Br. J . Surg. 1990, Vol. 77, November, 1233-1 237
M. Anselmi, S. Lancberg*, M. Deakin, E. Lanchbury*, 2.Drolc*, F. Burrows?, E. Elias and P. M c M a s t e r The Liver Unit and Departments of *Nuclear Medicine and ?Radiology, The Queen Elizabeth Medical Centre, Birmingham, UK Correspondence to: Mr P. McMaster, The Liver Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
Assessment of the biliary tract after liver transplantation: T tube cholangiography or IODIDA scann ing Biliary tract obstruction or anastomotic leakage are common problems following liver transplantation. In a sequential study, 31 patients with a liver transplant were investigated by 99mTc-IODIDA ( I O D I D A ) scanning and T tube cholangiography ( T T C ) and the results tc-ere compared with clinical outcome. Seven patients had an extrahepatic biliary obstruction and one patient had a biliary leak. In the detection of biliary complications T T C and IODIDA scanning were similar in terms of sensitivity (63per cent,for both) but T T C hada better specificitji (79 per cent versus 60 per cent) and accuracy (74 per cent versus 6 0 per cent) than IODIDA scanning. When liver ,function was taken into account, the diagnostic efii'cacy of both tests in patients with bilirubin levels of < 200 ,umol/l was similar. With levels > 200 ,umol/l there M ~ a greater number of' false positive results with IODIDA scanning (12 per cent versus 54 per cent). The only significant biliary leak MUY clearly detected by T T C but not IODIDA scanning. T T C remains the more effective way of evaluating the biliary tract ufter transplantation. IODIDA scanning has limited value when bilirubin levels are elevated, but may provide additional information about blood supply, hepatocyte func tion and intrahepat ic cholestasis. I
Keywords: Liver transplantation, biliary tract reconstruction, biliary anastomosis, biliary tract complication. T tube cholangiography, biliary scintigraphy, yy'"Tc-IODIDA scintiscanning
Improvement in surgical technique and use of low-dose steroids have resulted in fewer biliary tract complications following liver transplantation'-'. Nearly one in five patients still experiences problem^^.^ and assessment of the biliary tract is performed routinely 7 days post-transplantation and later if clinically indicated. Cholangiography (T tube, percutaneous transhepatic or endoscopic retrograde) has proved useful in demonstrating the type, location and severity of most early and late biliary complications'. Injection of contrast into the biliary tree is frequently related to episodes of cholangitis and bacteraemia'.' which are especially dangerous in immunosuppressed patients. Investigation is more diffcult when the patient does not have a T tube. Hepatobiliary scintigraphy is a non-invasive, widely available and inexpensive technique. In addition to identifying biliary t.ract complications it may provide useful information about hepatocyte function and intrahepatic cholestasis".' ' . IODIDA (Solcoscint IODIDA, SOLCO Balse Ltd., Birsfelden, Switzerland) has a lower urinary and higher biliary excretion and has significant advantages over other "'Tc-Sn-2,6diethylacetanilidoiminodiacetate (HIDA) derivatives especially when bilirubin levels exceed 80-100 p m ~ l / l ' ~ - 'It~ .has been advocated in diagnostically diffcult situations such as liver transplant, bile duct reconstruction, suspicion of bile leakage, examination in small children or of critically ill patients. A prospective study was undertaken to compare the diagnostic value of ""Tc-IODIDA isotope (IODIDA) scanning and T tube cholangiography (TTC) in patients after liver transplantation.
Patients and methods Thirty-one liver transplant patients underwent T T C and hepatobiliary scintigraphy. Assessment was performed routinely in 26 patients between 7 and 10 days after surgery and at a later stage in five patients
when suspicion of complications arose. Both tests were performed within 48 h and no change in the clinical situation was observed in any patient between the two studies. The indications for liver transplantation were: primary biliary cirrhosis in 12 patients, fulminant hepatic failure in six, chronic active hepatitis in four, sclerosing cholangitis in three, a-I-antitrypsin deficiency in two, cryptogenic cirrhosis in two, hepatocellular carcinoma in one and alcoholic cirrhosis in one. There were two children (aged 1.5 and 7 years) and 29 adults whose mean age was 40.7 years (range 18-68 years). Seventeen of the group were female. The types of biliary reconstruction used are shown in Figure I : ( a ) end-to-end choledochocholedochostomy in 16 patients (52 per cent); (b) gallbladder conduit in 11 (36 per cent); (c) choledocho-Roux-enY-jejunostomy in three cases (10 per cent); and ( d ) in one case a conduit to jejunum (3 per cent). Bilirubin levels at investigation ranged from 25 to 883 pnol/l and were less than 100pmol/l in nine patients, less than 200pmolll in 15 patients and greater than 200 pmol/l in the remaining 17 cases. Both imaging studies were reported independently by an examiner who did not have access either to clinical information or to the results of the other test. The T tube cholangiograms wcre read 'blind' by a consultant radiologist with experience in examining transplanted patients (F.B.) and the IODIDA scans were read 'blind' by a nuclear medicine specialist (S.L.). Scin/igruphj.nw/l7ocl After fasting for at least 4 h 80 MBq of commercially available IODIDA were injected as a bolus into an antecubital vein. With the patient supine, images of the abdomen were obtained using a large field of view y a m e r a with a low-energy all-purpose collimator and recorded on computer at one frame/min for 40 min. Analogue film images were also obtained every 5 min over this same interval. Further images were obtained at 1-4 h and 18-24 h if IODIDA clearance was delayed or a leak suspected. The results of scintigraphy were cl
I. 2.
No ohsrruciion; no obstruction to the flow of the scintigraphic agent into the bowel taking into account normal variation". Purrid ohs/ruc/ion; detection of partial obstruction was primarily
-~ 0007-1 323/90/1 I123345
(
1990 Butterworth-Heinemann Ltd
1233
Y
Biliary tract assessment after liver transplantation:
M.Anselmi et al.
a
C
d
\ Figure 1 Schematic diagram showing the types of biliary reconstrucrion used in 31 liver transplant recipients: choledochocholcystcholedochostoniy; c Roux-en- Y choledochojejunostomy; d Rour-en- Y choledochocholecys~ejunostomy a end-to-end choledochocholedocho.~tomy;b
3. 4.
based on filling defects in narrowing of the common bile duct or bile pooling in the segmental ducts” and a delay of 60min or more before appearance of isotope into the bowel. Complete obstruction: absence of the scanning agent from the small bowel after 24 h regardless of whether or not images of the bile duct had appearedI8-”. Presence or absence of biliary leak: the appearance of radioactivity in the peritoneal cavity.
