Doppler ultrasound (DUS) has the potential to

Transcription

1 Prospective Evaluation of the Role of Quantitative Doppler Ultrasound Surveillance in Liver Transplantation David Stell, Donal Downey, Paul Marotta, Edward Solano, Anand Khakhar, Douglas Quan, Cam Ghent, Vivian McAlister, and William Wall Doppler ultrasound (DUS) is able to measure parameters of blood flow within vessels of transplanted organs, and vascular complications are associated with abnormal values. We analyzed the findings of 51 consecutive patients who underwent DUS on 2 occasions in the first postoperative week following liver transplantation for cirrhosis to determine the range of values in patients following liver transplantation. Three patients developed early vascular thromboses that were detected by the absence of a Doppler signal. In patients making an uneventful recovery, the arterial velocity tended to increase and the resistive index (RI) to decrease during the first postoperative week. All recipients were shown to have high-velocity segments within the hepatic artery, without an increase in flow resistance. Assessment of the portal vein revealed narrowing at the anastomosis, associated with a segmental doubling of flow velocity, and the mean portal venous flow decreased by approximately 20% in the first postoperative week. In conclusion, a wide range of abnormalities occurs in the vessels of liver transplant recipients, which were not associated with the development of vascular complications or affect patient management. (Liver Transpl 2004; 10: ) Doppler ultrasound (DUS) has the potential to identify vascular pathology in liver transplant recipients, including stenosis and thrombosis of the hepatic artery and portal vein. Quantitative DUS has been shown to be useful in the care of kidney transplant recipients, 1 but its value in liver transplant recipients has been assessed largely by the retrospective correlation of DUS and angiography in patients with abnormal graft function. 2 5 DUS has a high specificity for the diagnosis of hepatic artery thrombosis. The absence of any arterial signal at the porta hepatis in the presence of abnormal liver function is highly suggestive of hepatic artery thrombosis. 2,5,6 DUS of the portal vein has also been performed, largely in patients presenting with abnormal graft function. 7,8 The value of quantitative DUS has not, however, been assessed in a prospective series of liver transplant recipients, and assessment of the range of arterial and venous Doppler indices and the prevalence of narrowing at the portal venous anastomosis following liver transplantation have not been established. DUS can measure the dimensions of blood vessels and parameters of arterial flow including the peak systolic velocity (Vs) and trough diastolic velocity (Vd), and the systolic acceleration time (SAT; the time taken to reach peak Vs after trough Vd). The resistive index (RI) describes the relationship between Vs and Vd (Vs Vd / Vs). A large difference between peak Vs and trough Vd indicates resistance to arterial flow and is reflected in a higher RI. In low resistance systems or in a vessel distal to an arterial stenosis, there may be little difference between the peak Vs and trough Vd, which is reflected as a low RI and prolonged SAT. The normal (nontransplant) hepatic artery values for peak Vs, RI, and SAT lie within the range cm/second,.6.8, and 80 ms, respectively. 9,10 The primary Doppler signal used in the detection of arterial stenosis in superficial vessels is a focal increase in arterial velocity associated with turbulent flow. 11 In deep abdominal vessels, such as those supplying the liver, precise localization of vascular lesions is more difficult 12 and the presence of proximal stenosis can be inferred by an assessment of the arterial waveform distal to the stenosis, which shows a characteristic pattern of a slow SAT and diminished Vs (the tardus et parvus waveform). 2 The RI will therefore be elevated proximal to, and decreased distal to, the high velocity segment. Similarly, the loss of pulsatile flow distal to an arterial stenosis is associated with a prolongation of the SAT. Complications affecting the portal vein after liver transplantation include thrombosis and anastomotic stenosis of the venous lumen, which are detected by the absence or acceleration of blood flow. 13 The aim of this study, therefore, was to assess the value of routine DUS in liver transplant recipients and Abbreviations: DUS, Doppler ultrasound; Vs, systolic velocity; Vd, diastolic velocity; RI, resistive index; SAT, systolic acceleration time. From the Multi-Organ Transplant Program, London Health Sciences Centre, London, Ontario, Canada. Address reprint requests to David Stell, Liver Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2TH, UK. Telephone: ; FAX: ; davidstell@doctors.org.uk Copyright 2004 by the American Association for the Study of Liver Diseases Published online in Wiley InterScience ( DOI /lt Liver Transplantation, Vol 10, No 9 (September), 2004: pp

