Effects of Gabexate Mesilate, a ProteaseInhibitor, on Human Sphincter of

Oddi MotilityVINCENZO DI FRANCESCO,* ALBERTO MARIANI,‡ GIAMPAOLO ANGELINI,† ENZO MASCI,‡LUCA FRULLONI,† GIORGIO TALAMINI,† SANDRO PASSARETTI,‡ PIERALBERTO TESTONI,‡

and GIORGIO CAVALLINI†

Gabexate mesilate is an antiprotease drug, which reduced the severity of pancreatitis andfrequency of post-ERCP pancreatitis. In dogs gabexate inhibits sphincter of Oddi motility butno data are available in humans. The aim of this study was to verify by manometry the actionof gabexate on human sphincter of Oddi motility. We enrolled 12 patients with idiopathicrecurrent pancreatitis (eight males, five females, mean age 46 � 8 years). Standard preen-doscopic sphincter of Oddi manometry was done in basal conditions and during infusion ofgabexate 20 mg/min: basal pressure, amplitude and frequency of phasic contractions, andmotility index (amplitude per frequency) were calculated before and after gabexate injection.Statistical analysis was performed by using Wilcoxon rank test for paired data. Six patientshad a manometric diagnosis of stenosis (basal pressure greater than 40 mm Hg); six hadnormal findings. Phasic activity was not evaluable in five patients with stenosis. Basal pressurewas unaffected by drug infusion, while gabexate caused a significant reduction of phasicactivity, both in terms of frequency (4.5 � 1 vs 3.6 � 1; P � 0.05) and amplitude (157.4 �44 vs 80.0 � 32; P � 0.05) of contractions. Motility index was reduced on average by 49%.In conclusion, this pilot study confirms, in patients with acute recurrent pancreatitis, theinhibitory action of gabexate on sphincter of Oddi motility already described in dogs. Thisaction needs to be revaluated at therapeutic dosages. On the other hand, prophylactic use ofthe drug should be avoided during sphincter of Oddi manometry, in order to avoid falsenegative results.

KEY WORDS: sphincter of Oddi; motility; antiprotease; recurrent pancreatitis.

Gabexate mesilate (GM) is a 417-dalton drug thatinhibits protease action; in particular it has beendemonstrated to be active against trypsin, kallikrein,plasmin, thrombin, phospholipase A2 and C1 esterase

(1). In animals, GM prevents acute pancreatitis (2, 3)and in humans it reduces complications and the needfor surgery in severe pancreatitis (4). It has beenreported recently that prophylactic GM infusion wasable to significantly reduce post-ERCP pancreatitiswhen compared with placebo (5). In dogs GM in-duced an inhibitory response on sphincter of Oddi(SO) motility, probably by stimulation of nonadren-ergic, noncholinergic inhibitory neurons. (6). No dataare available so far on the action of GM on humanSO.

Manuscript received January 18, 2001; accepted July 24, 2001.From the *Department of Biomedical and Surgical Sciences, and

†Department of Surgery and Gastroenterology, University of Ve-rona; and ‡Department of Biomedical Science, Division of Gas-troenterology, IRCCS San Raffaele, University of Milano, Italy.

Address for reprint requests: Vincenzo Di Francesco, Depart-ment of Biomedical and Surgical Sciences, University of Verona,P.le Stefani 1, 37126-Verona, Italy.

Digestive Diseases and Sciences, Vol. 47, No. 4 (April 2002), pp. 741–745 (© 2002)

741Digestive Diseases and Sciences, Vol. 47, No. 4 (April 2002)0163-2116/02/0400-0741/0 © 2002 Plenum Publishing Corporation

Aim of this study was to determine by manometrythe action of GM on SO motility.

MATERIALS AND METHODS

We enrolled patients with previous episodes of acuterecurrent pancreatitis who had been scheduled for sphinc-ter of Oddi manometry for suspected sphincter dysfunction.Any known cause, in particular the presence of bile stones,alcohol abuse, metabolic disorders, or pancreatic duct ab-normalities, had been ruled out by functional (pancreolau-ryl test and fecal fat assay) and morphological examinations(US and/or CT and ERCP). Hereditary pancreatitis andcystic fibrosis were ruled out as well. No patient had adilated common bile duct. Patients were symptom-free atthe time of examination and at least one month passed fromthe last recurrence of pancreatitis. Informed consent wasobtained from all the subjects. After an overnight fast,patients were sedated with diazepam 10 mg intravenously,then submitted to SO manometry. The procedure was per-formed endoscopically using a perfused three-lumen cath-eter (0.125 ml/min/channel, Wilson Cook) connected viapressure transducers to a computerized recording system(Polygram; Synectics).

The central port of the SO catheter was continuouslyaspirated. Intraduodenal pressure was taken as a zero ref-erence. The catheter was inserted deep into the pancreaticduct (aspiration or fluoroscopic check) or, when selectivepancreatic cannulation was not achieved, into the bile ductand retracted, measuring SO pressures at 1-mm intervals,via the station pull-through technique. After a stable basalrecording at least 2 min long was obtained at the level of thehigh-pressure zone, GM 100 mg (FOY, Hoechst) diluted in20 ml saline was injected intraneously over 5 min andregistration continued for few more minutes until a 2 min ofstable recording was obtained during drug infusion. In nocase did cannulation exceeded 10 min. Sphincter of Oddibasal pressure above duodenal zero, amplitude above basalpressure, and frequency of phasic contractions were calcu-lated as well as the motility index (amplitude per frequency)before and after gabexate injection. Manometry failed dueto the inability to deeply cannulate the sphincter in twopatients. The procedure was successful in 12 patients (eightmales, five females, mean age 46 � 8 years, range 37–55).

Manometric parameters were compared by using Wil-coxon rank test for paired data (software SPSS for Windowsrel. 8.0; SPSS Inc., Chicago, Illinois, USA).

RESULTS

Six patients had a manometric diagnosis of stenosis(basal pressure greater than 40 mm Hg). In five ofthese subjects it was not possible to clearly identifyphasic contractions. In five of six subjects with SOstenosis, and in the remaining six patients with nor-mal manometry, GM had no effect on sphincter ofOddi basal pressure. In one patient with stenosis andno clear evidence of phasic contractions, the druginfusion caused a further rise of basal pressure. In thesix patients with normal basal pressure and in onepatient with stenosis it was possible to evaluate phasiccontractions. After a latency period of about 1 min,GM administration caused a statistically significantreduction of phasic activity, both in terms of ampli-tude and frequency of contractions (Table 1,) (Figure1). Motility index was reduced on average by 49%(Figure 2). In one patient GM induced a completeinhibition of sphincter of Oddi motility (Figure 3).

Following manometry, three patients suffered frompancreatic pain accompanied by lipase elevation, butthe episodes resolved in all three within 24 hr withconservative therapy.

DISCUSSION

This pilot study confirmed on patients with acuterecurrent pancreatitis the inhibitory action of GM onSO motility, already described in dogs (6). In order toreduce the risk of complications following SO cannu-lation, we used a bolus of 100 mg GM infused over afew minutes. The therapeutic dose of the drug (up to4 g/24-hr infusion) needs to be tested against placeboto better define the clinical significance of this action.Some considerations may be based on our results.GM did not influence basal SO pressure that is reg-ulated by cholinergic tone (7). Only one patient withstenosis presented a rise of basal pressure followingGM administration. A similar paradoxical responsealso was reported in a few cases in experimental

TABLE 1. SO MANOMETRIC PARAMETERS BEFORE AND AFTER GM INJECTION*

N Before GM After GM P

Basal pressure (mm Hg) 12 52.7 � 11 48.2 � 11 0.16Frequency (contractions/min) 7 4.5 � 1 3.6 � 1 0.027Amplitude (mm Hg) 7 157.4 � 44 80.0 � 32 0.018

Motility index (mmHg �contraction/min)†

7 627 � 149 293 � 105 0.018

*Data are shown as mean � SE.†Motility index � frequency � amplitude of phasic contractions.

