J Hepatobiliary Pancreat Sci (2014) 21:818–823 DOI: 10.1002/jhbp.138
ORIGINAL ARTICLE
Laparoscopic versus open cystgastrostomy for pancreatic pseudocysts: a case-matched comparative study Yazan S. Khaled · Deep J. Malde · Jessica Packer · Thomas Fox · Prodromos Laftsidis · Tolulope Ajala-Agbo · Nicola De Liguori Carino · Rahul Deshpande · Derek A. O’Reilly · David J. Sherlock · Basil J. Ammori
Published online: 31 July 2014 © 2014 Japanese Society of Hepato-Biliary-Pancreatic Surgery
Abstract Background Cystgastrostomy is the commonest method of internal drainage of pancreatic pseudocysts (PPs). While large and persistent retrogastric pancreatic pseudocysts are amenable to laparoscopic cystgastrostomy, the potential benefits of this minimally invasive laparoscopic approach over open surgery remain to be demonstrated. The aim of this study was to compare the outcomes of the laparoscopic and open approaches for cystgastrostomy. Methods Patients who underwent laparoscopic cystgastrostomy (LCG) were matched on a 3:1 basis to those who underwent open cystgastrostomy (OCG) according to age, sex distribution, and size of pseudocyst. The outcomes of the two approaches were compared on an intention-to-treat basis. Data shown represent medians. Results A total of 54 patients underwent cystgastrostomy (35 LCG, 19 OCG) between 1997 and 2011. The final case matched cohort consisted of 40 patients (12 female and 28 male) of which 30 underwent LCG (two converted to open surgery) and 10 underwent OCG. The laparoscopic and open groups were comparable for age (55 vs. 59 years, P = 0.80), sex distribution, and size of pseudocyst (10 vs. 13 cm, P = 0.51). The laparoscopic approach had a significantly shorter operating time (62 vs. 95 min, P = 0.035) and carried a significantly lower risk of postoperative morbidity (10%
Y. S. Khaled · D. J. Malde · J. Packer · T. Fox · P. Laftsidis · T. Ajala-Agbo · N. De Liguori Carino · R. Deshpande · D. A. O’Reilly · D. J. Sherlock · B. J. Ammori (*) Hepato-Pancreato-Biliary Unit, North Manchester General Hospital, Delaunays Road, Crumpsall, Manchester M8 5RB, UK e-mail: [email protected] Y. S. Khaled · D. A. O’Reilly · B. J. Ammori The University of Manchester, Manchester, UK
vs. 60%, P = 0.024) and shorter postoperative hospital stay (6.2 vs. 11 days, P = 0.038). There was one operative death after OCG (10%). Conclusion The laparoscopic approach to cystgastrostomy for large and persistent retrogastric pancreatic pseudocysts is associated with a shorter operating time, smoother and more rapid recovery, and a shorter hospital stay compared with open surgery. The laparoscopic approach should be considered the preferable approach where expertise is available. Keywords Cystgastrostomy · Laparoscopic · Pancreatic · Pancreatitis · Pseudocyst Introduction Pancreatic pseudocysts (PPs) complicate acute inflammatory necrosis of the pancreas with disruption of the pancreatic duct that leads to an encapsulated collection of extravasated pancreatic exocrine secretions. It accounts for four fifths of all pancreatic cysts and it requires no intervention in up to 85% due to their spontaneous resolution within 4–6 weeks [1, 2]. However, persistent PPs can cause life threatening complications in 5–10% such as rupture, abscess, jaundice, and hemorrhage [3]. It is generally accepted that internal drainage is indicated for persistent PPs (>6 weeks), large (>6 cm) and/or when they are symptomatic, whereas percutaneous drainage is the choice of management for immature or infected pseudocysts, and in patients who cannot tolerate surgical internal drainage [4]. PPs are traditionally managed by open surgical approach with good long-term outcomes, and the surgical options include cystgastrostomy, cystduodenostomy and Roux-en-Y cystjejunostomy depending on their location and the best
J Hepatobiliary Pancreat Sci (2014) 21:818–823
approach to achieve dependent drainage. The reported morbidity, mortality and recurrence rates of these procedures are 10–30%, 1–5%, and 5–20% respectively [5–7]. However, minimally invasive endoscopic [8] and laparoscopic [9–11] approaches to internal drainage of PPs have been increasingly reported with comparable postoperative results and some advantages associated with the minimal surgical approach. The laparoscopic methods of drainage can be achieved by cystgastrostomy, cystjejunostomy, or cystduodenostomy and through transgastric, endogastric, exogastric or lesser sac, and infracolic access [9]. On the other hand, endoscopic drainage of PPs can be achieved via a transmural or transpapillary approaches. The indications and techniques for laparoscopic and endoscopic drainage have been refined and discussed in recent review articles [1, 12]. Cystgastrostomy is, by far, the commonest method for internal drainage. In a recent randomized clinical trial (RCT), the surgical cystgastrostomy for PPs drainage was found to offer similar favorable outcomes in comparison to the endoscopic approach [13]. However, to date, no comparative trials have evaluated the efficacy of laparoscopic against open or endoscopic approaches. The aim of this study was to compare the open and laparoscopic approaches for internal drainage of large and persistent retrogastric PPs by cystgastrostomy.
