Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2372041295
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Oei, T.
Right arrow Articles by Silverman, S. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Oei, T.
Right arrow Articles by Silverman, S. G.
(Radiology 2005;237:709-717.)
© RSNA, 2005

Vascular and Interventional Radiology

Radiofrequency Ablation of Liver Tumors: A New Cause of Benign Portal Venous Gas1

Tamara Oei, BS, Eric vanSonnenberg, MD, Sridhar Shankar, MD2, Paul R. Morrison, MS, Kemal Tuncali, MD and Stuart G. Silverman, MD

1 From the Departments of Radiology, Dana Farber Cancer Institute (T.O., E.V., S.S., K.T., S.G.S.) and Brigham and Women's Hospital (E.V., S.S., P.R.M., K.T., S.G.S.), Harvard Medical School, Boston, Mass. Received July 25, 2004; revision requested September 30; revision received November 30; accepted January 3, 2005. Address correspondence to E.V., Department of Radiology, St Joseph's Hospital and Medical Center, 350 W Thomas Rd, Phoenix, AZ 85013 (e-mail: Eric.vanSonnenberg{at}CHW.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
PURPOSE: To retrospectively describe and categorize the presence of portal venous gas (PVG) from radiofrequency (RF) ablation of hepatic tumors.

MATERIALS AND METHODS: The study was HIPAA compliant, and informed consent was waived. Thirty-four consecutive computed tomography (CT)-guided percutaneous RF ablations of liver tumors in 26 patients (13 men, 13 women; mean age, 69 years) with five hepatocellular carcinomas and 21 metastatic liver tumors (13 colon, five other, and three unknown primary tumors) were performed with an institutional review board–approved protocol. Two treatment modalities were used: RF ablation alone (13 procedures) and combined RF ablation and ethanol injection (21 procedures). Presence of PVG was quantified with three parameters: maximum length of a portal venous branch with gas, number of Couinaud segments in which PVG was seen, and total number of portal venous branch points with gas. Then an overall PVG score from 0 to 5 was determined. Also, when tumoral gas was seen on CT scans, the largest cross-sectional area of gas was measured. The two ablation methods were compared for quantities of PVG and tumoral gas. The role of N2O anesthetic in PVG and tumoral gas formation during ablation also was studied. Statistical analyses were performed with Wilcoxon rank sum and Student t tests.

RESULTS: In 25 procedures (74%), gas was found in portal vein branches; in 30 procedures (88%), gas was also found in tumoral and peritumoral tissues. There was no significant difference in frequency of PVG between the ablation methods. Combined therapy yielded significantly greater lengths of PVG (P < .002) and more portal venous branch points (P < .001) than did RF ablation alone. Mean PVG score was 2.4 ± 0.4 (standard error of the mean) for combined therapy and 0.9 ± 0.2 for RF ablation alone (P < .004). N2O anesthetic was associated with greater amounts of tumoral gas (P < .008) and PVG (P < .03). Tumoral gas, peritumoral gas, and PVG dissipated within 20 minutes after ablation in all patients. No morbidity or mortality was associated with PVG.

CONCLUSION: RF ablation is a common yet benign cause of transient PVG, tumoral gas, and peritumoral gas. Combined RF and ethanol ablation was associated with more PVG than was RF ablation alone.

© RSNA, 2005


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Historically, the presence of portal venous gas has been most commonly associated with bowel necrosis and abdominal sepsis. The finding demanded exploratory laparotomy but typically resulted in a grave prognosis (1,2). In one review of 64 cases of portal venous gas, a mortality of 75% was reported (3). The 25% survival correlated with the lack of extensive bowel necrosis (3). However, a more recent review of 182 cases of portal venous gas indicated a much lower mortality rate (39%) owing to earlier detection by means of more sensitive diagnostic techniques, including computed tomography (CT) and ultrasonography (US), as well as by means of better management of the underlying disease process (4). A 1997 report of a separate study cites an overall mortality as low as 29% in patients with portal venous gas (5).

