Surgical Complications


Long-term outcomes after pediatric liver transplantation are nowadays characterized by their excellence. Remarkably, when compared with older age group results, graft and patient midterm survival rates are higher, with a remarkably low loss in the longer term (for both the grafts and the patients). The survival curves are also close to a stable line after the second year. This relates to the fact that most of the candidates are very young recipients, with indications for the transplantations being nonrecurring diseases. In addition, the incidence of secondary major abdominal complications is relatively low, with most of them having been managed with success. This is together with a moderately low need for a late retransplantation. Overall, the most important cause of a patient’s later death is, in fact, infection, whereas a late graft loss often relates to the immune events, not to the surgical complications.

In this setting of a rare incidence of late surgical complications, the transplant clinician must keep his/her attention on the usually subtle symptoms of these complications. A high index of suspicion is necessary for early diagnosis, with this allowing for an adequate course of management before the condition worsens or before any clinical complications develop. Such a strategy increases the chances of successful treatment and a possible cure.

Late Vascular Problems

Late vascular complications (often defined as complications occurring more than 3 months after the transplantation) may relate to any hepatic vascular anastomosis or reconstruction, for instance, a vena cava, a hepatic vein (HV), a portal vein (PV), or a hepatic artery (HA). Late problems are typically less clinically relevant than the graft—which is already settled after the transplantation and has normalized its function—which may then better tolerate the reduction, or even the cessation, of the blood flow. This is in addition to the fact that the compensatory flow may be progressively restored by the development of a new collateral vascular supply. In many cases, the stricture of a vessel would be silent until there is an incidental diagnosis—usually by a Doppler ultrasound examination at one of the follow-up checkups. Even though a complete thrombosis may be silent for some time, the symptoms develop in a majority of patients, with the complete occlusion of a major vessel in the long term.

Inferior Vena Cava

An isolated obstruction of the vena cava is very rare. It can be observed with any type of graft (left lateral segment graft [LLSG] or full-size liver graft [FSLG]), but in an absence of the involvement of the HV outflow, it is well tolerated in children who remain asymptomatic. The diagnosis is suggested by a Doppler ultrasound or simply by the thorax x-ray showing an enlarged azygos system. An angio-magnetic resonance (MR)-computed tomography (CT) may help in defining the anatomy of the problem whenever necessary. The curing of an inferior vena cava (IVC) obstruction is rarely necessary, but this can be achieved by a radiological angioplasty with stenting ( Fig. 25.1 ).

Fig. 25.1

Complete thrombosis of the retrohepatic portion of the vena cava with redistribution of the flow through retroperitoneal collaterals and azygos veins ( left ). Antero-retrograde radiological interventional wiring of the thrombosed segment of the vena cava ( middle ). Successful recanalization of the vena cava by balloon angioplasty and internal stent positioning ( right ).

An unpublished personal experience has suggested that thrombosis of the IVC is symptomatic, but only in those cases where it is associated with severe portal hypertension. This association was observed in three patients who presented ascites and lower body edema. An IVC complete thrombosis was associated with a PV thrombosis in the first case, severe graft fibrosis and a partially functioning transjugular intrahepatic portosystemic shunt (TIPSS) in the second case, and an HV anastomosis stricture in the third case.

Hepatic Veins

The HVs represent the outflow of the graft (HVO), which is equally important for the other vessels for good functioning of the graft. A stricture or a thrombosis of the HVO is rare; most of the cases are observed after an LLSG transplantation, as this type of graft has a single HV for the HVO. An FSLG graft has not only a large HVO that is composed of three separate veins, but it also has multiple accessory veins along the vena cava that can compensate for the outflow in the case of a partial thrombosis of the HVO. The incidence of HVO problems is lower when compared with other types of graft.

An HVO obstruction (stricture or thrombosis) is rare and only occurs in less than 2% in most series. The favoring factors are a retransplantation (when the fibrous anastomotic ring of a previous transplant is reused for implanting the new graft) and the use of a nontriangular piggyback technique (up to 12% of obstruction). The incidence has now dropped close to zero after the introduction of the triangular anastomosis technique, which is now the gold standard.

