記住我
Abstract
Congenital portosystemic venous connections are rare vascular anomalies that connect the portal and systemic venous circulations. These vascular lesions can lead to complex and varied physiologic manifestations in single-ventricle patients as they progress through the various stages of palliation in the Fontan pathway. Specifically, these connections may be unmasked after a superior cavopulmonary anastomosis operation, but then “re-masked” after Fontan completion. We describe the complex physiology with an illustrative case report and present a novel method to unmask these pathologic connections after Fontan completion to facilitate transcatheter occlusion and resolve downstream symptoms.
Keywords: Congenital portosystemic venous connections, Fontan complications, hepatic factor
How to cite this article: Introduction 
Congenital portosystemic venous connections (PSVCs) are rare vascular anomalies, first described by Abernethy, that connect the portal and systemic circulations.[1] Although the primary systemic and hepatic veins all ultimately drain into the atrium in patients who have not undergone cardiac surgery, the hepatic parenchyma does increase resistance to flow. Thus, a PSVC creates the potential for hepatofugal flow (i.e., away from/bypassing the liver), whereby blood from the portal vein flows directly into the systemic venous system, bypassing the liver without undergoing hepatic metabolism [Figure 1]. As a result, patients with PSVC may present with hyperammonemia-related delirium, pulmonary arteriovenous malformations (pAVM) related to a lack of the putative hepatic factor, and pulmonary hemorrhage, among other problems. Importantly, some patients may present later in life, often in the teenage years or later, as some physiologic consequences take time to develop and manifest.[2],[3]
Figure 1: Physiology with an iIVC and PSVC before cavopulmonary interventions. Since both the hepatic and azygos veins drain to the heart, portal venous flow is hepatofugal (i.e., away from the liver) given the increased resistance to flow through the hepatic parenchyma. PSVC: Portosystemic venous connection; iIVC: Interrupted inferior vena cava; SMV: Superior mesenteric vein; SpV: Splenic vein; PV: Portal vein; HV: Hepatic vein; AzV: Azygos veinPSVC has been associated with other congenital anomalies and syndromes, most notably the polysplenic variant of heterotaxy (HTXp), in as many as 8% of those patients.[4] Similarly, a study of patients with univentricular heart disease and HTXp found that 19% had PSVCs.[4] Given the anticipated variance from “normal” in those patients (e.g., chronic cyanosis), coupled with the natural history of presentation after early childhood, PSVC may be underrecognized at the time of early cardiac presentation and management. In addition, in our experience, computed tomography (CT) angiography may show abnormal abdominal visceral venous connections in HTXp patients. However, and especially in cases with complex PSVC, the CT imaging is typically insufficient to provide clear anatomic detail of each vessel's course and connections. Moreover, CT does not allow for interventions. Therefore, we recommend routinely assessing for PSVC in all patients with univentricular heart disease and HTXp at the time of pre-Fontan catheterization. We describe a transcatheter technique to identify and treat occult PSVC in the setting of complex congenital heart disease, which can be tailored to meet patient-specific anatomy and physiology.
Procedure Description As described above, the presence of PSVC allows hepatofugal flow from the portal vein into the systemic venous circulation [Figure 1]. Thus, the standard angiographic technique in a systemic vein is unlikely to densely opacify the PSVC since the flow of blood is directed toward the vessel being injected. Hence, the physiologic construct underpinning this technique is the temporary exponential elevation of resistance to PSVC flow into the systemic venous circulation, without alteration of resistance through the hepatic veins. Accomplished through the use of balloon occlusion, this physiologic manipulation results in the temporary reversal of PSVC flow direction to be hepatopedal (i.e., toward/through the liver), thus facilitating easier identification through conventional venography. Creating this obstruction is accomplished with a compliant balloon, expanded to completely occlude the inferior vena cava (IVC) (or azygos vein in the setting of congenital IVC interruption) between its connection with the PSVC and the hepatic vein. Special attention must be paid to patients in whom the hepatic venous drainage is anomalous or has been altered surgically (e.g., interrupted IVC status post Fontan completion with hepatic venous connection to the azygos vein) [Figure 1], [Figure 2], [Figure 3].