Information was available on impaired hepatic uptake of radionuclide indicating primary hepatic dysfunction and intrahepatic cholestasis manifested by relatively good uptake of IODIDA, delayed visualization of intestine between 4 and 24 h and visualization of the bile duct adequate to rule out extrahiliary obstruction2’.z2. T tube cholangiography ( T T O TTC was performed with antibiotic cover. Approximately 2C30 ml urografin 325 were slowly injected during screening to display as much of the duct system as possible and to avoid discomfort, overfilling of the duct or bacteraemia. If contrast ran freely through the lower duct (or conduit) into the bowel without showing the upper ducts, a repeat injection was performed with the patient in head-down position. The results of TTC were classified as follows:
I. 2. 3.
N o obstruction: no obstruction to the flow of contrast into the bowel with non-dilated intrahepdtic ducts. Purtiul ohstruction: the presence of strictures or filling defects within the extrahepatic or intrahepatic bile ducts causing delay in passage of contrast into the bowel. Complete obstruction: obstruction to the passage of contrast either
1234
4.
into the bowel or alternatively from common hepatic duct to intrahepatic ducts. Presence or absence o f a leak.
In addition to scintiscanning and radiological examination all patients underwent detailed clinical monitoring, biochemical assessment. protocol liver biopsy on day 7 post-transplant and later if required. The ultimate assessment of the biliary tract and final diagnosis were based o n the patient’s clinical course and information which became available either through reoperation or autopsy.
Results No biliary tract complications were found in 23 patients (74per cent), seven had extrahepatic obstruction (23 per cent) and one had a biliary leak (3 per cent). The process of final diagnosis is shown in Table 1. T tube cholangiography (TTC) In the 23 patients without biliary complications T tube cholangiograms were reported normal in 18. There were five false positive studies in which the diagnosis of either complete o r partial obstruction was made. Three of these patients had a gallbladder conduit: in two of these the diagnosis of complete obstruction of the upper choledochocholecystanastomosis had been made (Figures2a and b ) and in one partial obstruction at the papilla with proximal dilatation was reported. The fourth patient had a duct-to-duct reconstruction and partial
Br. J. Surg.. Vol. 77, No. 11, November 1990
Biliary tract assessment after liver transplantation: M. Anselmi et al.
Table 1 Means of’ oeri/ication of the,final diagnosis in 31 cases Final diagnosis Autopsy
Operation
Clinical course
(n=5 )
(n=7)
( n = 19)
Biliary leak ( n = 1 )
1
Obstruction ( n = 7 ) Inflammation of conduit Anastornotic stenosis Biliary sludge
1 1 3
No biliary complications ( n = 23)
4
Again no changes in the diagnosis were suggested after unblinded review of the images when clinical information was available. The diagnostic accuracy of TTC and IODIDA scanning in predicting eventual clinical outcome is shown in Table 2 . Both
~
2 -
2
17
obstruction of the anastomosis was suspected. In the remaining patient with a choledochojejunal reconstruction, a proximal non-anastomotic obstruction was diagnosed. In none of these patients did the liver function tests, histology or clinical outcome indicate any clinically relevant biliary obstruction during 2-24 months of follow-up. The patient with the biliary leak was correctly diagnosed on T tube cholangiogram. In the seven patients with extrahepatic biliary obstruction, four were shown on the T tube cholangiograms. There were three false negative studies. Two were in patients with a duct-to-duct anastomosis. In the first of these patients a hepatic artery thrombosis was diagnosed on ultrasonography and confirmed angiographically. Reoperation showed a n anastomotic stricture and the anastomosis was converted to a Roux-en-Y choledochojejunostomy . In the second patient a papillotomy was performed and biliary sludge was removed 2 weeks after the TTC (Figures3a and b ) . The other patient with a false negative study had a gallbladder conduit. Reoperation because of clinically increasing cholestasis and abdominal pain revealed obstruction of the common bile duct due to inflammation of the conduit and after cholecystectomy a Roux-en-Y choledochojejunostomy was performed. When the studies were reassessed in the light of full clinical information no changes in the radiological diagnosis were made ruling out an error in the interpretation of the images. IOD ID A scanning In one of the 23 patients without proven biliary complications there was poor hepatic uptake of IODIDA indicating pure hepatocellular dysfunction and the biliary tract was not visualized. In the other 22 normal patients, 13 were reported to have normal scans. There were nine false positive scans. Six were reported as completely obstructed. Four of these patients had histologically proven moderate rejection, one had mild rejection and pneumonia and one had severe sepsis. Three scans were reported as partially obstructed on the basis of delayed appearance of radioactivity in the gut and poor delineation of the biliary tract. One of these patients had moderate rejection on biopsy, one had sepsis and mild rejection and one had an intra-abdominal abscess. The mean bilirubin level of the nine patients with false positive results was more than twice that of patients with true negative IODIDA scanning results (Figure 4 ) . The false positive rates for IODIDA scanning and T T C were 54 per cent and 14 per cent when the bilirubin level was >200pmol/l and 12 per cent and 18 per cent when the level was < 200 pmol/l. Scintiscanning failed to detect leakage of isotope into the abdominal cavity in the patient with a biliary leak. In the seven patients who were eventually shown to have biliary obstruction IODIDA scanning correctly diagnosed five. There were two false negative studies. Both patients had a duct-to-duct anastomosis and were diagnosed as having enlarged but unobstructed ducts and showed appearance of radioactivity in the gut within 35 min. Evidence of obstruction was found in both and corrected by reoperation in one and endoscopic dilatation in the other.
Br. J. Surg., Vol. 77, No. 11, November 1990
b
111
IV
Figure 2a Esaniple of’ a ,false po.sitir.e T tuhe cholunyioyram prohubiJ due to undetfilliny of flit hiliarj. systeni in a patient w i t h a gullhludder conduit reconstruction. The studv II‘US perfiwnied 27 d a . ~after ~ transplantation clue t o dcteriorution of the liver ,function tests (bilirubin ~i’as60 pniol/l and alkaline phosphatase II’CIS 970 unitsll). In this puticvit thc cliokmyiogram II’US reported U S a complete obstruction of the donor coninion hile duct to donor gullhiadder anasloniosi,s. N(J i~b.struclifJliIW.Y confirmedin the,follo~c~-upperiod. b Selectedanterior r.icw:s.froniIODIDA stucly ohtained U I (02-5 niin. (10 17-20 niin, ( I l l ) 27-30 niin utid ( 1 0 35-38 niin after injeerion. Thy,*s l m r good estrucrion nf IODIDA ,from the Mood pool (acticily within the Iieart is ,seen o n l ~ on , the iniuge ( i t 2 min) arid nornial clearance ,from parench~~niu into the ya.stroiiire.stiiiri1 tract
1235
Biliary tract assessment after liver transplantation: M. Anselmi et al. 700
--
1
l
0
o
600
O
pk o s i t i v es n e g a t Fi v e
l
False False positive negative
Figure 4 Comparison of the resulrs of IODIDA scunning study und hiliruhin levels Table 2 Compurison of the sensitinitj, spec$'city and uccurucy of TTC and IODIDA scanning in the diagnosis of hiliury truct complications after liver rransplantation ~
TTC (all patients) TTC (EECC, n= 16) TTC (GBCD, ! I = 11) IODIDA (all patients) IODIDA (EECC, n= 16) IODIDA (GBCD, n= 11)
Sensitivity
Specificity
Accuracy
(YO)
(YO)
(YO)
63
78
74
61 50
92 57
88 55
63 50 100
59 67 88
60 62 90
TTC, T t u b e cholangiography; IODIDA, 99"Tc-IODIDA scanning; EECC, end-to-end anastomosis of common bile duct; GBCD, gallbladder conduit reconstruction
tests had similar sensitivity but TTC had a better specificity and accuracy than IODIDA scanning. The only biliary leak was clearly shown on TTC but not identified on IODIDA scanning. The diagnostic value of the tests in the detection of biliary obstruction was influenced by the type of biliary reconstructive procedure. TTC was more accurate than IODIDA scanning in detecting obstruction in patients with end-to-end choledochocholedochostomy, but in those with a gallbladder conduit IODIDA scanning was more efficient (Table 2 ) . When the liver function tests did not show cholestasis, combination of the results ofTTC and IODIDA increased the sensitivity to 85 per cent and specificity to 91 per cent with an accuracy of90 per cent.
b
Discussion
111
IV
Figure 3a A n o.wniple u / X s e negutiw rutiiologicul studj.. Prutocol T tuhe cholungiogrup1i.v perfornird 6 rlu!.s ulier rrunspluntution in u patient with end-to-end unosroniosis qf the coninion hile duct. The .seruni hiliruhin IVUS 131 pmolll mid ulkuline phosphuruse II~US303 units//. TIW weeks hrer this putient ii'us suhniitted to an endoscopic papil1otoni.s und hiliurj.sluc!qe i t u s rmiowd,fiom the conimion hile durl. b Selected anterior iieaxfrom IODIDA stud). ohtuincrl ut (0 2-5 niin. (10 27-30 min. ( I l l ) I h und ( I V ) 24 h ufter injection. Prolonged retention u/' uctirity in the hlood pool. heart und other r.usculur swuctures, re1utii;eI). good u p t u h hj. the y r u o und uhsmce of'r.\-crrtion into gustrointesrinul tract indicutrs oh.rtri~tioii
1236
Liver transplant recipients are susceptible to a wide variety of complications which may threaten allograft and patient ~ u r v i v a l ~ ,Biliary '~. complications, hepatic preservation injury, rejection, hepatic artery thrombosis and infection may all lead to a rapid deterioration in liver function. Differentiating between the causes may prove difficult. Early diagnosis of biliary complications is important in planning treatment and because other post-transplant problems are often diagnosed by exclusion24. Oral cyclosporin is poorly and unpredictably absorbed from the intestine during bile diversion". Exclusion of biliary complications as a cause of deteriorating liver function allows early clamping of the T tube, with re-establishment of the enterohepatic circulation. Our results have shown that the sensitivity of both tests is similar but TTC has a better specificity and overall accuracy than IODIDA scanning. In patients with bilirubin levels above 200 pmol/l a greater number of false positive results appeared with IODIDA scanning because of misdiagnosed intrahepatic cholestasis due
Br. J . Surg., Vol. 77, No. 11. November 1990
Biliary tract assessment after liver transplantation: M. Anselmi et al.
t o rejection or sepsis. Hepatobiliary scintigraphy can provide accurate information in the evaluation of choIestasis22.26but this is difficult in patients with underlying intrahepatic disease where extrahepatic obstruction is suspected. In the transplant patients rejection affects small bile ducts leading to intrahepatic c h o l e ~ t a s i swhich ~ ~ . ~ can ~ mimic complete biliary obstruction". Six out of nine of o u r false positive studies showed this pattern. In the remaining three patients, poor visualization of the biliary tract and delayed intestinal appearance made scintigraphic interpretation equivocal as ductal delineation was inadequate. We found that in this g r o u p of patients, with bilirubin levels over 200 pmol/l, IODIDA scanning had a very low specificity and overall accuracy. Discrimination between high grade obstruction and severe intrahepatic disease based on the image pattern was not possible a n d suggests that the level of obstruction in such a case must be made by other modalities. The accuracy of TTC is not influenced by a high serum bilirubin level or by graft dysfunction. TTC does, however, require meticulous technique if cholangitis a n d bacteraemia ar e to be avoided. Required low pressure injection c a n result i n underfilling of the biliary tract as c a n rapid emptying of contrast into the intestine through a widely patent biliary+nteric anastomosis. Underfilling may explain a t least three of five false positive studies especially in the patients with gallbladder conduits a n d combination of the results of TTC and IODIDA scanning in these cases may be helpful if severe cholestasis is not present. The many pitfalls encountered with scintigraphy in the differential diagnosis of jaundice with high bilirubin levels undermines its usefulness as a first choice diagnostic procedure after transplantation. Scintigraphy c a n play a n adjuvant role and be of considerable clinical value in cases where technical difficulties yield indeterminate TTC results such as with a gallbladder conduit anastomosis. I n spite of its limitations, TTC still remains the most effective means of evaluating the biliary tract after transplantation. IODTDA scintiscanning may provide important information a b o u t the biliary tract a n d graft function with bilirubin levels below 200 pmol/l, but is unreliable in the presence of rejection, sepsis o r when bilirubin levels exceed 200 pmol/l.
6. 7. 8. 9. 10. 11.
12.
13. 14.
15.
16. 17. 18. 19. 20.
Acknowledgements Mr Mario Anselmi is a Fellow of the British Council. We would like to thank the medical, nursing and paramedical staff involved in the care and management of these patiens.
21. 22.
References 1.
2. 3.
4. 5.
Calne RY. A new technique for biliary drainage in orthotopic liver transplantation utilising the gall bladder as a pedicle graft conduit between the donor and recipient common bile duct. Ann Surg 1976; 184: 605-9. Wolff H, Otto G, David H. Biliary tract reconstruction in liver transplant. Transplant Proc 1985; 17: 274-5. Bismuth H, Castaing D, Gugenheim J, Traynor 0,Ciardullo M. Roux-en-Y hepatojejunostomy : a safe procedure for biliary anastomosis in liver transplantation. Transplunt Proc 1987; 19: 241 3- 15. Neuhaus P, Brolsch C, Ringe B, Lauchart W, Pichlmayr R. Results of biliary reconstruction after liver transplantation. Transplant Proc 1984; 16: 1225-7. Koneru B, Tzakis AG, Bowman J, Cassavilla A, Zajko AB, Starzl TE. Liver transplantation. Post operative surgical complication. Gastroenterol Clin North A m 1988; 17: 71-91,
Br. J. Surg., Vol. 77. NO. 11, November 1990
23. 24. 25. 26. 27.
Berchstein WO, Blumhardt G, Ringe B et a/. Surgical complications in 200 consecutive liver transplants. Trun.splun/ Proc 1987; 19: 3830-1. Zajko AB, Campbell WL, Logsdon GA ef a/. Cholangiography and interventional biliary radiology in adult liver transplantation. AJR 1985; 144: 127-33. Gillatt DA, May RE, Kennedy R, LongstaB AJ. Complications of T tube drainage of the common bile duct. Ann R Coll Surg 1985; 67: 370-1. Pitt H A , Postier RG, Camoeron JL. Postoperative T tube cholangiography. Is antibiotic coverage necessary? Ann Surg 1980; 191: 30-4. Fonseca C, Rosenthal L, Greenberg D et d .Differential diagnosis of jaundice by ""Tc-IDA hepatobiliary imaging. Clin Nucl Med 1979; 4: 135-72. Pauwels S, Piret L, Schouteus A er ul. Tc-99m-diethyl-IDA imaging: clinical evaluation in jaundiced patients. J Nucl Med 1980; 21: 1022-8. Schwarzzock R, Kotzerke J, Hunderschagen H, Bocker K, Ringer B. ""Tc-diethyl-IODO-HIDA (IODIDA): a new hepato-biliary agent in clinical comparison with ""Tcdisopropyl HIDA (DISIDA) in jaundiced patients. Eur J Nucl Med 1986; 12: 346-50. Chervu LR, Nunn AD, Lobero UD. Radio pharmaceuticals for hepatobiliary imaging. Sem Nucl Med 1982; 12: 5-17. Nunon AD, Loberg MD, Conly RA. A structure distribution relationship approach leading to the development of Tc-99m mebrofenim: an improved cholescintigraphic agent. J Nucl Med 1983; 24: 423-30. Klingersmith WC, Fritzbege AR, Spitzer VU, Kuni CC. Shanahan WSM. Clinical comparison of diisopropyl IDA-Tc99m and diethyl-IDA-Tc-99m for evaluation of the hepatobiliary system. Rudioloyj* 1981; 140: 791-5. Williams W, Krishnamurthy GT, Brar H et ul. Scintigraphic variability in normal biliary physiology. J Nucl Med 1984; 25: 16C-5. Krishnamurthy GT, Lieberman DA, Brar HS. Detection, localisation and quantisation of degree of common bile duct obstruction by scintigraphy. J Nucl Med 1985; 26: 726-35. Blue PW. Biliary scanning interpretation using Tc-99m DISIDA. Clin Nucl Med 1985; 10: 742-5. Blue PW. Hyperacute complete common bile duct obstruction demonstrated wih Tc-99m IDA cholescintography. Nucl Med Commun 1985; 6: 275-9. Klingensmith WC, Whitney WP, Spitzer VU et a/. Effect of complete biliary tract obstruction on serial hepatobiliary imaging in an experimental model (concise communication). J Nucl Med 1981; 22: 866-8. Lee AW, Ram UD, Shih W-J, Murphy K. Technetium-99m BlDA biliary scintigraphy in the evaluation of the jaundiced patient. J Nucl Med 1986; 27: 1407-12. Kuni CC, Klingensmith 111 WC, Fritzbeso A R . Evaluation of intrahepatic cholestasis with radionuclide hepatobiliary imaging. Gustrointest Rudiol 1984; 9: 163-6. Kirby RM, McMaster P, Clements D ef ul. Orthotopic liver transplantation: postoperative complications and their management. Br J Surg 1987; 74: 3-11. Hubscher SG, Clements D. Elias E, McMaster P. Biopsy findings in cases of rejection of liver allograft. J Clin Puthol 1985; 38: 1366-73. Jain AB, Elias E, Gunson BK et a/. Is elimination of cyclosporin (CyA) in bile dependent on liver graft function and biliary drainage? Transplant Proc 1988; 20: 516-22. Hubscher S, Clements D, Elias E, McMaster P. Liver graft rejection and /l,-microglobulin. Lancet 1985; 15: 1391. Rosenthal L. Cholescintigraphy in the presence of jaundice utilising Tc-IDA. Sem Nucl Med 1982; i: 53-63.