Portal venous gas has both potentially lethal and benign causes. Clinically important portal venous gas typically results from noniatrogenic causes, such as bowel necrosis, focal disruption of the bowel wall from carcinoma, ulcerative colitis, gastric ulcer, Crohn disease (4), infection with gas-producing bacteria (6), intestinal or gastric distention (7), or blunt trauma (8–10). Conversely, the survival rate in patients with benign portal venous gas approaches 100% (5,11). It is usually associated with iatrogenic causes, such as barium enema examination (7), endoscopy (7,12), colonoscopy (13), or hepatic transplantation (14). It is thought that the isolated finding of portal venous gas is no longer a reliable indicator of prognosis. The history and clinical context need to be considered to identify benign portal venous gas and thus avoid an unnecessary exploratory laparotomy (15,16). The purpose of our study was to retrospectively describe and categorize the presence of portal venous gas produced by radiofrequency ablation of hepatic tumors.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Patients
We retrospectively reviewed CT scans in all 26 patients who underwent 34 consecutive percutaneous hepatic radiofrequency ablation procedures with a three-pronged radiofrequency ablation probe (Cool-Tip Cluster Electrode; Radionics, Burlington, Mass) for primary or secondary liver tumors between January 2001 and December 2003. The patients included in the study were 13 men and 13 women aged 47–90 years (mean age, 69 years). Hepatic tumor types in the 26 patients included five hepatocellular carcinomas and 21 metastatic liver tumors (13 colon, five other primary tumors, and three unknown primary tumors). Two ablation methods were compared—radiofrequency ablation alone versus radiofrequency ablation and percutaneous ethanol injection combined—to determine whether the latter had any synergistic effect on the efficacy of ablation and the presence and characteristics of portal venous and tumoral gas. Radiofrequency ablation alone was used in the first 13 procedures, while the combined therapy was used in the latter 21 procedures. This Health Insurance Portability and Accountability Act–compliant study was conducted in accordance with an institutional review board–approved protocol, and informed consent for the study was waived. Medical records and images were reviewed, and the clinical features of these 34 procedures are summarized in Table 1.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Clinical Features in Patients Undergoing Radiofrequency Ablation of Liver Tumors

 
Radiofrequency Ablation Procedure
All procedures were performed after patients were administered general anesthetic. The type of volatile anesthetic (isoflurane, desflurane, or sevoflurane) and the use of N2O were selected and administered by the anesthesiologist on a case-by-case basis (Table 1). All ablations were performed by three authors (S.S., E.V., K.T., each with 5–15 years of experience). Intravenous contrast material was not used until immediately after the ablation procedure. At the beginning of each ablation, a 22-gauge needle was guided by means of CT into the liver lesion, and a biopsy specimen of the tissue was retrieved to confirm the diagnosis. Radiofrequency ablation was conducted by using a three-needle configuration of electrodes (Cool-Tip Cluster Electrode; Radionics). The radiofrequency probe was guided by means of CT into the tumor in tandem with the already placed biopsy and localizing needles. Radiofrequency energy (approximately 150 W) was deposited interstitially into the tissue for 12 minutes at each of several sites within the tumor (one to eight applications; Table 1).

For the combination therapy, 98% (volume/volume) ethanol was injected percutaneously into the tumor immediately prior to radiofrequency ablation. The volume of ethanol injected ranged from 2 to 15 mL; this volume was determined by means of several factors, including the proximity to surrounding structures that might be injured by ethanol, such as the bile ducts, liver capsule, gallbladder, and porta hepatis. We checked for the potential presence of gas in the hepatic arteries, hepatic veins, inferior vena cava, and right side of the heart on postablation CT scans.

Imaging and Image Review
All CT examinations were performed with a single channel, single–detector row CT scanner (Somatom Plus 4; Siemens, Malvern, Pa). Selected imaging parameters included 120 kVp, 110–260 mA, a detector configuration of 5 mm, a beam pitch of 1.5, a table speed of 7.5 mm per gantry rotation (0.75-second gantry rotation time), and the acquisition of 5-mm images reconstructed at 5-mm intervals. Intraprocedural monitoring of the ablation was performed without intravenous contrast material through the area of interest of the liver at 3–6-minute intervals; evaluation of these CT images for any other unexpected gas was performed as well. Intravenous contrast material was administered at a dual-phase hepatic CT examination performed 5–15 minutes after the completion of the ablation procedure.

All images were reviewed together in consensus by five authors (S.S., E.V., T.O., K.T., S.G.S.) with 1–15 years of ablation experience. All unenhanced transverse CT images were reviewed on picture archiving and communication system workstations by using both soft-tissue and lung windows. Lung windows facilitated the distinction of low-attenuation radiofrequency needle-tip artifacts from gas. The presence, quantity, and distribution of portal venous gas were ascertained from each of the 34 liver ablations by reviewing CT scans obtained prior to, during, and immediately after the ablation procedures. For procedures that involved multiple radiofrequency applications during the same treatment session, only the first 12-minute ablation was analyzed for the production of gas to avoid bias in quantification of gas from subsequent ablations (Table 1). The number of radiofrequency applications is recorded in Table 1.

The tumor location was identified by using the Couinaud system. The amount of gas at the tumor ablation site was approximated as the largest cross-sectional area of gas. This area was estimated as an ellipse: area = {pi}r1r2 = {pi}(d1 · d2)/4, where r1r2 is the radius in two 90° planes, d1 is the maximum longitudinal length, and d2 is the minor axis.