Materials and methods Patient population All patients who underwent open cystgastrostomy (OCG) for PPs by surgeons at our tertiary referral unit from 1997 were considered for the study. All patients who underwent laparoscopic cystgastrostomy (LCG) from the beginning of our learning curve in 2002 were included. One surgeon (BJA) performed laparoscopic drainage in all comers, while others performed OCG in all cases. Patients were excluded if they had laparoscopic or open pancreatic cystjejunostomy or cystduodenostomy. The indications for drainage typically involved symptoms related to acute and chronic PPs. These PPs represented fluid collections of more than 4 weeks duration that were surrounded by a clearly defined wall on imaging. All acute PPs were persistent (>6 weeks duration) and were large (≥6 cm in diameter) and symptomatic thus requiring surgery. The diagnosis of a PP was established by preoperative abdominal computed tomography (CT) or on an abdominal ultrasound scan (USS). The CT findings were used to differentiate a PP, in which fluid predominates its contents, from a walled-off pancreatic necrosis in which necrotic tissue predominates.
819
Study design Patients who underwent OCG or LCG between January 1997 and January 2011 were identified through screening the theatre register and our prospective hospital database, and their clinical records were reviewed for data collection. Patients who underwent LCG were matched to patients who have had OCG on a 1:3 basis. The matching criteria included age (±5 years), sex distribution, and size of pseudocyst (±4 cm). When more than one subject met the matching criteria, controls were chosen at random. Disease characteristics and perioperative outcomes were recorded in all patients. Comparisons between the two surgical approaches included operative time, intraoperative blood loss, rate of blood transfusion, length of postoperative hospital stay and postoperative morbidity and mortality rates.
Surgical technique We have previously described a methodological approach to the operative technique of LCG [9]. Briefly, PPs were laparoscopically drained via an anterior approach (endogastric or transgastric) or posterior (exogastric) approach [9]. The anterior approach is applicable to large retrogastric pseudocysts that are readily visible and palpable at laparoscopy, while the posterior approach is useful for smaller not readily palpable pseudocysts. We endeavor to create a large fenestration between the PP and the gastric lumen with a minimum diameter of 4–5 cm to ensure adequate drainage and debridement and to minimize the potential of future recurrence of the PP. Open surgery was carried out through an upper midline or transverse laparotomy. The margin of the cystgastrostomy was routinely over-sewn at OCG but selectively at LCG.
Postoperative management Patients were allowed on oral fluids postoperatively and later light diet as soon as tolerated. Abdominal drains and nasogastric decompression of the stomach were not used routinely, and when used they were removed as early as judged suitable by the caring surgical team. Patients were discharged from hospital when adequately mobile and tolerating a liquid diet. They were advised to continue with liquid diet for 10–14 days post-surgery and to resume a solid diet thereafter. All patients were followed up regularly in the surgical clinic and regular cross-sectional imaging was undertaken as appropriate.
820
J Hepatobiliary Pancreat Sci (2014) 21:818–823
Table 1 Patient characteristics
Sex: F/M Age (years)a ASA gradea BMI (kg/m2)a Size of pseudocyst (cm)a Etiology (No.): Gallstones Alcohol Other Acute/chronic pancreatitis (No.)
LCG (n = 30)
OCG (n = 10)
P-value
9/21 55 (26–75) 2 26 (22–70) 10 (6–19)
3/7 59 (34–79) 2 28 (20–37) 13 (6–22)
0.89 0.80 0.81 0.62 0.51
13 11 6 25/5
3 5 2 6/4
ASA American Society of Anesthesiologists, BMI body mass index, LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy a Data shown represent median (range) Table 2 Operative outcomes
Surgical approach: Trans-gastric (anterior) Exo-gastric (posterior) Operating time (min)a Blood loss (ml)a Blood transfusion (units)a Conversion to open surgery: No. (%)
LCG (n = 30)
OCG (n = 10)
P-value
23 7 62 (25–250) 91 (50–120) 0 2 (6.7)
9 1 95 (40–220) N/A 0 N/A
0.653 0.035
LCG laparoscopic cystgastrostomy, N/A not available, NS not significant, OCG open cystgastrostomy a Data shown represent median (range)
Definitions Postoperative mortality was defined as the number of deaths that occurred for any reason within 30 days after surgery or in-hospital regardless of duration of hospital stay. Postoperative morbidity rate included all complications that developed within 30 days of surgery. Complications were graded according to the Dindo-Clavien classification [14, 15]. The length of postoperative hospital stay was defined as the interval from surgery to discharge. Operative time was defined as the interval (minutes) from first incision to closure of the last skin wound. Statistical analysis Categorical variables were presented in proportions and continuous variables as median (range). Data was analyzed using the software package SPSS version 17 (Chicago, IL, USA). Analyses were carried out on an intention-to-treat basis; patients converted from laparoscopic to open surgery were therefore included in the laparoscopic group. Comparison between groups was performed using the Mann–
Whitney U-test and the independent t-test as appropriate. Significance was accepted at the 5% level.
Results Patient characteristics A total of 54 patients underwent cystgastrostomy (35 LCG, 19 OCG) during the study period. The final case matched cohort consisted of 40 patients (30 LCG vs.10 OCG). The median (range) age of the whole matched population was 53.4 (26–79) years and the overall male to female ratio was 2.5:1.0. The two groups were comparable with respect to preoperative demographic: sex, age, American Society of Anesthesiologists (ASA) score, size of the PP, and body mass index (Table 1).
Intraoperative outcomes These are shown in Table 2. Blood loss in both groups was minimal (less than 100 ml per procedure) and none required
J Hepatobiliary Pancreat Sci (2014) 21:818–823
821
Table 3 Postoperative outcomes
Postoperative morbidity: No. (%) Postoperative hospital stay (days)a Postoperative mortality: No. (%)
LCG (n = 30)
OCG (n = 10)
P-value
1 (10) 6 0 (0)
6 (60) 11 1 (10)
0.024 0.038 0.061
LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy a Data shown represent median (range) Table 4 Details of postoperative complications DC classification
No. patients
Complication
Management
1 2
LCG (n = 1) OCG (n = 4)
Abdominal pain COPD exacerbation (n = 1) Pneumonia (n = 1) Hematemesis (n = 1) Low hemoglobin (n = 1) Pulmonary sepsis (n = 1) – Abdominal pain Incisional hernia – – ARDS
Oral analgesics Medical therapy IV antibiotics Conservative Transfusion 2 units IV antibiotics – Laparoscopic Washout Hernia repair – – Death in ITU
3a 3b 4a 4b 5
LCG (n = 1) 0 LCG (n = 1) OCG (n = 1) 0 0 OCG (n = 1)
ARDS acute respiratory distress syndrome, DC Dindo-Clavien, ITU intensive therapy unit, IV intravenous, LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy
blood transfusion. The median operating time of LCG was significantly shorter than that of OCG. There were two conversions (6.7%) to open surgery due to uncontrolled intraoperative bleeding from the PP. Postoperative outcomes These are shown in Table 3. LCG was associated with significantly lower morbidity and with a lower mortality compared to OCG. There was only one death in the entire cohort of patients in a 56-year-old man who underwent OCG for a PP that complicated alcoholic pancreatitis and subsequently developed an acute respiratory distress syndrome on the 9th POD. The details of the postoperative complications are listed in Table 4. There was one patient in the LCG group who developed an acute abdominal pain on the 2nd POD and underwent a re-laparoscopy and abdominal washout, though no gastric leak was found; he made a subsequent uneventful recovery. LCG was also associated with a significantly shorter duration of postoperative hospital stay. Follow up The two groups had a comparable duration of postoperative follow up (Table 5). Recurrence of a PP was diagnosed in
Table 5 Follow up outcomes
Patients who attended: No. (%) Follow up duration (months)a Cyst resolution: No. (%) Recurrent symptomatic PP: No. (%)
LCG (n = 30)
OCG (n = 10)
30 (100) 13 (12–72) 29 (97) 1 (3.3)
10 (100) 18 (8–120) 12 (100) 0
LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy, PP pseudocyst a Data shown represent median (range)
one patient (3.3%) at 53 months after laparoscopic transgastric drainage; this was re-drained laparoscopically through the transgastric approach, and has not recurred on a further follow up of 12 months. There was no symptomatic recurrence of PP in the OCG group on follow up. Discussion This is the first comparative study in the literature, to date, of laparoscopic versus open cystgastrostomy for the treatment of symptomatic PPs. Our data showed that the laparoscopic approach offered significantly better outcomes than those observed with open surgery in terms of reductions in
822
operating time, operative morbidity and postoperative hospital stay while demonstrating comparative cyst resolution during long-term follow up. In our study, the longer operating time in the open group might be explained by the added time of opening and closing the abdomen at laparotomy and the higher frequency of a consultant surgeon grade carrying out the laparoscopic drainage. In a systematic review of the literature, LCG (n = 40) of PPs was associated with low morbidity (7%), short postoperative hospital stay (median of 4 days), and a recurrence rate of 6.7% that was comparable to recurrence rates reported after open surgery [12]. The current series of 30 LCG illustrates these observations with very low morbidity (10%), a median postoperative hospital stay of 6 days, and a recurrence rate of 3.3% at a median follow-up of 13 months. Our findings are also in keeping with the largest and most recently published data on LCG [16, 17]. In their published series of 90 LCG, Palanivelu et al. reported a mean hospital stay of 5.6 days while the rate of conversion, morbidity, mortality and recurrence were 0%, 3.3%, 0% and 1% respectively. Our relatively small series does not allow for a meaningful comparison of operative mortality, and in this regard we note that the previously published literature review of OCG [17] reported a mortality rate of 5% in contrast to 0% in the LCG [12]. The rate of long-term recurrence of PPs in the published series of OCG (5–20%) [18, 19] and LCG (1–6.7%) [9, 18] is variable due to the incomplete information regarding follow up duration and intensity of re-imaging. The principle of surgical treatment is to create a wide stoma between the PP and the gastrointestinal tract as well as the selection of a suitable site within the gastrointestinal tract that would achieve dependent drainage. Therefore, PPs could re-occur due to small stoma size formation and subsequently inadequate drainage or due to further episodes of pancreatic inflammation. While Palanivelu et al. treated the recurrent PP following LCG with open drainage [17], our patient was managed successfully with repeat LCG with no further evidence of recurrence on follow up. Although the recurrence rate may not be expected to be different from that reported with open surgery, as similar operative principles are applied, longer-term follow-up data are necessary to draw firm conclusions. Although LCG has been widely accepted by skilled laparoscopic surgeons, open surgery has a definite role in the management of PP such as in patients with recurrent PPs following OCG, the presence of multiple PPs and patients who had multiple open abdominal surgeries previously. Endoscopic drainage, especially when aided by endoscopic ultrasound (EUS), is also an important alternative management option for PPs, and is favored when the PP is indenting the stomach or duodenum and in the absence of necrotic pancreatic tissue. In selected patients, endoscopic drainage
J Hepatobiliary Pancreat Sci (2014) 21:818–823
was reported to be successful at 65–100% [20, 21] and was associated with failure to drain, morbidity and recurrence rates of 15.4%,13.3% and10.7%, respectively [12, 22]. A recent RCT endoscopic drainage (n = 20) versus OCG (n = 20) for the treatment of PPs reported significantly shorter hospital stay (2 vs. 6 days; P < 0.001) and lower cost with comparable morbidity and mortality rates and no recurrences with the endoscopic approach [13]. There are no RCTs, to date, to compare the outcomes of endoscopic drainage with LCG. However, endoscopic drainage is limited when the stomach does not share a common wall with the PP and the distance between stomach and the PP wall is >1 cm. In addition, ED requires a highly skilled endoscopist and it does not facilitate necrotic tissue debridement or adjuvant procedures such as cholecystectomy at the same stage. The limitations of our study are those inherent to data collection, patient population and selection bias. Because of retrospective data collection, some of the information was missing especially with regard to intraoperative blood loss in the OCG group and follow up evaluation. The etiology of acute pancreatitis in our patient population was quite heterogeneous. The relative infrequency of PPs that require internal drainage renders the conduction of a RCT only achievable in a multicentre setting. However, we attempted to circumvent selection bias through cohort matching. Conclusions In conclusion, this case-matched comparative study demonstrated that the laparoscopic approach to internal drainage of PPs through cystgastrostomy offered advantages over open surgery in terms of reductions in operating time, operative morbidity and postoperative hospital stay. While the recurrence rates are comparable in the short-term, longer follow up is required to enable firm conclusions. Conflict of interest
None declared.
References 1. Aljarabah M, Ammori B. Laparoscopic and endoscopic approaches for drainage of pancreatic pseudocysts: a systematic review of published series. Surg Endosc. 2007;21:1936–44. 2. Warshaw AL, Rattner DW. Timing of surgical drainage for pancreatic pseudocyst. Clinical and chemical criteria. Ann Surg. 1985;202:720–4. 3. Corfield A, Williamson R, McMahon M, Shearer M, Cooper M, Mayer A, et al. Prediction of severity in acute pancreatitis: prospective comparison of three prognostic indices. Lancet. 1985; 326:403–7. 4. Melman L, Azar R, Beddow K, Brunt LM, Halpin VJ, Eagon JC, et al. Primary and overall success rates for clinical outcomes after laparoscopic, endoscopic, and open pancreatic cystgastrostomy for pancreatic pseudocysts. Surg Endosc. 2009;23:267–71.
J Hepatobiliary Pancreat Sci (2014) 21:818–823 5. Köhler H, Schafmayer A, Lüdtke F, Lepsien G, Peiper HJ. Surgical treatment of pancreatic pseudocysts. Br J Surg. 1987;74:813– 15. 6. Huibregtse K, Schneider B, Vrij A, Tytgat G. Endoscopic pancreatic drainage in chronic pancreatitis. Gastrointest Endosc. 1988; 34:9–15. 7. Hauters P, Weerts J, Navez B, Champault G, Peillon C, Totte E, et al. Laparoscopic treatment of pancreatic pseudocysts. Surg Endosc. 2004;18:1645–8. 8. Kahaleh M, Shami V, Conaway M, Tokar J, Rockoff T, De La Rue S, et al. Endoscopic ultrasound drainage of pancreatic pseudocyst: a prospective comparison with conventional endoscopic drainage. Endoscopy. 2006;38:355–9. 9. Hamza N, Ammori BJ. Laparoscopic drainage of pancreatic pseudocysts: a methodological approach. J Gastrointest Surg. 2010;14:148–55. 10. Oida T, Mimatsu K, Kawasaki A, Kano H, Kuboi Y, Aramaki O, et al. Long-term outcome of laparoscopic cystogastrostomy performed using a posterior approach with a stapling device. Dig Surg. 2009;26:110–4. 11. Teixeira J, Gibbs K, Vaimakis S, Rezayat C. Laparoscopic Rouxen-Y pancreatic cyst-jejunostomy. Surg Endosc. 2003;17:1910– 13. 12. Bhattacharya D, Ammori BJ. Minimally invasive approaches to the management of pancreatic pseudocysts: review of the literature. Surg Laparosc Endosc Percutan Tech. 2003;13: 141–8. 13. Varadarajulu S, Bang JY, Sutton BS, Trevino JM, Christein JD, Wilcox CM. Equal efficacy of endoscopic and surgical
823
14.
15.
16.
17.
18.
19. 20. 21.
22.
cystogastrostomy for pancreatic pseudocyst drainage in a randomized trial. Gastroenterology. 2013;145:583–90. Clavien P-A, Sanabria JR, Strasberg SM. Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery. 1992;111:518–26. Dindo D, Demartines N, Clavien P-A. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240:205. Oida T, Mimatsu K, Kano H, Kawasaki A, Kuboi Y, Fukino N, et al. Laparoscopic cystogastrostomy via the posterior approach for pancreatic pseudocyst drainage. Hepatogastroenterology. 2011;58:1771–5. Palanivelu C, Senthilkumar K, Madhankumar MV, Rajan PS, Shetty AR, Jani K, et al. Management of pancreatic pseudocyst in the era of laparoscopic surgery–experience from a tertiary centre. Surg Endosc. 2007;21:2262–7. Imrie C, Shearer M. The diagnosis and management of pancreatic pseudocyst. In: Johnson CD, Imrie CW (eds). Pancreatic disease. London: Springer; 1991. p. 299–309. Moran B, Rew D, Johnson C. Pancreatic pseudocyst should be treated by surgical drainage. Ann R Coll Surg Engl. 1994;76:54. Lawson J, Baillie J. Endoscopic therapy for pancreatic pseudocysts. Gastrointest Endosc Clin N Am. 1995;5:181. Smits ME, Rauws EA, Tytgat GN, Huibregtse K. The efficacy of endoscopic treatment of pancreatic pseudocysts. Gastrointest Endosc. 1995;42:202–7. Rosso E, Alexakis N, Ghaneh P, Lombard M, Smart HL, Evans J, et al. Pancreatic pseudocyst in chronic pancreatitis: endoscopic and surgical treatment. Dig Surg. 2003;20:397–406.
ORIGINAL ARTICLE
Laparoscopic versus open cystgastrostomy for pancreatic pseudocysts: a case-matched comparative study Yazan S. Khaled · Deep J. Malde · Jessica Packer · Thomas Fox · Prodromos Laftsidis · Tolulope Ajala-Agbo · Nicola De Liguori Carino · Rahul Deshpande · Derek A. O’Reilly · David J. Sherlock · Basil J. Ammori
Published online: 31 July 2014 © 2014 Japanese Society of Hepato-Biliary-Pancreatic Surgery
Abstract Background Cystgastrostomy is the commonest method of internal drainage of pancreatic pseudocysts (PPs). While large and persistent retrogastric pancreatic pseudocysts are amenable to laparoscopic cystgastrostomy, the potential benefits of this minimally invasive laparoscopic approach over open surgery remain to be demonstrated. The aim of this study was to compare the outcomes of the laparoscopic and open approaches for cystgastrostomy. Methods Patients who underwent laparoscopic cystgastrostomy (LCG) were matched on a 3:1 basis to those who underwent open cystgastrostomy (OCG) according to age, sex distribution, and size of pseudocyst. The outcomes of the two approaches were compared on an intention-to-treat basis. Data shown represent medians. Results A total of 54 patients underwent cystgastrostomy (35 LCG, 19 OCG) between 1997 and 2011. The final case matched cohort consisted of 40 patients (12 female and 28 male) of which 30 underwent LCG (two converted to open surgery) and 10 underwent OCG. The laparoscopic and open groups were comparable for age (55 vs. 59 years, P = 0.80), sex distribution, and size of pseudocyst (10 vs. 13 cm, P = 0.51). The laparoscopic approach had a significantly shorter operating time (62 vs. 95 min, P = 0.035) and carried a significantly lower risk of postoperative morbidity (10%
Y. S. Khaled · D. J. Malde · J. Packer · T. Fox · P. Laftsidis · T. Ajala-Agbo · N. De Liguori Carino · R. Deshpande · D. A. O’Reilly · D. J. Sherlock · B. J. Ammori (*) Hepato-Pancreato-Biliary Unit, North Manchester General Hospital, Delaunays Road, Crumpsall, Manchester M8 5RB, UK e-mail: [email protected] Y. S. Khaled · D. A. O’Reilly · B. J. Ammori The University of Manchester, Manchester, UK
vs. 60%, P = 0.024) and shorter postoperative hospital stay (6.2 vs. 11 days, P = 0.038). There was one operative death after OCG (10%). Conclusion The laparoscopic approach to cystgastrostomy for large and persistent retrogastric pancreatic pseudocysts is associated with a shorter operating time, smoother and more rapid recovery, and a shorter hospital stay compared with open surgery. The laparoscopic approach should be considered the preferable approach where expertise is available. Keywords Cystgastrostomy · Laparoscopic · Pancreatic · Pancreatitis · Pseudocyst Introduction Pancreatic pseudocysts (PPs) complicate acute inflammatory necrosis of the pancreas with disruption of the pancreatic duct that leads to an encapsulated collection of extravasated pancreatic exocrine secretions. It accounts for four fifths of all pancreatic cysts and it requires no intervention in up to 85% due to their spontaneous resolution within 4–6 weeks [1, 2]. However, persistent PPs can cause life threatening complications in 5–10% such as rupture, abscess, jaundice, and hemorrhage [3]. It is generally accepted that internal drainage is indicated for persistent PPs (>6 weeks), large (>6 cm) and/or when they are symptomatic, whereas percutaneous drainage is the choice of management for immature or infected pseudocysts, and in patients who cannot tolerate surgical internal drainage [4]. PPs are traditionally managed by open surgical approach with good long-term outcomes, and the surgical options include cystgastrostomy, cystduodenostomy and Roux-en-Y cystjejunostomy depending on their location and the best
J Hepatobiliary Pancreat Sci (2014) 21:818–823
approach to achieve dependent drainage. The reported morbidity, mortality and recurrence rates of these procedures are 10–30%, 1–5%, and 5–20% respectively [5–7]. However, minimally invasive endoscopic [8] and laparoscopic [9–11] approaches to internal drainage of PPs have been increasingly reported with comparable postoperative results and some advantages associated with the minimal surgical approach. The laparoscopic methods of drainage can be achieved by cystgastrostomy, cystjejunostomy, or cystduodenostomy and through transgastric, endogastric, exogastric or lesser sac, and infracolic access [9]. On the other hand, endoscopic drainage of PPs can be achieved via a transmural or transpapillary approaches. The indications and techniques for laparoscopic and endoscopic drainage have been refined and discussed in recent review articles [1, 12]. Cystgastrostomy is, by far, the commonest method for internal drainage. In a recent randomized clinical trial (RCT), the surgical cystgastrostomy for PPs drainage was found to offer similar favorable outcomes in comparison to the endoscopic approach [13]. However, to date, no comparative trials have evaluated the efficacy of laparoscopic against open or endoscopic approaches. The aim of this study was to compare the open and laparoscopic approaches for internal drainage of large and persistent retrogastric PPs by cystgastrostomy.
Materials and methods Patient population All patients who underwent open cystgastrostomy (OCG) for PPs by surgeons at our tertiary referral unit from 1997 were considered for the study. All patients who underwent laparoscopic cystgastrostomy (LCG) from the beginning of our learning curve in 2002 were included. One surgeon (BJA) performed laparoscopic drainage in all comers, while others performed OCG in all cases. Patients were excluded if they had laparoscopic or open pancreatic cystjejunostomy or cystduodenostomy. The indications for drainage typically involved symptoms related to acute and chronic PPs. These PPs represented fluid collections of more than 4 weeks duration that were surrounded by a clearly defined wall on imaging. All acute PPs were persistent (>6 weeks duration) and were large (≥6 cm in diameter) and symptomatic thus requiring surgery. The diagnosis of a PP was established by preoperative abdominal computed tomography (CT) or on an abdominal ultrasound scan (USS). The CT findings were used to differentiate a PP, in which fluid predominates its contents, from a walled-off pancreatic necrosis in which necrotic tissue predominates.
819
Study design Patients who underwent OCG or LCG between January 1997 and January 2011 were identified through screening the theatre register and our prospective hospital database, and their clinical records were reviewed for data collection. Patients who underwent LCG were matched to patients who have had OCG on a 1:3 basis. The matching criteria included age (±5 years), sex distribution, and size of pseudocyst (±4 cm). When more than one subject met the matching criteria, controls were chosen at random. Disease characteristics and perioperative outcomes were recorded in all patients. Comparisons between the two surgical approaches included operative time, intraoperative blood loss, rate of blood transfusion, length of postoperative hospital stay and postoperative morbidity and mortality rates.
Surgical technique We have previously described a methodological approach to the operative technique of LCG [9]. Briefly, PPs were laparoscopically drained via an anterior approach (endogastric or transgastric) or posterior (exogastric) approach [9]. The anterior approach is applicable to large retrogastric pseudocysts that are readily visible and palpable at laparoscopy, while the posterior approach is useful for smaller not readily palpable pseudocysts. We endeavor to create a large fenestration between the PP and the gastric lumen with a minimum diameter of 4–5 cm to ensure adequate drainage and debridement and to minimize the potential of future recurrence of the PP. Open surgery was carried out through an upper midline or transverse laparotomy. The margin of the cystgastrostomy was routinely over-sewn at OCG but selectively at LCG.
Postoperative management Patients were allowed on oral fluids postoperatively and later light diet as soon as tolerated. Abdominal drains and nasogastric decompression of the stomach were not used routinely, and when used they were removed as early as judged suitable by the caring surgical team. Patients were discharged from hospital when adequately mobile and tolerating a liquid diet. They were advised to continue with liquid diet for 10–14 days post-surgery and to resume a solid diet thereafter. All patients were followed up regularly in the surgical clinic and regular cross-sectional imaging was undertaken as appropriate.
820
J Hepatobiliary Pancreat Sci (2014) 21:818–823
Table 1 Patient characteristics
Sex: F/M Age (years)a ASA gradea BMI (kg/m2)a Size of pseudocyst (cm)a Etiology (No.): Gallstones Alcohol Other Acute/chronic pancreatitis (No.)
LCG (n = 30)
OCG (n = 10)
P-value
9/21 55 (26–75) 2 26 (22–70) 10 (6–19)
3/7 59 (34–79) 2 28 (20–37) 13 (6–22)
0.89 0.80 0.81 0.62 0.51
13 11 6 25/5
3 5 2 6/4
ASA American Society of Anesthesiologists, BMI body mass index, LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy a Data shown represent median (range) Table 2 Operative outcomes
Surgical approach: Trans-gastric (anterior) Exo-gastric (posterior) Operating time (min)a Blood loss (ml)a Blood transfusion (units)a Conversion to open surgery: No. (%)
LCG (n = 30)
OCG (n = 10)
P-value
23 7 62 (25–250) 91 (50–120) 0 2 (6.7)
9 1 95 (40–220) N/A 0 N/A
0.653 0.035
LCG laparoscopic cystgastrostomy, N/A not available, NS not significant, OCG open cystgastrostomy a Data shown represent median (range)
Definitions Postoperative mortality was defined as the number of deaths that occurred for any reason within 30 days after surgery or in-hospital regardless of duration of hospital stay. Postoperative morbidity rate included all complications that developed within 30 days of surgery. Complications were graded according to the Dindo-Clavien classification [14, 15]. The length of postoperative hospital stay was defined as the interval from surgery to discharge. Operative time was defined as the interval (minutes) from first incision to closure of the last skin wound. Statistical analysis Categorical variables were presented in proportions and continuous variables as median (range). Data was analyzed using the software package SPSS version 17 (Chicago, IL, USA). Analyses were carried out on an intention-to-treat basis; patients converted from laparoscopic to open surgery were therefore included in the laparoscopic group. Comparison between groups was performed using the Mann–
Whitney U-test and the independent t-test as appropriate. Significance was accepted at the 5% level.
Results Patient characteristics A total of 54 patients underwent cystgastrostomy (35 LCG, 19 OCG) during the study period. The final case matched cohort consisted of 40 patients (30 LCG vs.10 OCG). The median (range) age of the whole matched population was 53.4 (26–79) years and the overall male to female ratio was 2.5:1.0. The two groups were comparable with respect to preoperative demographic: sex, age, American Society of Anesthesiologists (ASA) score, size of the PP, and body mass index (Table 1).
Intraoperative outcomes These are shown in Table 2. Blood loss in both groups was minimal (less than 100 ml per procedure) and none required
J Hepatobiliary Pancreat Sci (2014) 21:818–823
821
Table 3 Postoperative outcomes
Postoperative morbidity: No. (%) Postoperative hospital stay (days)a Postoperative mortality: No. (%)
LCG (n = 30)
OCG (n = 10)
P-value
1 (10) 6 0 (0)
6 (60) 11 1 (10)
0.024 0.038 0.061
LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy a Data shown represent median (range) Table 4 Details of postoperative complications DC classification
No. patients
Complication
Management
1 2
LCG (n = 1) OCG (n = 4)
Abdominal pain COPD exacerbation (n = 1) Pneumonia (n = 1) Hematemesis (n = 1) Low hemoglobin (n = 1) Pulmonary sepsis (n = 1) – Abdominal pain Incisional hernia – – ARDS
Oral analgesics Medical therapy IV antibiotics Conservative Transfusion 2 units IV antibiotics – Laparoscopic Washout Hernia repair – – Death in ITU
3a 3b 4a 4b 5
LCG (n = 1) 0 LCG (n = 1) OCG (n = 1) 0 0 OCG (n = 1)
ARDS acute respiratory distress syndrome, DC Dindo-Clavien, ITU intensive therapy unit, IV intravenous, LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy
blood transfusion. The median operating time of LCG was significantly shorter than that of OCG. There were two conversions (6.7%) to open surgery due to uncontrolled intraoperative bleeding from the PP. Postoperative outcomes These are shown in Table 3. LCG was associated with significantly lower morbidity and with a lower mortality compared to OCG. There was only one death in the entire cohort of patients in a 56-year-old man who underwent OCG for a PP that complicated alcoholic pancreatitis and subsequently developed an acute respiratory distress syndrome on the 9th POD. The details of the postoperative complications are listed in Table 4. There was one patient in the LCG group who developed an acute abdominal pain on the 2nd POD and underwent a re-laparoscopy and abdominal washout, though no gastric leak was found; he made a subsequent uneventful recovery. LCG was also associated with a significantly shorter duration of postoperative hospital stay. Follow up The two groups had a comparable duration of postoperative follow up (Table 5). Recurrence of a PP was diagnosed in
Table 5 Follow up outcomes
Patients who attended: No. (%) Follow up duration (months)a Cyst resolution: No. (%) Recurrent symptomatic PP: No. (%)
LCG (n = 30)
OCG (n = 10)
30 (100) 13 (12–72) 29 (97) 1 (3.3)
10 (100) 18 (8–120) 12 (100) 0
LCG laparoscopic cystgastrostomy, OCG open cystgastrostomy, PP pseudocyst a Data shown represent median (range)
one patient (3.3%) at 53 months after laparoscopic transgastric drainage; this was re-drained laparoscopically through the transgastric approach, and has not recurred on a further follow up of 12 months. There was no symptomatic recurrence of PP in the OCG group on follow up. Discussion This is the first comparative study in the literature, to date, of laparoscopic versus open cystgastrostomy for the treatment of symptomatic PPs. Our data showed that the laparoscopic approach offered significantly better outcomes than those observed with open surgery in terms of reductions in
822
operating time, operative morbidity and postoperative hospital stay while demonstrating comparative cyst resolution during long-term follow up. In our study, the longer operating time in the open group might be explained by the added time of opening and closing the abdomen at laparotomy and the higher frequency of a consultant surgeon grade carrying out the laparoscopic drainage. In a systematic review of the literature, LCG (n = 40) of PPs was associated with low morbidity (7%), short postoperative hospital stay (median of 4 days), and a recurrence rate of 6.7% that was comparable to recurrence rates reported after open surgery [12]. The current series of 30 LCG illustrates these observations with very low morbidity (10%), a median postoperative hospital stay of 6 days, and a recurrence rate of 3.3% at a median follow-up of 13 months. Our findings are also in keeping with the largest and most recently published data on LCG [16, 17]. In their published series of 90 LCG, Palanivelu et al. reported a mean hospital stay of 5.6 days while the rate of conversion, morbidity, mortality and recurrence were 0%, 3.3%, 0% and 1% respectively. Our relatively small series does not allow for a meaningful comparison of operative mortality, and in this regard we note that the previously published literature review of OCG [17] reported a mortality rate of 5% in contrast to 0% in the LCG [12]. The rate of long-term recurrence of PPs in the published series of OCG (5–20%) [18, 19] and LCG (1–6.7%) [9, 18] is variable due to the incomplete information regarding follow up duration and intensity of re-imaging. The principle of surgical treatment is to create a wide stoma between the PP and the gastrointestinal tract as well as the selection of a suitable site within the gastrointestinal tract that would achieve dependent drainage. Therefore, PPs could re-occur due to small stoma size formation and subsequently inadequate drainage or due to further episodes of pancreatic inflammation. While Palanivelu et al. treated the recurrent PP following LCG with open drainage [17], our patient was managed successfully with repeat LCG with no further evidence of recurrence on follow up. Although the recurrence rate may not be expected to be different from that reported with open surgery, as similar operative principles are applied, longer-term follow-up data are necessary to draw firm conclusions. Although LCG has been widely accepted by skilled laparoscopic surgeons, open surgery has a definite role in the management of PP such as in patients with recurrent PPs following OCG, the presence of multiple PPs and patients who had multiple open abdominal surgeries previously. Endoscopic drainage, especially when aided by endoscopic ultrasound (EUS), is also an important alternative management option for PPs, and is favored when the PP is indenting the stomach or duodenum and in the absence of necrotic pancreatic tissue. In selected patients, endoscopic drainage
J Hepatobiliary Pancreat Sci (2014) 21:818–823
was reported to be successful at 65–100% [20, 21] and was associated with failure to drain, morbidity and recurrence rates of 15.4%,13.3% and10.7%, respectively [12, 22]. A recent RCT endoscopic drainage (n = 20) versus OCG (n = 20) for the treatment of PPs reported significantly shorter hospital stay (2 vs. 6 days; P < 0.001) and lower cost with comparable morbidity and mortality rates and no recurrences with the endoscopic approach [13]. There are no RCTs, to date, to compare the outcomes of endoscopic drainage with LCG. However, endoscopic drainage is limited when the stomach does not share a common wall with the PP and the distance between stomach and the PP wall is >1 cm. In addition, ED requires a highly skilled endoscopist and it does not facilitate necrotic tissue debridement or adjuvant procedures such as cholecystectomy at the same stage. The limitations of our study are those inherent to data collection, patient population and selection bias. Because of retrospective data collection, some of the information was missing especially with regard to intraoperative blood loss in the OCG group and follow up evaluation. The etiology of acute pancreatitis in our patient population was quite heterogeneous. The relative infrequency of PPs that require internal drainage renders the conduction of a RCT only achievable in a multicentre setting. However, we attempted to circumvent selection bias through cohort matching. Conclusions In conclusion, this case-matched comparative study demonstrated that the laparoscopic approach to internal drainage of PPs through cystgastrostomy offered advantages over open surgery in terms of reductions in operating time, operative morbidity and postoperative hospital stay. While the recurrence rates are comparable in the short-term, longer follow up is required to enable firm conclusions. Conflict of interest
None declared.
