Stent versus surgery

Abstract

Following the introduction of percutaneous and endoscopic biliary drainage there has been an ongoing debate about the indications and outcomes of endoscopic versus surgical drainage in a variety of bilio-pancreatic disorders. The evidence-based literature concerning four different areas of pancreatobiliary diseases have been reviewed. Preoperative endoscopic biliary drainage in patients with obstructive jaundice should not be used routinely but only in selected patients. For patients with biliary leakage and bile duct strictures after a laparoscopic cholecystectomy, endoscopic stent therapy might be first choice and surgery should be used for failures of endoscopic treatment. Surgery is the treatment of choice after transection of the bile duct (the major bile duct injuries). The majority of patients with obstructive jaundice due to advanced pancreatic cancer will undergo endoscopic drainage but for relatively fit patients with a prognosis of more than 6 months, surgical drainage or even palliative resection might be considered. For patients with persistent pain due to chronic pancreatitis surgical drainage combined with limited pancreatic head resection might be first choice for pain relief. Most importantly, the management of patients with these pancreatobiliary diseases should be performed by a multidisciplinary HPB approach and teamwork consisting of gastroenterologists, radiologists and surgeons.

Introduction

Following the introduction of percutaneous and endoscopic drainage techniques for the biliary tract and more recently the pancreatic duct there has been an ongoing debate worldwide regarding the indications for endoscopic/percutaneous versus surgical drainage in a variety of bilio-pancreatic disorders.

Traditionally the AMC in Amsterdam had a very active Department of Gastroenterology under the leadership of Professor Guido Tytgat and Professor Kees Huibregtse and excellent interventional radiology under Professor Han Laméris, and therefore both minimally invasive techniques and surgery were used with good cooperation during the past decades.

This presidential lecture is an opportunity to review the evidence-based literature concerning these different drainage approaches for several HPB disorders such as:

• obstructive jaundice and the role of preoperative biliary drainage;

• endoscopic and surgical management of biliary stricture and bile duct injury;

• endoscopic versus surgical palliative treatment of pancreatic carcinoma; and finally,

• the use of both drainage procedures in patients with chronic pancreatitis.

We should realize that the implementation of these findings in daily HPB practice is mainly dependent on the local expertise of the different partners in the multidisciplinary HPB approach.

Obstructive jaundice and preoperative biliary drainage

Surgery in patients with obstructive jaundice caused by a periampullary (pancreas, papilla, distal bile duct) tumor as well as a proximal bile duct tumor (Klatskin tumor) is associated with a higher risk of postoperative complications than in non-jaundiced patients. The increased risk of surgery in jaundiced patients had already been recognized in 1935 by Allen O. Whipple, who proposed a two-stage procedure for surgery in deeply jaundiced patients ¹. After the introduction of percutaneous and endoscopic drainage, ERCP/PTC and subsequent drainage was included in the routine diagnostic work-up in several countries ²,³,⁴.

Internal biliary drainage has been shown to improve liver function and nutritional status, to reduce systemic endotoxaemia and cytokine release, and subsequently to improve immune response in multiple experimental models ⁵,⁶,⁷,⁸,⁹ (Table I). A number of non-randomized studies on internal drainage reported a reduced mortality and morbidity. However, other clinical studies and small randomized trials could not confirm the positive effect of preoperative biliary drainage (PBD) on outcome ¹⁰,¹¹,¹². Some studies even reported a deleterious effect, partly due to complications associated with the drainage procedure ⁴,¹³,¹⁴. We found that around 90% of patients with obstructive jaundice in The Netherlands currently undergo preoperative drainage.

Table I. The potential benefits and adverse effects of preoperative biliary drainage in patients with obstructive jaundice.

Potential benefit: - Decreases bilirubin level/improves liver function - Improves nutritional status/immune function - Reduces endotoxaemia and TNF/IL-6 release - Treatment of biliary infection (cholangitis) - Assessment of intraductal extent of tumor (proximal bile duct carcinoma) Adverse effects: - Cholangitis (prolongs hospital stay) - Morbidity of drainage procedure - Secondary infection (postoperative) leading to postoperative sepsis and leakage

In the light of the ongoing controversy regarding preoperative drainage, a meta-analysis of randomized clinical trials and comparative studies was performed ¹⁵. No difference in mortality could be detected between the two strategies, but the overall complication rate in patients who underwent preoperative drainage was significantly higher compared with direct surgical treatment, 57% and 42%, respectively. Unfortunately, most of these studies have methodological flaws. A few studies used external (percutaneous) drainage only (no restoration enterohepatic cycle), different tumors and levels of obstruction (Klatskin tumors and pancreatic tumors) were included, there was a wide range of drainage period (from 10 to 32 days), and different types of operation were compared (bypass vs resection). Therefore, a prospective randomized trial addressing the effects of PBD on patients with obstructive jaundice due to distal obstruction is currently being performed as a multicentre study in The Netherlands ¹⁶. The study design and protocol have been published and 210 patients should be included. An independent safety committee recently analysed the data after inclusion of 105 patients (50%). The treatment in both arms pylorus-preserving pancreaticoduodenectomy (PPPD) (71% vs 64%) and bypass (30% vs 33%) was not different. Mortality and complications rate in both groups did not lead to early closure of the trial and we are awaiting the final results. The outcome of this study will probably have consequences for the time interval for diagnostic work-up, the waiting time for surgery and referral pattern.

The strategy for proximal lesions is even more difficult. Most authors agree that at least extended liver resection should not be performed in severely jaundiced patients and they need drainage of the ‘remnant lobe’. The percutaneous or endoscopic approach to these patients is still a matter of local expertise and might be an important subject for a trial for the future, because there are no data to compare both techniques.

Endoscopic and surgical management of biliary stricture and bile duct injury

Bile duct injury (BDI) after laparoscopic cholecystectomy remains a major problem in current surgical practice. BDI is associated with poor survival, increased morbidity, and impaired quality of life ¹⁷,¹⁸. Timing of treatment and treatment strategy in terms of a surgical reconstruction versus endoscopic and percutaneous drainage and dilatation procedures are still subject to debate ¹⁹,²⁰. Management is of course partly dependent on the type of injury ²¹.

According to the Amsterdam classification, type A and B lesions (leakage of cystic duct and leakage of the bile duct) will primarily be treated by endoscopy. Type D lesions (transection of the bile duct) nearly always need surgical reconstruction ¹⁹,²²,²³,²⁴. Controversy still exists as to the management of type C lesions (strictures) ²⁰,²¹,²²,²³,²⁴,²⁵. Due to the nature of the lesions, the unpredictable diagnostic work-up and initial management in regional hospitals and variable referral pattern, as well as different types of injury, a randomized study might not be expected.

In a prospective cohort study 500 patients referred to our centre were analysed in terms of internal referral pattern and final treatment ²⁶. The referral pattern of BDI patients from the initial hospital to the tertiary center is summarized in Figure 1. The initial referral rate to the Departments of Gastroenterology, Surgery and Radiology was 66%, 29% and 5%, respectively. The referral rate within the tertiary centre, between different departments, ranged from 7% (from gastroenterology to radiology) to 40% (from radiology to surgery). In all, 160 patients (32%) underwent a definitive surgical treatment, whereas endoscopy was the definitive treatment in 264 patients (53%) and a radiologic intervention in 58 patients (12%). Eighteen patients (4%) did not receive additional interventional treatment after referral.

Figure 1. .

Referral pattern of BDI patients within the AMC (adapted from Surgery 2007, 42).

The hospital mortality was 0.4% (n=2), and after a mean follow-up period of 6.7±4.1 years, 42 patients had died (8.4%), a relatively low mortality compared with the study from Flum et al. ¹⁷. In 10 of 42 patients who died, death was related to the biliary injury.

Endoscopic stent therapy as a final treatment was performed in 93 patients with persistent leakage of the bile duct (including leakage of the cystic duct, n=67) and 110 patients underwent stenting for a bile duct stricture ²⁷. The overall long-term success rate (mean follow-up 4.5 years) was 95% for patients with bile leakage and 74% for patients with strictures. The mean duration of stents in situ was 2 months and 11.5 months, respectively (Table II). Independent predictors for outcome were injuries classified as Bismuth III and IV type, endoscopic stenting before referral and the number of stents inserted at the first procedure. We concluded that endoscopic stenting is the treatment of first choice for these lesions. Surgery is indicated after failure (6 months) of stenting.

Table II. Endoscopic stent therapy in bile duct injury patients with biliary leakage and bile duct stricture.

	Bile duct leakage		Bile duct stricture
Parameter	n=93	%	n=110	%
Number of stent changes, median (range)	1 (0–5)		4 (0–12)
Mean duration of stents in situ, months (±SD)	2 (1.8)		11.5 (9.4)
Number of patients with a stent-related complication	12	13	36	33
Referred for surgery	1	1	22	20
Subsequent stenting for recurrence of stenosis	2	2	6	6
Mortality related to BDI	1	1	2	2
Successful endoscopic stenting	88	95	81	74

As mentioned, surgical reconstruction by a hepaticojejunostomy has to be performed for nearly all type D lesions as well as failures of endoscopic management. Recently the long-term outcome of 151 patients (mean follow-up 5.3 years) who underwent reconstruction by a hepaticojejunostomy was analysed ²⁸. The in-hospital mortality was zero, surgical complications were found in 29 patients (19%) and 14 patients (3%) developed a stricture at the anastomosis after a mean follow-up period of 4.5 years.

In summary, reconstructive surgery has excellent short-term as well as long-term outcome for patients with major BDI (transection of the bile duct). Endoscopic stenting is the primary treatment for patients with bile leakage and strictures; surgery is still indicated after failure of endoscopy.

Stenting versus bypass or resection as palliative treatment of pancreatic carcinoma

Unfortunately the majority of patients with pancreatic carcinoma will have palliative treatment and the three most important symptoms that should be treated in advanced pancreatic and periampullary cancer are obstructive jaundice, duodenal obstruction and pain. Biliary drainage can be achieved non-surgically by placement of a biliary stent (endoscopically or percutaneously) or surgically by performing a biliary bypass. The success rate for short-term relief of biliary obstruction is comparable for both surgical and non-surgical drainage procedures and varies between 90 and 100%. Randomized studies (relatively older studies performed between 1988 and 1994) comparing surgical biliary drainage and endoscopic drainage showed that surgical treatment is associated with higher early morbidity, a longer hospital stay and higher mortality, but long-term results are better. Endoscopic treatment is associated with more long-term complications such as cholangitis, clotting of stents and gastic outlet obstruction ²⁹,³⁰,³¹,³²,³³.

In a more recent randomized study Nieveen et al. compared a Wallstent versus a surgical bypass (hepaticojejunostomy and gastroenterotomy) in patients with pathology proven metastasis after a diagnostic laparoscopy ³⁴. Survival and hospital-free survival were longer after surgery compared with stenting (192 days vs 116 days, respectively), but this was a selected group of relatively fit patients. The new development of duodenal stenting for gastric outlet obstruction might change the indication for bypass surgery ³⁵.

Currently palliative resection is also performed in a selected group of patients with limited liver or peritoneal metastasis, and a recent study showed a mean survival of 15 months and acceptable morbidity and mortality ³⁶. These encouraging results of palliative resection might also be due to a selection bias. A well conducted controlled trial of the role of palliative resection is still not available.

In summary, endoscopic treatment will still be indicated in the majority of patients; however, surgical palliation with a bypass procedure or even with a palliative resection might be preferred in a selected group of relatively fit patients.

Endoscopic and surgical drainage in chronic pancreatitis

Management of pain in chronic pancreatitis, particularly chronic pain, that is insufficiently relieved by medication (requiring opiates) remains a therapeutic dilemma.

In patients with ductal obstruction, as well as an inflammatory mass of >4 cm, a combined resection and drainage procedure is generally accepted ³⁷,³⁸. In patients with obstruction and a stricture/stones in the pancreatic duct but without an inflammatory mass in the pancreatic head, endoscopic and surgical drainage have both been used frequently ³⁹.

Reviewing the endoscopic treatment and summarizing the results from the literature from 11 studies with more than 30 patients (n=2319), stenting is associated with 74% complete or partial pain relief (mean follow-up 40 months) and 11% proceeded to surgery.

Surgical drainage by lateral pancreaticojejunostomy summarizing 16 studies with more than 20 patients (n=889) is associated with 80% complete or partial pain relief, after a mean follow-up period of 63 months, with a mortality of 1.1%.