Figure 2: Physiology s/p superior cavopulmonary anastomosis. (a). In patients with iIVC s/p Kawashima, the resistances through the azygos (i.e., pulmonary arterial circulation) and hepatic veins are different. As a result, PSVC flow becomes hepatopedal (i.e., toward/through the liver). (b). With normal caval anatomy, completion of the Glenn does not alter the relative resistance through the IVC and hepatic veins. Thus, PSVC flow is not “unmasked,” remaining hepatofugal, coursing through the hepatic portion of the IVC but bypassing the portal triad and liver parenchyma. Abbreviations: PSVC: Portosystemic venous connection; iIVC: Interrupted inferior vena cava; SMV: Superior mesenteric vein; SpV: Splenic vein; PV: Portal vein; HV: Hepatic veins; AzV: Azygos vein
Figure 3: (a) Physiology s/p Fontan completion with iIVC. The Fontan has reverted to initial physiology, with the resumption of hepatofugal flow. (b) Schematic of the technique to unmask the PSVC. The IVC/azygos is completely occluded between the PSVC and hepatic vein confluence connections. This occlusion forces hepatopedal flow through the PSVC and PV. Abbreviations: PSVC: Portosystemic venous connection; iIVC: Interrupted inferior vena cava; SMV: Superior mesenteric vein; SpV: Splenic vein; PV: Portal vein; HV: Hepatic veins; AzV: Azygos veinAfter hemodynamic assessment, angiography is performed in the IVC/azygos and its diameter is measured [[Figure 4] and Video 1 [Additional file 1]]. An appropriate-sized, compliant balloon (our choice is the Tyshak II [B. Braun, Bethlehem, PA USA]) is then advanced just caudal to the hepatic vein connection and inflated to completely occlude the IVC/azygos vein, without obstructing the hepatic vein confluence, or the PSVC connection. Repeat venography caudal to the occlusion balloon will opacify the PSVC [[Figure 5] and Video 2 [Additional file 2]]. Should the occlusion balloon also obstruct hepatic venous egress, all PSVC flow will cease. Other venovenous collaterals may be present and potentially impact angiographic interpretation. Opacification of the distal portal venous tree may not occur in the absence of selective venography, but when present, is often most easily identifiable in the lateral projection [Figure 6]. Hence, this initial screen for PSVCs involves one additional angiogram with a minor resultant increase in contrast dose and radiation.
Figure 4: Frontal (a) and lateral (b) projections of angiography in an unobstructed azygos of a patient with iIVC s/p Kawashima with subsequent “Fontan” completion via a left-sided hepatic vein conduit connection to the hemiazygos (*). The location of this connection is crucial, as occlusion must be performed caudal to the connection (†) for this technique to be successful
Figure 5: Frontal projections of angiography in the azygos of a patient with hepatic vein conduit into the central PAs. (a) The azygos is unobstructed, with contrast flow cephalad into the left cavopulmonary anastomosis. * = renal vein, arrow = renal pelvis. (b) The azygos is occluded by a Tyshak balloon (#). Note that the azygos is occluded cephalad since the hepatic veins are baffled directly into the pulmonary arteries. With balloon occlusion, a meshwork of PSVCs (^), coursing between the renal and superior mesenteric (†) veins, are unmasked; specific vessel anatomy will be better visualized via selective injection in the renal vein (i.e., the systemic vein from which the PSVCs originate). The portal vein ({) and intrahepatic portal system are clearly visualized
Figure 6: Portal system. Direct injection in the renal vein of the same patient from Figure 5, better delineates the PSVCs. A = portal vein; B = SMV; arrow = insertion of a splenic vein; arrowheads = multiple, serpentine PSVCs, arising from the renal vein and coursing to the SMV. The arborizing portal venous system is visualized in the hepatic parenchyma (*), SMV: Superior mesenteric veinIf evidence of PSVC is demonstrated, each feeder vessel connected to a PSVC is selectively engaged and the balloon occlusion maneuver is repeated, with selective venography, to define the anatomy of each PSVC. With IVC/azygos balloon occlusion, contrast opacification of the PSVC should flow hepatopedal, outlining the entire length of the PSVC, and ultimately connecting to the portal system [[Figure 7] and Video 3a [Additional file 3], Video 3b [Additional file 4]]. With the IVC/azygos egress unobstructed, the contrast will backfill the proximal PSVC during injection, but ongoing hepatofugal flow will wash out the contrast and reflux to the IVC/azygos vein, through the unobstructed PSVC [Video 4a [Additional file 5] and Video 4b [Additional file 6]]. Following each angiographic interrogation, the occlusion balloon should be deflated, to prevent sequelae of prolonged IVC/azygos venous hypertension.
Figure 7: Frontal (a) and lateral (b) projections of a selective injection in a PSVC (*) with the azygos occluded, forcing hepatopedal flow through the PSVC, ultimately draining into the splenic (^) and then portal (#) veins, PSVCs: Portosystemic venous connectionOnce the anatomy of each vessel has been defined, vascular occlusion can be performed. Given that these vessels are often dilated and tortuous, coupled with the need to ensure targeted embolization, our approach is to advance a microcatheter as deep into the PSVC as possible, define the portal and mesenteric venous anatomy, and then deliver detachable embolization coils. Care must be taken to ensure that occlusion devices/coils are placed in the PSVC and not in the splenic or superior mesenteric veins. A final broad interrogation is performed by performing IVC/azygos angiography during balloon occlusion; the goal is lack of flow through PSVCs into the portal venous system [Video 5a [Additional file 7] and Video 5b [Additional file 8]].
Discussion In patients with PSVC, the direction of portal venous blood flow ultimately depends on circulation status, vascular connections, and downstream resistances. In single-ventricle circulation following Fontan completion, PSVC flow is typically hepatofugal, and therefore not subject to routine identification with traditional caval venography. Further, and similar to the issue with pre-catheterization computed tomography angiography, superior mesenteric artery angiography with delayed venous visualization does not provide clear anatomic delineation of the portal venous system and PSVCs due to the complex nature of these connections and the hepatofugal portal venous flow. Our technique manipulates PSVC flow by altering the relative resistance of the systemic vein connecting to the PSVC and the hepatic veins. The presence of a PSVC can be “unmasked” if the resistance through the PSVC is increased while maintaining the same resistance through the hepatic venous limb of the circuit.
We developed this technique based on observations in a subset of single-ventricle patients with interrupted IVC and PSVC who have undergone the Kawashima operation as well as experience with a critically ill older Fontan patient. That older patient had PSVCs identified at the pre-Fontan catheterization which were not treated. She was lost to follow-up for a long period after the Fontan, and ultimately presented 15 years later with profound cyanosis, polycythemia, and was unable to walk 50 m. Catheterization demonstrated a massive PSVC network with a severely hypoplastic portal venous system and severe pAVMs bilaterally. We employed this technique to occlude all the PSVCs, but her condition unfortunately deteriorated and she ultimately expired. Thus, early recognition of these PSVCs is critical to prevent the development of potentially irreversible portal venous hypoplasia and pAVMS.
The key anatomic issue is that performing the Kawashima creates different resistances between the systemic and hepatic veins that are connected by the PSVC. Specifically, the resistance through the azygos (i.e., the systemic vein connecting to the PSVC) becomes considerably higher following connection to the pulmonary arteries at the time of superior cavopulmonary anastomosis (Kawashima), while the hepatic venous resistance remains low as this pathway still connects directly to the systemic atrium. At this stage, PSVC flow reverses and normalizes (becomes hepatopedal), which should be overt during catheterization with inferior caval venography. Importantly, the subsequent Fontan completion will “re-mask” the PSVC, depending on the vascular connections, when the hepatic veins are surgically directed to the pulmonary arteries through the Fontan circuit, thereby increasing resistance to hepatic venous egress [Figure 1], [Figure 2], [Figure 3]. Therefore, the pre-Fontan catheterization presents a unique opportunity to readily identify and occlude these anomalous portocaval connections before the development of complications which can ultimately result in chronic morbidity and mortality (e.g., hypoxemia related to pAVM, delirium secondary to hyperammonemia, etc.).
The putative “hepatic factor” plays a significant role in the development of PSVC-related morbidity. Awareness of this factor first came from observations that pAVMs developed in pulmonary vascular beds that were not exposed to hepatic venous effluent.[5] Growing evidence suggests the existence of a substance, as yet unidentified, within the hepatic venous effluent which inhibits the recruitment and dilation of preexisting pulmonary arteriovenous connections.[6] Notably, single-ventricle patients who have undergone superior cavopulmonary anastomosis, where the pulmonary arteries are devoid of hepatic venous effluent, often develop pAVM. Typically, when no PSVC is present, Fontan completion re-introduces hepatic venous effluent to the pulmonary vasculature with resultant regression of pAVM over time.[7] Although a case has been reported of pAVM regression after 9 years with Glenn physiology, the ability for pAVMs to regress once hepatic factor is re-delivered remains unknown, hence the importance of recognizing these lesions.[8]
In conclusion, PSVC is associated with complex anatomy and dynamic physiology, compounded by changes in vascular circuit resistances in patients undergoing staged single-ventricle palliation. Although additional data are needed, early recognition and treatment of PSVC before the development of irreversible portohepatic pathology is likely to have a positive impact on long-term patient outcomes. Given the occlusion maneuvers required to unmask these vessels, congenital interventional cardiologists are optimally suited to diagnose and treat these vascular lesions. Greater awareness of these lesions, a better understanding of their evolving physiology, and increased attention at the time of pre-Fontan catheterization should improve detection and therapeutic result.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References 
Correspondence Address:
Dr. Sarosh P Batlivala
Department of Pediatrics, The Heart Institute, Cincinnati Children's Hospital Medical Center, University of Cincinnati College of Medicine, 3333 Burnet Ave, MLC 2003, Cincinnati, OH 45229
USA
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/apc.apc_34_22
留言 (0)