures used in these patients are listedin Table 3. All complicated cysts except one infectedcyst were drained externally. Internal drainage was car-ried out in another 12 symptomatic matured pseudocystsand a distal pancreatectomy was done in a patient with

Table 3. Operative Procedures

Procedures Number (%)

External drainage 9 (40.9%)Internal drainage

Cystogastrostomy 8 (36.4%)Cystojejunostomy 4 (18.2%)

Distal pancreatectomy 1 (4.5%)Total 22 (100%)

3A

Fig. 3. Cystogastrostomy. (A) Localization of the PP bytransgastric approach. (B) Cystogastrostomy wasdone with non-absorbable interlocking sutures.

3B

PP at the tail. All biopsies showed a thick fibrous walldevoid of mucosa. Cystogastrostomy was performed inmore than half of the patients (66.6%, Fig. 3). Sutureligation outside the cyst for a bleeding psueudoaneurysmwas done in two cases and partial gastrectomy was addedfor a case presenting with a cyst perforation into the stom-ach Significant morbidity occurred in 10 of 22 patients(45.5%), six of which (60%) occurred in patients withexternal drainage: three cases of pancreaticocutaneousfistula, two cases of recurrence of pseudocyst, and onecase of colon perforation. All three fistulas closed spon-taneously with conservative management which includedtotal parenteral nutrition followed by enteral feeding.One of the recurrent pseudocysts was treated first withpercutaneous drainage. Endoscopic stenting at the ini-tial attempt failed.

4B

4A

Pancreatic pseudocyst

Tzu Chi Med J 2004 � 16 � No. 6 �� ! PSP

4CFig. 4. Recurrence of PP. (A) Sonographic findings after ED.

(B) CT findings after ED. (C) CT findings after per-cutaneous drainage.

5A

5B

5C

Fig. 5. Pancreaticopleural fistula following ED. (A) Mas-sive pleural effusion. (B) CT findings of markedlycollapsed right lung. (C) ERCP shows pancreatic duc-tal leakage. (D) Resolution of PP after transpapillarydrainage.

5D

W. Y. Yin, H. T. Chen, S. M. Huang, et al

Tzu Chi Med J 2004 � 16 � No. 6PSQ �� !

Unfortunately, it recurred (Fig. 4A, 4B, 4C) and wascomplicated by a pancreaticopleural fistula associatedmassive pleural effusion and respiratory failure eightmonths later (Fig. 5A, 5B, 5C). However, it finally re-sponded well to transpapillary stent insertion (Fig. 5D).Urgent proximal colostomy was done for colon perfora-tion and reconstruction was carried out a half year later.Two recurrences occurred in the cystojejunostomy (CJ)group and only one of them needed repeat internal en-teric drainage 9 years later because of biliary obstruction.The other patient had percutaneous aspiration and didnot require repeat surgery. One patient in the cystoga-strostomy (CG) group had delayed massive bleeding,

prompting surgery for cessation of anastomotic bleed-ing on the seventh postoperative day. A case with a gi-ant cyst (Fig. 6A) treated by CJ was complicated withpancreatic abscess which was finally resolved by per-cutaneous drainage with a large bore drain tube (Fig.6B, 6C). No operative deaths occurred.

DISCUSSION

A PP is an extravasated collection of exocrine pan-creatic juice surrounded by a fibrous membrane madeof adjacent viscera and parietal wall devoid of an epithe-lial lining [8-10]. Controversies in the surgical treatmentof PP include the optimal time for intervention, thechoice of surgical technique and the choice of manage-ment of complications. In addition to the surgery, ap-propriate use of diagnostic imaging studies and non-operative techniques are the fundamentals of success inthe management of the PP.

PP is seen in over 10% of pancreatitis cases as a

6A

6B

Fig. 6. Clinical course of a giant PP. (A) A huge PP occupy-ing half of the abdomen including the pelvis. (B) Per-cutaneous drainage with a pigtail catheter in the pan-creatic abscess. (C) Multiple drains including widebore tube in different locations for better drainage.