Paper accepted 12 May 1990
1237
M. Anselmi, S. Lancberg*, M. Deakin, E. Lanchbury*, 2.Drolc*, F. Burrows?, E. Elias and P. M c M a s t e r The Liver Unit and Departments of *Nuclear Medicine and ?Radiology, The Queen Elizabeth Medical Centre, Birmingham, UK Correspondence to: Mr P. McMaster, The Liver Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
Assessment of the biliary tract after liver transplantation: T tube cholangiography or IODIDA scann ing Biliary tract obstruction or anastomotic leakage are common problems following liver transplantation. In a sequential study, 31 patients with a liver transplant were investigated by 99mTc-IODIDA ( I O D I D A ) scanning and T tube cholangiography ( T T C ) and the results tc-ere compared with clinical outcome. Seven patients had an extrahepatic biliary obstruction and one patient had a biliary leak. In the detection of biliary complications T T C and IODIDA scanning were similar in terms of sensitivity (63per cent,for both) but T T C hada better specificitji (79 per cent versus 60 per cent) and accuracy (74 per cent versus 6 0 per cent) than IODIDA scanning. When liver ,function was taken into account, the diagnostic efii'cacy of both tests in patients with bilirubin levels of < 200 ,umol/l was similar. With levels > 200 ,umol/l there M ~ a greater number of' false positive results with IODIDA scanning (12 per cent versus 54 per cent). The only significant biliary leak MUY clearly detected by T T C but not IODIDA scanning. T T C remains the more effective way of evaluating the biliary tract ufter transplantation. IODIDA scanning has limited value when bilirubin levels are elevated, but may provide additional information about blood supply, hepatocyte func tion and intrahepat ic cholestasis. I
Keywords: Liver transplantation, biliary tract reconstruction, biliary anastomosis, biliary tract complication. T tube cholangiography, biliary scintigraphy, yy'"Tc-IODIDA scintiscanning
Improvement in surgical technique and use of low-dose steroids have resulted in fewer biliary tract complications following liver transplantation'-'. Nearly one in five patients still experiences problem^^.^ and assessment of the biliary tract is performed routinely 7 days post-transplantation and later if clinically indicated. Cholangiography (T tube, percutaneous transhepatic or endoscopic retrograde) has proved useful in demonstrating the type, location and severity of most early and late biliary complications'. Injection of contrast into the biliary tree is frequently related to episodes of cholangitis and bacteraemia'.' which are especially dangerous in immunosuppressed patients. Investigation is more diffcult when the patient does not have a T tube. Hepatobiliary scintigraphy is a non-invasive, widely available and inexpensive technique. In addition to identifying biliary t.ract complications it may provide useful information about hepatocyte function and intrahepatic cholestasis".' ' . IODIDA (Solcoscint IODIDA, SOLCO Balse Ltd., Birsfelden, Switzerland) has a lower urinary and higher biliary excretion and has significant advantages over other "'Tc-Sn-2,6diethylacetanilidoiminodiacetate (HIDA) derivatives especially when bilirubin levels exceed 80-100 p m ~ l / l ' ~ - 'It~ .has been advocated in diagnostically diffcult situations such as liver transplant, bile duct reconstruction, suspicion of bile leakage, examination in small children or of critically ill patients. A prospective study was undertaken to compare the diagnostic value of ""Tc-IODIDA isotope (IODIDA) scanning and T tube cholangiography (TTC) in patients after liver transplantation.
Patients and methods Thirty-one liver transplant patients underwent T T C and hepatobiliary scintigraphy. Assessment was performed routinely in 26 patients between 7 and 10 days after surgery and at a later stage in five patients
when suspicion of complications arose. Both tests were performed within 48 h and no change in the clinical situation was observed in any patient between the two studies. The indications for liver transplantation were: primary biliary cirrhosis in 12 patients, fulminant hepatic failure in six, chronic active hepatitis in four, sclerosing cholangitis in three, a-I-antitrypsin deficiency in two, cryptogenic cirrhosis in two, hepatocellular carcinoma in one and alcoholic cirrhosis in one. There were two children (aged 1.5 and 7 years) and 29 adults whose mean age was 40.7 years (range 18-68 years). Seventeen of the group were female. The types of biliary reconstruction used are shown in Figure I : ( a ) end-to-end choledochocholedochostomy in 16 patients (52 per cent); (b) gallbladder conduit in 11 (36 per cent); (c) choledocho-Roux-enY-jejunostomy in three cases (10 per cent); and ( d ) in one case a conduit to jejunum (3 per cent). Bilirubin levels at investigation ranged from 25 to 883 pnol/l and were less than 100pmol/l in nine patients, less than 200pmolll in 15 patients and greater than 200 pmol/l in the remaining 17 cases. Both imaging studies were reported independently by an examiner who did not have access either to clinical information or to the results of the other test. The T tube cholangiograms wcre read 'blind' by a consultant radiologist with experience in examining transplanted patients (F.B.) and the IODIDA scans were read 'blind' by a nuclear medicine specialist (S.L.). Scin/igruphj.nw/l7ocl After fasting for at least 4 h 80 MBq of commercially available IODIDA were injected as a bolus into an antecubital vein. With the patient supine, images of the abdomen were obtained using a large field of view y a m e r a with a low-energy all-purpose collimator and recorded on computer at one frame/min for 40 min. Analogue film images were also obtained every 5 min over this same interval. Further images were obtained at 1-4 h and 18-24 h if IODIDA clearance was delayed or a leak suspected. The results of scintigraphy were cl
I. 2.
No ohsrruciion; no obstruction to the flow of the scintigraphic agent into the bowel taking into account normal variation". Purrid ohs/ruc/ion; detection of partial obstruction was primarily
-~ 0007-1 323/90/1 I123345
(
1990 Butterworth-Heinemann Ltd
1233
Y
Biliary tract assessment after liver transplantation:
M.Anselmi et al.
a
C
d
\ Figure 1 Schematic diagram showing the types of biliary reconstrucrion used in 31 liver transplant recipients: choledochocholcystcholedochostoniy; c Roux-en- Y choledochojejunostomy; d Rour-en- Y choledochocholecys~ejunostomy a end-to-end choledochocholedocho.~tomy;b
3. 4.
based on filling defects in narrowing of the common bile duct or bile pooling in the segmental ducts” and a delay of 60min or more before appearance of isotope into the bowel. Complete obstruction: absence of the scanning agent from the small bowel after 24 h regardless of whether or not images of the bile duct had appearedI8-”. Presence or absence of biliary leak: the appearance of radioactivity in the peritoneal cavity.
