Prophylactic pancreatic duct stenting: a panacea?

Abstract

Fazel A, Quadri A, Catalano MF, Meyerson SM, Geenen JE (St. Luke’s Hospital, Milwaukee, Wisconsin). Does a pancreatic duct stent prevent post-ERCP pancreatitis? A prospective randomized study. Gastrointest Endosc 2003;57:291–294.Acute pancreatitis is the most common complication of endoscopic retrograde cholangiopancreatography (ERCP). Although there exists a risk of pancreatitis regardless of the underlying pathology or indication for the procedure, patients with sphincter of Oddi dysfunction (SOD) and those with papilla that are difficult to cannulate are at exceptionally high risk. Such patients are also more likely than others to develop severe pancreatitis. Fazel et al. report their experience with pancreatic duct stenting to reduce the risk of post-ERCP pancreatitis in high-risk patients.Over a 6-year period, patients with difficult cannulation at ERCP, or who had undergone sphincter of Oddi manometry (SOM) or endoscopic sphincterotomy were prospectively randomized to either short-term placement of a pancreatic duct stent or no stent on completion of ERCP. Eight percent of the patients had sphincterotomy alone as the risk factor. Exclusion criteria included acute or chronic pancreatitis at the time of the procedure and failure of cannulation of the pancreatic duct.Seventy-six patients at high risk for post-ERCP pancreatitis were enrolled and randomized. Two patients in the stent group were excluded because cannulation of the pancreatic duct deep enough to allow stent placement was not achieved. Thirty-six patients underwent stent placement and 38 patients were assigned to the control group without a pancreatic duct stent. Post-ERCP pancreatitis occurred in 2 (5%) of the patients in the stent group and 10 (28%) of the patients in the control group (P < 0.05). Mean duration of hospitalization was also shorter in the stent group at 2.5 days versus 5.5 days (P < 0.20). Both cases of pancreatitis in the stent group were mild, whereas half of the cases of pancreatitis in the control group were moderate or severe.The investigators conclude that a pancreatic duct stent reduces the frequency and severity of post-ERCP pancreatitis and recommend a pancreatic duct stent be placed after ERCP in patients at high risk for procedure-related pancreatitis.CommentThe end result of pancreatitis regardless of cause including post-ERCP pancreatitis involves the activation of trypsinogen to trypsin. Trypsin in turn activates other digestive enzymes including chymotrypsinogen, proelastase, prolipase, procarboxypeptidase, and prophospholipase. The inciting factor(s) that trigger this cascade in the post-ERCP setting are unclear.Several risk factors for post-ERCP pancreatitis have been identified. In a prospective, multicenter study (Gastrointest Endosc 2001:54:425–434), previous episodes of ERCP-induced pancreatitis, suspected SOD, female gender, difficult cannulation, pancreatic sphincterotomy, normal serum bilirubin, and opacification of the pancreatic duct all were identified as significant risks with adjusted odds ratios ranging from 2.6 to 5.4. These data suggest that patient-specific characteristics are important in the pathogenesis of pancreatitis after ERCP. Pancreatic duct hypertension probably also plays a role in the pathogenesis of pancreatitis after ERCP given the increased risk associated with intraprocedural factors such as difficult cannulation or overinjection. Further analysis of the data in the same study revealed a cumulative increase in risk when multiple factors coexisted.There are no data available regarding the optimal timing of a pharmacologic intervention directed at preventing progression of the autodigestion cascade in this setting. The ideal drug for prophylaxis of post-ERCP pancreatitis would be safe, inexpensive, administered as a single dose shortly before the ERCP, and efficacious when administered to patients believed to be at high risk. Various drugs have been tried in differing doses and regimens, including somatostatin, octreotide, steroids, interleukin 10, gabexate mesylate, heparin, allopurinol, and nifedipine. The results of these studies are either disappointing, inconclusive, or discordant (Gastrointest Endosc 2002;56:488–495, Gastrointest Endosc 2000;51:1–7, N Engl J Med 1996;335:919–923). At this time, there is no pharmacologic agent that is available for use in clinical practice for the prevention of post-ERCP pancreatitis.Endoscopic prophylaxis involves the placement of a pancreatic stent prior to terminating the ERCP procedure or during needle knife sphincterotomy. The presumed mechanism is decompression of the pancreatic duct. There are no data regarding the optimal timing of stent placement—is it advantageous to place the stent early in the procedure as opposed to the end of the procedure? When exactly does the autodigestion cascade commence, and is there a small window of opportunity for such interventions? If efficacious, endoscopic prophylaxis has its advantages. Even in patients without underlying inherent risk factors such as SOD, intraprocedural events help stratify a patient as high risk for post-ERCP pancreatitis. These events include difficult cannulation, multiple pancreatic duct injections, and acinarization. In contrast to pharmacotherapy, which requires initiation before the ERCP, pancreatic duct stent placement can be selectively used in these patients.Several studies are available suggesting efficacy of pancreatic stent placement to prevent the occurrence of post-ERCP pancreatitis in selected populations. In 1993, the Indiana University group reported a study of 98 patients with SOD, common bile duct diameter less than 10 mm, and patients requiring precut sphincterotomy, who were randomized to either placement or nonplacement of a main pancreatic duct stent. Although not statistically significant, a trend towards more severe pancreatitis and longer hospital stay was noted in the no-stent group (Gastrointest Endosc 1993;39:652–657). In another study, 80 patients with pancreatic sphincter hypertension were randomized to pancreatic duct stent or no stent after biliary sphincterotomy. The risk of pancreatitis decreased from 26% to 7% with stenting (Gastrenterology 1998;115:1518–1524). The Indiana group subsequently reported a retrospective review of patients with SOD who underwent pancreaticobiliary sphincterotomy with pancreatic stent placement compared with biliary sphincterotomy alone. The risk of pancreatitis was 10.7% with the use of a needle knife over a pancreatic stent, 19.2% in patients with a pull-type sphincterotomy plus pancreatic stent, and 28.3% in the biliary sphincterotomy alone group (Endoscopy 2002;34:280–285). Although the studies are not uniform with regard to patient selection and risk factors, the data supporting use of a prophylactic stent are stronger than for pharmacologic intervention.What are the pitfalls of placing or attempting to place a pancreatic stent? First, pancreatic stent placement is not always successful. Repeated unsuccessful attempts may result in further trauma and papillary edema. Second, stent-induced changes in the pancreatic duct are well described. Following long-term stent placement, iatrogenic strictures may occur, and in a small number of individuals result in chronic pancreatitis. The incidence of pancreatic duct strictures after short-term stenting is unknown, but generally is believed to be low. Using stents made of more pliable and softer material or using a stent without internal flaps may mitigate this further. Third, the stent may not pass spontaneously. As in this study, if a radiograph reveals the stent to be present after a week, the patient needs a repeat endoscopic procedure to extract the stent. However, assuming that a protective effect is truly present, the costs of a repeat procedure to remove the stent are insignificant compared with the expense associated with management of pancreatitis (Gastroenterology 1998;115:1518–1524).The current study by Fazel et al. includes a broader range of patients than in the studies discussed above. However, 63% of the patients in the no-stent group and 74% of patients in the stent group had undergone sphincter of Oddi manometry. The risk of post ERCP pancreatitis after biliary sphincterotomy alone is uncertain (Gastrointest Endosc 2002;55:50–54), and the appropriateness of including such patients in the study is debatable. The study population was recruited over 6 years, raising the question whether all consecutive patients in whom the risk factors occurred were included. The study would have been more rigorous if patients were blinded to the placement of the pancreatic duct stent for at least the first 24 hours, the time interval when post-ERCP pancreatitis typically occurs. The number of patients needing a second procedure for stent removal was not mentioned, and, for broader extrapolation to the endoscopic community, it would have been helpful to analyze individuals with difficult cannulation separately. Given the predominance of patients with sphincter of Oddi manometry in the study group, the investigators have presented data that are consistent with the previously discussed observations.In conclusion, there is currently no pharmacologic prophylaxis available for post-ERCP pancreatitis. I believe that short-term placement of a prophylactic pancreatic duct stent in patients with SOD and after sphincter of Oddi manometry prevents or reduces the severity of post-ERCP pancreatitis. The stent should be small in caliber, possibly without internal flaps, and removed within a few days if it does not pass spontaneously. On the other hand, in patients with other risk factors such as difficult cannulation, there is insufficient data to justify routine placement of a prophylactic pancreatic duct stent—and the debate should continue. Fazel A, Quadri A, Catalano MF, Meyerson SM, Geenen JE (St. Luke’s Hospital, Milwaukee, Wisconsin). Does a pancreatic duct stent prevent post-ERCP pancreatitis? A prospective randomized study. Gastrointest Endosc 2003;57:291–294. Acute pancreatitis is the most common complication of endoscopic retrograde cholangiopancreatography (ERCP). Although there exists a risk of pancreatitis regardless of the underlying pathology or indication for the procedure, patients with sphincter of Oddi dysfunction (SOD) and those with papilla that are difficult to cannulate are at exceptionally high risk. Such patients are also more likely than others to develop severe pancreatitis. Fazel et al. report their experience with pancreatic duct stenting to reduce the risk of post-ERCP pancreatitis in high-risk patients. Over a 6-year period, patients with difficult cannulation at ERCP, or who had undergone sphincter of Oddi manometry (SOM) or endoscopic sphincterotomy were prospectively randomized to either short-term placement of a pancreatic duct stent or no stent on completion of ERCP. Eight percent of the patients had sphincterotomy alone as the risk factor. Exclusion criteria included acute or chronic pancreatitis at the time of the procedure and failure of cannulation of the pancreatic duct. Seventy-six patients at high risk for post-ERCP pancreatitis were enrolled and randomized. Two patients in the stent group were excluded because cannulation of the pancreatic duct deep enough to allow stent placement was not achieved. Thirty-six patients underwent stent placement and 38 patients were assigned to the control group without a pancreatic duct stent. Post-ERCP pancreatitis occurred in 2 (5%) of the patients in the stent group and 10 (28%) of the patients in the control group (P < 0.05). Mean duration of hospitalization was also shorter in the stent group at 2.5 days versus 5.5 days (P < 0.20). Both cases of pancreatitis in the stent group were mild, whereas half of the cases of pancreatitis in the control group were moderate or severe. The investigators conclude that a pancreatic duct stent reduces the frequency and severity of post-ERCP pancreatitis and recommend a pancreatic duct stent be placed after ERCP in patients at high risk for procedure-related pancreatitis. CommentThe end result of pancreatitis regardless of cause including post-ERCP pancreatitis involves the activation of trypsinogen to trypsin. Trypsin in turn activates other digestive enzymes including chymotrypsinogen, proelastase, prolipase, procarboxypeptidase, and prophospholipase. The inciting factor(s) that trigger this cascade in the post-ERCP setting are unclear.Several risk factors for post-ERCP pancreatitis have been identified. In a prospective, multicenter study (Gastrointest Endosc 2001:54:425–434), previous episodes of ERCP-induced pancreatitis, suspected SOD, female gender, difficult cannulation, pancreatic sphincterotomy, normal serum bilirubin, and opacification of the pancreatic duct all were identified as significant risks with adjusted odds ratios ranging from 2.6 to 5.4. These data suggest that patient-specific characteristics are important in the pathogenesis of pancreatitis after ERCP. Pancreatic duct hypertension probably also plays a role in the pathogenesis of pancreatitis after ERCP given the increased risk associated with intraprocedural factors such as difficult cannulation or overinjection. Further analysis of the data in the same study revealed a cumulative increase in risk when multiple factors coexisted.There are no data available regarding the optimal timing of a pharmacologic intervention directed at preventing progression of the autodigestion cascade in this setting. The ideal drug for prophylaxis of post-ERCP pancreatitis would be safe, inexpensive, administered as a single dose shortly before the ERCP, and efficacious when administered to patients believed to be at high risk. Various drugs have been tried in differing doses and regimens, including somatostatin, octreotide, steroids, interleukin 10, gabexate mesylate, heparin, allopurinol, and nifedipine. The results of these studies are either disappointing, inconclusive, or discordant (Gastrointest Endosc 2002;56:488–495, Gastrointest Endosc 2000;51:1–7, N Engl J Med 1996;335:919–923). At this time, there is no pharmacologic agent that is available for use in clinical practice for the prevention of post-ERCP pancreatitis.Endoscopic prophylaxis involves the placement of a pancreatic stent prior to terminating the ERCP procedure or during needle knife sphincterotomy. The presumed mechanism is decompression of the pancreatic duct. There are no data regarding the optimal timing of stent placement—is it advantageous to place the stent early in the procedure as opposed to the end of the procedure? When exactly does the autodigestion cascade commence, and is there a small window of opportunity for such interventions? If efficacious, endoscopic prophylaxis has its advantages. Even in patients without underlying inherent risk factors such as SOD, intraprocedural events help stratify a patient as high risk for post-ERCP pancreatitis. These events include difficult cannulation, multiple pancreatic duct injections, and acinarization. In contrast to pharmacotherapy, which requires initiation before the ERCP, pancreatic duct stent placement can be selectively used in these patients.Several studies are available suggesting efficacy of pancreatic stent placement to prevent the occurrence of post-ERCP pancreatitis in selected populations. In 1993, the Indiana University group reported a study of 98 patients with SOD, common bile duct diameter less than 10 mm, and patients requiring precut sphincterotomy, who were randomized to either placement or nonplacement of a main pancreatic duct stent. Although not statistically significant, a trend towards more severe pancreatitis and longer hospital stay was noted in the no-stent group (Gastrointest Endosc 1993;39:652–657). In another study, 80 patients with pancreatic sphincter hypertension were randomized to pancreatic duct stent or no stent after biliary sphincterotomy. The risk of pancreatitis decreased from 26% to 7% with stenting (Gastrenterology 1998;115:1518–1524). The Indiana group subsequently reported a retrospective review of patients with SOD who underwent pancreaticobiliary sphincterotomy with pancreatic stent placement compared with biliary sphincterotomy alone. The risk of pancreatitis was 10.7% with the use of a needle knife over a pancreatic stent, 19.2% in patients with a pull-type sphincterotomy plus pancreatic stent, and 28.3% in the biliary sphincterotomy alone group (Endoscopy 2002;34:280–285). Although the studies are not uniform with regard to patient selection and risk factors, the data supporting use of a prophylactic stent are stronger than for pharmacologic intervention.What are the pitfalls of placing or attempting to place a pancreatic stent? First, pancreatic stent placement is not always successful. Repeated unsuccessful attempts may result in further trauma and papillary edema. Second, stent-induced changes in the pancreatic duct are well described. Following long-term stent placement, iatrogenic strictures may occur, and in a small number of individuals result in chronic pancreatitis. The incidence of pancreatic duct strictures after short-term stenting is unknown, but generally is believed to be low. Using stents made of more pliable and softer material or using a stent without internal flaps may mitigate this further. Third, the stent may not pass spontaneously. As in this study, if a radiograph reveals the stent to be present after a week, the patient needs a repeat endoscopic procedure to extract the stent. However, assuming that a protective effect is truly present, the costs of a repeat procedure to remove the stent are insignificant compared with the expense associated with management of pancreatitis (Gastroenterology 1998;115:1518–1524).The current study by Fazel et al. includes a broader range of patients than in the studies discussed above. However, 63% of the patients in the no-stent group and 74% of patients in the stent group had undergone sphincter of Oddi manometry. The risk of post ERCP pancreatitis after biliary sphincterotomy alone is uncertain (Gastrointest Endosc 2002;55:50–54), and the appropriateness of including such patients in the study is debatable. The study population was recruited over 6 years, raising the question whether all consecutive patients in whom the risk factors occurred were included. The study would have been more rigorous if patients were blinded to the placement of the pancreatic duct stent for at least the first 24 hours, the time interval when post-ERCP pancreatitis typically occurs. The number of patients needing a second procedure for stent removal was not mentioned, and, for broader extrapolation to the endoscopic community, it would have been helpful to analyze individuals with difficult cannulation separately. Given the predominance of patients with sphincter of Oddi manometry in the study group, the investigators have presented data that are consistent with the previously discussed observations.In conclusion, there is currently no pharmacologic prophylaxis available for post-ERCP pancreatitis. I believe that short-term placement of a prophylactic pancreatic duct stent in patients with SOD and after sphincter of Oddi manometry prevents or reduces the severity of post-ERCP pancreatitis. The stent should be small in caliber, possibly without internal flaps, and removed within a few days if it does not pass spontaneously. On the other hand, in patients with other risk factors such as difficult cannulation, there is insufficient data to justify routine placement of a prophylactic pancreatic duct stent—and the debate should continue. The end result of pancreatitis regardless of cause including post-ERCP pancreatitis involves the activation of trypsinogen to trypsin. Trypsin in turn activates other digestive enzymes including chymotrypsinogen, proelastase, prolipase, procarboxypeptidase, and prophospholipase. The inciting factor(s) that trigger this cascade in the post-ERCP setting are unclear. Several risk factors for post-ERCP pancreatitis have been identified. In a prospective, multicenter study (Gastrointest Endosc 2001:54:425–434), previous episodes of ERCP-induced pancreatitis, suspected SOD, female gender, difficult cannulation, pancreatic sphincterotomy, normal serum bilirubin, and opacification of the pancreatic duct all were identified as significant risks with adjusted odds ratios ranging from 2.6 to 5.4. These data suggest that patient-specific characteristics are important in the pathogenesis of pancreatitis after ERCP. Pancreatic duct hypertension probably also plays a role in the pathogenesis of pancreatitis after ERCP given the increased risk associated with intraprocedural factors such as difficult cannulation or overinjection. Further analysis of the data in the same study revealed a cumulative increase in risk when multiple factors coexisted. There are no data available regarding the optimal timing of a pharmacologic intervention directed at preventing progression of the autodigestion cascade in this setting. The ideal drug for prophylaxis of post-ERCP pancreatitis would be safe, inexpensive, administered as a single dose shortly before the ERCP, and efficacious when administered to patients believed to be at high risk. Various drugs have been tried in differing doses and regimens, including somatostatin, octreotide, steroids, interleukin 10, gabexate mesylate, heparin, allopurinol, and nifedipine. The results of these studies are either disappointing, inconclusive, or discordant (Gastrointest Endosc 2002;56:488–495, Gastrointest Endosc 2000;51:1–7, N Engl J Med 1996;335:919–923). At this time, there is no pharmacologic agent that is available for use in clinical practice for the prevention of post-ERCP pancreatitis. Endoscopic prophylaxis involves the placement of a pancreatic stent prior to terminating the ERCP procedure or during needle knife sphincterotomy. The presumed mechanism is decompression of the pancreatic duct. There are no data regarding the optimal timing of stent placement—is it advantageous to place the stent early in the procedure as opposed to the end of the procedure? When exactly does the autodigestion cascade commence, and is there a small window of opportunity for such interventions? If efficacious, endoscopic prophylaxis has its advantages. Even in patients without underlying inherent risk factors such as SOD, intraprocedural events help stratify a patient as high risk for post-ERCP pancreatitis. These events include difficult cannulation, multiple pancreatic duct injections, and acinarization. In contrast to pharmacotherapy, which requires initiation before the ERCP, pancreatic duct stent placement can be selectively used in these patients. Several studies are available suggesting efficacy of pancreatic stent placement to prevent the occurrence of post-ERCP pancreatitis in selected populations. In 1993, the Indiana University group reported a study of 98 patients with SOD, common bile duct diameter less than 10 mm, and patients requiring precut sphincterotomy, who were randomized to either placement or nonplacement of a main pancreatic duct stent. Although not statistically significant, a trend towards more severe pancreatitis and longer hospital stay was noted in the no-stent group (Gastrointest Endosc 1993;39:652–657). In another study, 80 patients with pancreatic sphincter hypertension were randomized to pancreatic duct stent or no stent after biliary sphincterotomy. The risk of pancreatitis decreased from 26% to 7% with stenting (Gastrenterology 1998;115:1518–1524). The Indiana group subsequently reported a retrospective review of patients with SOD who underwent pancreaticobiliary sphincterotomy with pancreatic stent placement compared with biliary sphincterotomy alone. The risk of pancreatitis was 10.7% with the use of a needle knife over a pancreatic stent, 19.2% in patients with a pull-type sphincterotomy plus pancreatic stent, and 28.3% in the biliary sphincterotomy alone group (Endoscopy 2002;34:280–285). Although the studies are not uniform with regard to patient selection and risk factors, the data supporting use of a prophylactic stent are stronger than for pharmacologic intervention. What are the pitfalls of placing or attempting to place a pancreatic stent? First, pancreatic stent placement is not always successful. Repeated unsuccessful attempts may result in further trauma and papillary edema. Second, stent-induced changes in the pancreatic duct are well described. Following long-term stent placement, iatrogenic strictures may occur, and in a small number of individuals result in chronic pancreatitis. The incidence of pancreatic duct strictures after short-term stenting is unknown, but generally is believed to be low. Using stents made of more pliable and softer material or using a stent without internal flaps may mitigate this further. Third, the stent may not pass spontaneously. As in this study, if a radiograph reveals the stent to be present after a week, the patient needs a repeat endoscopic procedure to extract the stent. However, assuming that a protective effect is truly present, the costs of a repeat procedure to remove the stent are insignificant compared with the expense associated with management of pancreatitis (Gastroenterology 1998;115:1518–1524). The current study by Fazel et al. includes a broader range of patients than in the studies discussed above. However, 63% of the patients in the no-stent group and 74% of patients in the stent group had undergone sphincter of Oddi manometry. The risk of post ERCP pancreatitis after biliary sphincterotomy alone is uncertain (Gastrointest Endosc 2002;55:50–54), and the appropriateness of including such patients in the study is debatable. The study population was recruited over 6 years, raising the question whether all consecutive patients in whom the risk factors occurred were included. The study would have been more rigorous if patients were blinded to the placement of the pancreatic duct stent for at least the first 24 hours, the time interval when post-ERCP pancreatitis typically occurs. The number of patients needing a second procedure for stent removal was not mentioned, and, for broader extrapolation to the endoscopic community, it would have been helpful to analyze individuals with difficult cannulation separately. Given the predominance of patients with sphincter of Oddi manometry in the study group, the investigators have presented data that are consistent with the previously discussed observations. In conclusion, there is currently no pharmacologic prophylaxis available for post-ERCP pancreatitis. I believe that short-term placement of a prophylactic pancreatic duct stent in patients with SOD and after sphincter of Oddi manometry prevents or reduces the severity of post-ERCP pancreatitis. The stent should be small in caliber, possibly without internal flaps, and removed within a few days if it does not pass spontaneously. On the other hand, in patients with other risk factors such as difficult cannulation, there is insufficient data to justify routine placement of a prophylactic pancreatic duct stent—and the debate should continue.