6C

Pancreatic pseudocyst

Tzu Chi Med J 2004 � 16 � No. 6 �� ! PSR

result of pancreatic inflammation, trauma, or ductobstruction. As in most other case series, alcohol-relatedpancreatitis was the most common association withpseudocysts in our series [5]. PP should be suspected inpatients with acute pancreatitis whose symptoms fail toresolve within 7 to 10 days or in patients with chronicpancreatitis who complain of persistent pain, nausea, orvomiting. Epigastric pain or fullness occurred in 95.5%of our patients, but a lower percentage (n=7, 31.8%) hada palpable abdominal mass.

The most reliable diagnostic tools are ultrasound andCT scanning. Hessel found ultrasound to be 90% accu-rate and 98% specific when the pancreas could bevisualized. Unfortunately gas obscures the pancreas innearly one third of patients. CT scan has higher sensi-tivity and specificity, reveals retroperitoneal extension,and shows the relationship between the cyst and the ad-jacent enteric lumen [11]. CT and ultrasound wereequally accurate for diagnosis of the cyst in our cases asmost cysts were >5 cm in diameter (Fig. 1). We ana-lyzed the diagnostic procedures in 80% of our non-emer-gency cases. An initial preoperative CT scan with cor-related sonography followed by serial sonographic ex-aminations was very effective in showing resolution (Fig.2) or recurrence of the cyst (Fig. 4) before and after theoperative treatment.

Endoscopic retrograde cholangio-pancreaticography(ERCP) demonstrates abnormalities of the pancreaticduct in up to 90% of patients with pseudocyst and nearlytwo-thirds of the pseudocysts communicate with thepancreatic duct. ERCP can provide valuable informa-tion concerning the natural course of pseudocysts andselection of suitable candidates for nonoperativemanagement. Several studies suggested routine ERCP

as it might change the operative management in nearly60% of cases with improved outcome [8,12]. But it mightexacerbate acute pancreatitis and infection, and there areother endoscopic related side effects such as bowel per-foration or bleeding. We use ERCP only in elective caseswithout maturation of the wall on CT image. Percutane-ous drainage may be considered if ERCP shows no com-munication between the cyst and pancreatic duct.Transpapillary pancreatic stenting may be used in caseswith a connection, as well as in patients suspect of hav-ing pancreatic cancer and in those reluctant to havesurgery. Otherwise, the appropriate operative procedureis straight forward for uncomplicated cases. It should beperformed within 24-48 hours before a planned drain-age procedure under the umbrella of antibiotics. Theadvent of sonography and CT scan of the abdomen in1970s clearly established that approximately 50% ofpseudocysts resolve spontaneously [6,13-15]. A reviewof the literature also reveals that up to 40% of patientswith untreated PP develop complications [11]. Resolu-tion and development of complications of PP were there-fore thought to be a function of time. As a matter of fact,it is a great challenge for a surgeon to decide the timingof an operation to avoid unnecessary consequences.However, there are few longitudinal studies ofpseudocyst. The Mayo clinic experience reported byVitas and Sarr [6] reviewed 68 asymptomatic patientswith a mean follow-up of 51 months and 63% remainedasymptomatic. A similar study by Yeo at el [15] at JohnHopkins University included 36 patients and spontane-ous resolution occurred in 69% of the cases by the 1year follow-up. However, both studies had a short fol-low-up duration (1 to 4 years) and the PP were small toaverage sized. On the other hand, Bradley et al [16] af-

Table 4. Postoperative Morbidity and Management with Different Operative Methods

Procedures Complications Number (%) Management

External drainage 1.Recurrence 1 Aspiration2.Recurrence with pancreaticopleura 1 Transpapillary drainage3.Colon perforation 1 Colostomy4.Pancreatic fistula 3 Conservative

Subtotal 6 (60%)