Information was available on impaired hepatic uptake of radionuclide indicating primary hepatic dysfunction and intrahepatic cholestasis manifested by relatively good uptake of IODIDA, delayed visualization of intestine between 4 and 24 h and visualization of the bile duct adequate to rule out extrahiliary obstruction2’.z2. T tube cholangiography ( T T O TTC was performed with antibiotic cover. Approximately 2C30 ml urografin 325 were slowly injected during screening to display as much of the duct system as possible and to avoid discomfort, overfilling of the duct or bacteraemia. If contrast ran freely through the lower duct (or conduit) into the bowel without showing the upper ducts, a repeat injection was performed with the patient in head-down position. The results of TTC were classified as follows:
I. 2. 3.
N o obstruction: no obstruction to the flow of contrast into the bowel with non-dilated intrahepdtic ducts. Purtiul ohstruction: the presence of strictures or filling defects within the extrahepatic or intrahepatic bile ducts causing delay in passage of contrast into the bowel. Complete obstruction: obstruction to the passage of contrast either
1234
4.
into the bowel or alternatively from common hepatic duct to intrahepatic ducts. Presence or absence o f a leak.
In addition to scintiscanning and radiological examination all patients underwent detailed clinical monitoring, biochemical assessment. protocol liver biopsy on day 7 post-transplant and later if required. The ultimate assessment of the biliary tract and final diagnosis were based o n the patient’s clinical course and information which became available either through reoperation or autopsy.
Results No biliary tract complications were found in 23 patients (74per cent), seven had extrahepatic obstruction (23 per cent) and one had a biliary leak (3 per cent). The process of final diagnosis is shown in Table 1. T tube cholangiography (TTC) In the 23 patients without biliary complications T tube cholangiograms were reported normal in 18. There were five false positive studies in which the diagnosis of either complete o r partial obstruction was made. Three of these patients had a gallbladder conduit: in two of these the diagnosis of complete obstruction of the upper choledochocholecystanastomosis had been made (Figures2a and b ) and in one partial obstruction at the papilla with proximal dilatation was reported. The fourth patient had a duct-to-duct reconstruction and partial
Br. J. Surg.. Vol. 77, No. 11, November 1990
Biliary tract assessment after liver transplantation: M. Anselmi et al.
Table 1 Means of’ oeri/ication of the,final diagnosis in 31 cases Final diagnosis Autopsy
Operation
Clinical course
(n=5 )
(n=7)
( n = 19)
Biliary leak ( n = 1 )
1
Obstruction ( n = 7 ) Inflammation of conduit Anastornotic stenosis Biliary sludge
1 1 3
No biliary complications ( n = 23)
4
Again no changes in the diagnosis were suggested after unblinded review of the images when clinical information was available. The diagnostic accuracy of TTC and IODIDA scanning in predicting eventual clinical outcome is shown in Table 2 . Both
~
2 -
2
17
obstruction of the anastomosis was suspected. In the remaining patient with a choledochojejunal reconstruction, a proximal non-anastomotic obstruction was diagnosed. In none of these patients did the liver function tests, histology or clinical outcome indicate any clinically relevant biliary obstruction during 2-24 months of follow-up. The patient with the biliary leak was correctly diagnosed on T tube cholangiogram. In the seven patients with extrahepatic biliary obstruction, four were shown on the T tube cholangiograms. There were three false negative studies. Two were in patients with a duct-to-duct anastomosis. In the first of these patients a hepatic artery thrombosis was diagnosed on ultrasonography and confirmed angiographically. Reoperation showed a n anastomotic stricture and the anastomosis was converted to a Roux-en-Y choledochojejunostomy . In the second patient a papillotomy was performed and biliary sludge was removed 2 weeks after the TTC (Figures3a and b ) . The other patient with a false negative study had a gallbladder conduit. Reoperation because of clinically increasing cholestasis and abdominal pain revealed obstruction of the common bile duct due to inflammation of the conduit and after cholecystectomy a Roux-en-Y choledochojejunostomy was performed. When the studies were reassessed in the light of full clinical information no changes in the radiological diagnosis were made ruling out an error in the interpretation of the images. IOD ID A scanning In one of the 23 patients without proven biliary complications there was poor hepatic uptake of IODIDA indicating pure hepatocellular dysfunction and the biliary tract was not visualized. In the other 22 normal patients, 13 were reported to have normal scans. There were nine false positive scans. Six were reported as completely obstructed. Four of these patients had histologically proven moderate rejection, one had mild rejection and pneumonia and one had severe sepsis. Three scans were reported as partially obstructed on the basis of delayed appearance of radioactivity in the gut and poor delineation of the biliary tract. One of these patients had moderate rejection on biopsy, one had sepsis and mild rejection and one had an intra-abdominal abscess. The mean bilirubin level of the nine patients with false positive results was more than twice that of patients with true negative IODIDA scanning results (Figure 4 ) . The false positive rates for IODIDA scanning and T T C were 54 per cent and 14 per cent when the bilirubin level was >200pmol/l and 12 per cent and 18 per cent when the level was < 200 pmol/l. Scintiscanning failed to detect leakage of isotope into the abdominal cavity in the patient with a biliary leak. In the seven patients who were eventually shown to have biliary obstruction IODIDA scanning correctly diagnosed five. There were two false negative studies. Both patients had a duct-to-duct anastomosis and were diagnosed as having enlarged but unobstructed ducts and showed appearance of radioactivity in the gut within 35 min. Evidence of obstruction was found in both and corrected by reoperation in one and endoscopic dilatation in the other.
Br. J. Surg., Vol. 77, No. 11, November 1990
b
111
IV
Figure 2a Esaniple of’ a ,false po.sitir.e T tuhe cholunyioyram prohubiJ due to undetfilliny of flit hiliarj. systeni in a patient w i t h a gullhludder conduit reconstruction. The studv II‘US perfiwnied 27 d a . ~after ~ transplantation clue t o dcteriorution of the liver ,function tests (bilirubin ~i’as60 pniol/l and alkaline phosphatase II’CIS 970 unitsll). In this puticvit thc cliokmyiogram II’US reported U S a complete obstruction of the donor coninion hile duct to donor gullhiadder anasloniosi,s. N(J i~b.struclifJliIW.Y confirmedin the,follo~c~-upperiod. b Selectedanterior r.icw:s.froniIODIDA stucly ohtained U I (02-5 niin. (10 17-20 niin, ( I l l ) 27-30 niin utid ( 1 0 35-38 niin after injeerion. Thy,*s l m r good estrucrion nf IODIDA ,from the Mood pool (acticily within the Iieart is ,seen o n l ~ on , the iniuge ( i t 2 min) arid nornial clearance ,from parench~~niu into the ya.stroiiire.stiiiri1 tract
1235
Biliary tract assessment after liver transplantation: M. Anselmi et al. 700
--
1
l
0
o
600
O
pk o s i t i v es n e g a t Fi v e
l
False False positive negative
Figure 4 Comparison of the resulrs of IODIDA scunning study und hiliruhin levels Table 2 Compurison of the sensitinitj, spec$'city and uccurucy of TTC and IODIDA scanning in the diagnosis of hiliury truct complications after liver rransplantation ~
TTC (all patients) TTC (EECC, n= 16) TTC (GBCD, ! I = 11) IODIDA (all patients) IODIDA (EECC, n= 16) IODIDA (GBCD, n= 11)
Sensitivity
Specificity
Accuracy
(YO)
(YO)
(YO)
63
78
74
61 50
92 57
88 55
63 50 100
59 67 88
60 62 90
TTC, T t u b e cholangiography; IODIDA, 99"Tc-IODIDA scanning; EECC, end-to-end anastomosis of common bile duct; GBCD, gallbladder conduit reconstruction
tests had similar sensitivity but TTC had a better specificity and accuracy than IODIDA scanning. The only biliary leak was clearly shown on TTC but not identified on IODIDA scanning. The diagnostic value of the tests in the detection of biliary obstruction was influenced by the type of biliary reconstructive procedure. TTC was more accurate than IODIDA scanning in detecting obstruction in patients with end-to-end choledochocholedochostomy, but in those with a gallbladder conduit IODIDA scanning was more efficient (Table 2 ) . When the liver function tests did not show cholestasis, combination of the results ofTTC and IODIDA increased the sensitivity to 85 per cent and specificity to 91 per cent with an accuracy of90 per cent.
b
Discussion
111
IV
Figure 3a A n o.wniple u / X s e negutiw rutiiologicul studj.. Prutocol T tuhe cholungiogrup1i.v perfornird 6 rlu!.s ulier rrunspluntution in u patient with end-to-end unosroniosis qf the coninion hile duct. The .seruni hiliruhin IVUS 131 pmolll mid ulkuline phosphuruse II~US303 units//. TIW weeks hrer this putient ii'us suhniitted to an endoscopic papil1otoni.s und hiliurj.sluc!qe i t u s rmiowd,fiom the conimion hile durl. b Selected anterior iieaxfrom IODIDA stud). ohtuincrl ut (0 2-5 niin. (10 27-30 min. ( I l l ) I h und ( I V ) 24 h ufter injection. Prolonged retention u/' uctirity in the hlood pool. heart und other r.usculur swuctures, re1utii;eI). good u p t u h hj. the y r u o und uhsmce of'r.\-crrtion into gustrointesrinul tract indicutrs oh.rtri~tioii
1236
Liver transplant recipients are susceptible to a wide variety of complications which may threaten allograft and patient ~ u r v i v a l ~ ,Biliary '~. complications, hepatic preservation injury, rejection, hepatic artery thrombosis and infection may all lead to a rapid deterioration in liver function. Differentiating between the causes may prove difficult. Early diagnosis of biliary complications is important in planning treatment and because other post-transplant problems are often diagnosed by exclusion24. Oral cyclosporin is poorly and unpredictably absorbed from the intestine during bile diversion". Exclusion of biliary complications as a cause of deteriorating liver function allows early clamping of the T tube, with re-establishment of the enterohepatic circulation. Our results have shown that the sensitivity of both tests is similar but TTC has a better specificity and overall accuracy than IODIDA scanning. In patients with bilirubin levels above 200 pmol/l a greater number of false positive results appeared with IODIDA scanning because of misdiagnosed intrahepatic cholestasis due
Br. J . Surg., Vol. 77, No. 11. November 1990
Biliary tract assessment after liver transplantation: M. Anselmi et al.
t o rejection or sepsis. Hepatobiliary scintigraphy can provide accurate information in the evaluation of choIestasis22.26but this is difficult in patients with underlying intrahepatic disease where extrahepatic obstruction is suspected. In the transplant patients rejection affects small bile ducts leading to intrahepatic c h o l e ~ t a s i swhich ~ ~ . ~ can ~ mimic complete biliary obstruction". Six out of nine of o u r false positive studies showed this pattern. In the remaining three patients, poor visualization of the biliary tract and delayed intestinal appearance made scintigraphic interpretation equivocal as ductal delineation was inadequate. We found that in this g r o u p of patients, with bilirubin levels over 200 pmol/l, IODIDA scanning had a very low specificity and overall accuracy. Discrimination between high grade obstruction and severe intrahepatic disease based on the image pattern was not possible a n d suggests that the level of obstruction in such a case must be made by other modalities. The accuracy of TTC is not influenced by a high serum bilirubin level or by graft dysfunction. TTC does, however, require meticulous technique if cholangitis a n d bacteraemia ar e to be avoided. Required low pressure injection c a n result i n underfilling of the biliary tract as c a n rapid emptying of contrast into the intestine through a widely patent biliary+nteric anastomosis. Underfilling may explain a t least three of five false positive studies especially in the patients with gallbladder conduits a n d combination of the results of TTC and IODIDA scanning in these cases may be helpful if severe cholestasis is not present. The many pitfalls encountered with scintigraphy in the differential diagnosis of jaundice with high bilirubin levels undermines its usefulness as a first choice diagnostic procedure after transplantation. Scintigraphy c a n play a n adjuvant role and be of considerable clinical value in cases where technical difficulties yield indeterminate TTC results such as with a gallbladder conduit anastomosis. I n spite of its limitations, TTC still remains the most effective means of evaluating the biliary tract after transplantation. IODTDA scintiscanning may provide important information a b o u t the biliary tract a n d graft function with bilirubin levels below 200 pmol/l, but is unreliable in the presence of rejection, sepsis o r when bilirubin levels exceed 200 pmol/l.
6. 7. 8. 9. 10. 11.
12.
13. 14.
15.
16. 17. 18. 19. 20.
Acknowledgements Mr Mario Anselmi is a Fellow of the British Council. We would like to thank the medical, nursing and paramedical staff involved in the care and management of these patiens.
21. 22.
References 1.
2. 3.
4. 5.
Calne RY. A new technique for biliary drainage in orthotopic liver transplantation utilising the gall bladder as a pedicle graft conduit between the donor and recipient common bile duct. Ann Surg 1976; 184: 605-9. Wolff H, Otto G, David H. Biliary tract reconstruction in liver transplant. Transplant Proc 1985; 17: 274-5. Bismuth H, Castaing D, Gugenheim J, Traynor 0,Ciardullo M. Roux-en-Y hepatojejunostomy : a safe procedure for biliary anastomosis in liver transplantation. Transplunt Proc 1987; 19: 241 3- 15. Neuhaus P, Brolsch C, Ringe B, Lauchart W, Pichlmayr R. Results of biliary reconstruction after liver transplantation. Transplant Proc 1984; 16: 1225-7. Koneru B, Tzakis AG, Bowman J, Cassavilla A, Zajko AB, Starzl TE. Liver transplantation. Post operative surgical complication. Gastroenterol Clin North A m 1988; 17: 71-91,
Br. J. Surg., Vol. 77. NO. 11, November 1990
23. 24. 25. 26. 27.
Berchstein WO, Blumhardt G, Ringe B et a/. Surgical complications in 200 consecutive liver transplants. Trun.splun/ Proc 1987; 19: 3830-1. Zajko AB, Campbell WL, Logsdon GA ef a/. Cholangiography and interventional biliary radiology in adult liver transplantation. AJR 1985; 144: 127-33. Gillatt DA, May RE, Kennedy R, LongstaB AJ. Complications of T tube drainage of the common bile duct. Ann R Coll Surg 1985; 67: 370-1. Pitt H A , Postier RG, Camoeron JL. Postoperative T tube cholangiography. Is antibiotic coverage necessary? Ann Surg 1980; 191: 30-4. Fonseca C, Rosenthal L, Greenberg D et d .Differential diagnosis of jaundice by ""Tc-IDA hepatobiliary imaging. Clin Nucl Med 1979; 4: 135-72. Pauwels S, Piret L, Schouteus A er ul. Tc-99m-diethyl-IDA imaging: clinical evaluation in jaundiced patients. J Nucl Med 1980; 21: 1022-8. Schwarzzock R, Kotzerke J, Hunderschagen H, Bocker K, Ringer B. ""Tc-diethyl-IODO-HIDA (IODIDA): a new hepato-biliary agent in clinical comparison with ""Tcdisopropyl HIDA (DISIDA) in jaundiced patients. Eur J Nucl Med 1986; 12: 346-50. Chervu LR, Nunn AD, Lobero UD. Radio pharmaceuticals for hepatobiliary imaging. Sem Nucl Med 1982; 12: 5-17. Nunon AD, Loberg MD, Conly RA. A structure distribution relationship approach leading to the development of Tc-99m mebrofenim: an improved cholescintigraphic agent. J Nucl Med 1983; 24: 423-30. Klingersmith WC, Fritzbege AR, Spitzer VU, Kuni CC. Shanahan WSM. Clinical comparison of diisopropyl IDA-Tc99m and diethyl-IDA-Tc-99m for evaluation of the hepatobiliary system. Rudioloyj* 1981; 140: 791-5. Williams W, Krishnamurthy GT, Brar H et ul. Scintigraphic variability in normal biliary physiology. J Nucl Med 1984; 25: 16C-5. Krishnamurthy GT, Lieberman DA, Brar HS. Detection, localisation and quantisation of degree of common bile duct obstruction by scintigraphy. J Nucl Med 1985; 26: 726-35. Blue PW. Biliary scanning interpretation using Tc-99m DISIDA. Clin Nucl Med 1985; 10: 742-5. Blue PW. Hyperacute complete common bile duct obstruction demonstrated wih Tc-99m IDA cholescintography. Nucl Med Commun 1985; 6: 275-9. Klingensmith WC, Whitney WP, Spitzer VU et a/. Effect of complete biliary tract obstruction on serial hepatobiliary imaging in an experimental model (concise communication). J Nucl Med 1981; 22: 866-8. Lee AW, Ram UD, Shih W-J, Murphy K. Technetium-99m BlDA biliary scintigraphy in the evaluation of the jaundiced patient. J Nucl Med 1986; 27: 1407-12. Kuni CC, Klingensmith 111 WC, Fritzbeso A R . Evaluation of intrahepatic cholestasis with radionuclide hepatobiliary imaging. Gustrointest Rudiol 1984; 9: 163-6. Kirby RM, McMaster P, Clements D ef ul. Orthotopic liver transplantation: postoperative complications and their management. Br J Surg 1987; 74: 3-11. Hubscher SG, Clements D. Elias E, McMaster P. Biopsy findings in cases of rejection of liver allograft. J Clin Puthol 1985; 38: 1366-73. Jain AB, Elias E, Gunson BK et a/. Is elimination of cyclosporin (CyA) in bile dependent on liver graft function and biliary drainage? Transplant Proc 1988; 20: 516-22. Hubscher S, Clements D, Elias E, McMaster P. Liver graft rejection and /l,-microglobulin. Lancet 1985; 15: 1391. Rosenthal L. Cholescintigraphy in the presence of jaundice utilising Tc-IDA. Sem Nucl Med 1982; i: 53-63.
Paper accepted 12 May 1990
1237
