Umbilical Portion Research Articles

Critical progressive small-graft injury caused by intrasinusoidal pressure elevation following living donor liver transplantation

In adult-to-adult living liver transplantation, small-for-size graft syndrome sometimes occurs. The relationship between the hemodynamic changes and histologic findings has not been studied in patients with failure of small-for-size grafts. We analyzed the relationship between the postoperative hemodynamic changes and pathologic findings in patients with small-for-size grafts that ended in graft failure. From March 1999 to December 2002, adult-to-adult living-donor liver transplantation with small-size grafts (graft volume/standard liver volume less than 40%) was performed in eight patients. Three patients died from graft failure caused by overperfusion, which was diagnosed from pathologic findings. We analyzed the relation between hepatic hemodynamic parameters, such as portal venous blood velocity or splenic arterial pulsatility index, and histologic changes in patients with graft failure. Severe portal hyperperfusion (90 cm/sec at the umbilical portion) was observed on postoperative day 1. Among patients with graft failure, critical hemodynamic changes, such as sudden onset of extremely deteriorated portal venous blood flow, occurred during the early postoperative period (postoperative day 5, 3, 6, respectively). Histologic examination revealed vacuolar changes in the cytoplasm of hepatocytes, and submassive necrosis indicated intrasinusoidal pressure elevation. These changes were not observed in the biopsy obtained soon after reperfusion. In conclusion, critically decreased vascular beds may cause intrasinusoidal pressure elevation and sinusoidal circulatory disturbances.

Liver inflammatory pseudotumor due to an intrahepatic wooden toothpick

Ingestion of foreign bodies is a common clinical problem. However, intrahepatic migration is an exceptional occurrence: it has caused a hepatic abscesses which did not heal with percutaneous drainage, and caused patient death because of sepsis [1]. Clinical history is not helpful and only 50% of ingested toothpicks are radio-opaque [2]. A 63-year-old man was admitted in our department because of a typical episode of cholangitis with upper right abdominal pain, fever and jaundice. Ultrasonography (US) revealed a 6.5 cm in diameter, hypoechoic hepatic mass in the left lobe, which showed undefined margins, and caused a slight dilation of the intrahepatic bile ducts of segments 2 and 3. Helical CT scan confirmed the presence of the lesion located mainly in segment 3 and extended to segment 2: the lesion showed delayed peripheral enhancement (arrows in left panel; UP, umbilical portion) and was interpreted as a mass-forming type cholagiocellular carcinoma (MFCCC). Magnetic resonance imaging (MRI), and magnetic resonance cholangiopancreatography (MRCP) confirmed this diagnosis. The serum level of AFP, CEA and Ca 19-9 were normal. The patient underwent surgery. An inverted T-shaped incision was accomplished and a palpable sticky mass was noted on the inferior surface of the left lobe with a thickened hepato-gastric ligament possible expression of a previous gastrointestinal perforation. At intra-operative US a linear high echogenic mass within the liver tumour was identified; vessels running through the tumor without dislodgement were also disclosed. A left hepatic lobectomy was performed and a 4 cm wooden toothpick was extracted from the resected liver (right panel). At histology the liver mass proved to be an inflammatory granuloma. Postoperative outcome was uneventful and patient was discharged at eighth postoperative day.

Meso-portal bypass in children with portal vein thrombosis: rapid increase of the intrahepatic portal venous flow after direct portal hepatic reperfusion

Background/Purpose: In children with portal vein (PV) thrombosis, hepatopetal portal flow can be restored by an innovative surgical procedure, the meso-portal-bypass (MPB), if the umbilical portion of the intrahepatic left PV and the superior mesenteric vein are patent. This is associated with resolution of symptoms related to extrahepatic portal hypertension (EHPH). However, no data are available yet on intrahepatic hemodynamic changes after MPB. The aim of this study was to evaluate morphologic adaptation and flow characteristics of the intrahepatic PV branches (ihPVb) after MPB. Methods: Prospective follow-up Doppler scans of the ihPVb were performed at 0.5 to 1, 3 to 6 and 12 months after MPB in 13 consecutive patients (2000-2002) and compared with preoperative findings. Results: Only small ihPVb were detected preoperatively on Doppler in 8 of 13 cases. Postoperatively (median follow up 12; range, 6 to 24 months), all 13 patients had patent MPB with hepatopetal flow, and ihPVb were easy to detect with satisfactory vessel diameters and flow velocities. Conclusions: The preoperatively small ihPVb increase rapidly in diameter and hepatopetal flow velocity in patients benefiting from MPB. This correlates well with progressive resolution of their symptoms related to EHPH and reflects rapid adaptation of ihPVb.

