Ultrasound-guided extracorporeal shock wave lithotripsy of pancreatic ductal stones: six years' experience.

Extracorporeal shock wave lithotripsy (ESWL) and endoscopic sphincterotomy (EST) was performed in 35 patients suffering from pancreatic duct stones. Calculi disintegration and resolution of obstruction were achieved in all cases. Completely stone-free ducts were achieved in 16 patients (46%) while some peripheral asymptomatic stone material remained in 19 (54%). Dilation of the main pancreatic duct was reduced in 29 patients (83%). Twelve patients (34%) became completely asymptomatic and 17 (49%) reported a marked reduction of pain. Pancreatogenic steatorrhea ceased and 18 patients (51%) gained weight. Pathological glucose tolerance returned to normal in one patient. No major complications were observed. The combination of ESWL and EST is a successful, nonoperative, new treatment in pancreatic stone disease.


PATIENTS
Thirty-five patients (17 men, 18 women) suffering from chronic pancreatitis complicated by an obstruction of the pancreatic duct system resulting from pancreaticolithiasis who were treated by ESWL were included in this six-year study. Average age of patients was 48 years (range 14 to 61). Nineteen patients (54%) suffered from chronic alcoholic pancreatitis; in 15 patients (43%) the etiology of the pancreatitis was not evident and in the case of a 13-year-old, idiopathic juvenile pancreatitis was assumed. All patients complained of upper abdominal pain, mostly with radiation to the back, which was classified as recurrent pain attacks in 24 (69%) and as continued pain in 11 (31%). Average history of symptoms was five years (range three to 12). Nine patients (26%) had solitary stones, 16 (46%) had one to five stones and 10 (29%) had multiple stones, in some cases completely filling Wirsung's duct (Figures 3,4a). Mean average diameter of the largest stone was 11 mm (range 5 to 25). In conventional abdominal ultrasound the mean average diameter of the dilated pancreatic duct was 9 mm (range 5 to 28).
Each patient underwent endoscopic retrograde pancreatography (ERP), which showed moderate to marked chronic inflammatory ductal changes. Significant strictures at the main pancreatic duct were found in 15 patients (43%). In 30 patients (86%) ERP confirmed pancreatic duct system dilation as observed by ultrasound; in five patients (14%) stone impaction prevented contrasting of the distal part of the duct. Two cases of pancreatic pseudocysts (4 and 5 cm in diameter) showed communication with the pancreatic ductal system. Shock wave treatment was administered because the stones were not extractable by applied endoscopic measures. A total of 29 patients (83%) presented with exocrine pancreatic dysfunction (reduced fecal chymotrypsin, steatorrhea) and weight loss; 28 of them had enzyme replacement. Five patients had overt diabetes mellitus and two presented with impaired glucose tolerance.

METHODS
Pancreatic ductal stones were fragmented using an electrohydraulic lithotripter (MPL 9000, Dornier Medizintechnik, Germany) (12,13) after exact sonographic targeting. All patients were treated in a prone position. Up to 2000 electrocardiogram-triggered shock waves were delivered per session under continuous ultrasound monitoring. If fragmentation   was not successful, ie, insufficient disintegration of the stone(s) visible in the ultrasound examination, shock wave lithotripsy was repeated. The average shockwave energy was 18 kV (range 14 to 22). Patients were given piritramide (available only in Europe through Janssen Pharmaceutica Inc) and midazolam for analgesia and sedation. Endoscopic sphincterotomy (EST) of the pancreatic portion of the sphincter was done during ERP in 34 patients. In the 13year-old, presumably presenting with idiopathic juvenile pancreatitis, EST was not done. Pancreatic duct diameters and stone fragmentation were controlled ultrasonographically. Fragments not passed spontaneously after ESWL were extracted as completely as possible using Dormia baskets (Schadlowsky; Voerde, Germany).
Average follow-up was 23 months (range three to 70). Patients were asked about their conditions and examined clinically and by ultrasonography every three months in the first year of follow-up, and thereafter at intervals of up to 12 months. If there was any doubt about fragment migration or impaction, or stone recurrence, an ERP was performed.

