Pre-, Intra- & Post-Operative Doppler Ultrasonographic Imaging of Portosystemic Venous Anomalies in Dogs
ACVIM 2008
Viktor Szatmári, DVM, PhD, ECVIM-CA (Cardiology)
Utrecht, the Netherlands

Diagnosis of Portosystemic Shunting with Ultrasound: Can it be Accurate?

Portosystemic shunting can occur via acquired portosystemic collaterals (APSCs) as a consequence of a sustained portal hypertension of hepatic or prehepatic origin, or via a congenital portosystemic shunt (CPSS) as a result of a developmental anomaly. Portosystemic shunting is suspected based on the medical history, and is justified by measuring a high blood ammonia level or obtaining an abnormal rectal ammonia tolerance test result. However, direct visualization of the aberrant vein(s) is essential for a definitive diagnosis. It is important to differentiate CPSSs non-invasively (i.e., without a laparotomy) from other conditions that cause hyperammonemia, such as APSCs and urea cycle enzyme deficiency, since only a CPSS requires surgical treatment while the other conditions do not. Diagnostic laparotomy causes unnecessary pain and anesthetic risks.

Although scintigraphy is the gold standard for the detection or exclusion of portosystemic shunting, it does not allow differentiation of CPSSs from APSCs, thus a positive scintigram has actually a similar information value to a high blood ammonia level. Portography does provide straightforward images, that make the differentiation of APSCs from CPSSs, and intra- from extrahepatic CPSSs possible. However, portography requires anesthesia and usually laparotomy. Excellent 3-dimensional reconstructions can be made by multiscale helical computed tomography, but it has several weak points: limited availability, need of anesthesia, and the evaluation of segments of the portal vein or the shunt vessel that are not filled with contrast agent is impossible. Color Doppler ultrasonography gives real time information about the anatomy and the flow directions of the portal vein and the shunt without the need of sedation.

We developed a protocol for systematic ultrasonographic evaluation of the canine portal system, which allowed us to accurately diagnose CPSSs and APSCs, and to identify the subtypes of CPSSs as well as to rule out portosystemic shunting.1-4 The high accuracy of ultrasonography in diagnosing portosystemic vascular anomalies is not only the result of this examination protocol and operator experience, but also of a logical diagnostic work up and patient selection. Ultrasonography for diagnosing portosystemic vascular anomalies should only be performed on patients that have a suggestive history and hyperammonemia (or an abnormal rectal ammonia tolerance test result). Although there are several anatomic variations of the anomalous veins, a CPSS is always a large-bore vein (i.e., a diameter that equals or exceeds that of the portal vein), which drains almost 100% of the portal blood to a systemic vein. The vast majority of the CPSSs in dogs has a consistent anatomy and can be categorized into one of the following groups.1-4 Intrahepatic CPSSs arise either from the left or right portal branch and terminate in the caudal vena cava. Extrahepatic CPSSs drain the blood of the portal vein via a short segment of the splenic or the gastroduodenal vein and terminate in the caudal vena cava slightly cranial to the celiac artery or enter the thorax. A dilated left gonadal vein is indicative for the presence of spleno-renal APSCs. Spleno-renal APSCs seem to develop consistently in dogs with sustained severe portal hypertension.1-4

Postoperative Complications: Are They Predictable and Preventable with the Use of Intraoperative Doppler Ultrasonography During Shunt-Ligation?

The definitive therapy for CPSSs would ideally be complete occlusion of the anomalous vein. However, in most dogs, only partial shunt ligation can be performed because attenuating the shunt would result in severe acute portal hypertension and subsequent shock, or development of APSCs as a result of sustained post-ligation portal hypertension. The reason why portal hypertension develops is that shunt-ligation forces portal blood to flow through the underdeveloped (i.e., hypoplastic) portal branches. Hypoplasia (i.e., small diameter thus insufficient vascular capacity) occurs in the portal vein segment that is cranial to the origin of the extrahepatic CPSS, because this segment has been hypoperfused since birth (Figure 1). In cases of intrahepatic CPSS, the contralateral portal branch is hypoplastic. The underdeveloped portal vein segment is unable to absorb the normal amount of portal blood that will be forced to flow through it after the shunt has been ligated. The degree of post-ligation portal hypertension depends on the degree of attenuation and the severity of portal vein hypoplasia. The following discussion will focus on the spleno-caval extrahepatic CPSSs, which is the most common type of CPSSs in dogs. The presence of shunt has an impact not only on the diameter of the portal vein, but also on the flow direction in it. The blood from the gastroduodenal vein may find a smaller resistance to flow caudally than towards the hepatic sinusoids creating a hepatofugal (away from the liver) flow direction in the portal vein segment between the gastroduodenal vein and the shunt (Figure 1). In dogs, in which this flow does have a normal (hepatopetal, i.e., towards the liver) direction, a well-developed portal system may be expected (resulting from a low intrahepatic resistance). In these cases a favorable surgical outcome may be predicted.

Several methods (e.g., intraoperative measurement of portal pressure, monitoring the color of intestines together with the magnitude of change in mean systemic arterial blood pressure and heart rate) have been used to overcome the development of post-ligation portal hypertension, however all of them failed to guarantee a consistently favorable surgical outcome. In most cases life-threatening acute portal hypertension can be successfully avoided; however, development of chronic portal hypertension with APSCs remains a frequent complication. Gradual attenuation of the shunt with using an ameroid constrictor does not prevent the development of APSCs either. The major disadvantage of an ameroid constrictor is that shunt-attenuation becomes an uncontrollable process and a hypoplastic portal system may not be able to adapt to the increased blood flow at the same rate as the contraction rate of the device would force it to.

We evaluated the flow-directions in the portal vein both cranial and caudal to the shunt origin and of the shunt itself with Doppler ultrasound during surgical ligation of CPSSs, using a sterile ultrasound probe placed directly on the portal vein at the point where the shunt originated from the portal vein (Figures). We found that shunt attenuation has an immediate effect on the flow directions in the various parts of the portal vein, and these flow direction changes can be used to guide the degree of shunt attenuation and predict the surgical outcome.4-6 We consider the surgical outcome excellent when there is no more portosystemic shunting, evidenced by a normal result of a rectal ammonia tolerance test. In dogs with an excellent surgical outcome the hepatofugal portal flow seen before shunt attenuation cranial to the shunt origin turned to hepatopetal (i.e., flow directed towards the liver) after shunt-attenuation (Figure 2). Post-operative follow-up studies 1 month after shunt attenuation revealed that the dogs, in which the portal flow remained hepatofugal failed to show clinical improvement after surgery, however the dogs, in which the portal flow became hepatopetal became clinically healthy. When the flow in the portal vein cranial to the shunt origin remained hepatofugal during a temporary applied complete occlusion of a CPSS, a suboptimal surgical outcome may be predicted because of the presence of a severe (probably a coinciding primary) portal vein hypoplasia.6 In these cases even complete shunt occlusion would be unable to create a greater resistance towards the ligature than the resistance of the severely underdeveloped portal branches. We found that shunt attenuation made not only the portal flow, but also the flow in the shunt hepatopetal (Figure 2). Hepatopetal flow in the shunt results from the anatomical fact that the splenic vein enters the shunting vessel and the blood from the splenic vein finds lower resistance to flow via the shunt to the portal vein than towards the ligature (toward the caudal vena cava) after shunt attenuation (Figure 2). Though a fraction of the splenic venous blood still flows through the partially attenuated shunt, this has no clinical significance, since the splenic vein contains equally low amount of toxins as any systemic veins, unlike the mesenteric veins. A great hemodynamic advantage of the hepatopetal flow in the shunt is that this prevents the toxin rich mesenteric venous blood from shunting (Figure 2).

We recommend that during surgical ligation of an extrahepatic CPSS, the largest shunt diameter should be found that ensures hepatopetal flow in the entire portal vein and in the shunt adjacent to the portal vein with the help of Doppler ultrasonography. A smaller shunt diameter due to exaggerated shunt occlusion would only increase the degree of portal hypertension, but would not reduce shunting, nor would it improve hepatic perfusion. Regardless of these flow directions, hepatopetal portal flow caudal to the shunt origin should always be maintained to prevent severe portal hypertension.4-6

Click on an image to see a larger view.

Figure 1.

Figure 1. The blood from the gastroduodenal vein (GDV) is responsible for the hepatofugal portal flow between the shunt-origin and GDV. The splenic vein (SPLV) seems to enter the shunt, however in reality the shunt originates from the splenic vein, and the shunting blood from the portal vein results in dilation of this segment of the splenic vein and generates hepatofugal flow in it (*).

Figure 2.

Figure 2. Shunt-attenuation adjacent to the CVC made the resistance towards the shunt higher than towards the portal branches. This results in hepatopetal flow both in the shunt and in the portal vein cranial to the shunt-origin. A portion of the splenic venous blood remains to flow towards the CVC via the narrowed segment of the shunt vessel (this flow is not indicated with an arrow on the figure).

Figures (with permission of the JAVMA5) Schematic drawings of a congenital extrahepatic spleno-caval shunt in a dog before (Figure 1) and after (Figure 2) partial shunt-ligation. Arrows indicate flow-directions. Black arrows indicate the flow that do not change directions subsequent to shunt-attenuation, open interrupted arrows indicate the flow that may change. To the left of the image is cranial, to the right is caudal. The squares indicate the regions that were imaged with ultrasound during surgery.

Abbreviations: SH shunt, PVcrSH portal vein cranial to the shunt-origin, PVcaudSH portal vein caudal to the shunt-origin, PVbrL & PVbrR left and right portal branches, GDV gastroduodenal vein, PVcrGDV portal vein cranial to the point where the gastroduodenal vein enters the portal vein

Partial Ligation Can Be Superior to Complete Shunt Ligation. When is a Second Surgery Necessary After a Partial Shunt Closure?

It had been suggested that complete shunt ligation would result in a better outcome than a partial one; and if complete shunt-occlusion is not feasible during the surgical ligation of a CPSS because it would cause the development of a fatal portal hypertension, a second surgery several months later should be performed to attempt a complete shunt-occlusion. The underlying idea is that an initial partial ligation would allow the portal system to adapt to an increased flow and the portal branches would have become gradually wider by the time of the second surgery. This speculation is based on the same theory as that of the ameroid constrictor.

Based on our studies, a superior clinical outcome after complete occlusion compared to partial ligation can only be expected in cases of intrahepatic CPSSs, but not in extrahepatic CPSSs. Furthermore, complete shunt ligation may result in a worse outcome than partial attenuation. The 2 aims of the shunt attenuation are: to cease shunt flow and to establish a hepatopetal flow in the entire portal vein. We found that dogs with hepatopetal portal flow in the entire portal vein are clinically healthy regardless of the presence of portosystemic shunting. Thus, the clinical signs seem to be more strongly associated with the reduced portal venous perfusion of the liver than the presence of hyperammonemia (i.e., persistent portosystemic shunting). Dogs, whose extrahepatic CPSS was partially ligated and the flow direction in the entire portal vein and in the shunt adjacent to the portal vein became hepatopetal, should not be re-operated on because hepatopetal flow in the shunt adjacent to the portal vein prevents the toxin-rich portal blood from shunting. Persistent shunting of the splenic venous blood via the attenuated segment of the shunt has no clinical significance, therefore requires no further narrowing. This is because though the shunt is patent, but not functional evidenced by the presence of low blood ammonia levels and normal ammonia tolerance test results. By this we showed that extrahepatic CPSSs do not have to, and actually should not, be completely occluded. A complete shunt ligation in these cases would not reduce the amount of shunting from the portal vein (because it is already zero), but would result in the development of portal hypertension.

A second surgery that aims further narrowing of a partially attenuated extrahepatic CPSS is recommended in dogs, only when one month after surgery the flow is hepatofugal in the portal vein segment cranial to the origin of the shunt. These dogs also have portosystemic shunting and persisting related clinical signs. A second surgery should only be considered when portosystemic shunting persists exclusively through the narrowed CPSS. When shunting occurs also through APSCs, further attenuation of the CPSS is contraindicated because this would only make the already existing portal hypertension more severe. To determine whether post-ligation portosystemic shunting occurs via the attenuated CPSS, or via APSCs or both is therefore critical. Color Doppler ultrasonography performed one month after surgical ligation of a shunt can reveal whether a persisting hyperammonemia is the result of a functional CPSS, APSCs, or both. Dogs with a dilated left gonadal vein should not undergo a second surgery because they already developed a portal hypertension.4-6


1.  Szatmári V, et al. J Am Vet Med Assoc 2004;224:717.

2.  Szatmári V, et al. J Am Vet Med Assoc 2004;224:713.

3.  Szatmári V, Rothuizen J. in WSAVA Standards for clinical and histological diagnosis of canine and feline liver diseases. 2006;15.

4.  Szatmári V.

5.  Szatmári V, et al. J Am Vet Med Assoc 2004;224:395.

6.  Szatmári V, et al. Vet Rec 2004;155:448.

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Viktor Szatmári, DVM, PhD, DECVIM-CA (Cardiology)
Utrecht University
Utrecht, CM, The Netherlands

MAIN : VUS : Portosystemic Anomalies
Powered By VIN