Internal drainageCystogastrostomy (CG) 1.internal bleeding 1 Reopened one week later

2.Persistent pain 1 ConservativeCystojejunostomy (CJ) 1.Recurrence 1 Cystojejunostomy

2.Pancreatic abscess 1 Percutaneous drainage

ResectionDistal pancreatectomy Nil Nil Nil

Subtotal 4 (40%)Total 10 (100%)

W. Y. Yin, H. T. Chen, S. M. Huang, et al

Tzu Chi Med J 2004 � 16 � No. 6PSS �� !

firmed the value of a waiting period. Spontaneous reso-lution occurs in 40% of acute pancreatitis with a dura-tion less than 6 weeks, and the complication rate is only20%. A waiting period longer than 7 weeks, however,was associated with a 56% complication rate and reso-lution was observed in only 1 of 25 patients from 7-18weeks. But a current review of recent studies byPitchumoni and Agarwal [17] revealed that one formerabsolute that a 6 cm pseudocyst present for 6 weeks ormore requires surgery, is absolutely untrue.

We had a 50% (11/22) complicated cases and al-though there was no mortality, they were often associ-ated with severe morbidity. As all PP were diagnosed atthe first visit, we did not know how long the patientshad a PP. Also, elective cases with larger pseudocystshad some complications. Two cases with huge PP (>15cm) sustained pancreatic infections, in one case afterERCP and in the other patient after percutaneous drain-age was followed by leakage of pancreatic juice withdiffuse peritonitis. A third patient had severe hemody-namic changes and a pancreatic abscess following in-ternal drainage due to sudden massive fluid loss and poordrainage, respectively. A huge PP may be more diffi-cult to handle than an average sized PP (6 cm to 10 cm).Another factor was that our patients with complicatedPP often had a history of alcoholism (7/11). One symp-tomatic patient with an immature cyst wall had a recur-rence after external drainage (Fig. 4A, 4B, 4C) and se-vere pulmonary effusion due to pancreaticopulmonaryfistula after percutaneous drainage (Fig. 5A, 5B, 5C). Itwas successfully treated with transpapillary drainage(Fig. 5D). Pitchumoni's suggestions are rational [17] butthe former 6 cm-6 weeks criteria should still be a rela-tive indicator rather than an absolute one.

Surgical drainage has been the gold standard formanaging PP. However, it may be associated with sig-nificant morbidity and mortality as mentioned above.Despite the relatively limited experience in the use ofpercutaneous drainage and recurrence and failure ratesranging as high as 25% to 79%, this method has beenincreasingly recommended as the first line of treatment[3,18-22]. VonSonnenberg et al reported a 94% cure ratein 51 patients with infected PP treated by percutaneousdrainage. The fistula usually closed in less than 30 days[21]. Adams and Anderson [22] reviewed 52 patientswith an average hospital stay of 40 days and the dura-tion of fistula drainage was 42.1 days. Sometimes thesepatients failed to respond and required surgery [23]. Weused percutaneous drainage in a patient with a large in-fected PP with systemic toxemia. But it was compli-cated with leakage around the catheter with diffuseperitonitis. On the other hand, a postoperative pancre-

atic abscess with sepsis can be controlled through a widebore drain tube. Theoretically, endoscopic drainage ofPP mimics operative internal drainage. But an inherentproblem with endoscopic drainage is that not all PP arecontinuous with the stomach or duodenum. In addition,it is technically demanding and widely known for initialsuccess but early recurrence [24]. One patient in the EDgroup with pancreaticopleural fistula even after percu-taneous drainage for operative recurrence was cured withtranspapillary drainage and has been stable until the mostrecent follow-up.

We generally use surgical intervention as an initialtreatment for cases fit for anesthesia and operation. Forselected cases such as those with infected PP with fri-able thin walls, thin PP without ductal leaks, or unstableclinical conditions, we may use percutaneous drainage.The endoscopic transpapillary approach is chosen forsymptomatic cases with thin walls or ductal leaks or inpatients who are reluctant to have surgery.