Portal venous gas was distinguished from biliary gas by means of its peripheral distribution, as well as by the more obvious branching structure of the portal veins. The Couinaud segments containing portal venous gas were documented. Three parameters were used to quantify the portal venous gas: (a) the maximum length of a portal venous branch, (b) the number of Couinaud segments with portal venous gas, and (c) the total number of portal venous branch points that contained gas. These measurements were obtained, and the mean and standard deviation were calculated. These measured values were then converted to a grade on a scale of 0–5; a grade of 0 indicated the absence of portal venous gas and a grade of 5 indicated a portal venous gas measurement more than two standard deviations from the mean. Intermediate integer grades (grades 1–4) were determined by means of the following equation: grade = 5x/(2 · SD), where x is the measurement, and SD is the standard deviation. The graded values of the three parameters were then averaged. These mean values were subsequently graded on a scale of 0–5 to yield an overall "portal venous gas score."

Statistical Analysis
The mean and standard deviation of data were determined for all measurements by using an Excel 2000 (version 9.0.2720; Microsoft, Redmond, Wash) spreadsheet. Furthermore, the relationship between tumor gas and portal venous gas, as well as the relationship between tumor gas and volume of ethanol used, was plotted with Excel, and the R2 values were determined.

Statistical analyses were performed by using JMP 5.0.1.2 software (SAS Institute, Cary, NC). Statistical comparisons of portal venous gas and tumor gas quantities between the two methods of ablation (radiofrequency ablation alone and radiofrequency ablation and ethanol injection combined) were made by using the Student t test for unpaired data. The significance of the presence of N2O on portal venous gas quantities during the two methods of ablation was determined by using the Student t test. Specifically, unpaired data used for the Student t test consisted of the cross-sectional area of gas at the ablation site, the maximum length of a portal venous gas branch, the number of portal venous branch points, and the number of Couinaud segments containing portal venous gas. The nonparametric Wilcoxon rank sum test was used to determine whether there was a significant difference between the portal venous gas scores with the two ablation methods, as well as the presence or absence of N2O. Further comparisons by means of the Student t test for unpaired data were conducted to determine if the presence of N2O had a significant effect on gas production. For all analyses, a statistically significant difference was defined at P < .05.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Portal Venous and Tumoral Gas
There were no instances in which differentiation of portal venous gas from biliary gas was difficult. In 25 (74%) of 34 ablations conducted in the 26 patients, gas was detected in the branches of the portal vein (Tables 2, 3). The mean portal venous gas score was 1.6 ± 0.2 (±standard error of the mean) out of maximum possible score of 5 (Fig 1). The mean maximum length of a portal venous branch with gas was 1.2 cm ± 0.2. The mean number of portal venous branch points containing gas was 1.7 ± 0.5, and the mean number of Couinaud segments containing portal venous gas was 1.2 ± 0.2.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Quantity of Gas with Radiofrequency Ablation–Only Procedure

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Quantity of Gas with Combined Radiofrequency and Percutaneous Ethanol Injection Ablation Procedure

 


View larger version (93K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a. Transverse CT scans show examples of scoring of radiofrequency ablation–induced portal venous gas. (a) Scans in 66-year-old man with metastatic lung cancer. Score of 3 was based on portal venous gas in one Couinaud segment, five branch points, and maximum branch length of 2.7 cm (long arrows). Several portal venous gas branches are shown (short arrows). (b) Scan in 61-year-old woman with metastatic pancreatic cancer. Score of 4 was based on portal venous gas in three Couinaud segments, four branch points, and maximum branch length of 1.2 cm. Scan shows ablation of a tumor with extensive intratumoral gas (*) in posteromedial aspect of right lobe and portal venous gas in peripheral aspects of left lobe (small arrows), as well as centrally (large arrow). The superior low-attenuation area in segment V (arrowheads) represents another metastasis that was ablated subsequent to current tumor ablation. (c) Scan in 81-year-old man with metastatic colon cancer. Score of 5 was based on portal venous gas in three Couinaud segments, four branch points, and maximum branch length of 2.2 cm. Scan shows multiple portal venous gas branches (arrows), as well as a large amount of gas in the tumor.

 


View larger version (156K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b. Transverse CT scans show examples of scoring of radiofrequency ablation–induced portal venous gas. (a) Scans in 66-year-old man with metastatic lung cancer. Score of 3 was based on portal venous gas in one Couinaud segment, five branch points, and maximum branch length of 2.7 cm (long arrows). Several portal venous gas branches are shown (short arrows). (b) Scan in 61-year-old woman with metastatic pancreatic cancer. Score of 4 was based on portal venous gas in three Couinaud segments, four branch points, and maximum branch length of 1.2 cm. Scan shows ablation of a tumor with extensive intratumoral gas (*) in posteromedial aspect of right lobe and portal venous gas in peripheral aspects of left lobe (small arrows), as well as centrally (large arrow). The superior low-attenuation area in segment V (arrowheads) represents another metastasis that was ablated subsequent to current tumor ablation. (c) Scan in 81-year-old man with metastatic colon cancer. Score of 5 was based on portal venous gas in three Couinaud segments, four branch points, and maximum branch length of 2.2 cm. Scan shows multiple portal venous gas branches (arrows), as well as a large amount of gas in the tumor.

 


View larger version (152K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c. Transverse CT scans show examples of scoring of radiofrequency ablation–induced portal venous gas. (a) Scans in 66-year-old man with metastatic lung cancer. Score of 3 was based on portal venous gas in one Couinaud segment, five branch points, and maximum branch length of 2.7 cm (long arrows). Several portal venous gas branches are shown (short arrows). (b) Scan in 61-year-old woman with metastatic pancreatic cancer. Score of 4 was based on portal venous gas in three Couinaud segments, four branch points, and maximum branch length of 1.2 cm. Scan shows ablation of a tumor with extensive intratumoral gas (*) in posteromedial aspect of right lobe and portal venous gas in peripheral aspects of left lobe (small arrows), as well as centrally (large arrow). The superior low-attenuation area in segment V (arrowheads) represents another metastasis that was ablated subsequent to current tumor ablation. (c) Scan in 81-year-old man with metastatic colon cancer. Score of 5 was based on portal venous gas in three Couinaud segments, four branch points, and maximum branch length of 2.2 cm. Scan shows multiple portal venous gas branches (arrows), as well as a large amount of gas in the tumor.

 
For 30 (88%) of 34 ablations, gas was also visualized in the tumor and tissue surrounding the radiofrequency probe tips during ablation (Fig 2; Tables 2, 3). The mean cross-sectional area of gas in the tumor in all 34 ablations was 2.5 cm2 ± 0.4. The quantity of portal venous gas showed a general positive relationship, but there was no obvious linearity (R2 = 0.18, P < .01) (Fig 3). In five of 34 ablations, gas was present at the tumor site, but no portal venous gas was detected on CT scans. Conversely, all ablations in which portal venous gas was present had gas at the tumor site.



View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Transverse CT scans of intraprocedural radiofrequency ablation–induced gas in hepatic tumors. (a) Scan in 67-year-old man with metastatic colon cancer shows gas surrounding tip of cluster probe within the tumor. (b) Scan in 66-year-old man with metastatic lung cancer shows a different pattern of gas in both tumor and tumor wall during ablation (thick arrows). A portal venous branch containing gas is seen immediately adjacent to the tumor (thin arrows).

 


View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Transverse CT scans of intraprocedural radiofrequency ablation–induced gas in hepatic tumors. (a) Scan in 67-year-old man with metastatic colon cancer shows gas surrounding tip of cluster probe within the tumor. (b) Scan in 66-year-old man with metastatic lung cancer shows a different pattern of gas in both tumor and tumor wall during ablation (thick arrows). A portal venous branch containing gas is seen immediately adjacent to the tumor (thin arrows).

 


View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Graph shows portal venous gas (PVG*) as a function of gas at tumor site. There is a general positive relationship, but no obvious linearity (R2 = 0.18, P < .01). Quantity of portal venous gas is expressed as an average of three parameters—maximum portal venous gas length, number of Couinaud segments, and number of branch points. Each parameter is recentered on a scale of 0–5.

 
Radiofrequency Ablation versus Combined Therapy
There was no significant difference in the incidence of portal venous gas between the use of radiofrequency ablation alone (nine of 13 cases, 69%) and the use of combined ablation (16 of 21 cases, 76%). The distribution graph of portal venous gas scores shows that single ablation therapy with radiofrequency produced a maximum portal venous gas score of 2 out of 5 (Fig 4), whereas the combined therapy produced portal venous gas scores ranging from 0 out of 5 to 5 out of 5 (Fig 5).



View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. Transverse CT scans show comparison of highest portal venous gas scores with the two ablation methods. (a) Scans in 67-year-old man with colon cancer metastases who had the highest portal venous gas score (score of 2) in the radiofrequency-only group on the basis of portal venous gas in two Couinaud segments, one branch point (arrow), and maximum branch length of 0.82 cm that traversed several CT sections. (b) Scans in 57-year-old woman with metastatic breast cancer who had the highest portal venous gas score (score of 5) in the combined ablation group on the basis of portal venous gas in four Couinaud segments, seven branch points, and maximum branch length of 3.6 cm. Scans show portal venous gas diffusely throughout the liver. The combined method produced, on average, more portal venous gas.

 


View larger version (67K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. Transverse CT scans show comparison of highest portal venous gas scores with the two ablation methods. (a) Scans in 67-year-old man with colon cancer metastases who had the highest portal venous gas score (score of 2) in the radiofrequency-only group on the basis of portal venous gas in two Couinaud segments, one branch point (arrow), and maximum branch length of 0.82 cm that traversed several CT sections. (b) Scans in 57-year-old woman with metastatic breast cancer who had the highest portal venous gas score (score of 5) in the combined ablation group on the basis of portal venous gas in four Couinaud segments, seven branch points, and maximum branch length of 3.6 cm. Scans show portal venous gas diffusely throughout the liver. The combined method produced, on average, more portal venous gas.

 


View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Graph of portal venous gas distribution. The combination of radiofrequency ablation (RFA) and ethanol (alcohol) ablation yielded higher portal venous gas scores of up to 5, whereas ablations with radiofrequency alone had a maximum portal venous gas score of 2.

 
The combined therapy produced a greater average cross-sectional area of gas at the tumor site (3.9 cm2± 0.7) compared with that produced by ablation with radiofrequency alone (2.2 cm2± 0.7, P < .12) (Fig 6). Among the 34 ablations, the combined therapy created significantly greater portal venous gas lengths (1.6 cm ± 0.3) than did radiofrequency ablation alone (0.6 cm ± 0.2, P < .002) (Fig 7). In addition, the number of portal venous branch points containing gas was significantly higher from combined ablations (2.7 ± 0.6) than from ablations performed with radiofrequency alone (0.2 ± 0.1, P < .001) (Fig 7). Portal venous gas appeared in more Couinaud segments when the combined ablation procedure was used (1.4 ± 0.2) than when radiofrequency ablation alone was used (0.8 ± 0.2, P < .06) (Fig 7). Portal venous gas was located in a Couinaud segment of the liver other than that of the tumor in 19 of the 25 ablation procedures with portal venous gas. Furthermore, the mean portal venous gas score in patients who underwent the combined ablation procedure was higher (2.4 ± 0.4) than that in patients who were treated with radiofrequency ablation alone (0.9 ± 0.2, P < .004) (Fig 8).



View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Graph compares average cross-sectional area of gas produced at ablation sites with the ablation methods—radiofrequency ablation (RFA) (n = 13) and combined ablation (combo) (n = 21) (P < .12). Results are mean ± standard error of the mean (error bars) and suggest no significant difference in amounts of tumoral gas produced by the two methods of ablation.

 


View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7a. Graphs compare portal venous gas produced by two ablation methods. Average values of parameters used to quantify portal venous gas at radiofrequency ablation (RFA) (n = 13) versus combined therapy (combo) (n = 21). Results are mean ± standard error of the mean (error bars). Graphs compare (a) average length of portal vein (P < .002), (b) average number of gas-filled portal venous branch points (P < .001), and (c) average number of Couinaud segments (P < .06). Findings suggest a significantly greater amount of portal venous gas with combination ablation for two of the three criteria.

 


View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7b. Graphs compare portal venous gas produced by two ablation methods. Average values of parameters used to quantify portal venous gas at radiofrequency ablation (RFA) (n = 13) versus combined therapy (combo) (n = 21). Results are mean ± standard error of the mean (error bars). Graphs compare (a) average length of portal vein (P < .002), (b) average number of gas-filled portal venous branch points (P < .001), and (c) average number of Couinaud segments (P < .06). Findings suggest a significantly greater amount of portal venous gas with combination ablation for two of the three criteria.

 


View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7c. Graphs compare portal venous gas produced by two ablation methods. Average values of parameters used to quantify portal venous gas at radiofrequency ablation (RFA) (n = 13) versus combined therapy (combo) (n = 21). Results are mean ± standard error of the mean (error bars). Graphs compare (a) average length of portal vein (P < .002), (b) average number of gas-filled portal venous branch points (P < .001), and (c) average number of Couinaud segments (P < .06). Findings suggest a significantly greater amount of portal venous gas with combination ablation for two of the three criteria.

 


View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8. Graph of mean portal venous gas (PVG) score with radiofrequency ablation alone (RFA) (n = 13) versus combined therapy (combo) (n = 21). Portal venous gas scores represent the averages of three parameters: maximum length of portal venous gas, number of Couinaud segments containing portal venous gas, and number of branch points containing portal venous gas (P < .004). Results are mean ± standard error of the mean (error bars) and suggest a significantly higher portal venous score from the combined ablation method.

 
Although radiofrequency ablation with percutaneous ethanol injection produced more gas in the tumor and in portal veins, there was no correlation between the amount of gas produced at the tumor site and the volume of ethanol injected (R2 < 0.04).

Complications and Time Course of Gas
There was no gas detected in the hepatic arteries or hepatic veins nor was gas detected in the inferior vena cava or right side of the heart. No patient experienced any untoward clinical or CT effect of the portal venous gas from radiofrequency ablation of the liver tumors, which was confirmed at routine postablation abdominal contrast material–enhanced CT.

Gas in the tumors accumulated during the course of the 12-minute ablation session was at a maximum at the end of the ablation procedure. The gas typically dissipated within 20 minutes. Almost all of the gas at the tumor ablation site had dissipated on the postprocedural contrast-enhanced CT scan, which was obtained 13 minutes after ablation (Fig 9). Patients who received multiple applications during the same procedure had little or no gas at the tumor site before the next application was initiated.



View larger version (164K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9a. Transverse CT scans show typical rapid dissipation of gas from tumor after radiofrequency ablation in an 82-year-old woman with hepatocellular carcinoma. (a) Intraprocedural scan obtained immediately before termination of the procedure shows extensive gas within the tumor. (b) Contrast-enhanced scan obtained 13 minutes after completion of ablation shows gas has almost completely resolved.

 


View larger version (169K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 9b. Transverse CT scans show typical rapid dissipation of gas from tumor after radiofrequency ablation in an 82-year-old woman with hepatocellular carcinoma. (a) Intraprocedural scan obtained immediately before termination of the procedure shows extensive gas within the tumor. (b) Contrast-enhanced scan obtained 13 minutes after completion of ablation shows gas has almost completely resolved.

 
Role of Anesthetic Gas
There was no correlation between the type of volatile anesthetic administered (isoflurane, sevoflurane, or desflurane) and the amount of tumoral gas or portal venous gas. N2O was used in 63% (12 of 19) of the combined radiofrequency ablation and percutaneous ethanol injection procedures but in only 8% (one of 13) of the radiofrequency ablation–only procedures. Within the group that underwent combined ablation therapy, patients who were administered N2O were found to have significantly larger cross-sectional areas of gas at the tumor site (P < .008) and higher portal venous gas scores (P < .03) than were those who underwent ablations without N2O (Fig 10). When patients receiving N2O were excluded from analysis, there was no significant difference in amount of gas at tumor site and portal venous gas between the two ablation methods.



View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 10. Graph shows effect of N2O on combined ablation. When N2O was administered, there was more gas produced at the ablation site (P < .03) and in the portal veins (P < .008). Results are mean ± standard error of the mean. PVG = portal venous gas.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Although it was common in our series, to our knowledge, portal venous gas secondary to radiofrequency ablation of hepatic tumors (25 of 34 ablations in our study, 74%) has not been reported previously. The portal venous gas produced during radiofrequency ablation of liver tumors has no clinical effects, and it disappears within 20 minutes after the end of the ablation. In a large multicenter study (17) that was specifically focused on complications from radiofrequency ablation of hepatic tumors, the presence of gas in the portal veins, hepatic veins, or inferior vena cava was not reported. However, one report (18) described severe postprocedural pain in a patient who developed a perihepatic rim of gas and fluid, both of which resolved spontaneously. On the basis of our data, radiofrequency ablation of liver tumors should be included with other iatrogenic causes of benign portal venous gas. The ephemeral nature of the tumoral and portal venous gas suggests that any gas seen 1 day or longer after ablation likely has a more ominous cause.

Tumoral gas has been described previously in articles on CT and US-guided radiofrequency ablations of liver tumors (18–27). Authors of prior studies have reported the use of this zone of "microbubbles" surrounding the radiofrequency probe to evaluate the extent of the ablation (23,24). However, tumoral gas has been found not to be an accurate indicator of tissue necrosis (26,27) and, in our series, was not consistently present in all ablations.

Our data show increased amounts of both tumoral and portal venous gas in patients treated with a combination of ethanol injection and radiofrequency ablation compared with patients treated with radiofrequency ablation alone. One explanation would be that the presence of ethanol improves the conduction of heat from the probe tips and thus enhances the combustion of tissues or increases vaporization of surrounding compounds. This phenomenon would be consistent with results from a study that demonstrated more effective ablations when ethanol was used with radiofrequency compared with when radiofrequency ablation was used alone (28).

The data show that within the combined treatment group, those procedures in which N2O anesthetic was used produced significantly more portal venous gas than did procedures in which N2O was not used. Hence, one explanation for the greater amount of gas during procedures with both radiofrequency ablation and percutaneous ethanol injection might be the use of N2O. N2O has a unique property in that it tends to fill and expand closed air spaces because it is 34 times more soluble in blood than is nitrogen (29). After a pocket of gas is produced during radiofrequency ablation, N2O quickly diffuses into the space and rapidly expands it. However, an argument against N2O being the predominant cause of the portal venous gas is the transient nature of the observed gas. In all ablations, the portal venous gas dissipated quickly after the radiofrequency probe was switched off even though N2O was still being administered as an anesthetic at a steady rate. Perhaps N2O has a synergistic effect with ethanol in producing larger amounts of tumoral and portal venous gas during radiofrequency liver ablations.

Limitations of this study included the small sample population size, the retrospective nature of the study, observer subjectivity, and measurement ambiguities. Portal venous gas frequently took tortuous paths in different planes, which meant that measurements were not always straightforward. Also, there may have been instances in which the three parameters (number of Couinaud segments, number of branch points, and maximum length) of quantifying portal venous gas overlapped and thus may not have accurately reflected the true volumes of portal venous gas. Quantification of gas at the tumor site was complicated by the often irregular areas that may not have been well approximated as ellipses. Furthermore, the gas at the tumor site may have been better approximated as a volume instead of a cross-sectional area. The boundaries of the gas formed at the probe tips were not always distinct, and distinguishing gas from artifact at the cluster tip was not always straightforward. However, lung windows were useful to distinguish gas from probe tip artifacts, both of which appear to be of very low attenuation.

Although the exact identity and origin of the portal venous gas is unknown, we speculate that it was produced either by combustion of tissues at high temperatures, which yielded CO2 and water vapor, or by vaporization of CO2 and other substances in blood. The fact that there was more gas produced in the patients who received combined therapy could be attributed to better heat conductance with ethanol or vaporization of ethanol itself. N2O also may play a role in the greater observed amounts of gas in the combined therapy group. The transient nature of the gas suggests that it is readily soluble in blood at physiologic temperatures and pressures. Further analysis of the gas with studies such as gas chromatography will help determine the exact nature of the gas, the mechanism by which it is produced, and its composition. In addition, a larger study population may help to better elucidate the role of N2O gas anesthetic in the formation of portal venous gas at radiofrequency ablation of liver tumors. Nevertheless, the presence of this portal venous gas phenomenon and its benign nature were documented in this study.


   FOOTNOTES
 
2 Current address: UMass Medical Center, Worcester, Mass Back

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, T.O., E.V., S.S.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, T.O., E.V., S.G.S.; clinical studies, T.O., E.V., S.S., P.R.M., K.T.; statistical analysis, T.O.; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Hong JJ, Gadaleta D, Rossi P, Esquivel J, Davis JM. Portal vein gas, a changing clinical entity: report of 7 patients and review of the literature. Arch Surg 1997;132:1071–1075.[Abstract/Free Full Text]
  2. See C, Elliott D. Images in clinical medicine: pneumatosis intestinalis and portal venous gas. N Engl J Med 2004;350:e3.
  3. Liebman PR, Patten MT, Manny J, Benfield JR, Hechtman HB. Hepatic–portal venous gas in adults: etiology, pathophysiology and clinical significance. Ann Surg 1978;187:281–287.[Medline]
  4. Kinoshita H, Shinozaki M, Tanimura H, et al. Clinical features and management of hepatic portal venous gas: four case reports and cumulative review of the literature. Arch Surg 2001;136:1410–1414.[Abstract/Free Full Text]
  5. Faberman RS, Mayo-Smith WW. Outcome of 17 patients with portal venous gas detected by CT. AJR Am J Roentgenol 1997;169:1535–1538.[Abstract/Free Full Text]
  6. Griffiths DM, Gough MH. Gas in the hepatic portal veins. Br J Surg 1986;73:172–176.[Medline]
  7. Brown MA, Hauschildt JP, Casola G, Gosink BB, Hoyt DB. Intravascular gas as an incidental finding at US after blunt abdominal trauma. Radiology 1999;210:405–408.[Abstract/Free Full Text]
  8. Kelly BS Jr, Meyers P, Choe KA, Hurst J, Luchette FA. Traumatic pneumatosis cystoides intestinalis with portal venous air embolism. J Trauma 1997;42:112–114.[Medline]
  9. Kingsley DD, Albrecht RM, Vogt DM. Gastric pneumatosis and hepatoportal venous gas in blunt trauma: clinical significance in a case report. J Trauma 2000;49:951–953.[Medline]
  10. Hamilton P, Rizoli S, McLellan B, Murphy J. Significance of intra-abdominal extraluminal air detected by CT scan in blunt abdominal trauma. J Trauma 1995;39:331–333.[Medline]
  11. Schulze CG, Blum U, Haag K. Hepatic portal venous gas: imaging modalities and clinical significance. Acta Radiol 1995;36:377–380.[Medline]
  12. Blind PJ, Oberg L, Hedberg B. Hepatic portal vein gas following endoscopic retrograde cholangiography with sphincterotomy: case report. Eur J Surg 1991;157:299–300.[Medline]
  13. Haber I. Hepatic portal vein gas following colonoscopy in ulcerative colitis: report of a case. Acta Gastroenterol Belg 1983;46:14–17.[Medline]
  14. Chezmar JL, Nelson RC, Bernardino ME. Portal venous gas after hepatic transplantation: sonographic detection and clinical significance. AJR Am J Roentgenol 1989;153:1203–1205.[Abstract/Free Full Text]
  15. Peloponissios N, Halkic N, Pugnale M, et al. Hepatic portal gas in adults: review of the literature and presentation of a consecutive series of 11 cases. Arch Surg 2003;138:1367–1370.[Abstract/Free Full Text]
  16. Little AF, Ellis SJ. ‘Benign’ hepatic portal venous gas. Australas Radiol 2003;47:309–312.[CrossRef][Medline]
  17. Livraghi T, Solbiati L, Meloni MF, Gazelle GS, Halpern EF, Goldberg SN. Treatment of focal liver tumors with percutaneous radio-frequency ablation: complications encountered in a multicenter study. Radiology 2003;226:441–451.[Abstract/Free Full Text]
  18. Gazelle GS, Goldberg SN, Solbiati L, Livraghi T. Tumor ablation with radio-frequency energy. Radiology 2000;217:633–646.[Abstract/Free Full Text]
  19. Grasso A, Watkinson AF, Tibballs JM, Burroughs AK. Radiofrequency ablation in the treatment of hepatocellular carcinoma: a clinical viewpoint. J Hepatol 2000;33:667–672.[CrossRef][Medline]
  20. Cha CH, Lee FT Jr, Gurney JM, et al. CT versus sonography for monitoring radiofrequency ablation in a porcine liver. AJR Am J Roentgenol 2000;175:705–711.[Abstract/Free Full Text]
  21. Nahum Goldberg S, Dupuy DE. Image-guided radiofrequency tumor ablation: challenges and opportunities. I. J Vasc Interv Radiol 2001;12:1021–1032.[Medline]
  22. Leyendecker JR, Dodd GD 3rd, Halff GA, et al. Sonographically observed echogenic response during intraoperative radiofrequency ablation of cirrhotic livers: pathologic correlation. AJR Am J Roentgenol 2002;178:1147–1151.[Abstract/Free Full Text]
  23. de Baere T, Elias D, Dromain C, et al. Radiofrequency ablation of 100 hepatic metastases with a mean follow-up of more than 1 year. AJR Am J Roentgenol 2000;175:1619–1625.[Abstract/Free Full Text]
  24. Siperstein A, Garland A, Engle K, et al. Laparoscopic radiofrequency ablation of primary and metastatic liver tumors: technical considerations. Surg Endosc 2000;14:400–405.[CrossRef][Medline]
  25. McGahan JP, Dodd GD 3rd. Radiofrequency ablation of the liver: current status. AJR Am J Roentgenol 2001;176:3–16.[Free Full Text]
  26. Solbiati L, Goldberg SN, Ierace T, et al. Hepatic metastases: percutaneous radio-frequency ablation with cooled-tip electrodes. Radiology 1997;205:367–373.[Abstract/Free Full Text]
  27. Goldberg SN, Charboneau JW, Dodd GD 3rd, et al. Image-guided tumor ablation: proposal for standardization of terms and reporting criteria. Radiology 2003;228:335–345.[Abstract/Free Full Text]
  28. Shankar S, vanSonnenberg E, Morrison PR, Tuncali K, Silverman SG. Combined radiofrequency and alcohol injection for percutaneous hepatic tumor ablation. AJR Am J Roentgenol 2004;183:1425–1429.[Abstract/Free Full Text]
  29. Stoelting R, Miller R, eds. Basics of anesthesia. New York, NY: Churchill Livingstone, 1994;17–31.



This article has been cited by other articles:


Home page
 Arch SurgHome page
A. L. Nelson, T. M. Millington, D. Sahani, R. T. Chung, C. Bauer, M. Hertl, A. L. Warshaw, and C. Conrad
Hepatic Portal Venous Gas: The ABCs of Management
Arch Surg, June 1, 2009; 144(6): 575 - 581.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Oei, T.
Right arrow Articles by Silverman, S. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Oei, T.
Right arrow Articles by Silverman, S. G.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE