Conservative treatment of hepatic portal venous gas consecutive to a complicated diverticulitis: A case report and literature review

Highlights

• •A Conservative treatment of hepatic portal venous gas is described.
• •The literature showed only one clinical trial establishing a scoring system.
• •A pneumoportogram (hepatic portal venous gas) is a very rare and from poor prognosis.

Abstract

Introduction and presentation of case

Eight days after being diagnosed with multiple small strokes a 71 year old male patient is readmitted with suspicion of a petit mal seizure also complained of diarrhoea and abdominal pain. The patient was stable, not febrile and neurologically intact with a slight tenderness in the left lower quadrant. An ultrasound revealed presence of air in the hepatic portal venous system and a suspicion for sigmoid diverticulitis. A CT-scan confirmed both diagnoses. We proceeded with a conservative regimen under close observation. The clinical course and laboratory results were unremarkable.

Discussion

The review of the literature (PubMed database) triggered 685 items with only one clinical trial establishing a scoring system to detect adult individuals, which need operation.

Conclusion

A pneumoportogram (hepatic portal venous gas, HPVG) is a very rare and usually associated with bowel ischemia and from poor prognosis. The last decades saw the emergence of numerous other aetiologies (also benign) with a shift of paradigm from systematic emergency laparotomies to individual patient selection.

1. Introduction and case presentation

Eight days after being diagnosed with multiple small strokes a 71 year old male patient is readmitted with suspicion of a petit mal seizure. He also complained of diarrhoea and abdominal pain. The patient was hemodynamically stable (HR 66/min, BP 114/71 mmHg), not febrile (36.2 °C), the neurological examination revealed no further symptoms compared with the examination eight days ago. The abdominal examination showed slight tenderness with no guarding in the left lower quadrant.

A follow-up cranial CT-scan showed no further ischemia or bleeding.

The laboratory data presented a normal WBC count of 5.3G/l, a CRP level of 93.5 mg/l with elevated GGT 80 U/l and a normal blood urea nitrogen (BUN).

ECG outcome was a first grade block (PQ 208 ms) and the chest X-ray was unremarkable.

An ultrasound of the abdomen revealed the presence of massive air in the hepatic portal venous system (Image 1 ) and a suspicion for sigmoid diverticulitis. These findings were confirmed by a subsequent CT scan (Image 2 and 3 ). The sigmoid diverticulitis was staged as a covered perforation without abscess formation (IIb according to Hansen and Stock).

Image 1.

sonography with suspicion of HPVG.

Image 2 and 3.

Pneumoportogram in the CT-scan.

We treated the patient with iv Piperacillin/Tazobactam (3 × 4.5 g/d) and fluid intake. To rule out a seizure an EEG was performed and showed no epileptical activity but without certain exclusion of a petit mal event.

Due to findings consistent with complicated but mild diverticulitis (no septic findings, no signs of a gangrenous process within the sigmoid colon) and the patients comorbidities we proceeded with the conservative regimen under close observation. After 48 h the patient was allowed on solid food with consecutive return to a light diet. The clinical course and laboratory results were unremarkable with a decline of the CRP and 8 days after admission the patient was discharged. The antibiotic therapy was changed to Amoxicillin and Clavulanic acid po for a total of 21 days since the onset of antibiotics. To rule out a secondary liver abscess or a portal venous thrombosis a CT scan was planned after 2 months or immediately in worsening conditions.

To rule out malignancy a colonoscopy was performed 6 weeks after discharge. The subsequent CT-scan revealed no signs of hepatic abscess formation and a resolved diverticulitis.

2. Review of the literature

The utilized search string was ‘portal venous gas’ and was exerted on PubMed database. In total it triggered 685 items. The results were limited to adult patients and clinical settings.

Besides retrospective review of HPVG or pneumatosis intestinalis or both to establish clinical algorithms based on clinical, labarotory and radiographic findings [1] there was only one multicenter clinical trial to detect risk factors for colic ischemia or necrosis if CT findings of pneumatosis intestinalis and portal venous gas were given [2]. Further 35 case reports and reviews were available.

3. Discussion

A pneumoportogram is described as the presence of air in the hepatic portal venous system (hepatic portal venous gas, HPVG), typically at the periphery of the liver in contrast to an aerobilie [3] and was first described in 1955 by Wolfe and Evans [4]. HPVG is a rare condition and can be associated with an overall mortality rate up to 40% [5] coming down from 75% as observed in a literature survey in 1978 [6]. Aetiologies are in most cases an ischemic bowel necrosis (43%), peritonitis or abscess (11%) as may be present in the case of a diverticulitis [7]^, [8]^, [9]^, [10]. The mortality risk is highest, if HPVG is present in combination with necrotic bowel (>50%) [6].

Though the presence of HPVG was formerly considered a mandatory indication for urgent exploratory laparotomy since the late 70s and early 80s cases of HPVG without ischemic necrosis were reported and the question arose, if it is still an indication for surgery or if there may be benign courses [2]. The increase of nonfatal and nonsurgical reported conditions associated with HPVG is on the one hand due to the increasing number of CT-scans as diagnostic tool and not any more on left lateral decubitus plain films [4]^, [11]^, [12]^, [13]. On the other hand sensitivity of gas detection in CT is much higher as with conventional X-ray [14]^, [15]. Thus maybe contributing as a bias to the decreasing mortality rate in contrast to a more restricted indication for surgery as outlined underneath.

Obviously the question of surgery indication would be complicated and probably impossible to be answered by randomized controlled trials. In absence of a larger cohort study parsing out implications of HPVG in series of laparotomies for abdominal catastrophe, remain only case reports and case series to make a therapeutic decision.

According to the fact that the presence of HPVG on CT could be associated with a range of non-surgical pathologies (Table 1), it’s sole presence does not mandate surgical exploration and has no prooven predictive power. Surgical intervention in all HPVG cases would result in approximately 30% non-therapeutic laparotomies [2]^, [6]^, [10].

Table 1.

summarizes various reasons for HPVG.

Reason	Author
Diverticulitis	Haak et al. [7]
	Sellner et al. (2007) [19]
Ileus	Quirke (1995) [20]
Chemotherapy/Enterocolitis	Tan et al. (2015) [21]
	Kanagushi et al. (2014) [22]
Tubo-ovarian abscess	Onyeabor et al. (2015) [23]
Colonoscopy	Shinagawa et al. (2015) [24]
Viral gastroenteritis	Mirmanesh et al. (2013) [25]
Crohn’ s disease	Kirsch et al. (1990) [26]
Incarcerated diaphragmatic hernia	Mittal et al. (2015) [27]
Gastroduodenal ulcer	Choi et al. (2014) [28]
Blunt abdominal trauma	Vauthey et al. (1992) [29]
Bickerstaff’ s encephalitis	Slawinski et al. (2015) [30]
Cocaine-induced intestinal ischemia	Patel et al. (2015) [31]
Hydrogen peroxide ingestion	Papafragkou et al.(2012) [32]
Acute pancreatitis	Park et al. (2007) [33]
Endoscopic retrograde cholangiopancreatography	Hong et al. (1997) [34]
Seizure	Chen et al. (1997) [35]
Cystic fibrosis	Mallens et al. (1995) [36]

A current evidence based indication for surgery is bowel ischemia/necrosis with or without mechanical obstruction. The significant risk factors delineated in a consecutive gathering of 88 cases by Wayne et al. to distinguish the patient towards surgery and develop an algorithm are pain, elevated lactate, presence of small bowel PI and calculated vascular disease score. This algorithm is based only on 70 cases of pneumatosis intestinalis and/or HPVG still with a sensitivity of 89%. Once distinguished into a subgroup the specificity and positive predictive value is 100% [2]. The Algorithm was tested only on a limited number of 14 cases and has to be proven on a larger population.

Keeping in mind that CT-scans alone are not always satisfactory in revealing bowel ischemia [1]^, [16]^, [17]^, [18]. Additional factors like older age, peritoneal signs and high BUN were found to be significantly associated with ischemia or necrosis [1].

If ischemia or necrosis is diagnosed the presence of HPVG has a predictive value since an increasing mortality could be observed [6].

According to the algorithm developed by Wayne et al. [2] our patient had a benign HPVG (score < 4), supporting our conservative treatment approach with a benign course.

4. Conclusion

HPVG is a very rare condition from poor prognosis when associated with bowel ischemia. The last decades saw the emergence of numerous other aetiologies including benign courses and a shift of paradigm from systematic emergency laparotomies to individual patient selection. Revealing bowel necrosis/ischemia is the most important step towards patient selection for surgery. Developed algorithms can aid the decision making as in our case but unfortunately are based on small populations and present a wide variety of variables. Establishing clinical guidelines by using a register would be desireable.

Conflict of interest

We have no conflict of interest and no source of funding to declare.

Funding

No sources of funding to declare.

Ethical approval

According to the Canton Bern (Switzerland) regulation, this manuscript doesn’t require ethical approval.

Consent

Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request (please ask: alexandre.moser@ssvnn.net)

Author contribution

All authors have made significant contribution to the submitted manuscript. Alexandre Moser and Alexander Radke wrote the manuscript and did the literature review. André Wyss wade significantly corrections to the script and was responsible for the radiologist point of view. Claudio Schneider and Anita Stauffer realized the US of he patient and participated at the literature review. Manfred ESSIG shared his expert opinion about this case.

Guarantor

Alexandre MOSER and Alexander RADKE.