2 1184 Stell et al. Table 1. Demographics of the Study Population (n 45) Age (yr; median interquartile range) 52 (45 58) Gender M:F 25:20 Diagnosis Viral liver disease 13 Alcoholic liver disease 8 Cholestatic liver disease 9 Cryptogenic cirrhosis 3 NASH 5 Autoimmune liver disease 2 Other 5 Abbreviation: NASH, non-alcoholic steatohepatitis. to determine the range of Doppler indices in patients with recovering graft function after liver transplantation in whom major vascular complications did not occur. Patients and Methods Between July 2002 and March 2003, 51 patients underwent routine DUS within 48 hours of receiving a primary cadaveric liver transplant, followed by a second DUS between 5 and 7 days postoperatively. Of the 51 liver transplant recipients evaluated in the DUS protocol, 3 patients developed major vascular complications requiring surgical intervention in the early postoperative period. One patient developed portal vein thrombosis and 2 patients developed hepatic artery thrombosis, which were detected by DUS before the development of major graft dysfunction. The patients with vascular thrombosis were explored surgically and successful revascularization was performed. In addition, 3 patients developed acute cellular allograft rejection within the first 7 postoperative days. The remaining 45 recipients did not develop significant graft dysfunction requiring treatment within the first postoperative week (shown by improving liver function tests) and these recipients form the study group for the determination of Doppler indices after liver transplantation. (Table 1). All patients were transplanted for cirrhosis; the patient demographics are shown in Table 1. All patients received whole grafts and transplantation was performed with caval replacement without the use of veno-venous bypass. Arterial anastomosis was performed using an arterial patch with reconstruction of accessory vessels where necessary. All exams were performed by qualified sonographers accredited by the American Society of Diagnostic Medical Sonographers. All exams were performed with HDI 5000 ultrasound machines (Philips/ATL, Bothell, WA). The highest frequency curved array ultrasound transducer that could resolve the Doppler signals was used, either the C7-4 or the C5-3. Doppler images were recorded and stored on hard copy films. These were reviewed by a transplant surgeon (D.S.) and a radiologist (D.D.). DUS characteristics, including maximum hepatic arterial Vs and Vd and the RI were recorded. The arterial velocity and RI were recorded at the site of peak velocity and at a site before and after the high flow segment in the largest hepatic artery detectable at the porta hepatis. The SAT was recorded only at a distal site within the hepatic artery, where it is of greatest value in identifying arterial stenosis. The narrowest diameter of the portal vein was recorded along with the diameter at sites proximal and distal to this point, and the maximum flow velocity (Vmax) measured at each point. The portal vein blood flow was calculated from recordings in the proximal portal vein and at the narrowest point (recordings from the distal portal vein were not used as turbulent blood flow produced a wider range of flow velocities), according to the formula 14 : Flow (L/minute) area.57 Vmax. Data were collected prospectively in a database format. Statistical analysis between 2 group medians was performed by Mann-Whitney test and Kruskal-Wallis analysis of variance for multiple groups. Results Graft recipients in general showed a wide variation in Doppler indices at both time points after transplantation, in both the hepatic artery and portal vein. Blood flow velocities in particular showed very high variability between patients. In many patients the data were incomplete. This usually occurred because of difficulty in identifying arterial flow signals on one side of the peak flow segment or the other. Arterial Velocity All 45 patients showed nonuniform flow within the hepatic artery, with a high velocity segment and lower velocity on either side. The median peak arterial velocity for the whole group was 103 cm/second at the first examination and 122 cm/second at the second examination (Fig. 1). A wide range of peak arterial velocities was noted, from 13.2 to 367 cm/second. A total of 4 patients had a peak hepatic arterial velocity greater than 200 cm/second at the first examination and 6 patients at the second examination. The arterial velocity tended to increase over the first postoperative week (Fig. 1). Resistive Indices The median RI measured before, at, and after the peak velocity segment in the extrahepatic portion of the