DI FRANCESCO ET AL

742 Digestive Diseases and Sciences, Vol. 47, No. 4 (April 2002)

animals. In the majority of cases both normal andabnormally high basal pressures were unaffected bythis drug.

In one other patient, with a SO manometry tracecharacterized by normal basal pressure and high am-plitude phasic contractions, GM caused rapid aboli-tion of phasic activity, which came back to the pre-treatment pattern immediately after drug infusionwas stopped. In the remaining six patients with de-tectable phasic activity, GM caused a rapid but lessprofound inhibition of spontaneous contractions bothin terms of amplitude and frequency, with a meanreduction of 35% and 40% respectively.

SO phasic contraction activity shows circadianmodifications, mostly correlated with feeding, withpeaks of activity in postprandial phases and duringphase II of the migrating motor complex (8), and

nearly complete inhibition of motility during gallblad-der contraction (9). This complex motor function iscoordinated by a neural network connecting the jeju-num, SO, and gallbladder (10). The role of neuropep-tides and hormones in physiologic and pathologicconditions has not been completely clarified, so it ishard to know how GM can influence SO motility inhumans. The very rapid response to infusion suggestsa neuromediated mechanism. Similar to our experi-ment, in dogs GM induced an inhibitory response ofSO motility, and the authors (6) concluded that thiswas probably obtained by stimulation of nonadrener-gic noncholinergic inhibitory neurons, but not byGM-induced cholecystokinin secretion, since the re-sponse was not reduced or terminated by pretreat-ment with atropine, guanethidine, hexamethonium,and/or proglumide.

Fig 1. SO basal pressure before and after injection of 100 mg GM (bars show mean value �SE).

Fig 2. SO motily index (calculated as the product of phasic contraction frequency andamplitude) before and after injection of 100 mg GM (bars show mean value � SE).

EFFECTS OF GABEXATE ON SPHINCTER OF ODDI

743Digestive Diseases and Sciences, Vol. 47, No. 4 (April 2002)

On the basis of our results on humans, we can onlyhypothesize that GM antiprotease activity may inter-fere with enzymatic production, activation, or inacti-vation of neuropeptides or mediators (such as nitricoxide derivated by proteic degradation) (11), involvedin SO motility control. In general this seems to causemild motor activity inhibition, but in some individualsor at different times GM may cause a more profoundinhibition or, on the contrary, excitation. In the caseswith high basal pressure and no evident phasic con-traction, an organic basis for stenosis (fibrosis) is themost probable cause of SO dysfunction, so the lack ofresponse to the drug can be explained.

Only further experimental models will clarify themechanisms of GM action on SO motility, but at leasttwo clinical conclusions may be drawn from this study.First, the reduction of SO activity causes, in humans,a reduction of resistance through the sphincter topancreatic outflow (9). GM could be usefully utilizedin treatment and prevention of acute pancreatitis notonly for its anti-protease activity (1) but also becauseit may decompress SO in obstructive forms such asthose caused by stones or by ERCP (12). Preventionof post-ERCP pancreatitis at therapeutic dosages ofGM may have been successful also because of SOinhibition (12). On the other hand, octreotide, along-acting somatostatin analog and, effective antise-cretive drug, failed to prevent post-ERCP pancreati-tis (13). One possible explanation of these results wasthe observation that octreotide excited SO phasic

activity in a series of patients with the same clinicalcharacteristics as those in the present study (14).

A second, minor aspect regards SO manometry.This is a relatively high-risk diagnostic procedure,frequently causing pancreatitis (15). If our resultscould be confirmed with therapeutic dosages of GM,prophylactic use of the drug should be avoided duringSO manometry, otherwise altered tracings could beobtained. As an alternative, due to its very shortduration of action (about 1 min) (16), GM infusionmay be just stopped few minutes before manometryand restarted immediately after, for prophylaxis ofpancreatitis following the procedure.

REFERENCES

1. Tamura Y, Hirado M, Okamura K, Minato Y, Fujii S: Syn-thetic inhibitors of trypsin, plasmin, kallikrein, thrombin, C1r,C1 esterase. Biochim Biophys Acta 484:417–422, 1977

2. Ito T, Kimura T, Furukawa M, Yamaguchi H, Goto M, NakanoI, Nawata H: Protective effects of gabexate mesilate on acutepancreatitis induced by tacrolimus (FK-506) in rats in whichthe pancreas was stimulated by caerulein. J Gastroenterol29(3):305–313, 1994

3. Wakayama T, Itoh T, Shibayama K, Idezuki Y: Prevention ofspread of experimental acute pancreatitis by intraductal ad-ministration of a synthetic inhibitor in dogs. Am J Gastroen-terol 84:272–278, 1989

4. Messori A, Rampazzo R, Scroccaro G, Olivato R, Bassi C,Falconi M, Pederzoli P, Martini N: Effectiveness of gabexatemesilate in acute pancreatitis. A metaanalysis. Dig Dis Sci40(4):734–738, 1995

5. Cavallini G, Tittobello A, Frulloni L, Masci E, Mariani A, Di

Fig 3. SO manometry recording from a patient with recurrent attacks of pancreatitis. Before GM injection, the tracing shows normalbasal pressure and low-frequency, high-amplitude phasic contractions. After GM injection, there is rapid and profound inhibition ofsphincter motility, which returns to normal few seconds after injection has stopped. Bar marks 2 min of GM infusion.

DI FRANCESCO ET AL

744 Digestive Diseases and Sciences, Vol. 47, No. 4 (April 2002)

Francesco V: Gabexate mesilate in the prevention of pancre-atic damage induced by endoscopic retrograde cholangio pan-creatography. N Engl J Med 335 (13):919–923, 1996

6. Yamasato T, Mori M, Mimura H, Hamazaki K, Kashino H,Gouchi A, Okabayashi T, Tsuge H, Miyake J, Suga M: Effectsof gabexatate mesilate (FOY) on the gallbladder, sphincter ofOddi and duodenum of the normal and gastrectomized dogs. JSmooth Muscle Res 27:87–96, 1991

7. Behar J, Biancani P: Neural control of the sphincter of Oddi.Gastroenterology 86:134–141, 1984

8. Worthley CS, Baker RA, Iannos J, Saccone GT, Toouli J:Human fasting and postprandial sphincter of Oddi motility.Br J Surg 76:709–714, 1989

9. Toouli J, Hogan WJ, Geenen JW, Dodds WJ, Arndorfer RC:Action of cholecystokinin octapeptide on sphincter of Oddipressure and phasic wave activity in man. Surgery 92:497–503,1982

10. Sand J, Tainio H, Nordback I: Peptidergic innervation ofhuman sphincter of Oddi. Dig Dis Sci 39:293–300, 1994

11. Sand J, Arvola P, Jantti P, Oja S, Singaram C, Baer G, Pasricha

PJ, Nordback I: The inhibitory role of nitric oxide in thecontrol of porcine and human sphincter of Oddi activity. Gut41 (3):375–380, 1997

12. Pasricha PJ: Prevention of ERCP-induced pancreatitis: successat last. Gastroenterology 112:1415–1417, 1997

13. Binmoeller KF, Harris AG, Dumas R, Grimaldi C, DelmontJP: Does the somatostatin analogue octreotide protect againstERCP induced pancreatitis? Gut 33:1129–1133, 1992

14. Di Francesco V, Angelini G, Bovo P, Casarini MB, Filippini M,Vaona B, Frulloni L, Rigo L, Brunori MP, Cavallini G: Effectof octreotide on sphincter of Oddi motility in patients withacute recurrent pancreatitis. A manometric study. Dig Dis Sci41:2392–2396, 1996

15. Rolny P, Anderberg B, Ihse I, Lindstrom E, Oilason G, ArvillA: Pancreatitis after sphincter of Oddi manometry. Gut31:821–824, 1990

16. Nishijima MK, Takezawa J, Taenaka N, Shimada Y, Yoshiya I:Application of HPLC measurement of plasma concentration ofgabexate mesilate. Thromb Res 31:279–284, 1983

EFFECTS OF GABEXATE ON SPHINCTER OF ODDI

745Digestive Diseases and Sciences, Vol. 47, No. 4 (April 2002)