There are many reports of the outcomes and adverse effects of EPLBD. Fortunately, most AEs are mild, and few serious AEs have been reported. Nevertheless, this does not mean that EPLBD is completely safe in patients with dilated CBD, considering the lack of large-scale multicenter clinical studies and possible reluctance of clinicians to report serious AEs. The outcomes and AEs of EPLBD are summarized in Tables 1, 2.1,3-19

During the initial trials of this procedure, the occurrence of pancreatitis was of much concern because a large balloon was used. In practice, however, no case of severe pancreatitis has been reported.15 One possible reason for this lack of pancreatitis occurrence is that the incision of the papillary orifice may shift the expansive force more toward the CBD than toward the pancreas orifice. In addition, the incidence and severity of pancreatitis were not related to the duration of ballooning. EPLBD without a preceding papillary incision did not cause severe pancreatitis.16 Therefore, ballooning itself may not be the culprit. Considering EPLBD, the cause of fatal pancreatitis during endoscopic papillary balloon dilation (EPBD) should be reconsidered. Instead of balloon compression of the pancreatic flow, edema or spasm caused by irritation of the pancreatic orifice while performing difficult selective cannulation and struggling to remove the stone may be the main determinants of severe pancreatitis during EPLBD and EPBD.

There is no need to worry about bleeding during partial EST. Even if a small amount of bleeding does occur, it will be controlled by the subsequent balloon inflation. The advantage of a small incision in comparison with a full EST is the reduced possibility of cutting a major vessel in the papillary roof. However, after the EPLBD, there is a possibility of delayed bleeding with a preceding large incision of the papilla. Delayed bleeding may occur if the procedure is terminated without confirming complete hemostasis when there is bleeding during the procedure, or if the balloon compresses the injured vessel for an insufficient time. EPLBD without a preceding small EST is feasible for patients with large bile duct stones who are at risk of bleeding. In practice, uremic patients and those with advanced liver cirrhosis with large CBD stones can be treated successfully with EPLBD without EST. Minimal oozing after balloon inflation can occur with microvascular rupture accompanied by stretching of the mucosa. However, this does not often cause significant bleeding similar to that caused by a vascular cutting injury after EST.

Is EPLBD always safe? EPLBD is already widely accepted in Western countries. We feel responsible for establishing guidelines regarding the indications for EPLBD and the technique. Our main concern is to identify invisible strictures to avoid a fatal AE. To answer this question, we conducted a large retrospective multicenter study. We enrolled 946 consecutive patients with large CBD stones. The patients were divided into two groups: those with and without AEs. Of the 946 patients, only 95 (10%) experienced AEs.

Among the patients who suffered AEs, most cases of bleeding and pancreatitis were mild to moderate, with the exception of one case of bleeding. By contrast, perforation was classified as moderate to severe. We experienced five severe cases in total, four of which were fatal: one from bleeding and three from perforations. Most of the AEs were classified as mild, while those classified as severe were mostly fatal.

In this study, perforation was the most serious AE of EPLBD. The multivariate analysis indicated that a distal CBD stricture was an independent predictor of perforation, and we suggest that the presence of a distal CBD stricture should be considered as an absolute contraindication to EPLBD. Looking more closely at the causes of mortality, over-inflation and rapid inflation in patients with an undetermined CBD stricture caused perforation. In two cases, a notch remained during dilation but the balloon was inflated further, which caused severe perforation. In one case, the balloon was dilated rapidly without checking for resistance, which caused perforation. Of the three patients with perforations, two died of septic shock and multi-organ failure and the other of cardiogenic shock. In one patient, the full EST and subsequent thrombocytopenia caused delayed massive bleeding. Therefore, from these experiences, we believe that the accompanying guidelines should be followed strictly.

The advantage of EPLBD is that it avoids the complications of full EST. Partial EST is preferred over full EST because it avoids large-vessel injury and perforation. Indeed, we experienced one death due to delayed massive bleeding after full EST. In general, any intra-procedure bleeding was controlled easily using balloon tamponade at the EST site. Delayed bleeding may still occur in such patients whose large vessel located at the proximal part of the papillary roof is injured and is controlled only transiently using balloon tamponade at the incision site. Partial EST itself did not increase the risk of bleeding compared to non-EST, which might be explained by decreased likelihood of severing a major vessel in the papillary roof.

Another advantage of partial EST is direct observation of the remaining intact papillary roof during gradual balloon inflation, which can help to avoid perforation.

Another option is EPLBD without a preceding EST. This can be used in patients with severe bleeding tendencies. An EPLBD study comparing groups with preceding partial EST and no EST should be performed to determine the efficacy and safety of EPLBD without EST.

While visible strictures are easily detectable, our main concern is to identify invisible strictures, which are detectable only during the procedure. Any marked resistance during balloon inflation is an absolute contraindication for continued inflation (Fig. 2A). To determine whether there is an invisible stricture, the balloon should not be inflated beyond 75% of the manufacturer's recommended maximum inflation capacity if resistance continues (Fig. 2B). EPLBD led to death of one of our patients in whom no obvious distal CBD stricture was identified before balloon inflation and the balloon waist did not disappear after inflating the balloon to 75% of the maximum recommended pressure. Although 75% was determined arbitrarily, more importantly, the remaining waist should not be inflated forcibly. In addition, the balloon should always be inflated gradually and never exceed the CBD diameter. When a tapered distal CBD or invisible stricture is identified, the operator should pay particular attention during balloon inflation to avoid fatal AEs due to a large perforation. The leakage of bile and pancreatic fluid through a large perforation and delay in the proper surgical intervention can lead to rapid deterioration in the patient.