3 Quantitative DUS in Liver Transplantation 1185 Figure 1. Median arterial velocity (interquartile range) (cm/second) at 3 sites (prepeak, peak, and postpeak arterial velocity) within the transplant hepatic artery, recorded within 2 days and 5 to 7 days posttransplantation. Differences not significant except *P <.05. hepatic artery were.76,.78, and.76, respectively, at the first examination and.71,.74, and.69, respectively, at the second examination (Fig. 2), all within the normal range for a visceral artery. The RI decreased significantly at each site over the first postoperative week (Fig. 2). A total of 38% of liver transplant recipients with good graft function, however, showed RI values greater than.8 in the early postoperative period, while only 4% showed values less than.6. Over the first 7 days, resistance to hepatic arterial flow tended to decrease, with only 10% of patients recording values greater than.8, and 11% less than.6, by day 7. Systolic Acceleration Time (SAT) The median SAT measured in the distal hepatic artery was 70 ms at the first examination (interquartile range: ms) in 29 recipients and 75 ms at the second Figure 3. Median arterial velocity and resistive index (interquartile range) in 17 patients recorded within 48 hours of liver transplantation. Median arterial velocity (diamonds) increased significantly in high flow segment (P.001), while resistive index (RI) (squares) did not vary significantly across this segment. examination (interquartile range ms) in 26 recipients, within the normal range for a visceral artery. However, 5 patients at the first examination and 4 at the second examination recorded SATs greater than 100 ms. Correlation between Hepatic Arterial Flow Velocity and Resistive Index A total of 17 patients had complete values available at the first examination for arterial velocity and RI at the segment with maximum flow velocity and in segments proximal and distal to this point, and 21 patients had complete values at the second examination. In these patients, high velocity segments were found in the absence of changes in resistance to flow. At the first examination, the arterial flow velocity increased significantly in a segment of hepatic artery (median flow velocity: 74 cm/second, cm/second, and 54 cm/second, proximal to, at, and distal to the high flow segment, respectively; n 17) although the resistive index did not change significantly at these sites (Fig. 3). Similarly, at the second examination, the arterial flow velocity increased significantly in a segment of hepatic artery (median flow velocity: cm/second, cm/second, and 70 cm/second, proximal to, at, and distal to the high flow segment, respectively; n 21) and the RI did not change significantly across these 3 sites (Fig. 4). Figure 2. Median RI (resistive index) (interquartile range) at 3 sites (prepeak, peak, and postpeak arterial velocity) within the transplant hepatic artery recorded within 2 days and 5 to 7 days posttransplantation. Differences significant; *P.04, **P.007, ***P.007. Portal Vein Diameter All recipients showed a narrowing in the portal vein lumen at the anastomosis. The recordings of portal vein diameter were very similar at both examinations, reduc-

4 1186 Stell et al. Figure 4. Median arterial velocity and resistive index (interquartile range) in 21 patients recorded between 5 and 7 days after liver transplantation. Median arterial velocity (diamonds) increased significantly in high flow segment on both days (P.001), while resistive index (RI) (squares) did not vary significantly across this segment. ing from a median of 1.2 cm in the proximal portal vein to.7 cm at its narrowest point (Fig. 5). The range of reduction in vein diameter at the anastomosis was.05 to.78 cm at the first examination and.12 to 1 cm at the second examination. Portal Venous Flow Velocity The median flow velocity increased significantly at the site of narrowest diameter, from to /second at the first examination and from 41 to /second at the second examination (Fig. 6). A very wide range of portal venous velocities were noted, from 15 to 400 /second. Turbulent flow was often detected in the portal vein distal to the anastomosis. There was a significant reduction in flow velocity over the first postoperative week (Fig. 6). The calculated mean portal venous blood flow decreased from 1,902 to 1,501 ml/minute between the first and second examination. Figure 5. Diameter of portal vein measured at 3 points on 2 occasions following liver transplantation (n 45). Figure 6. Median velocity of portal blood flow (interquartile range) measured at 3 points in the portal vein on 2 occasions following liver transplantation (* n 45), ** P <.05. Doppler Indices in Patients with Acute Cellular Rejection Three patients in the series developed acute rejection within the first postoperative week. These patients all showed maximum arterial velocities below 200 /second and RIs within the range of values described for patients without graft dysfunction. Similarly, portal vein flow velocity was within the range described for patients without graft dysfunction. Discussion This study illustrates the range of Doppler indices recorded in the hepatic vasculature of uncomplicated liver transplant recipients and also provides information on the physiology of hepatic blood flow following liver transplantation. The most striking findings were the wide range of flow velocities in both the hepatic artery and portal vein in patients with improving graft function. The peak flow velocity, RI, and shape of the waveform recorded by DUS are dependent on many factors. These include the arterial pressure, the lumen diameter, the upstream resistance, the compliance of the vessel wall, and the blood viscosity. The high arterial velocities recorded in these patients may be caused by a high arterial in-flow secondary to pretransplant portal hypertension, which remains transiently elevated following liver transplantation. 15 In this series, a wide range of peak arterial velocities is compatible with graft recovery, from 13 to 367 /second, indicating that detection of an isolated arterial segment with high flow velocity may not indicate the presence of arterial stenosis. These high-velocity segments may be