Complete dependent drainage is critical in any in-ternal drainage procedure. Cystogastrostomy is usedwhen the cyst wall is adherent to the posterior gastricwall (Fig. 4). If the cyst wall is adherent to the duodenalwall, cystoduodenostomy should be performed.Cystojejunostomy is used when the cyst is not closelyadherent to the upper gastrointestinal tract and is locatedat the base of the transverse mesocolon. It is also sug-gested for extremely large PP (>15 cm) to achieve de-pendent drainage [11,17].

Infected pseudocyst presents as a secondary infec-tion of a previously sterile pseudocyst. The chance ofinfection increases with time and occurs in approxi-mately 10%-15% of patients [16]. Infection may be dueto translocation of bacteria from the GI tract, or secon-dary infection from an intracystic hematoma or it maybe iatrogenic after puncture or ERCP [25]. Externaldrainage is the traditional surgical therapy for infectedpseudocyst, since the creation of an anastomosis in aninfected field is thought to be unsafe. But developmentof pancreaticocutaneous fistula occurs in 12%-20% ofpatients [26]. Djamila Boerma [27] proved internal drain-age is safe and effective in selected cases with a firmcyst wall which can hold an anastomosis well. We usedthis technique in only one of 5 patients with infectedPP. That patient had an uneventful outcome. The remain-ing 4 patients had diffuse peritonitis or associated sys-temic toxemia and one of whom was complicated witha colon perforation and two of whom had prolongedhospital stays because of pancreatic fistulas. We sug-gest ID for an infected PP if the patient is stable andthere is a mature cyst wall.

The mortality and recurrence rate for ED and ID

Pancreatic pseudocyst

Tzu Chi Med J 2004 � 16 � No. 6 �� ! PST

are 10% vs 3% and 18% vs 8% respectively [17]. Tenpercent of persistent pancreatic fistulas treated by EDmay require a distal pancreatectomy or drainage into theRoux limb of the bowel. In our series, ED was associ-ated with higher morbidity than internal drainage (n=6,60% vs n=4, 40%) (Table 4). This was significantlyhigher than the 23% reported by Balfour [28] and quitesimilar to 72.7% and 61% noted in Brian et al [4] andShatney's [29] series respectively. This might basicallybe due to the complex state of the disease itself and lackof a chance to evaluate the pancreatic ductal conditionin an emergency.

Only one patient in each group (ID or ED) neededrepeat surgery during the same hospitalization, one dueto bleeding and one due to inadvertent colon injury withperforation. Endoscopic intervention (transpapillarydrainage) in a patient with a pancreaticopleural fistulain the ED group and radiologic intervention (percutane-ous drainage) in a patient with postoperative pancreaticabscess in the ID group prevented repeat surgery. There-fore, a multidisciplinary approach by a well-experiencedteam composed of surgeon, endoscopist, and radiolo-gist is fundamental for the best outcomes. Seventy toeighty percent of the fistulas close spontaneously within4 months. But high output (>200 cc/day) appears to ex-tend the duration of closure [30]. Three in our ED grouphad pancreatic fistulas and they were managed with to-tal parental nutrition (TPN) in the early days followedby jejunostomy feeding in all patients. Nobody requiredrepeat surgery or prolonged TPN. All fistulas closedwithin 1-2 months. Consequently, less intervention wasneeded in the ED group (33.3%) in contrast to the IDgroup (50%). There was a significantly lower incidenceof repeat surgeries, 16.6% (one out of 6 cases) in the EDgroup compared to 75% (3 out of 4) in ID group. EarlyTPN combined with routine use of early jejunostomyfeeding may be the most important strategy. We encoun-tered massive bleeding from the anastomosis in acystogastrotomy patient 7 days after surgery which re-quired emergency- repeat surgery for hemostasis. Theincidence ranged from a few cases as in our study andanother study [31] up to 50% [32] though there havealso been a few studies without this life-threatening com-plication [4]. This has been attributed to a reflux of di-gestive juices with enzymatic activation [33] on absorb-able sutures. Non-absorbable sutures with an interlock-ing suturing method may secure hemostasis better alonga cyst-enteric anastomosis. Interestingly, pseudocystrecurred in one patient 9 years after an initial cystojeju-nostomy and it was associated with biliary obstruction.

A giant cyst should be dependently drained by ID[11,17]. But this led to severe complications with sud-

den unstable hemodynamic changes and pancreaticabscess. (Fig. 6A, 6B) in immediate postoperativeperiod. Massive fluid loss and inadequate drainage withcontamination of the PP by intestinal contents from thenon-functional limb of the Roux loop into the cavity ex-plains this complication. Johnson et al reported this com-plication in three of four patients with giant PP (>15cm) and two of the patients died. They concluded thatCJ may not be an appropriate treatment for giant PP be-cause it might not provide dependent drainage [34].Usage of transpapillary drainage in combination withpercutaneous drainage may be a good alternative in thesecases if we can prove there is a communication betweenthe PP and the pancreatic duct. If surgery is chosen, ex-ternal drainage should be considered.

Formerly, operative mortality was high (8.6% to16%) but it has been reduced significantly in recent yearsto 0% to 5% [35,36]. We had no deaths in our seriesalthough complicated cases with emergency surgerywere included.

In summary, we surgically treated 22 symptomaticcases of PP with or without complications from 1989through 2003. They presented with abdominal pain orfullness. Ultrasonography and CT were equally effec-tive for diagnosis. Sonography was favored because itis noninvasive, cheaper, and convenient for follow-up.CT scanning was also helpful for evaluation of local-ization, extent, and correlation with surrounding struct-ures, and as an image guide for drainange. External drain-age of the PP was associated with higher morbidity thanID. But only one patient in each group needed repeatsurgery. Percutaneous drainage or endoscopic drainageis a good tool in the perioperative period. Thesetechniques, when coordinated with a team approach,result in the best outcome.

REFERENCES

1. LeDentu C: Rapport sur I' observation precedente. BullSoc Anat Paris 1865; 40:197-214.

2. Vitale GC, Lawhon JC, Larson GM, Harrell DJ , ReedDN Jr, MacLeods S: Endoscopic drainge of the pancre-atic pseudocyst. Surgery 1999; 126:616-623.

3. Heider R, Meyer AA, Galanko JA, Behrns KE: Percuta-neous drainage of pancreatic pseudocysts is associ-ated with a higher failure rate than surgical treatment inunselected patients. Ann Surg 1999; 229:781-789.

4. Boggs BR, Potts JR 3rd, Postier RG: Five year experi-ence with pancreatic pseudocysts. Am J Surgery 1982;144:685-688.

5. Mullins RJ, Malangoni MA, Bergamini TM, Casey JM ,Richardson JD: Controversies in the management ofpancreatic pseudocysts. Am J Surgery 1998; 155:165-172.

W. Y. Yin, H. T. Chen, S. M. Huang, et al

Tzu Chi Med J 2004 � 16 � No. 6PSU �� !

6. Vitas GJ, Sarr MG: Selected management of pancre-atic pseudocysts: Operative versus expectantmanagement. Surgery 1992; 111:123-130.

7. Pavlovsky M, perejaslov A, Chooklin S, Dovgan Y: Cur-rent management of pancreatic pseudocysts. Hepato-gastroenterology 1998; 45:846-848.

8. Nealon WH, Townsend CM Jr, Thompson JC: Preop-erative endoscopic retrograde cholangiopancreato-graphy (ERCP) in patients with pancreatic pseudocystassociated with resolving acute and chronic pancreatitis.Ann Surg 1989; 209:532-540.

9. Kozarek RA, Ball TJ, Paterson DJ, Freeny PC, RyanJA, Traverso LW: Endoscopic transpapillary therapy fordisrupted pancreatic duct and peripancreatic fluidcollection. Gastroenterology 1991; 100:1362-1370.

10. Deviere J, Bueso H, Baize M, et al: Complete disrup-tion of the main pancreatic duct: Endoscopic manage-ment. Gastrointest Endosc 1995; 42:445-451.

11. Andrew LW, Carlos FC, David WR: Pancreatic cysts,pseudocysts, and fistulas. In: Michael JZ, Seymour IS,Harold E, eds. Mainogt's abdominal operations. 10thed. London: Appleton & Lange 2000, pp 1917-1940.

12. Sugawa C, Walt AJ: Endoscopic retrograde pancrea-tography in the surgery of pancreatic pseudocysts. Sur-gery 1979; 86:639-647.

13. Aranha GV, Prinz RA, Esguerra AC, Greenlee HB: Thenature and course of cystic pancreatic lesions diagnosedby ultrasound. Arch Surg 1983; 118:486-488.

14. Beebe DS, Burburick MP, Onstad GR, et al: Manage-ment of pancreatic pseudocysts. Surg Gynecol Obstet1989; 159:562-564.

15. Yeo CJ, Bastides JA, Lynch-Nyhan A, Fishman EK,Zinner MJ, Cameron JL: The natural history of pancre-atic pseudocysts documented by computed tomography.Surg Gynecol Obstet 1990; 170:411-417.

16. Bradley EL III, Clemcnrs JL, Gonzalez AC: The naruralhistory of pancreatic pseudocysts: A unified concept ofmanagement. Am J Surg 1979; 137:135-141.

17. Pitchumoni CS, Agarwal N: Pancreatic pseudocysts .When and how should drainage be performed? Gastro-entero Clin North Am 1999; 28:615-639.

18. Torres WE, Evert MB, Baumgartner BR, Bernadino ME:Percutaneous aspiration and drainage of pancreaticpseudocysts. AJR 1986; 147:1007-1009.

19. Aurell Y, Forsberg L, Hederstrom E, Anderson R: Per-cutaneous puncture and drainage of pancreatic pseu-docysts. A retrospective study. Acta Radiol 1990; 31:177-180.

20. Adams DB, Harvey TS, Anderson ML: Percutaneouscatheter drainage of pancreatic pseudocysts. Am Surg1991; 57:29-33.

21. Van Sonnenburg E, Wittich GR, Casola G, et al: Percu-taneous drainage of infected and noninfected pancre-atic pseudocysts: Experience in 101 cases. Radiology1989; 170:757-761.

22. Adams DB, Anderson MC: Percutaneous catheter drain-age compared with internal drainage in the manage-ment of pancreatic pseudocyst. Ann Surg 1992; 215:571-578.

23. Rao R, Fedorak I, Prinz RA: Effect of failed computedtomography-guided and endoscopic drainage on pan-creatic pseudocyst management. Surgery 1993; 114:843-849.

24. Cremer M, Deviere J, Engelholm L: Endoscopic man-agement of cysts and pseudocysts in chronicpancreatitis: Long-term flow-up after 7 years ofexperience. Gastrointest Endosc 1989; 35:1-9.

25. Fedorak IJ, Ko TC, Djuricin G, McMahon M, ThompsonK , Prinz RA: Secondary Pancreatic infections: Are theydistinct clinical entities? Surgery 1992; 112:824-831.

26. Martin EW, Catalano P, Cooperman M, Hecht C, CareyLC: Surgical decision-making in the treatment of pan-creatic pseudocysts. Internal versus external drainage.Am J Surg 1979; 138:821-824.

27. Boerma D, van Gulik TM, Obertop H, Gouma DJ: Inter-nal drainage of infected pancreatic pseudocysts: Safeor sory ? Dig Surg 1999; 16:501-505.

28. Balfour JH: Pancreatic pseudocysts: Complications andtheir relationship to the timing of treatment. Surg ClinNorth Am 1970; 50:395.

29. Shatney CH, Lillihei RC: Surgical treatment of pancre-atic pseudocysts. Ann Surg 1797; 189:386-394.

30. Zinner MJ, Barker RR, Cameron JL: Pancreatic cuta-neous fistulas. Surg Gynecol Obstet 1974; 138:710.

31. Rosenberg IK, Kahn JA, Walt AJ: Surgical experiencewith pancreatic pseudocsts. Am J Surg 1969; 117:11-17.

32. Huston DG, Zeppa R, Warren WD: Prevention of post-operative hemorrhage after pancreatic cystogastros-tomy. Ann Surg 1973; 193:176-179.

33. Maroutti RA, Spina GP, Zannini P, et al: Current man-agement of pancreatic pseudocysts. Int Surg 1982; 67:299-305.

34. Johnson LB, Rattner DW, Warshaw AL: The effect ofsize of giant pancreatic pseudocysts on the outcome ofinternal drainage procedures. Surg Gynecol Obstet1991: 173:171-174.

35. Becker WF, Pratt HS, Ganji H: Pseudocysts of thepancreas. Surg Gynecol Obstet 1968; 127:744-747.

36. Thomford NR, Jesseph JE: Pseudocyst of the pancreas.A review of fifty cases. Am J Surg 1969; 118:86-94.

Pancreatic pseudocyst

Tzu Chi Med J 2004 � 16 � No. 6 �� ! PSV

�� !"#$%&'()*+,-.

�� �� N��

N��

O�� �� �� ��

N��

P

�� P��

N

�� !"#$%&'()*+ �� !O

�� !P

�� !"#$%&'()N

��

�� �� !"#$%&'()*+,-./01'23456789 !"3:;2<=>?"@AB

�� !"#$� %&'()*� +,-./0123456789:;<=>?@ABCD(E100%

�� !"#$%�&'()*+,-. 14�� !"#$%&'()*�+,-./0123�� !

���� ! 14�E1989~2003F�� !"#$%&'("#$%)*+,-$%./ OO�� !"#$%

�� !"#$%&'()*+,-.��� �� !14�E63.6%F�� 8�E36.4%F�� !15-79��

�E�� ! 38.2 �F�� !"#$%&'()*+,-./01/23/45/67889:;<%=>

�� !"#$%&'()*+,-./01234.56789:';<=>?1@ABCD"EFGH

�� !"#$%&'()*+,-./-012314567819:1;<1-=>1-?;<@AB

�� !"#$%&'(�)*+,Eexternal drainage, ED: 9�F�� !"#$%&'()*+Ecystojejunos-

tomy, CJ: 4 �F�� !"#$%&'Ecystogastrostomy, CG: 8 �F�� !"#$E1 �F�� !"#E45.5%F

�� !"#E1 in ED and 1 in CJF�� ! �E1 in EDF�� !"E1 in EDF�� !"#$%E1 in CGF��

� �E3 in EDF�� !"E1 in CJF�� !"#E1 in CGF�� !"#$%&'()*�+,-� ./E1

in GDF�� !"#$%&'(!")*+,-E1 in EDF�� !"#$%&'()*+,-. /0E1 in

CJF�� !"#$%&'()*+,-./0123456789:;<=>?@ABCDE13$%FG

�� !"#$%&'()*+,-./0123456789:;<=>?@A"B7�� �� !"

�� !"#$%&'()*+,-./0123456789:;<=>!?@AB4CD�� EFG

�� !"#$%&'()*+,%-�./012� 3"4567� 898:;*<=>?@ABC

�� !"E���� 2004; 16:359-369F

�� !�� !"#$%&'()$%&*()$+,-.()$+/01%()

�� !"VO� U� T�� !�"#VO� NO� O�� !�"#VP� P� OS�

�� !"#$%&'()*+,-./01 O� �� !"#$%&'()*+ �� !"

��

��