An HVO stricture may be clinically silent for long periods, and this makes early diagnosis difficult. There is, however, an increasing amount of graft damage with time by progressive fibrosis. If undiagnosed and left untreated, an HVO obstruction evolves into Budd-Chiari type symptoms (hepatomegaly, ascites, and transaminitis) and secondary portal hypertension. Although the latter symptoms lead to the diagnosis, at that stage, the graft damage may not be reversible anymore, and even if the HVO obstruction is treated, the damage in the graft may progress on its own, with the risk of a fatal evolution.

Any early diagnosis can only be conducted incidentally by a high index of suspicion and by a Doppler ultrasound that shows typical observations, that is, a monophasic waveform in a venous flow pattern with a low peak velocity within the graft (both in the HV and in the portal branches) and with a significant acceleration of the flow at the outflow anastomosis, where an anatomical stricture may be seen. At an advanced stage, the PV flow may be reversed. Because many patients present mild liver test abnormalities at that stage, a biopsy may have to be performed at some point, which typically would show perivenous congestion and progressing fibrosis, or even areas of hemorrhage and necrosis. An angio-CT may help define the anatomy of the problem.

Because the evolution is moving toward severe graft damage followed by hepatic dysfunction, even in the relatively silent cases, any case with a high clinical suspicion must be assessed as early as possible. A precise diagnosis requires conventional venography and a measurement of the transanastomotic pressure gradient, with the possibility of extemporaneous management. The treatment would consist of a retrograde endovascular approach and a balloon angioplasty to dilate the venous stricture. In selected cases, a percutaneous transhepatic orthograde maneuver can be combined. A recurrence is, however, not rare, and repeated procedures, or the use of balloon-expandable stents, may be necessary. When inserting a stent, attention must be given to the positioning of the stent within the liver, with only a minimal extension into the IVC. This would be to allow for a possible surgical removal at a retransplantation if the latter would become necessary.

Many series have shown that the radiological approach is safe and very effective, providing long-lasting patency and with a good long-term outcome if the graft was not excessively damaged before the cure.

A retransplantation may be necessary in those cases with severe graft dysfunction, where the transplantation procedure helps in solving both liver disease and vascular occlusion. A TIPSS may be useful as a bridge to the transplantation to treat portal hypertension and in those cases of complications.

Hepatic Artery

Hepatic Artery Stricture

HA anastomotic strictures were found in 6.9% of the cases in a large recent series. They are often suspected incidentally at a routine Doppler ultrasound examination, whereas the liver tests and a clinical examination are unremarkably prognostic. Typically, the Doppler features are a high peak of systolic velocities (> 200 cm/s) at some point along the arterial route (usually the anastomosis), whereas the resistive index is low (< 0.5) and the systolic peak is reduced at the level of the intrahepatic vessels (“tardus parvus waveform”) ( Fig. 25.2 ). An angio-CT/MR can be useful to confirm the suspicion and define the anatomy of the lesion. The patient can be managed temporarily with an antiaggregant therapy, but a conventional angiographic assessment with a possible radiological intervention is recommended (see Fig. 25.2 ) to avoid the progressions toward a thrombosis, a graft ischemic damage, and a possible graft loss.

Fig. 25.2

Typical arterial anastomotic stricture: aspect at conventional angiography before angioplasty ( above ), and Doppler ultrasonographic observations ( below ) at the level of the stricture ( left ) and of the intrahepatic arteries ( right ) with tardus parvus aspect.

Although thrombosis, as well as an intimal dissection, or an arterial rupture have been reported after angioplasty for an HA stricture, a transluminal balloon angioplasty by experienced teams is nowadays associated with satisfactory results and a positive outcome. In the case of an unsuccessful plasty and a recurrence, or primarily in selected cases, a stent may be positioned. An indication for a surgical revision and revascularization are limited to those cases with complex lesional anatomy (i.e., a tortuosity of the arterial route), or an acute thrombosis after a radiological maneuver that did not respond to the thrombolytic treatment.

Hepatic Artery Late Thrombosis

The late thrombosis of an artery (LHAT) is often clinically silent at the event itself. With time, collaterals develop, and they reconnect to the intrahepatic arteries. In children, this phenomenon often happens at the level around the bilio-jejunal anastomosis, feeding on the mesentery of the Roux-en-Y loop that is used for a biliary reconstruction ( Fig. 25.3 ).

Fig. 25.3

Development of arterial neocollaterals after thrombosis of the graft hepatic artery (1). The intrahepatic arterial system (5) is reperfused through collaterals around the bilio-jejunal anastomosis (4), feeding from the artery of the Roux-en-Y jejunal loop (3) arising from the superior mesenteric artery (2). Because of ischemic biliary stricture, a biliary catheter (BC) has been positioned transhepatically in this patient. (0, recipient common hepatic artery).

Although the arterial reflowing from the neocollaterals into the transplanted liver contributes to minimizing the consequences of ischemia in a few patients, the LHAT is frequently associated with a secondary biliary pathology (diffuse cholangiopathy and multiple intrahepatic strictures) because the artery is the sole blood supply of the transplanted bile ducts. Typically, the symptoms consist of mild abnormalities in the liver tests (1.5–2 times normal levels) that are found at routine follow-up checks in otherwise initially healthy patients. Of these tests, the gamma-glutamyl transferase levels are sensitive indicators of biliary ischemic sequelae from the LHAT.

With time and with the evolution of the sequelae, patients develop recurrent sepsis (cholangitis) and progressive cholestasis, although this is not constant. A liver biopsy is often aspecific, but the sequelae are evolving toward progressing fibrosis and a biliary obstructional feature.

Because of the compensatory collateral arterial circulation to the liver graft, the radiologist may find—at a distance from the thrombotic event itself—a relatively normal Doppler signal within the liver. However, a detailed ultrasonographic study will not show a patent hepatic arterial trunk at the porta hepatis and outside of the graft. Because the diagnosis of LHAT is difficult and comes with uncertainty, more invasive imaging (angio-CT, angio-MR, or angiography) may be necessary in those cases where a precise diagnosis is necessary (i.e., to explain the chronic liver test abnormalities or the recurrent cholangitis of a dubious etiology).

Because a late thrombosis of the HA is often diagnosed incidentally while searching for the cause of the liver graft abnormalities or the biliary problems, there is often no opportunity for a direct corrective action (i.e., an interventional radiology procedure, a surgical exploration, or thrombolysis). The management remains palliative and symptomatic, with the use of ursodeoxycholic acid, to optimize the bile flow and with antibiotics to treat the bouts of cholangitis. At a later stage, with biliary obstructions, the role of interventional radiology is also limited, as a secondary ischemic cholangiopathy is characterized by multiple intrahepatic strictures that are usually not amenable to a cure. Many patients will slowly evolve through chronic obstructive cholestasis and recurrent cholangitis toward secondary biliary cirrhosis and, ultimately, the need for a retransplantation.

Portal Vein

With an incidence of up to 11% of cases, a late PV stenosis (LPVS) or a thrombosis (LPVT) is a relatively common complication after pediatric liver transplantation. Anastomotic strictures can be observed in any type of graft, whereas an LPVT is more typically observed after a split or a living donor liver transplantation. There is a strong association between a small weight/age and biliary atresia on one side and an occurrence of portal venous problems on the other side. This relates to the fact that biliary atresia patients are often transplanted at a young age/small weight, and that many present a hypoplasia of the PV (diameter < 5 mm) and/or anatomical variations of the portal tract (a preduodenal or an absent PV trunk).

Because of a PV hypoplasia, many of these patients have a variant PV reconstruction at the transplantation, with them mostly using venous grafts (either from the donor or from a third party). Although the use of vascular grafts has been associated in the past with an increased risk of vascular thrombosis (especially when using cryopreserved venous homografts), there is evidence that the actual type of reconstruction matters—not the use of a venous graft/reconstruction per se—and excellent results have been reported with a longitudinal PV plasty.

Portal Vein Stricture

Those patients with a significant anastomotic stricture of the PV anastomosis or the trunk often remain clinically silent for months or years. The diagnosis is often made at one of the follow-up checks, either as an incidental observation by ultrasonography and by a Doppler examination, or by a clinical finding of a recurrence of splenomegaly and thrombopenia in an otherwise unremarkable patient. The major risk is, of course, an evolution toward a complete thrombosis of the PV.

In patients with an LPVS, the typical observation is a fibrous ring at the level of the anastomosis that is visible on a Doppler ultrasound, with flow velocities greater than 150 cm/sec at the level of the stricture ( Fig. 25.4 ) and with splenomegaly. Aneurysmal dilatation of a portion of the vein downstream is frequently associated (in the case of an LLSG, the dilated portion is usually in the Rex recessus) (see Fig. 25.4 ).

Fig. 25.4

Typical imaging of a significant stricture of the portal vein anastomosis (small arrow) after transplantation of a left lateral segment graft. Pseudoaneurysmal dilatation of the recessus of Rex (large arrow) and high flow velocities at Doppler examination of the level of the anastomosis.

In the case of a relevant stricture, a clinical examination can confirm the presence of the symptoms of portal hypertension. The logical strategy is then to propose a cure of the problem (and avoiding at best, further complications, such as the thrombosis of a vein) by a radiological intervention. The latter consists of a percutaneous transhepatic portography, with a measurement of the pressure gradient through the stricture, followed by a balloon venoplasty.

A percutaneous balloon plasty is associated with excellent outcomes and long-term success in more than 90% of cases. In a few cases, the procedure must be repeated at a later time, and stenting may be necessary in selected cases. Stents are useful to treat a highly resistant or recurring stricture or a positional pseudostricture that typically would expand under ballooning but would immediately reform itself.

When using a stent, the radiologist should be very cautious in choosing a stent that can be expanded to at least 10 mm in the future (when the child has further grown) so that the flow can be adjusted to the physiological needs. The stent should also be short enough to be positioned in the portal trunk but without encompassing the spleno-mesenteric confluence or entering the Rex recessus. This may be important for any future interventions (i.e., a meso-Rex bypass or a retransplantation) ( Fig. 25.5 ).

Fig. 25.5

Management of a stricture of the portal vein anastomosis with stent insertion: a short stent is carefully positioned high enough to preserve the spleno-mesenteric confluence and low enough to not encompass the Rex recessus. This strategy preserve the venous structures for possible future need for either a meso-Rex bypass or liver retransplantation.

Portal Vein Thrombosis

Those patients with a complete thrombosis of the PV present unique features, along with what is usually a long and complicated clinical story because a chronic portal hypertension is secondary to that condition, and it drives various complications recurrently and successively over a very long period (esophageal and gastric varices, portal enteropathy, chronic anemia, ascites, and gastrointestinal hemorrhage). When left to its spontaneous clinical course and in the absence of a cure, an LPVT is an important cause of morbidity, a decreased quality of life, a late graft loss, and even the patient’s death in the long term after pediatric transplantation.

Often, the flow of the PV has been slow for a prolonged time before the thrombosis, and the latter event occurs in an absence of acute symptoms. The diagnosis is usually incidental at a routine Doppler ultrasound check. Typically, the graft is unremarkable and a portal flow is observed, but the velocity is lower than usual (flow velocities of around 10 cm/sec). More evident is the absence of an identifiable extrahepatic PV trunk. Although this might seem surprising, thrombosis of the PV is only limited to the extrahepatic trunk of a PV, with the spleno-mesenteric confluence on the one side and the recessus of Rex on the other side remaining patent. Within the liver, some parenchymal areas present a retrograde portal flow that feeds the other areas with an orthograde portal flow, and this maintains some flow into the intrahepatic portal system and, thereby, its patency.

With time, venous collaterals develop from the splanchnic area toward the liver—through the adhesions and along the biliary reconstruction—that reconnect and refeed the portal branches within the liver. These collaterals then increase in size and number, and they create a real “cavernoma” at the porta hepatis. These cavernomas, however, never develop sufficient venous flow to achieve a significant reduction of the portal pressure. In the setting of an LLSG, the cavernoma usually develops along the Roux-en-Y loop, and the typical presentation is that of a cavernoma that develops through the mesentery and the mucosa of the Roux-en-Y loop ( Fig. 25.6 ). Directly related to that particular anatomical condition, these children present a digestive hemorrhage in an absence of esophageal varices or a gastropathy—in fact, this happens from the mucosal varices at the distal part of the Roux-en-Y loop (see Fig. 25.6 ). Because of the collateral feed to the liver around the biliary anastomosis, some patients present mild cholestasis that is secondary to extrinsic compression of the biliary anastomosis, similar to a portal biliopathy in those patients with an idiopathic cavernoma.

Fig. 25.6

Development of a venous collateral network after thrombosis of the portal vein in a patient with a left lateral segment graft. The Roux-en-Y jejunal loop used for the biliary reconstruction is a natural route for reperfusion of the graft: from the superior mesenteric vein (1), the splanchnic blood follows the jejunal vein (2) to the end of the Roux loop where a cavernoma develops with time (3). This cavernomatous transformation of the blind end of the loop reconnects with the rex recessus (4) and refeeds the intrahepatic portal veins.

Because the major complications usually only occur late in the evolution of this chronic condition, most teams have opted in the past—and still do—for a “wait and see” strategy over many years. This, however, lasts for a time, until the condition worsens and the patients accumulate comorbidities (chronic anemia, recurrent gastrointestinal bleedings, ascites, abdominal distension and discomfort, asthenia, and sport limitations) over the years, with an increasingly lower quality of life. At some point, the child’s condition usually becomes a concern and is a challenge for any type of intervention, when, in fact, the only alternatives are opting for a difficult portosystemic shunt or a high-risk retransplantation.

A better option is to consider curing the problem at an early stage because curing the problem allows for them to maintain a good quality of life in the absence of a progression of portal hypertension. Because of the Rex recessus, the intrahepatic portal branches and also the mesenteric vein are usually patent in these patients, and these patients can be considered for a meso-Rex bypass intervention. A meso-Rex bypass has been shown to be a cure of the condition, with physiological hepatic revascularization and with a reversal of all of the abnormalities that were caused by the portal hypertension.

Because it is obvious that the evolution of the patient is slowly increasing toward the symptoms of morbidity and also because chronic portal hypertension is associated with intraabdominal anatomical changes that make a reintervention more demanding with time (a worsening of the venous stasis, a collateral development, a thickening of the mesentery, a growing perihepatic cavernoma), the Meso-Rex bypass must be proposed earlier rather than later after an LPVT is diagnosed. A recommendation is to proceed within the first 2 to 3 years after diagnosis. With time and in a more severe portal hypertension setting, this type of surgery is increasingly demanding, and it can become hazardous in those cases that display 10 or more years of portal hypertension.

The preparation for a meso-Rex intervention includes an angio-CT and a retrograde portography for assessing the abdominal vasculature and the patency of the intra-HVs. A Doppler ultrasound of the neck is useful to check the procurability of one of the internal jugular veins. Because other problems alter the quality of the hepatic parenchyma (i.e., a chronic rejection, biliary cirrhosis) that would compromise the flow (owing to the higher resistance of a fibrotic liver), for the success of the intervention, a liver biopsy is recommended to assess the grade of the fibrosis.

Technically speaking, the meso-Rex bypass procedure is as standardly described when it comes to managing an LPVT after FSLG, but it must be adapted in the case of an LLSG. In the latter case, the anatomy is modified by the size of the graft and its position in the abdomen, which shifts the Rex recessus into a more anterior position that is more caudal and more to the right. For these reasons, the bypass is best routed straight between the duodeno-pancreas and the mesocolon, from the superior mesenteric vein to the Rex recessus. This route is, in fact, copying the natural drainage that develops through the jejunal vein of the Roux loop in LLSG patients after LPVT ( Fig. 25.7 ). Resecting the edge of the lower aspect of segment III is useful to allow a straight routing of the bypass from the mesenteric vein toward the Rex recessus.

Feb 23, 2021 | Posted by in HEPATOPANCREATOBILIARY | Comments Off on Surgical Complications

Full access? Get Clinical Tree

Get Clinical Tree app for offline access