5 Quantitative DUS in Liver Transplantation 1187 explained by variations in the internal diameter of the arterial lumen. In this series, arterial anastomosis was performed using an arterial patch in the manner described by Carrel and Guthrie, 16 to prevent anastomotic strictures from occurring in the early posttransplant period. Following anastomosis of the transplanted hepatic artery, however, there is often some redundancy of the hepatic artery, which may form a kink and narrow the arterial lumen without a decrease in arterial wall compliance, resulting in accelerated flow velocity. In addition, accurate determination of peak velocity requires correct determination of the angle at which the segment of vessel being evaluated lies relative to the skin surface, which was evaluated by color Doppler in this series. As color Doppler has poorer resolution than B mode imaging, it is possible that kinks in the hepatic artery may not have been detected and spuriously high readings might have occurred. Despite the high arterial velocity, the median RI lay within the normal range for a visceral artery at all sites within the posttransplant hepatic artery, although both elevated and decreased arterial resistance are compatible with graft recovery. In contrast to arterial velocity, the RI did not vary at points within the hepatic artery, but did decrease at all sites within the transplant hepatic artery during the first postoperative week. These findings suggest that the nonuniform arterial velocity is not caused by a functional stenosis. Correlation of arterial velocity with RIs confirms that high velocity flow in the transplant hepatic artery is associated with normal arterial resistance, with antegrade flow during diastole. The finding of normal SAT at a distal point within the transplant hepatic artery also suggests that the high velocity segments are not caused by increased resistance to arterial flow. The decline in RIs during the first week may be due to changes in the graft parenchyma. The arterial resistance within the liver is determined by tone within hepatic arterioles 17 and graft edema due to preservation and cold ischemia have been shown to increase resistance to arterial blood flow. 18 Doppler assessment of the portal vein revealed a higher incidence of anastomotic narrowing than expected. Clinically significant portal vein stenosis is more common following pediatric liver transplantation than adult liver transplantation, and this has been presumed to be due to a size discrepancy between recipient and donor portal vein. 19,20 This study revealed a narrowing of the portal vein with accelerated flow velocity in all liver transplant recipients, despite construction of the portal vein anastomosis using a redundant length of suture to allow for a growth factor. The finding of a median reduction in portal vein diameter of greater than 40% at the site of the anastomosis without adverse clinical effect suggests that adults may have a greater capacity to tolerate the consequences of portal vein stenosis than pediatric patients. The range of normal values of portal venous velocity following liver transplantation has not previously been established. The high flow velocities in these recipients may be caused by the combination of portal hypertension and portal venous narrowing, although very low velocities are also compatible with normal graft function. The fall in portal venous velocity in the first week posttransplantation, despite the unchanging diameter of the portal vein, allows an estimate of the change in portal blood flow. Portal blood flow has been shown to decline gradually for up to 2 years after liver transplantation, 21 whereas our study reveals a rapid fall in portal blood flow of 20% in the 7 days following recovery of graft function. The increased arterial flow velocity noted over the first week may be due in part to the fall in portal flow during the same period, as the reciprocal regulation of portal and hepatic arterial flow is intact in transplanted livers, 15 but it may also be influenced by the fall in resistance to arterial flow. The small number of patients in the series who developed acute cellular rejection did not allow assessment of the value of DUS in predicting this complication, although the finding of Doppler values within the range described for patients without acute rejection suggests that DUS may be of little value in diagnosing this complication. This study reveals a range of arterial indices outside those previously recorded in visceral arteries, both elevated and reduced, in patients making an uneventful recovery following liver transplantation. Narrowing of the portal vein with accelerated venous flow is also common in these patients and appears to have no detrimental effect on graft function. Where a Doppler signal was detectable, abnormal indices did not affect management or prognosis of any patient in our series. However, patients with abnormal graft function may still benefit from quantitative DUS to aid in the assessment of arterial stenosis. Our findings support the view that the lack of a detectable Doppler signal in the hepatic artery or portal vein is highly predictive of vascular thrombosis and warrants surgical intervention. Where the transplant vessels are patent, however, abnormal Doppler values must be interpreted with caution, as these findings generally will not be clinically significant and may improve with time. For these reasons, we believe that detailed Doppler examination of the transplanted hepatic ves-

6 1188 Stell et al. sels to obtain flow values in patients with improving graft function is not warranted and a more appropriate use of Doppler surveillance after liver transplantation is to perform frequent examinations with portable ultrasound to detect only the presence of arterial and portal venous blood flow. References 1. Rifkin MD, Needleman L, Pasto ME, Kurtz AB, Foy PM, Mc- Glynn E, et al. Evaluation of renal transplant rejection by duplex Doppler examination: value of the resistive index. AJR Am J Roentgenol 1987;148: Dodd GD 3rd, Memel DS, Zajko AB, Baron RL, Santaguida LA. Hepatic artery stenosis and thrombosis in transplant recipients: Doppler diagnosis with resistive index and systolic acceleration time. Radiology 1994;192: Abbasoglu O, Levy MF, Vodapally MS, Goldstein RM, Husberg BS, Gonwa TA, Klintmalm GB. Hepatic artery stenosis after liver transplantation incidence, presentation, treatment, and long term outcome. Transplantation 1997;63: Platt JF, Yutzy GG, Bude RO, Ellis JH, Rubin JM. Use of Doppler sonography for revealing hepatic artery stenosis in liver transplant recipients. AJR Am J Roentgenol 1997;168: De Gaetano AM, Cotroneo AR, Maresca G, Di Stasi C, Evangelisti R, Gui B, Agnes S. Color Doppler sonography in the diagnosis and monitoring of arterial complications after liver transplantation. J Clin Ultrasound 2000;28: Kok T, Slooff MJ, Thijn CJ, Peeters PM, Verwer R, Bijleveld CM, van den Berg AP, et al. Routine Doppler ultrasound for the detection of clinically unsuspected vascular complications in the early postoperative phase after orthotopic liver transplantation. Transpl Int 1998;11: Kok T, Peeters PM, Hew JM, Martijn A, Koetse HA, Bijleveld CM, Slooff MJ. Doppler ultrasound and angiography of the vasculature of the liver in children after orthotopic liver transplantation: a prospective study. Pediatr Radiol 1995;25: Malassagne B, Soubrane O, Dousset B, Legmann P, Houssin D. Extrahepatic portal hypertension following liver transplantation: a rare but challenging problem. HPB Surg 1998;10: [Discussion ]. 9. Vassiliades VG, Ostrow TD, Chezmar JL, Hertzler GL, Nelson RC. Hepatic arterial resistive indices: correlation with the severity of cirrhosis. Abdom Imaging 1993;18: Hubner GH, Steudel N, Kleber G, Behrmann C, Lotterer E, Fleig WE. Hepatic arterial blood flow velocities: assessment by transcutaneous and intravascular Doppler sonography. J Hepatol 2000;32: Pourcelot L, Tranquart F, De Bray JM, Philippot M, Bonithon MC, Salez F. Ultrasound characterization and quantification of carotid atherosclerosis lesions. Minerva Cardioangiol 1999;47: Berland LL, Koslin DB, Routh WD, Keller FS. Renal artery stenosis: prospective evaluation of diagnosis with color duplex US compared with angiography. Work in progress. Radiology 1990;174: Settmacher U, Nussler NC, Glanemann M, Haase R, Heise M, Bechstein WO, Neuhaus P. Venous complications after orthotopic liver transplantation. Clin Transplant 2000;14: Moriyasu F, Ban N, Nishida O, Nakamura T, Miyake T, Uchino H, et al. Clinical application of an ultrasonic duplex system in the quantitative measurement of portal blood flow. J Clin Ultrasound 1986;14: Henderson JM, Gilmore GT, Mackay GJ, Galloway JR, Dodson TF, Kutner MH. Hemodynamics during liver transplantation: the interactions between cardiac output and portal venous and hepatic arterial flows. Hepatology 1992;16: Carrel A, Guthrie C. Functions of a transplanted kidney. Science 1905;22: Lautt WW. Mechanism and role of intrinsic regulation of hepatic arterial blood flow: hepatic arterial buffer response. Am J Physiol 1985;249:G549 G Seifalian AM, Mallet SV, Rolles K, Davidson BR. Hepatic microcirculation during human orthotopic liver transplantation. Br J Surg 1997;84: Millis JM, Seaman DS, Piper JB, Alonso EM, Kelly S, Hackworth CA, et al. Portal vein thrombosis and stenosis in pediatric liver transplantation. Transplantation 1996;62: Sieders E, Peeters PM, TenVergert EM, de Jong KP, Porte RJ, Zwaveling JH, et al. Early vascular complications after pediatric liver transplantation. Liver Transpl 2000;6: Bolognesi M, Sacerdoti D, Bombonato G, Merkel C, Sartori G, Merenda R, et al. Change in portal flow after liver transplantation: effect on hepatic arterial resistance indices and role of spleen size. Hepatology 2002;35: