Absence of the Horizontal Portion of the Left Portal Vein Diagnosed by Ultrasound and Computed Tomography

: The absence of the horizontal portion of the left portal vein is rare, and it is often not diagnosed by ultrasound. This article reports a case where ultrasound identified a structural anomaly between the gallbladder and the round ligament of the liver, and then computed tomography confirmed the absence of the horizontal portion of the left portal vein. We proposed the possible etiologies from an embryonic development perspective, and described the questions to be further investigated after a literature review.

T he portal vein is an important indicator of liver segmentation because it participates in the segmentation with hepatic fissure. We should identify them in the ultrasound examination. Although not uncommon, variations in the portal vein are difficult to identify and can easily be missed; especially if a sonographer is unfamiliar with the condition and readily confuses it with normal liver anatomy. Variations in the left portal vein (LPV) are rare [1]; many reports that ultrasound examinations have not found liver variations [2]. Variation of portal vein is often accompanied by liver fissure mutation. Normally the gallbladder is located in the median fissure, which is also an important indicator of liver fissure mutation. Therefore, if the anatomical relationship between the gallbladder, liver fissure, and the course of the portal vein is closely observed, many variations can be identified. In this case, it was the unusual location of the gallbladder in the round ligament fissure that led to the finding of the portal vein variation. Informed consent was obtained for anonymous publication of patient's clinical data.

Case Report
A 36-year-old male patient was admitted to the hospital due to intermittent dizziness for 20 days. He had no previous medical history. He underwent routine admission to our department for abdominal ultrasound examination (GE's LOGIQ E9, C3-5 probe). The size and shape of the liver was normal. The extrahepatic portal vein ran and branched normally. After giving off right posterior branch at the hepatic hilus, the portal vein coursed to the upper right. The LPV and horizontal portion were not displayed. The vertical portion ran in the left interlobular fissure, branched to the left inferolateral segment and left superolateral segment, also a thick vessel to the right (Fig. 1A). Three hepatic veins ran normally. CT and intensive CT examinations were performed on the second and third days respectively (Aquilion one320, the thickness 0.5cm, and ioverol 70ml). The portal vein bifurcated to the right posterior and anterior branches at the hepatic hilus: the right posterior branch arc to the right posterior and finally divided into two small branches downwards; the right anterior branch kept going to the right superior for 2 cm then made a sudden anterior turn to form the vertical portion ( Fig. 2A). Branches were given off to the left and right at the corner, the right one coursing for 1cm followed by three branches to right anterior, right superior and right posteroinferior (Fig. 2B). The vertical portion coursed 4cm then terminated as a cul de sac, which attached with the round ligament of the liver. The round ligament fissure zigzagged from the left posterior to the right anteroinferior. The angle between fissure and the horizontal plane was 13° at last third, 43° at middle third, and 74.1° at first third. The gallbladder was located in the ligament fissure with an angle of 79.7°between the long axis and the horizontal plane.

Discussion
The classic portal vein at the hepatic hilus and is divided into left and right branches. The right branch then is divided into right anterior and right posterior branches. The left branch first runs horizontally to the left liver, then turns anteriorly through the corner to form a vertical portion, which runs within the left round ligament fissure. Branches, including left inferior lateral, left superolateral and left medial segmental branches, given off from the vertical portion formed a typical "工"shaped structure. In this case, an approximate "工"shaped structure was found in the right lobe during the first ultrasound examination. However, the structure rotated counterclockwise slightly, as a result, the vertical portion tilts to the right, and the two left laterals form an angle. The horizontal segment of the LPV was not shown. Because the patient is young without surgical history nor abdominal symptom, the seemingly minor "工"-shaped structure abnormality did not raise any concerns. The doctor continued the examination by turning the probe upwards on the transverse plane under the xiphoid process. It was found that the gallbladder was located at the end of the sac, and the long axis of gallbladder coincided with the round ligament of the liver (Fig. 1B) instead of the common right side. This caught our attention. It was not easy to scan the gallbladder on the intercostal plane due to the slightly medial deviation from the usual position. A large but short blood vessel similar to the right anterior branch of the portal vein was seen near the unclear neck. A CT examination was obtained in suspect of absence of the horizontal portion of the portal vein.
CT scan revealed a generally normal liver. The bottom of the gallbladder seemed to be located below the right lobe, and the neck was located in front of the MPV (main portal vein). The gallbladder neck and bottom went from superior medial to inferior lateral in the ligament fissure. Due to the abnormal coincidence of the long axis of the gallbladder with the round ligament, the patient underwent an enhanced CT examination the next day. The results showed that the MPV bifurcated into right posterior branch and right anterior branch which was the large but short vessel seen in ultrasound. After running its way of right superior for a short length, the right anterior branch turned down anteriorly, and branched to supply the left liver lobe, which was similar to the structure beyond the corner of the normal LPV. There are two possible etiologies for the variations in this case -the absence of the horizontal portion or the right vertical. One is the dysplasia of the left branch; the other is the RRL (right round ligament). Indeed, the multiplicity of the embryologic origins of the left branch makes the agenesis of LPV less likely than the right one. However, LPV dysplasia is possible. The horizontal portion arises from the anastomotic venous channel between the caudal portion of the right vitelline vein with the left umbilical vein, while the vertical portion comes from the left umbilical vein [2]. When the anastomosis between the right vitelline vein and the left umbilical vein fails, dysplasia of the horizontal portion may occur while the vertical portion is present. The left umbilical vein cannot be anastomosed with the horizontal portion, so it moves to the right to be anastomosed with the right anterior branch. In this case, the right anterior branch was followed by the vertical portion, and round ligament fissure deviated to right side. The position of the middle hepatic vein is normal, and the round ligament is still on the left side of the middle hepatic vein. Thus, according to the Couinaud segmentation principle, the left medial lobe is absent.  [3,4]. The common feature was that the right anterior branch sent out an aberrant large blood vessel to cross the left medial lobe and connect to the vertical portion of LPV. The vertical portion was still located in the left round ligament fissure. This aberrant blood vessel was not found in the present case. Jin et al. [5] reported a case of lack of horizontal segment in CT review. Zhang et al. [6] found a case of lack of horizontal segment in autopsy. Both cases did not describe this aberrant blood vessel, instead, MPV bifurcated at the left hilus, and the LPV ran as the vertical portion. In this case, the round ligament fissure zigzagged from left to right, till the anteroinferior end reached the gallbladder fossa. The criteria of the liver fissure show the Cantlie line is the plane from the midpoint of the gallbladder fossa to the left edge of the inferior vena cava. Therefore, there was fusion of the middle fissure and left round ligament fissure at the anteroinferior end in this case. Lucidarme et al. [7] support the theory of right ward-deviated round ligament and the fusion of the left and midplane. But Rajesh et al. [8] had different opinion that the midplane is located on the right side of the vertical portion and runs along the middle hepatic vein.
The second possible reason is the RRL. In the embryonic development when the cranial segment of the right umbilical vein and the left umbilical vein are obstructed, the caudal segment of the right umbilical vein expands and anastomoses with the right vitelline vein to form a PRUV (permanent right umbilical vein), which will turn into the RRL after birth. The incidence of RRL is 0.2% -1.2% [9], 0.1% -1.2% [10], and PRUV is 0.09% -0.2% [2], 0.2% -1% [11] and 0.08% -0.23% [9]. If all the RRL comes from the PRUV, the incidence of the RRL should be less than or equal to the PRUV.
Since persons living with RRL may not experience any symptoms in adulthood, they may not seek medical attention; even if they undergo medical examination, it may not be identified. Therefore, the incidence of the RRL in the actual population must be lower than the current statistics. Based on the fact that the current incidence of RRL is higher than that of PRUV, we speculate that some patients diagnosed with RRL may not come from the PRUV. Shen et al. [12] followed up five newborns who were diagnosed with PRUV prenatally. Their intrahepatic portal vein runs like that before birth. There is no trace of adulthood. Many cases of these structures (no LPV and horizontal portion, the right anterior branch or the right main stem attaching to an umbilical portion and coursing anteroinferior which gives off some branches) were regarded as RRL.  The structure in this case is basically the same but an anatomic relationship. The middle hepatic vein in all the above cases were located on the left side of the round ligament, which was just opposite in this case. And some people suspect [10] RSLT must follow a left gallbladder, and the middle hepatic vein must be on the left side of RSLT. The fact in this case is that the gallbladder is located beneath RSLT, and the middle hepatic vein is on the right side of RSLT. Something interesting is that the inferior edge of RRL fissure in the articles of Shen et al [12]. Lin et al. [10] runs from right to left, which is vertically downward in this case. Although we still do not know the significance of the difference between these two directions. The umbilical vein enters the abdomen and moves cranially along the abdominal wall for a short extrahepatic course and enters the liver through the falciform ligament. We do not know whether the rightward deviated left round ligament and RRL will cause the difference between the edge of the ligament fissures.

Conclusion
The ultrasonic sign that makes portal vein variation found in this case is the coincidence of the gallbladder and the round ligament. This sign calls for attention from every radiologist to further investigate potential portal vein variations. Portal vein variation is of great significance in liver surgery, especially in the field of precise hepatectomy. We need to diagnose the hepatic portal vein variation correctly. Ultrasound can detect most of portal vein structural variations, such as the absence of horizontal portion in this case. Combined CT scan can demonstrate the course and structure of the portal vein accurately.