Transient portal venous gas

An 18-year-old female patient presented to the emergency room with epigastric pain for 2 months with acute exacerbation for the past 3 days. The pain was resistant to oral antacids. They was no history of analgesic abuse or steroid intake. No other comorbidities.

Abdominal ultrasound was unremarkable. The patient was started on intravenous (iv) proton pump inhibitor (PPI).

A contrast-enhanced computed tomography (CECT) of the abdomen on the next day revealed extensive gas within the intrahepatic portal vein branches. The extrahepatic portal vein, superior mesenteric vein (SMV) and splenic vein (SV) were unremarkable. The posterior wall of the fundus of the stomach showed discontinuous, suboptimal mucosal enhancement with pneumatosis. There was a speck of air within the lesser sac. Bowel loops were unremarkable. An upper gastrointestinal endoscopy confirmed multiple fundic erosions.

The patient responded to iv PPI and antibiotics. CECT abdomen was repeated after three days, showing near complete resolution of portal venous gas, except for a tiny focus in the left portal vein with a small thrombus adjacent to it. The thrombus was a new finding. The fundus of the stomach showed normal mucosal enhancement with no obvious mucosal discontinuity.

The patient remained asymptomatic at follow up after 2 months.

The presence of gas only in the portal vein, sparing the SMV and SV along with the fundic mucosal abnormality on CECT and endoscopy suggest that those mucosal erosions might have resulted in intraluminal air gaining access to the portal venous system via the left gastric vein. The simultaneous resolution of gastric mucosal pathology and portal venous gas along with no recurrence of symptoms at follow up strengthens our hypothesis.

Hepatic portal venous gas (PVG) could occur either due to i) entry of gas from bowel lumen or abscess into the portal venous radical or ii) Infection of the portal venous system by gas-forming organisms [1]. PVG can be differentiated from biliary gas by the presence of air within 2 cm of liver capsule [1]. The PVG predominantly enters the left lobe branches because of its non-dependent location in supine position [2]. The presence of PVG could represent a benign finding or an ominous sign of an underlying pathology [3]. The common causes of PVG include bowel ischaemia (most common [4]), necrotising enterocolitis, diverticulitis, inflammatory bowel disease, intraabdominal abscesses, cholangitis and iatrogenic causes like colonoscopy, gastric dilatation etc [1, 5]. Transient PVG has been described in cases with reversible aetiologies like bowel ischaemia which recovered, acute gastric dilatation, diverticulitis and temporary hypotension [3, 4, 6, 7]. The majority of the patients with PVG have mucosal discontinuity, mural ischaemia or raised intraluminal pressure [2].

The presence of pneumatosis intestinalis can help in localising the cause for PVG. Also, the distribution of PVG can point towards the site of aetiology. Like in our case, the presence of gas selectively within the portal vein with sparing of SMV and SV, relatively narrows the likely site of aetiology to stomach, duodenum or pancreas.

PVG can be detected on X-ray, ultrasound, CT scan or MRI, the latter not commonly used for this purpose. X-ray has very low sensitivity, with the highest being seen in ultrasound [1, 2]. The ultrasound in our case was unremarkable, could be because the gastric erosions were yet to communicate with the left gastric vein radicals. The increasing use of imaging modalities have resulted in detection of minimal amount of PVG and detection of PVG in increasing proportion of benign cases [3].

PVG is not a diagnosis, but a radiological sign. Hence, treatment depends on the underlying cause. In our case, treatment of the gastric erosions resulted in resolution of PVG within days.

Written informed patient consent for publication has been obtained.