Portal hypertension — a potentially life-threatening rise in pressure within the vein that drains blood from the intestines and spleen into the liver — is rare in children, yet it behaves so differently from the adult disease that pediatric specialists have long improvised with guidelines that never quite fit. That era may now be ending. A sweeping review published on 9 July 2026 in the journal Pediatric Radiology lays out, in one authoritative document, how modern imaging should be used to detect, classify, and track portal hypertension in children, and how image-guided procedures can sometimes restore normal liver blood flow without transplantation. The international team, led by radiologists Julian Jürgens of University Medical Center Hamburg-Eppendorf in Germany, Paolo Marra of the University of Milano-Bicocca and Papa Giovanni XXIII Hospital in Bergamo, Italy, and Stéphanie Franchi-Abella of Hôpital Bicêtre, Université Paris-Saclay, in France, argues that ultrasound, Doppler, elastography, CT, and MRI now form a complete diagnostic chain — from the first suspicion of disease to lifelong surveillance of rerouted circulation.

The technical definitions explain why children cannot simply be treated as small adults. Portal hypertension is diagnosed when pressure in the portal vein exceeds 10 mmHg and the hepatic venous pressure gradient — the difference between wedged and free hepatic venous pressure, measured with a catheter — exceeds 5 mmHg, with complications becoming far more likely once the gradient passes 10 mmHg. Measuring that gradient in a child, however, requires hepatic vein catheterization under general anesthesia, which is precisely why it is rarely performed in this age group. The review therefore endorses the operational concept of clinically evident portal hypertension, recognized through an enlarged spleen with hypersplenism, gastroesophageal varices, and ascites. Causes fall into prehepatic, intrahepatic, and posthepatic categories, and in children the dominant culprits are extrahepatic portal vein obstruction and biliary cirrhosis, most often a consequence of biliary atresia. Rarer entities — porto-sinusoidal vascular disease, cystic fibrosis liver disease, and the hepatic venous outflow block of Budd–Chiari syndrome — complete a differential diagnosis in which imaging is decisive.

Ultrasound remains the cornerstone of evaluation, and the reviewers insist on a systematic, multiparametric protocol rather than a casual look at the portal vein. Under normal conditions, flow in the portal and splenic veins is hepatopetal — directed toward the liver — and continuous, gently modulated by respiration. As portal hypertension worsens, flow velocities and phasicity progressively decline; a to-and-fro waveform may emerge in advanced disease, while sustained, complete hepatofugal flow signals severe hypertension. Hepatofugal flow in the splenic vein is an especially ominous finding, although it can occasionally reflect segmental hypertension confined to the splenic circuit. Pulsatile, arterialised flow anywhere in the portal system should raise the possibility of arterio-portal communications. Caliber changes tell their own story: a dilated main portal vein supports the diagnosis, whereas a small-calibre vessel points to chronic obstruction or portal vein hypoplasia, as seen in biliary atresia. Because mesenteric inflow rises after meals, Doppler findings differ between fasting and postprandial states — healthy fasting individuals can show transient flow reversal in anterior subsegmental branches — so fasting status must always be recorded. Body habitus and distorted, atrophic livers remain the main technical limits.

The technique also has treacherous pitfalls. After chronic extrahepatic portal vein obstruction, the blocked vessel is typically replaced by cavernous transformation — a spongy network of tortuous periportal collaterals. The commonest error is mistaking a large, winding cavernous vein for a patent main portal vein; the genuine vessel runs straight at the liver hilum, while cavernous veins are always tortuous. Conversely, when cavernous collaterals coexist with a patent main portal vein or patent intrahepatic branches, the authors urge suspicion of porto-sinusoidal vascular disease rather than simple obstruction — a distinction with major therapeutic consequences, confirmed by CT and histology in reported cases. The hepatic veins and inferior vena cava demand equal scrutiny, because narrowed or occluded vessels, loss of the normal triphasic waveform, continuous low-velocity flow, and abnormal venous pathways betray Budd–Chiari syndrome. Collaterals themselves confirm the diagnosis and gauge severity: thickening of the lesser omentum in infants, a persistent patent ductus venosus beyond the first month of life, para-umbilical veins within the ligamentum teres, and spleno-renal shunts best seen in coronal planes. Splenomegaly, reported in up to 98 percent of affected children, is often the earliest or only clue — when it appears without explanation, color Doppler of the portal and splenic veins becomes mandatory, interpreted against age- and size-adjusted normative values.

Beyond moving pictures, ultrasound can now quantify the disease. Elastography measures tissue stiffness: liver stiffness reflects fibrosis and congestion but is confounded by inflammation and cholestasis, while spleen stiffness mainly tracks portal venous congestion. Using transient or shear-wave techniques, several pediatric studies show that spleen stiffness consistently outperforms liver stiffness in predicting clinically significant varices, with diagnostic thresholds typically between 28 and 40 kilopascals and areas under the receiver-operating-characteristic curve above 0.85 to 0.90. Spleen stiffness also falls significantly after a portosystemic shunt is created in children with extrahepatic portal hypertension, making it a practical treatment-response biomarker, and the spleen-to-liver stiffness ratio has been proposed to separate presinusoidal from sinusoidal disease, albeit with moderate accuracy. The adult Baveno VII consensus already endorses both measurements; pediatric equivalents are pending, reference ranges remain fragmented across vendors and probe types, and the reviewers advise that serial measurements be obtained on the same system with the same transducer.

When ultrasound is inconclusive, the pathway escalates. Contrast-enhanced ultrasound, although not formally approved for intravascular use in children in many countries, can delineate the extrahepatic portal vein when color Doppler fails, though not yet routine. Contrast-enhanced CT delivers superb portal-phase vascular mapping for preoperative planning, yet ionising radiation confines it to specific indications — a caution underscored by recent analyses of pediatric cancer risk — although child-specific low-dose protocols and emerging photon-counting scanners may broaden its role. MRI is the workhorse for complex cases: dynamic contrast-enhanced sequences resolve arterial, portal venous, and systemic venous phases separately, improving visualization of collaterals and varices; magnetic resonance cholangiopancreatography exposes portal biliopathy; diffusion-weighted imaging probes parenchymal health without contrast; and hepatocyte-specific agents, though off-label, add indirect functional insight. Magnetic resonance elastography reproduces well across platforms but stays technically demanding, less available, and harder in young children.

Therapeutically, the review draws a principled line between physiological and non-physiological solutions. Initial management relies on medical therapy and endoscopic control of variceal bleeding; when these measures fail or hypertension remains severe, imaging decides everything. If the block sits outside the liver — extrahepatic portal vein obstruction or Budd–Chiari syndrome — the goal is to restore hepatopetal portal flow, either by endovascular recanalization of the obstructed vein or by a Meso-Rex bypass, a surgical graft linking the superior mesenteric venous system to the umbilical segment of the left portal vein inside the Rex recess. These strategies are called physiological because they preserve the liver’s first-pass circulation of nutrient-rich splanchnic blood, supporting normal liver growth and metabolism. Feasibility hinges on anatomy: patency of the Rex recess, assessed by wedged hepatic venous portography via a jugular route — the gold standard when non-invasive imaging is inadequate — and classified by a pediatric scheme first proposed in 2014. For recanalization, a straight remnant main portal vein within the cavernoma and preserved straight intrahepatic branches are favorable signs; access may be transhepatic, transsplenic, or transmesenteric, with balloon angioplasty and, in most patients, stent placement. Success falls as thrombosis extends, but a 2025 preliminary cohort suggests percutaneous recanalization could become a first-line physiological option in experienced centers.

When the block sits inside the liver, or physiological routes fail, the only recourse short of transplantation is decompression: diverting portal blood into the systemic circulation. Surgeons construct shunts — mesocaval from the superior mesenteric vein to the inferior vena cava, portocaval, or the selective distal splenorenal Warren shunt that preserves hepatopetal flow in the superior mesenteric vein — using autologous jugular vein or prosthetic grafts. Interventional radiologists offer the transjugular intrahepatic portosystemic shunt, or TIPS, in which a stented tract is created between a hepatic vein and a portal branch through the liver parenchyma. Devised originally for intrahepatic disease, TIPS is now also used in selected prehepatic cases, most often as a bridge to transplantation; the small caliber of pediatric vessels makes it technically hard, and failure rates remain relatively high. Two recent meta-analyses conclude that TIPS effectively controls refractory bleeding and ascites but helps hypersplenism and thrombocytopenia less. Liver failure and complications of pre-existing shunts — portopulmonary hypertension, hepatopulmonary syndrome, hepatorenal syndrome, or encephalopathy — contraindicate further diversion. For varices beyond endoscopic reach, retrograde obliteration techniques such as BRTO and its coil- or plug-assisted variants, and partial splenic artery embolization for severe cytopenias, are adjunct options, though pediatric data remain thin.

Surveillance is where imaging earns its keep. Thrombosis tends to strike early after surgery or intervention and can present acutely; stenosis develops weeks to months later, silently rebuilding the pressure it was meant to relieve. Children with TIPS face a unique problem: as their bodies grow, the stent can become too short to maintain the hepatic venous outflow, so elective stent extension may be needed, while early stenosis usually reflects kinking, malposition, or neointimal hyperplasia. Color and pulsed-wave Doppler, performable at the bedside even perioperatively, remain the primary tools, though covered stents shadow the sound beam for at least 72 hours after placement and early postoperative windows may hide deep anastomoses. Interpretation demands knowing the operation: after a Meso-Rex bypass, reversed flow in the left portal vein is the expected sign of success, and successful recanalization enlarges intrahepatic branches alongside growing liver volume. Because absolute velocities vary between patients, relative change rules: a focal threefold velocity jump with downstream turbulence indicates significant stenosis, while velocities below 15 centimeters per second suggest dysfunction. Suspected thrombosis or shunt failure triggers CT in emergencies or magnetic resonance venography when time allows; salvage ranges from catheter-directed thrombolysis and mechanical thrombectomy to balloon angioplasty with stenting, and embolization of collaterals that steal flow.

The authors close on a forward-looking note. Standardization of pediatric elastography, viscoelastic magnetic resonance elastography, photon-counting CT, and 4D-flow MRI — which maps true volumetric hemodynamics — are expected to sharpen diagnostic accuracy, while prospective pediatric studies must validate imaging biomarkers and refine treatment algorithms. Progress in miniature interventional devices and image guidance should expand minimally invasive therapy and its safety, steering care toward individually tailored strategies and better long-term outcomes. Published open access, the review signals that pediatric portal hypertension, long a blind spot managed with borrowed adult rules, now has a complete roadmap — from the first gray-scale sweep of an ultrasound probe in a child with an unexplained enlarged spleen to the last follow-up Doppler of a restored, growing portal vein.