References 1. Aljarabah M, Ammori B. Laparoscopic and endoscopic approaches for drainage of pancreatic pseudocysts: a systematic review of published series. Surg Endosc. 2007;21:1936–44. 2. Warshaw AL, Rattner DW. Timing of surgical drainage for pancreatic pseudocyst. Clinical and chemical criteria. Ann Surg. 1985;202:720–4. 3. Corfield A, Williamson R, McMahon M, Shearer M, Cooper M, Mayer A, et al. Prediction of severity in acute pancreatitis: prospective comparison of three prognostic indices. Lancet. 1985; 326:403–7. 4. Melman L, Azar R, Beddow K, Brunt LM, Halpin VJ, Eagon JC, et al. Primary and overall success rates for clinical outcomes after laparoscopic, endoscopic, and open pancreatic cystgastrostomy for pancreatic pseudocysts. Surg Endosc. 2009;23:267–71.
J Hepatobiliary Pancreat Sci (2014) 21:818–823 5. Köhler H, Schafmayer A, Lüdtke F, Lepsien G, Peiper HJ. Surgical treatment of pancreatic pseudocysts. Br J Surg. 1987;74:813– 15. 6. Huibregtse K, Schneider B, Vrij A, Tytgat G. Endoscopic pancreatic drainage in chronic pancreatitis. Gastrointest Endosc. 1988; 34:9–15. 7. Hauters P, Weerts J, Navez B, Champault G, Peillon C, Totte E, et al. Laparoscopic treatment of pancreatic pseudocysts. Surg Endosc. 2004;18:1645–8. 8. Kahaleh M, Shami V, Conaway M, Tokar J, Rockoff T, De La Rue S, et al. Endoscopic ultrasound drainage of pancreatic pseudocyst: a prospective comparison with conventional endoscopic drainage. Endoscopy. 2006;38:355–9. 9. Hamza N, Ammori BJ. Laparoscopic drainage of pancreatic pseudocysts: a methodological approach. J Gastrointest Surg. 2010;14:148–55. 10. Oida T, Mimatsu K, Kawasaki A, Kano H, Kuboi Y, Aramaki O, et al. Long-term outcome of laparoscopic cystogastrostomy performed using a posterior approach with a stapling device. Dig Surg. 2009;26:110–4. 11. Teixeira J, Gibbs K, Vaimakis S, Rezayat C. Laparoscopic Rouxen-Y pancreatic cyst-jejunostomy. Surg Endosc. 2003;17:1910– 13. 12. Bhattacharya D, Ammori BJ. Minimally invasive approaches to the management of pancreatic pseudocysts: review of the literature. Surg Laparosc Endosc Percutan Tech. 2003;13: 141–8. 13. Varadarajulu S, Bang JY, Sutton BS, Trevino JM, Christein JD, Wilcox CM. Equal efficacy of endoscopic and surgical
823
14.
15.
16.
17.
18.
19. 20. 21.
22.
cystogastrostomy for pancreatic pseudocyst drainage in a randomized trial. Gastroenterology. 2013;145:583–90. Clavien P-A, Sanabria JR, Strasberg SM. Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery. 1992;111:518–26. Dindo D, Demartines N, Clavien P-A. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240:205. Oida T, Mimatsu K, Kano H, Kawasaki A, Kuboi Y, Fukino N, et al. Laparoscopic cystogastrostomy via the posterior approach for pancreatic pseudocyst drainage. Hepatogastroenterology. 2011;58:1771–5. Palanivelu C, Senthilkumar K, Madhankumar MV, Rajan PS, Shetty AR, Jani K, et al. Management of pancreatic pseudocyst in the era of laparoscopic surgery–experience from a tertiary centre. Surg Endosc. 2007;21:2262–7. Imrie C, Shearer M. The diagnosis and management of pancreatic pseudocyst. In: Johnson CD, Imrie CW (eds). Pancreatic disease. London: Springer; 1991. p. 299–309. Moran B, Rew D, Johnson C. Pancreatic pseudocyst should be treated by surgical drainage. Ann R Coll Surg Engl. 1994;76:54. Lawson J, Baillie J. Endoscopic therapy for pancreatic pseudocysts. Gastrointest Endosc Clin N Am. 1995;5:181. Smits ME, Rauws EA, Tytgat GN, Huibregtse K. The efficacy of endoscopic treatment of pancreatic pseudocysts. Gastrointest Endosc. 1995;42:202–7. Rosso E, Alexakis N, Ghaneh P, Lombard M, Smart HL, Evans J, et al. Pancreatic pseudocyst in chronic pancreatitis: endoscopic and surgical treatment. Dig Surg. 2003;20:397–406.