Relationship between left biliary duct system and left portal vein: evaluation with three-dimensional portocholangiography.

To investigate the relationship between the left biliary duct system and the left portal vein by using in vivo analysis of the human liver with three-dimensional portocholangiography. Twenty-seven patients underwent helical computed tomography during both arterial portography and cholangiography. Three-dimensional portocholangiograms were reconstructed to evaluate the relationship between the left biliary duct system and the left portal vein, in particular the umbilical portion. In 16 (59%) patients, segment II and III ducts united just above the umbilical portion or laterally to it, and the segment IV duct joined medially to the umbilical portion. In eight (30%) patients, segment III and IV ducts united medially to the umbilical portion and the segment II duct joined at a point close to the hepatic hilum. In three (11%) patients, three ducts from segments II-IV united at a position immediately medial to the umbilical portion. Three types of left bile duct anatomy were seen in relation to the left portal vein.

Portal vein aneurysm of the umbilical portion: imaging features and the relationship with portal vein anomalies.

We evaluated the imaging features of portal vein aneurysm occurring at the umbilical portion and the relation between portal vein aneurysm of the umbilical portion and portal vein anomaly. Retrograde analysis was performed in 18 patients with portal vein aneurysm of the umbilical portion of the portal vein. We analyzed the size, shape, and imaging findings of portal vein aneurysm. In addition, we evaluated associated portal vein anomalies, specifically, the branching pattern of the intrahepatic portal vein and the position of the umbilical portion of the left portal vein. The diameter of the portal vein aneurysm ranged from 1.5 to 2.2 cm and the shape of the portal vein aneurysm was saccular (n = 14) or fusiform (n = 4). Imaging findings of the portal vein aneurysm of the umbilical portion were typical for its vascular nature. Portal vein anomalies were seen in eight of 18 patients (44%): the right anterior segmental portal vein (n = 7) or the right anterior and posterior segmental portal veins (n = 1) originating from the umbilical portion of the left portal vein and a rightward deviation of the umbilical portion of the left portal vein (n = 4). Due to its vascular nature, noninvasive diagnostic modalities can enable the detection of portal vein aneurysm of the umbilical portion, and there is a close relation between portal vein aneurysm of the umbilical portion and portal vein anomaly.

Arantius' ligament approach to the left hepatic vein and to the common trunk.

During fetal development, the ductus venosus (ductus venosus Arantii) connects the umbilical portion of the left branch of the portal vein to the inferior vena cava, shunting oxygenated umbilical cord blood away from the liver. After birth the duct obliterates and persists as the ligamentum venosum or Arantius’ ligament. Because of its functional role, it is commonly believed that the ligament runs from the left branch of the portal vein to the vena cava itself. But attention to anatomical detail demonstrates that the fibers of the ligament insert either on the left hepatic vein or at the junction between the left hepatic and the middle hepatic veins (Figs. 1, 2). We describe a maneuver that takes advantage of the insertion of the ligamentum venosum to isolate and tape the left hepatic vein, or the common trunk of the middle and left hepatic veins. This maneuver is useful to pass a vessel loop to prepare for in situ left lobe splitting or for pediatric living donor liver transplantation. With further dissection, the maneuver can be used to encircle the common trunk of the left and middle hepatic veins to prepare for harvesting of the left liver for living donor or split liver transplantation or for selective hepatic vein occlusion during liver resections.

Poststenotic dilatation and helical flow in the umbilical portion of the portal vein

Poststenotic dilatation (PSD) occurs in an artery distal to a site of stenosis. Usually, PSD occurs in humans in regions distal to coarctation of the aorta, abdominal aorta, or pulmonary arteries [1]. On the other hand, PSD in the portal vein is relatively rare. We herein present a case of PSD in the umbilical portion of the portal vein. The patient was a 60-year-old man with hilar cholangiocarcinoma. The right portal vein was invaded by the cancer, and blood flow in the right portal vein could not been detected by color Doppler ultrasonography (CDUS). The left branch of the portal vein was invaded by the cancer and an area of severe stenosis was revealed (arrows; panel 1a , computed tomography and panel 1b, percutaneous transhepatic portography), while the umbilical portion of the portal vein was dilated (panel 1a,b, arrowheads). CDUS showed parallel red and blue bands in the dilated umbilical portion (panel 2, arrowheads; epigastric transverse section). Although portal blood flow velocity is usually about 10 cm/s, pulsed Doppler showed that the red band was a hepatopetal high-velocity flow (75 cm/s) and the blue band a hepatofugal high-velocity flow (33 cm/s). We consider this to be a phenomenon in which the jet flow, formed by both the stenosis in the left branch and the obstruction in the right portal vein, turned at the top of the umbilical portion and swirled in the right side of the umbilical portion where it had caused the internal pressure to become negative (curved arrow in panel 2 indicates the direction of the blood flow). Finally, the hemodynamic forces of the jet flow caused PSD in the portal vein. These parallel red and blue bands shown by CDUS have been called a helical flow. Rosenthal et al. [2] reported that such helical flows are seen most commonly during the early postoperative period in liver transplant recipients, and persist in patients with anastomotic portal vein stenosis. PSD in the intrahepatic portal vein is relatively rare, but the dilation of the portal vein and the helical flow shown by CDUS are clinically important and characteristic signs that indicate the proximal stenosis of the portal vein.

Analysis of portal venous waveform after living-related liver transplantation with pulsed Doppler ultrasound.

We evaluate the portal venous waveform (PVW) with pulsed Doppler ultrasound (US) after living-related liver transplantation (LRLT) and correlate it with subsequent abnormalities. In the first 2 wk after LRLT, pulsed Doppler US demonstrated three types of PV waveform (PVW) in 33 recipients: non-phasic PVW in 19 patients, pulsatile in 10, and turbulent in 4. In the pulsatile PVW group, arterio-portal (A-P) shunt was confirmed in three grafts by either arteriograhy or computed tomography during hepatic arteriography. A severe stenosis in the grafted vein was confirmed in one case by both US and venography. The pulsatile PVW in the remaining six cases spontaneously disappeared and turned to the non-phasic PVW without treatment. The graft volume/liver standard volume (GV/SV) ratio was significantly smaller in the pulsatile PV waveform group than in the non-phasic PVW group (p<0.01). In the turbulent PVW group, aneurysmal-like dilatation of the portal vein at the umbilical portion was formed in 3 of the 4 patients. The pulsatile waveform in the PV is frequently observed with pulsed Doppler after LRLT, especially in patients that received small grafts. We should keep in mind that they often disappear without any treatment, although some examples of pulsatile waveforms represent vascular complications.

Intrahepatic inflow areas of the drainage vein of the gallbladder: Analysis by angio-CT

Background. The drainage veins of the gallbladder (cystic veins) are important hepatic metastatic routes of gallbladder cancer. We studied the cystic vein inflow portions of the intrahepatic vessels by computed tomography during angiography (angio-CT). Methods. The subjects were 27 unoperated pancreatic-biliary patients. After superselective catheterization of the cystic artery, angio-CT was performed, and cystic veins and sinusoidal filling were visualized. We identified the cystic vein inflow routes and the vessels they fed. Results. We found 72 cystic veins in 26 patients. All cystic veins flowed into the intrahepatic portal branches or sinusoids. The cystic veins took either of two routes: one into the liver through the hepatic hilum (17 patients, 21 veins), taken mainly by the portal branch for subsegment 4a (P4a), the anterior portal branch, and the umbilical portion of the portal branch; and the other through the hepatic bed (23 patients, 51 veins), taken mainly by S4a sinusoid, S5 sinusoid, P4a, and P5. Conclusions. Angio-CT is useful for detecting the cystic vein inflow portions of the intrahepatic vessels. It makes possible identification of areas where there is a possibility for micrometastasis of the gallbladder cancer. (Surgery 2000;128:417–21)

A new operation for noncorrectable biliary atresia.

An improved operative technique to transect the fibrous cord by dividing the ligamentum venosum (Arantius' canal) is described for noncorrectable biliary atresia. The Arantius' canal is situated cranial and posterior side to the bifurcation of the umbilical portion and the portal branch of the Quinous' segment 3. The portal vein is fully mobile and the porta hepatis can be widely exposed by dividing the Arantius' canal. The fibrous cord of the porta hepatis can easily be dissected posteriorly and laterally where there is an extensive number of bile ducts. Eight patients with biliary atresia underwent this procedure. Jaundice resolved completely (serum total bilirubin concentration: < or = 1 mg/dl) in 7 patients within 40 days. Postoperative cholangitis did not occur. By dividing the Arantius' canal, the portal vein comes free from the portal fissure to make the hepatic hilum wider, and surgeons are able to work within a larger porta hepatic space without causing portal vein compression. Free drainage of the bile from the porta hepatis may prevent postoperative cholangitis and promote resolution of jaundice.

S4a + S5 with caudate lobe (S1) resection using the Taj Mahal liver parenchymal resection for carcinoma of the biliary tract

Recently we have been performing S4a + S5 with total resection of the caudate lobe (S1) by using a dome-like dissection along the root of the middle hepatic vein at the pinnacle, which we refer to as the Taj Mahal liver parenchymal resection, for carcinoma of the biliary tract. This procedure offers the following advantages: (1) It allows total resection of the caudate lobe, including the paracaval portion (S9), and (2) because the cut surface of the liver is large, it allows intrahepatic jejunostomy to be performed more easily with a good field of view. The indications for this procedure include hilar bile duct carcinoma, gallbladder carcinoma, and choledochal cyst (type IVA). Because of the high rate of hilar liver parenchyma and caudate lobe invasion associated with hilar bile duct carcinoma, the liver must be resected. The Taj Mahal procedure is indicated in cases where extended liver resection is impossible. The dissection limits of this procedure are, on the left side, the B2+3 bifurcation at the right margin of the umbilical portion of the portal vein and, on the right side, the B8 of the anterior branch and the B6+7 bifurcation of the right posterior branch. This procedure could also be described as a reduced form of extended right hepatectomy and extended left hepatectomy. For gallbladder carcinoma, this procedure is indicated to ensure an adequate surgical margin and eradicate transvenous liver metastasis, particularly in cases of pT2 lesions. Hilar and caudate lobe invasion also occurs in liver bed-type gallbladder carcinoma, and bile duct resection and caudate lobe resection are required for the surgery to be curative. We performed this procedure in four cases of hilar bile duct carcinoma, five cases of gallbladder carcinoma, and one case each of choledochal cyst (type IVA) with carcinoma of the bile duct and gallbladder adenomyomatosis. Curative resection was possible in all except the patient with adenomyomatosis, and all of the patients are alive and recurrence free 10 to 37 months postoperatively. This procedure, in addition to preserving liver function, provides a wide field of view and facilitates reconstruction of multiple intrahepatic bile ducts. Thus it can be said to be a curative operation not only in patients considered high risk but also in those whose hilar bile duct carcinoma is limited to the bifurcation area (Bismuth type IIIa and IIIb) and in gallbladder carcinoma up to pT2 with slight extension on the hepatic side.

CT Findings of Portal Vein Aneurysm

Purpose : To describe the CT findings of portal vein aneurysm in eight patients. Materials and Methods: Allpatients included in this study (two men and six women) under went CT examinations between October 1996 and June1998. Of these eight, three were suffering from hepatic disease and portal hypertension. We determined thelocation, shape, size, and characteristics of the lesions, and the presence or ab-sence of portal vein anomaly.Results : S even patients had intrahepatic portal vein aneurysm (at the umbilical por-tion of the left portal veinin five patients, between the transverse and umbilical por-tion of the left portal vein in one, and at thebifurcation of the anterior and posterior branch of the right portal vein in one), while extrahepatic portal veinaneurysm, at the confluence of the superior mesenteric and splenic vein was found in only one. Lesions werecyst-shaped in seven cases and saccular in one, and showed well - circum scribed, markedly enhanced mass, whichcommunicated with the portal vein and/or gives off major branches. Portal vein anomaly, in which the rightanterior segmental portal vein originated from the umbilical portion of the left portal vein, was seen in threepatients. In all three, intrahepatic portal vein aneurysm was present at the umbilical portion of the left portalvein, and in one, the umbilical portion of the left portal vein was located to the right of the Cantlie line.Conclusion : CT examination can help reveal portal vein aneurysm by detectinga well - circumscribed, markedlyenhanced mass which communicates with the portal vein and/or gives off major branches.

Primary lymphoma of the liver with bile duct invasion and tumoral occlusion of the portal vein: report of a case

A 55-year-old woman presented to hospital with epigastric pain and jaundice. Diagnostic imaging studies revealed a biliary stricture of the hepatic confluence and a hepatic tumour of the left and caudate lobes with a portal tumour thrombus, which occupied the main portal trunk, the umbilical portion of the left portal vein, and the right anterior and posterior portal branches. Left hepatic trisegmentectomy, caudate lobectomy, portal tumour thrombectomy, bile duct resection and bilioenteric anastomosis were performed. There were no other lesions, and so it was diagnosed as a primary lumphoma of the liver (B-cell, diffuse, large cell type). The patient underwent postoperative chemotherapy and has remained well for 4.5 years after surgery. Primary lymphoma of the liver is very rare, and this is the first case report with bile duct invasion and tumoral occlusion of the portal vein.

Portal vein anomaly associated with deviation of the ligamentum teres to the right and malposition of the gallbladder.

To develop a complete clinical picture of a portal vein anomaly associated with a rightward-deviated ligamentum teres. Three hundred twenty-seven consecutive living donor candidates for partial liver transplantation were evaluated for hepatic vascular abnormalities with computed tomography supplemented with intraoperative examinations. Special attention was given to the branching pattern of the portal system and the relative positions of the umbilical portion of the portal vein, the ligamentum teres, and the gallbladder. Four donor candidates (1.2%) had a portal vein anomaly associated with a rightward-deviated ligamentum teres. In all four, the umbilical portion of the portal vein (and hence the ligamentum teres) was abnormal, lying above the gallbladder bed and giving off major branches to the right anterior segment. If unrecognized, these might have been ligated, leading to major complications in the donors. A portal vein anomaly associated with a rightward-deviated ligamentum teres is an important clinical entity that is not as rare as indicated in the literature. This anomaly should be kept in mind during preoperative examination of the liver.

Division of the left hemiliver in man--segments, sectors, or sections.

A major obstacle to coherent terminology for liver anatomy and resections has been that American and French anatomists have divided the left side of the liver through different planes. Couinaud divided the left hemiliver into "sectors" by a plane through the left hepatic vein. Healey and Schroy divided it into "segments" through the umbilical fissure. One anatomic justification for Couinaud's system of sectors is that the transverse portion of the left portal vein was said to terminate by dividing into the umbilical portion of the left portal vein and the vein to segment II. However, corrosion cast studies fail to consider the position of the ligamentum venosum, the structure defining the end of the transverse portion of the left portal vein. Therefore, it is uncertain whether the branch to segment II is a terminal branch of the transverse portion of the left portal vein or a branch of the first part of the umbilical portion. Ten cadaver livers were dissected to determine the position of the branch to segment II in relation to the ligamentum venosum. In all, the branch to segment II came off downstream to the ligamentum venosum, showing that it is a branch of the umbilical portion of the left portal vein. This study does not support the "sectoral" system of Couinaud on the left side of the liver. Division of the left side of the liver through the umbilical fissure as proposed by Healey is anatomically logical and fits well with common surgical resections.

Segment 4 (the quadrate lobe): a barometer of cirrhotic liver disease at US.

To evaluate the dimensions of segment 4 of the liver in patients with cirrhosis by using ultrasonography (US). The transverse diameter of segment 4 was measured in 125 control subjects without liver disease and 167 patients with cirrhosis. The size of segment 4 was measured on oblique subcostal US scans obtained between the left wall of the gallbladder (or the main fissure after cholecystectomy) and the ascending or umbilical portion of the left portal vein at the point where it gives rise to the branch to segment 4. In the control subject group, the mean diameter of segment 4 was 43 mm +/- 8 (standard deviation). In the patient group, the mean diameter of segment 4 was 28 mm +/- 9. The cause or severity of cirrhosis had no influence on the size of segment 4. A decreased diameter of segment 4 may be a helpful adjunct sign of cirrhosis in the US investigation of chronic liver disease.

Anatomic dissociation between the intrahepatic bile duct and portal vein: risk factors for left hepatectomy.

The anatomic variations of the intrahepatic portal vein and bile duct were analyzed to evaluate the potential risk of left hepatectomy. A total of 210 cholangiograms and hepatic arterioportograms were performed in which the ramifications of the intrahepatic portal vein and bile duct were investigated. The orientation of the intrahepatic duct and portal vein were classified into five types. In 175 patients (83.33%), the intrahepatic portal vein and bile duct had the same anatomic classification. In 24 patients (11.43%), the right anterior or posterior intrahepatic duct drained into the left hepatic duct at the umbilical portion (type IV); there were only 15 patients (7.14%) whose portal veins fell into this category. All patients with type IV portal veins had type IV hepatic ducts, but there were 9/49 patients (18.36%) whose hepatic duct distribution belonged to type IV but their portal veins belonged to type II (6 cases) or III (3 cases). Without complete knowledge of the intrahepatic portal and biliary anatomy, insufficient portal perfusion and bile duct complications may result from the left hepatectomy operation. Preoperative portal vein evaluation or left portal vein clamping can provide significant information, but there are still 18.36% of patients where type IV biliary ducts were not detected in those with type II and III portal veins. Cholangiography is of paramount importance in these two groups of patients, as it can prevent inadvertent injury to the right intrahepatic ducts, which drain into the left intrahepatic duct. On the other hand, intraoperative ultrasonography is recommended to identify or exclude an aberrant portal vein if type VI biliary anatomy is detected during intraoperative cholangiography.

Are left-sided gallbladders really located on the left side?

The aim of this investigation was to establish the association between left-sided gallbladders and right-sided round ligaments. The left-sided gallbladder is a rare anomaly and has been classified into two types: 1) gallbladder migration to the left side and 2) development of a second gallbladder with atrophy of the original one. Recently, left-sided gallbladders were reported to be associated with right-sided round ligaments. The authors reviewed 3 patients treated in their departments and 15 patients reported in the literature diagnosed as having left-sided gallbladders accompanied by right-sided round ligaments. Although the gallbladders of all 18 patients were located at the normal site, they were diagnosed as being left sided because of the right-sided round ligaments. This anomaly was accompanied by abnormal intrahepatic portal venous branching, which could be classified into two types. In eight patients, the first branch of the portal vein ran to the posterior segment and then the portal vein formed a trunk of the left and right anterior portal veins. The latter portal vein formed the umbilical portion and finally joined the right-sided round ligament (trifurcation type). In five, the portal vein diverged normally to form the left and right portal veins, then the latter branched to form the anterior and posterior segments, and finally the anterior branch joined the round ligament (bifurcation type). In the other five, the branching type could not be determined. A right-sided round ligament causes a gallbladder at the normal site to be located on the left side. This anomaly should not be diagnosed as a left-sided gallbladder but as a right-sided round ligament. Recognition of this anomaly clinically is important when performing hepatectomy, because it is always associated with abnormal intrahepatic portal venous branching.

Laparoscopic abdominal anatomy of foals positioned in dorsal recumbency.

Eleven foals (nine cadavers and two anesthetized) positioned in dorsal recumbency were examined to describe normal abdominal anatomy viewed by laparoscopy. The foals ages ranged from 1 to 150 days. Insertion sites were selected for a Verres needle, laparoscope and instrument portals to avoid trauma to the umbilical structures. The abdominal cavity was insufflated to a pressure of 10 to 12 mm Hg using an automatic carbon dioxide insufflator. Laparoscopic examination allowed excellent observation of the umbilical structures, the bladder, the genital tract, the inguinal area, the liver and segments of the intestinal tract in all foals. However, the duodenum and ileum were not identified. An embryonic vestige of the yolk stalk (ductus pedunculi vitellini) was identified in six foals. In four foals, this appeared as a fibrous band extending from the umbilicus to the cranial abdominal cavity where it inserted on the small intestine. In two foals only the fibrous umbilical portion of the band was observed.

Longitudinal study of intrahepatic portal flow: a novel approach using Doppler technique.

The present study investigates whether the application of the summation of portal blood flow (SPBF), measured from the left umbilical portion and the right anterior branch by Doppler, could be of use in the longitudinal study of intrahepatic portal flow or not. Forty-five subjects were divided into two groups. Group I subjects included males and post-menopausal females. Thirty-seven subjects (19 normal volunteers, 12 subjects with chronic hepatitis, six cirrhotics) were investigated by the more experienced doctor A at 9 day intervals for an intra-observer study. Among them, seven subjects (four normal volunteers, three with chronic hepatitis) were also investigated consecutively at 10 min apart by doctor B on days 1 and 10 for an interobserver study. Group II subjects included eight normal subjects with a regular menstrual cycle. They were investigated by doctor A during menstruation, mid-follicular, ovulatory and mid-luteal phases of the menstrual cycle to evaluate the possible impact of the menstrual cycle on SPBF measurement. The reproducibility of measurements was assessed by the coefficient of variation (CV) and the reproducibility coefficient. Data are expressed as mean +/- s.d. The CV of the SPBF for intra-observer and interobserver study were 4.88 +/- 3.36 and 4.19 +/- 2.53, respectively. All differences of SPBF were within their corresponding reproducibility limits. The results of SPBF measurements obtained from different phases of the menstrual cycle were variable. Only the comparison between the data obtained from mid-follicular and mid-luteal phases could demonstrate acceptable CV values (3.98 +/- 3.12%) and all the differences in SPBF measurements were within the repeatability limit. In conclusion, Doppler SPBF measurement may be useful in the longitudinal evaluation of intrahepatic portal flow changes.