Recently we conducted a randomized trial comparing endoscopic and surgical drainage with respect to the outcome of pain relief, morbidity, quality of life and pancreatic function.

This study showed that patients who underwent surgery, as compared with those who were treated endoscopically, had lower Izbicki pain scores (25 vs 51, p<0.001) and better physical health summary scores on the Medical Outcomes Study 36-Item Short-Form General Health Survey questionnaire (p=0.003). At the end of follow-up, complete or partial pain relief was achieved in 32% of patients assigned to endoscopic drainage as compared with 75% of patients assigned to surgical drainage (p=0.007). Patients receiving endoscopic treatment required more procedures than did patients in the surgery group ⁴⁰. The benefits of surgery were demonstrated by more rapid, effective and sustained pain relief (Figure 2). Even with more aggressive endoscopic management patients will suffer pain during the relatively long treatment period, as shown in our study. This study led to comments about the surgical and endoscopic treatment. We should realize that surgical treatment was limited to drainage of the duct and not the pancreatic head area and the uncinate process. Indeed four randomized studies comparing different and more extended drainage procedures with limited resection or standard PPPD showed further improved results after more extended treatment up to pain relief between 94% and 100% ⁴¹,⁴²,⁴³,⁴⁴ (Table III). However, during the trial we were limited to drainage by pancreaticojejunostomy instead of the more commonly used Frey procedure that we are using routinely after closing the trial.

Figure 2. .

Endoscopic versus surgical PD duct drainage: mean Izbicki pain score during follow-up. (Adapted from NEJM, 2007; 356)

Table III. Prospective randomized studies comparing surgical drainage and resection techniques for patients with chronic pancreatitis.

Authors	Year	No. of patients	Surgical procedures	Follow-up (months)	Results
Klempa et al. 42	1995	43	HR vs PD	36–66	HR: pain relief 100% (vs 70%), better pancreatic function. Equal mortality and morbidity
Büchler et al. 43	1995	40	HR vs PPPD	6	HR: pain relief 94% (vs 67%), better pancreatic function, morbidity 15% (vs 20%)
Izbicki et al. 44	1998	61	Frey vs PPPD	24	Frey: pain relief 94% (vs 95%), in-hospital complications 19% (vs 53%)
Strate et al. 41	2005	74	Frey vs HR	104	Equal pain relief and pancreatic function

HR, pancreatic head resection (Beger); (PP)PD (pylorus-preserving) pancreaticoduodenectomy.

This study was also led to comments about the drawbacks of the endoscopic treatment in the trial, such as the relatively short period of endoscopic stenting ⁴⁵.

Fortunately M. Delhaye and J. Devière, who also performed the lithotripsy in our patients (n=16) recently published another randomized trial ⁴⁶. An even less invasive treatment protocol than ours (ESWL only) showed that ESWL was superior in terms of symptom relief and costs to prolonged and aggressive stenting combined with ESWL. They concluded that adding endoscopy and stenting on top of ESWL adds to the costs of care without improving outcome.

In the light of the results from that study the criticism of our trial should be reconsidered ⁴⁵.

We therefore conclude that surgery in patients with advanced symptomatic chronic pancreatitis is more effective than endoscopic treatment. Pain relief is immediate and consistent. Combined surgical drainage with limited pancreatic head resection might further improve results in the near future.

Conclusions

Summarizing the role of endosopic versus surgical drainage in four different areas of pancreatobiliary disorders it can be concluded that preoperative endoscopic intervention (preoperative biliary drainage) in patients with obstructive jaundice should not be used routinely but only in selected patients. For patients with biliary leakage and bile duct strictures after a laparoscopic cholecystectomy endoscopic stent therapy might be first choice and surgery should be used for failures and is the treatment of choice after transection of the bile duct (the major bile duct injuries).

The majority of patients with obstructive jaundice due to advanced pancreatic cancer will undergo endoscopic drainage, but for relatively fit patients with a prognosis of more than 6 months survival, surgical drainage or even palliative resection might be considered. For patients with persistent pain due to chronic pancreatitis, surgical drainage combined with limited pancreatic head resection might be the first choice for pain relief. Most importantly, management of patients with these bilio-pancreatic disorders should be performed by a multidisciplinary HPB approach and teamwork consisting of gastroenterologist, radiologists and surgeons.