RESULTS
Targetting of pancreatic stones within the shock wave focus and complete therapeutic monitoring were possible in all treatment sessions under ultrasonographic control.
Stone disintegration was achieved in all patients; 13 (37%) required one treatment session, nine (26%) required two, eight (23%) required three, and five (14%), with multiple stones completely filling the duct of Wirsung, required between four and seven. A total of 5000 (range 500 to 13,500) shock waves were used per patient. Under analgesia and sedation with piritramide and midazolam ESWL was well tolerated. There were no cardiopulmonary complications.
After sufficient stone fragmentation, controlled by ultrasound examination (Figure 4b), follow-up ERP was performed. Based on ultrasound and ERP, 16 patients (46%) were shown to be completely stone-free. Nineteen patients (54%) had small residual fragments in the main ductal system, mostly located in the tail portion of the pancreatic duct (Table 1). Complete endoscopic extraction of these fragments was not possible because of strictures, kinking of the  Figure 2 pancreatic duct or both. The 13-year-old with idiopathic chronic pancreatitis (without sphincterotomy) and three other patients showed complete stone clearance spontaneously.
In 29 cases (83%) the diameter of Wirsung's duct was reduced to more than 50% of the baseline value ( Figures  4c,5). After treatment, mean diameter of the main pancreatic duct was 2.8 mm (range 0 to 6). In patients with pancreatic pseudocysts the diameter of pseudocysts decreased within two to three days after removal of the stone obstruction. After three months they were no longer detectable ultrasonographically.
After treatment 12 patients (34%) were completely free from complaints and a further 17 (49%) reported significantly less pain (Table 2). Eighteen patients (51%) gained an average of 5.5 kg (range 2 to 15). Pathological glucose tolerance returned to normal in one patient. Five of six patients (17%) who reported no improvement despite successful ESWL had a filiform stenosis of the pancreatic duct and underwent pancreatic surgery. Six of nine patients with solitary stones, 10 of 26 patients with several stones and six of 15 patients with significant strictures of the pancreatic duct became stone-free (Tables 1,3). Eight of 12 patients who became pain-free were also stone-free and only three of the 12 had a stricture of the pancreatic duct (Table 2).
ESWL had to be repeated in four patients (11%) because of pain recurrence due to the migration of residual fragments or stones from the tail section of the pancreas. Three patients were successfully retreated for recurrence of symptomatic stones 18 to 28 months after successful primary therapy; they remained free from stones and complaints for up to 26 months. Diagnosis of fragment migration and recurrent concrements was made ultrasonographically and confirmed by ERP.
There were no serious complications associated with shock wave therapy. In five patients (14%) serum amylase and lipase were slightly elevated immediately after ESWL, without evidence of acute pancreatitis. Clinically overt pancreatitis occurred in three patients and subsided within two days under symptomatic therapy. Repeated ultrasound examinations revealed no additional morphological changes in the pancreatic parenchyma or the peripancreatic region. As a complication of EST, a localized retroduodenal perforation was found in one patient (managed conservatively). Two patients had acute pancreatitis.

DISCUSSION
Because of high perioperative mortality and generally poor long term results of surgical resection or drainage procedures in patients with chronic calcific pancreatitis (1,(14)(15)(16), alternative forms of treatment are needed. Endoscopic procedures involving division of the pancreatic sphincter and extraction of ductal stones are often unsuccessful due to the incongruity between the size of the stone and the anatomy of the pancreatic duct. Stone fragmentation by ESWL permits clearance of the duct by spontaneous passage or endoscopic extraction of stone fragments (11,12,(17)(18)(19)(20)(21)(22)(23)(24).
Fragmentation of the occluding pancreatic ductal stones, with reduction of the stone volume and clearance of the obstruction, was achieved in all patients. In 16 patients (46%) treatment resulted in complete stone clearance. Twelve patients (34%) were free from complaints after treatment and pain diminished considerably in 17 (49%). The diameter of Wirsung's duct was reduced by more than 50% of the baseline value, indicating the clearance of obstruction with a decrease of pressure in the pancreatic duct system, in 29 patients (83%) (25) (Figures 4b,4c,5). Importantly, five of six patients who complained of unchanged symptoms after ESWL and who subsequently required pancreatic surgery had a filiform stenosis in the distal portion of the main pancreatic duct. In patients with fragments not endoscopically extractable, these fragments were primarily located behind strictures or in especially narrow segments of the duct. However, there were some patients with a large stone volume, a stricture of the pancreatic duct or both, who became stone-free and asymptomatic (Tables  1-3). Thus, neither stone characteristics nor pancreatic duct morphology seems to be of predictive value regarding the therapeutic outcome.   We found that complete stone clearance was not necessary for complete abolition of symptoms. Remaining residual fragments were small, caused no obstruction and were usually situated in the tail portion of the gland. Four patients who suffered a further episode of pain due to fragment migration after initial success of ESWL were pain-free after repeated ESWL. Stone recurrence was observed in three patients after 18 to 28 months, and they were again treated successfully with ESWL.
Results equivalent to or better than those discussed have been reported by others (18,19,21,22,26,27) using electrohydraulic or electromagnetic lithotripters with radiographic stone location. Three groups used ultrasonography alone (28,29) or in combination with radiography (24) for targetting pancreatic ductal concrements; freedom from stones and complete freedom from pain were achieved in up to 70% of patients.
The success of ultrasonography in indication, therapeutic monitoring and follow-up examination renders it the method of choice in our opinion in pancreatic ESWL. By using ultrasound to locate stones, patients avoid exposure to radiation (12,20). With continuous treatment monitoring by real-time ultrasound, we can also spare patients the nasopancreatic tube necessary for administration of contrast medium during x-ray-guided ESWL.
In our experience, ESWL is not contraindicated in the presence of pancreatic pseudocysts; in fact, removal of duct obstruction can improve drainage of the pseudocysts if they communicate with the duct system. Further controlled prospective studies are needed to determine whether the weight gain observed in 18 patients in our study was due simply to adequate intestinal enzyme replacement and improved appetite after pain relief, or whether it also reflects a significant reduction in pancreatic exocrine dysfunction. Improved endocrine function may occur in individual cases. However, in view of the highly variable spontaneous course of chronic pancreatitis our results have to be interpreted with caution.

CONCLUSIONS
The combination of ultrasound-guided ESWL, EST and fragment extraction is a new nonoperative approach -with a low rate of complications -for the treatment of pancreatic ductal stones. Although most patients showed improvement in their general condition and especially their pain, controlled prospective studies comparing the spontaneous course of the disease with the results of interventional endoscopy and surgical methods are needed.