| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Invited Commentary |
Department of Radiology, UMDNJ–New Jersey Medical School, Newark, New Jersey
The relationship between a particular radiographic finding and its clinical implications is almost always a continually changing "work in progress," altered by a growing number of reported cases, the evolution of the mix of cases displaying the same abnormality, and development of newer modalities that allow improved detection of disease. The history of the correlation between portal vein gas and patient outcome demonstrates this dynamic interaction between an imaging observation and its import for diagnosis and prognosis.
Air in the portal venous system was first reported 45 years ago by Wolf and Evans (1) and was initially regarded as a dire event presaging imminent death. Actually, air or gas in the portal vein and its branches is innocuous in and of itself. However, its intravascular presence seemed to occur predominantly in patients with sepsis or ischemia, often in association with mesenteric or intestinal intramural gas. In the earliest reports, all of these disease entities resulted from damage to the bowel mucosa caused by inflammation or insufficient blood flow (2,3). As the preceding article by Sebastià et al reveals, portal vein gas is no longer regarded as always being a harbinger of impending doom, and the list of possible causes has expanded to include a wide range of benign entities. Moreover, even in patients with hypotension, the prospect for recovery is very real.
In the 1950s and 1960s, portal vein gas was observed at conventional radiography presumably only when it was of massive volume and therefore easily discernible. Since the advent of CT, focused cross-sectional views of the hepatic substance have demonstrated smaller intravascular deposits, sometimes confined to only one or two portal vein branches.
One could argue that the survival rate is now better because the roster of inciting causes includes some conditions that are readily treatable or entirely asymptomatic. Moreover, minimal amounts of venous gas are generally considered detectable because of the greater sensitivity of CT compared with conventional imaging methods. This technology-based explanation is an almost irresistible conclusion given the versatility and diagnostic incisiveness of CT. Undoubtedly, CT allows improved recognition of gas in mesenteric vessels because it enables the differentiation of intramural from intraluminal areas of hyperlucency, a capability that conventional radiography lacks.
However, that same argument should not be uncritically applied to gas within the liver itself because the homogeneous gray shadow of the hepatic substance affords an excellent background for the delineation of small amounts of air or gas within the liver or in the peritoneal or extraperitoneal spaces. A recent report has shown that supine, unenhanced conventional imaging can help detect less than 1 mL of free air in the right upper quadrant (4). We have used both patient radiographs and images of phantoms to demonstrate that CT and conventional radiography have a similar sensitivitity in the detection of air overlying the liver (5). Sebastià et al assert the relative insensitivity of conventional radiography but provide no documentation of the claim that CT is superior in the depiction of gas in the portal venous system.
Nevertheless, there are possible explanations for why conventional unenhanced imaging has not demonstrated portal vein gas as readily as CT:
1. If the gas is situated in the left lobe of the liver, it may be positioned over the spine, and the combined shadow of hepatic tissue and bone may disguise superimposed small areas of hyperlucency. In contrast, overlying osseous structures do not usually impede CT observation.
2. A kidney-ureter-bladder (KUB) image is often the only image from the supine abdominal study that is used for evaluation, so that in most cases only a part of the hepatic shadow will be included. In most adults, the abdominal expanse is too long to be encompassed on a single supine image obtained at the standard source-to-film distance of 40 inches. This widespread use of a KUB image as a surrogate for a thorough abdominal examination tends to exclude the uppermost part of the liver, often "cutting off" its rostral half from view.
3. Because portal vein gas may accumulate preferentially in peripheral venous radicles, it is often poor positioning rather than technical limitations that support the notion of the insensitivity of conventional radiography.
4. Although portal vein gas is often more clearly seen at CT, it is still frequently demonstrated (albeit perhaps more subtly) at conventional radiography, although it must be looked for carefully. However, the decline in interest in conventional image interpretation has likely played a role in the denigration of conventional radiography with respect to the discernment of rostrally positioned intrahepatic portal vein gas.
Of course, a separate abdominal examination may be unnecessary if CT is to be performed. However, a digital scout image is almost always included in a CT study, and in most cases, this image encompasses the entire abdomen including the upper and lateral margins of the hepatic shadow. If we were to compare the digital scout image with CT sections through the liver, it would be possible once and for all to measure the actual differences in sensitivity (if any) between cross-sectional and conventional imaging. Such a comparison can, of course, be made routinely with CT because this modality involves nearly simultaneously generated images, thereby eliminating possible variations in venous gas volume, which may occur when two studies are performed at separate times.
A final point that bears on such a comparison relates directly to the evanescence of the gas in the portal system. It has been noted that imminent death is more likely in cases involving tiny intrahepatic venous areas of hyperlucency than in those involving massive quantities of gas in the portal venous tree. Gaseous accumulation depends on a patient's ability to swallow air, which traverses the intestinal wall to enter the portal venous system. A moribund individual may cease the act of swallowing. Because portal vein gas is rapidly absorbed, the amount of gas in the vascular radicles will promptly decrease. In contrast, a patient who is alert and oriented but in severe distress will tend to ingest copious amounts of air, which can then make its way to veins within the liver.
Portal vein gas remains a fascinating entity. The final chapter on the interplay between its formation, absorption, and clinical correlation has yet to be written. The article by Sebastià et al, although a valuable addition to the literature on the subject, underscores the need for further investigation.
References
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOGRAPHICS | RADIOLOGY | RSNA JOURNALS ONLINE |
