Is ERCP a Safe Procedure, but for Experts Only?

Dear Sir,

the demonstration that a complication can occur at a determined rate is certainly of interest for our everyday endoscopic practice and also for medicolegal reasons.

The knowledge of risk factors can help us to lower the incidence of complications in our current practice. Thus there is much interest today in the complications rate and risk factors associated with endoscopic retrograde cholangiopancreatography (ERCP) and a paper, such as that of Enns et al., is welcome [1].

In recent years, ERCP has increasingly become a complex therapeutic procedure; in fact nowadays ERCP alone is rarely used to make a diagnosis, given the safety and efficacy of competing imaging technologies. Both the number of ERCPs and the indications for the procedure have significantly changed in the past few years. The lack of diagnostic and “easy” ERCPs makes it difficult to teach the procedure today, and it is obvious that performance of a sufficient volume of cases is extremely important for ensuring an acceptable outcome [2] [3] [4]. The increasing competition from other diagnostic or therapeutic procedures emphasizes the need to maintain an optimal outcome level for ERCPs. If this is the future for ERCP, it would be better to have data about the best potential use of this difficult endoscopic procedure, i. e., not multicentric assessment from endoscopic units with different levels of experience, but data provided by single tertiary referral centres [5] [6] [7]. Enns et al. report the experience at Duke University, where 9314 ERCPs were performed over about 12 years by a team of three expert endoscopists. This paper is very reassuring with regard to perforative complications as a perforation rate of 0.35 % is reported with only seven patients requiring surgery. I have read this paper with great attention and would like to make some comments.

As the authors stated, two types of perforation can complicate ERCP: visceral, which can occur reaching the papilla, and retroperitoneal. Esophageal perforation has been rarely reported during ERCP. Subcutaneous emphysema has also been reported in a Billroth II patient which was not related to esophageal perforation, and it is interesting that the esophageal perforation was not found at surgery in the patient described by Enns et al.

Gastric and duodenal perforation are severe complications which always require surgery. Enns et al. report two cases of gastric perforation, occurring only in patients who had undergone Billroth II gastrectomy and never in patients with normal anatomy. Billroth II gastrectomy is a risk factor for perforation when reaching the papilla, with an incidence which is very high in some series; can Enns et al. report the incidence of this complication for their subgroup of Billroth II patients? Duodenal perforation is reported by Enns et al. only in two patients with abnormal conditions (duodenal stricture and periampullary diverticulum) and never in patients with normal duodenal anatomy. In my recent experience of 2193 ERCPs, performed in the last 7 years, there have been three (0.13 %) visceral perforations reaching the papilla which occurred only in patients with abnormal anatomy: two were Billroth II patients and one had a duodenal neoplastic stenosis. Thus these patients had the same risk factors as those in the series of Enns et al. In the multicentre study of Loperfido et al., which also include small, low-throughput centres, a greater visceral perforation rate of 0.57 % (16 among 2769 patients) is reported, with four related deaths [6].

Retroperitoneal perforations can complicate ERCP, as Enns et al. stated, but with less clinical impact than visceral ones. Only three patients required surgery and no mortalities we re observed. Near half of the complications observed were related to sphincterotomy and half to the use of the guide wire. Two patients in the group with sphincterotomy-related perforation required surgery, one for hemodynamic instability, possibly for associated severe hemorrhage, and the other on post-ERCP day 3 for fever and pain, possibly because of a retroperitoneal abscess. Information is lacking about the number of and the indications for endoscopic sphincterotomies. The second group, of 12 perforations, was related to the use of the guide wire. More information would be useful, concerning for instance the guide wire used, how the authors diagnosed guide wire perforation after ERCP, and whether perforation also occurred when a soft-tipped guide wire was used.

Finally it is not clear what constituted the indication for surgery following retroperitoneal perforation if no delayed complication or abscess was shown. In two multicentre studies with low-throughput centres, retroperitoneal perforation was reported in 0.6 % of therapeutic ERCPs by Loperfido et al. [6] with six patients referred for surgery and one related death, and in 0.57 % by Masci et al. [7], with five patients referred for surgery and no deaths. These two studies do not report perforation related to the use of a guide wire. No perforation related to the use of a guide wire is seen in may own series, also: all four retroperitoneal perforations observed (0.17 %) were related to needle pre-cutting, and among these only one required surgery, because of retroperitoneal abscess. Freeman et al. report eight perforations (0.34 %), including visceral and retroperitoneal, with three patients referred for surgery, in a multicentre study of 2347 ERCPs [5]. Finally, why did the authors select a group with normal ERCPs in order to evaluate the role of some risk factors for complication when normal ERCPs do not usually require sphincterotomy or guide wires?

For the endoscopic community, the main message of the paper by Enns et al. may then be: ERCP is a safe procedure, but in expert hands. The fundamental dilemma concerns who should perform ERCP [8] [9] [10] [11] [12]? Following some large multicenter studies [5] [6] [7], I feel there is currently a need for papers from single referral centers in order to describe the optimal outcome of any endoscopic procedure in this era of managed and competitive health care. It is important to known how safe and efficacious endoscopic procedures can be when performed at maximum capability. Such data would obviously allow selection between endoscopic centres and an improvement in widely used endoscopic techniques. Nowadays it is essential to be able to offer highly successful and safe ERCP which can compete with alternative therapeutic procedures.

References

• 1 Enns R, Eloubeidi M A, Mergener K. et al . ERCP-related perforations: risk factors and management.  Endoscopy. 2002;  34 293-298
• 2 Siegel J, Cohen S A, Kasmin F E. Experience and volume: the ingredients for successful therapeutic endoscopic outcome especially ERCP in postgastrectomy patients.  Am J Gastroenterol. 2000;  95 2133-2134
• 3 Mosca S. Which scope and which technique will enhance the success rate in ERCP for the Billroth II patients?.  Am J Gastroenterol. 2000;  95 553-555
• 4 Jimenez-Perez P J, Pastor G, Aznares R. et al . Decrease of diagnostic ERCP: implication on ERCP training programmes (abstract).  Gastrointest Endosc. 2001;  53 AB79
• 5 Freeman M, Nelson D B, Sherman S. et al . Complications of endoscopic biliary sphincterotomy.  N Engl J Med. 1996;  335 909-918
• 6 Loperfido S, Angelini G, Benedetti G. et al . Major early complications from diagnostic and therapeutic ERCP. A prospective multicenter study.  Gastrointest Endosc. 1998;  48 1-10
• 7 Masci E, Toti G, Mariani A. et al . Complications of diagnostic and therapeutic ERCP: a prospective multicenter study.  Am J Gastroenterol. 2001;  96 417-423
• 8 Mosca S. How can we reduce the complication rate and enhance the success rate in Billroth II patients during endoscopic retrograde cholangiopancreatography?.  Endoscopy. 2000;  32 589-590
• 9 Costamagna G. ERCP and endoscopic sphincteromy in Billroth II patients: a demanding technique for experts only?.  Ital J Gastroenterol. 1998;  30 306-309
• 10 Mosca S. Is ERCP a procedure for all, the majority, or just a few endoscopists? A dilemma.  Gastrointest Endosc. 2001;  54 140-142
• 11 Baillie J. ERCP for all.  Gastrointest Endosc. 1995;  42 373-376
• 12 Mosca S. What sort of endoscopist for the endoscopy of the future?.  Endoscopy. 2002;  34 742-743