| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Memorial Sloan-Kettering Cancer Center, 1275 York
Ave., New York, NY 10021.
2 Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY
10021.
Received October 12, 2004;
accepted after revision December 10, 2004.
Address correspondence to A. M. Covey
(coveya{at}mskcc.org).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Fifty-eight consecutive patients with small predicted future liver remnants (FLRs) underwent PVE with PVA particles to induce hypertrophy of the contralateral hemi-liver before surgical resection of liver metastases. Total liver, right hemi-liver, and left hemi-liver volumes were calculated before and after embolization using a 3D workstation.
RESULTS. Eight patients underwent left PVE; 47, right PVE; and three, right and segment IV PVE. There were no major complications of the procedure. The mean increases in the ratio of the FLR to the total estimated liver volume after right, right and segment IV, and left PVE were 9%, 10%, and 3%, respectively; the corresponding mean hypertrophy ratios were 24.3%, 31.9%, and 1.5%, respectively.
CONCLUSION. Right PVE using PVA particles alone as the embolic agent is safe and effective in achieving left hemi-liver hypertrophy. In contrast, left PVE did not induce significant right hemi-liver hypertrophy in this patient population.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The utility of preoperative PVE with gelatin particles (Gelfoam, Upjohn) to improve the safety of hepatic resection for patients with hilar cholangiocarcinoma was proposed by Makuuchi et al. [7] in 1989, when they reported results of this procedure in 14 patients. Since that time, several authors have described different methods of performing PVE for cholangiocarcinoma and hepatocellular carcinoma using various agents, including N-butyl cyanoacrylate with iodized oil [8]; fibrin glue with Lipiodol (iodized oil, Andre Guerbet) [9]; absolute ethanol [10]; and a mixture of Gelfoam, thrombin, contrast material, Lipiodol, and gentamicin [11]. At our institution, we perform a large number of arterial embolizations for control of hepatic tumors using polyvinyl alcohol (PVA) particles. When we began performing preoperative PVE, we found the use of absolute ethanol to be cumbersome and time-consuming, which led us to explore the use of PVA particles for PVE [12].
The purpose of our study was to evaluate the safety and efficacy of preoperative PVE with PVA in patients with hepatic metastases.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PVE was performed on 57 patients with metastatic colon cancer and one patient with metastatic squamous cell carcinoma of the anus. None of these patients was known to have underlying liver disease or cirrhosis at the time of PVE based on CT findings and laboratory values. The absence of cirrhosis in all patients was confirmed on pathologic review of resected specimens.
|
|
|
|
|
|
|
|
The diameter of the largest tumor was documented on preembolization CT and was reassessed on follow-up CT after embolization. The hypertrophy ratio was calculated by the following formula: (contralateral lobar volume postembolization - contralateral volume preembolization/contralateral lobar volume preembolization) x 100%. The atrophy ratio was calculated with a similar equation substituting the ipsilateral lobar volume for the contralateral lobar volume, with the term "lobar" referring to the ipsilateral or contralateral hemi-liver.
Patient charts were retrospectively reviewed to assess for complications related to embolization, and hospital stay after the procedure was recorded for all patients.
Preembolization (within 1 month), postembolization (on discharge), and postoperative (postoperative day 10) biochemical analyses included serum aspartate aminotransferase, serum alanine aminotransferase, serum alkaline phosphatase, serum albumin, serum bilirubin, prothrombin time, blood urea nitrogen, and creatinine.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All patients tolerated the procedure well. Twenty-one patients (36%) had a single temperature spike (> 38.5°C) within 24 hr of embolization, and four patients (7%) had fevers lasting 2-4 days, but no positive blood cultures were documented. No patient experienced abdominal pain that required prescription medication after discharge from the hospital. Patients were discharged from the hospital 1-5 days after embolization, with a mean hospital stay of 1.6 days.
Eight patients were excluded from volumetric analysis because either the pre- or postembolization CT scans were not available in a suitable digital format that would allow accurate calculation of liver volumes. Changes in hepatic volumetric measurements for the remaining 50 patients were calculated. The mean preembolization TLV was 2,199 cm3 (range, 1,252-4,386 cm3). The FLR before and after embolization and mean hypertrophy and atrophy ratios for right PVE, right and segment IV PVE, and left PVE are shown in Table 1. Representative CT images before and after right and left PVE are shown in Figures 1E, 1F, 1G, 1H, 2A, 2B, 2C, and 2D.
|
|
|
|
|
All patients had normal laboratory values both before and after embolization. No patient had significant changes in liver synthetic function, transaminases, total bilirubin, or renal function after embolization. No patient experienced puncture site complications, such as pneumothorax or hematoma.
Thirty-eight patients (65.5%) ultimately underwent surgical resection a mean of 44 days after PVE. This included left trisegmentectomy in two, left hepatic lobectomy in three (two of whom also underwent wedge resection of tumors in segments I and V), right hepatic lobectomy with additional wedge resections or ablation in 14, and right trisegmentectomy in 19 (six of whom also underwent wedge resections or ablation of tumor). There were no instances of sustained postoperative liver failure.
The diameter of the largest measurable lesion by contrast-enhanced CT before PVE was 0.5-12 cm (mean, 3.8 cm; median, 3.1 cm). After embolization, the corresponding tumors measured 1.1-14.4 cm (mean, 4.2 cm; median, 3.5 cm). The planned surgery was not changed in any patients after embolization on the basis of an increase in hepatic tumor burden.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PVE was described in humans in the early 1980s [7, 13], more than 50 years after it was discovered that portal vein ligation in rabbits induces contralateral lobar hypertrophy [14]. Recently, PVE has gained acceptance among hepatobiliary and oncologic surgeons as a means of inducing hypertrophy of the FLR preoperatively in an effort to increase the number of patients who are candidates for hepatic resection [7-11]. Nearly every commercially available embolic agent has been used for PVE, including N-butyl cyanoacrylate; fibrin glue with Lipiodol; absolute ethanol; coils; and a mixture of Gelfoam, thrombin, contrast material, Lipiodol, and gentamicin with varying results.
Ethanol is a cytotoxic agent that causes inflammation and sclerosis of the vascular endothelium on contact. As a low-viscosity liquid, ethanol achieves distal embolization but has been documented histologically in rats and dogs to cause dose-dependent periportal fibrosis and necrosis [10, 15], as evidenced chemically in humans by elevated levels of transaminases after PVE [2]. In addition, because ethanol is not radiopaque, the embolization is technically challenging and requires the use of balloon occlusion catheters to prevent reflux into nontarget vessels. Empiric dosing based on patient weight is somewhat inexact, and patients may have systemic effects including intoxication, abdominal pain, and thrombocytopenia.
Cyanoacrylate, like ethanol, is not radiopaque, but can be mixed with tantalum powder that renders it visible fluoroscopically. Like ethanol, cyanoacrylate causes immediate and irreversible cross-sectional occlusion and is cytotoxic, causing a variable degree of hepatic necrosis [8]. Because the U.S. Food and Drug Administration (FDA) only recently approved it for intravascular use, it is not readily available in the United States, and many interventional radiologists are not familiar with its use.
Nagino et al. [9] used fibrin glue mixed with iodinated oil for PVE via an ipsilateral approach. In their series, 16 patients underwent right trisegment PVE (right lobe and segment IV) and 41 patients underwent standard right PVE. The mean increase in FLR after right trisegment PVE was 55% compared with 27% in the right PVE group.
Coils have been used for PVE in dogs [4] but were found to have an increased risk of clot propagation into nontarget vessels that may complicate or preclude liver resection. Indeed, we used coils in segmental right portal vein branch vessels after embolizing to stasis with PVA to prevent recanalization in three patients (not included in this series), and propagation of thrombus into the left portal vein was found at surgery in one patient. The presence of stainless steel coils also results in significant artifact if MRI or MR spectroscopy is performed after PVE. Consequently, the use of coils for PVE was abandoned at our institution.
Imamura et al. [11] reported their experience with PVE using a transileocolic approach and a temporary embolic agent, gelatin sponge powder (Gelfoam, Upjohn) mixed with iodinated oil, thrombin, and gentamicin in 84 patients. In contrast to ethanol and cyanoacrylate, embolization with this mixture showed little change in hepatic biochemical parameters before and after embolization. The mean hypertrophy in the nonembolized liver was 30%.
We have used PVA extensively for hepatic artery embolization and began using it for PVE because it is easy to use, is easily suspended in contrast material to make it radiopaque, is a permanent agent, and results in sustained terminal vessel occlusion. In addition, PVA is neither absorbed systemically nor cytotoxic and therefore should not result in systemic effects or cause periportal inflammatory changes hindering hepatic resection. In other words, because PVA causes less hepatic necrosis than other agents, more viable hepatocytes are preserved. This is significant because inevitably some patients who undergo PVE will ultimately not be candidates for resection [5], and the ability to induce contralateral hypertrophy while preserving maximal hepatic function is valuable. We believe that PVA combines the desirable properties of Gelfoam, which does not induce hepatic necrosis, with those of the cytotoxic agents, which are permanent; therefore, we began using it for PVE in November 1999.
We found a hypertrophy ratio after right PVE of 24.3% in the left liver, which increased to 31.9% in the three patients who underwent embolization of segment IV in addition to the right hemi-liver. Our experience with left PVE was less fruitful. In the eight patients who underwent preoperative left PVE, the mean hypertrophy ratio of the right side was 1.5% and the mean FLR/TELV ratio increase was only 3%. Despite the suboptimal imaging results in these cases, five of these patients (62.5%) ultimately underwent resection and none experienced postoperative liver failure. We believe the discrepancy between our results for right and left PVE is based on the fact that the mean FLR/TELV ratio before PVE of patients undergoing left PVE was 67%. Therefore, the embolized liver represented less than one third of the TELV.
Overall, our data for right PVE are comparable to achieved by Imamura et al. [11] and de Baere et al. [8] using Gelfoam and Madoff et al. [16] using a combination of PVA and coils. Madoff et al. reported a series of 16 patients, nine of whom underwent PVE before resection of liver metastasis. The mean increase in the FLR/TELV ratio in that subset of patients was 7.1%, although their overall increase (including both patients with primary liver tumors and liver metastases) was 8%. In comparison, the mean increase in FLR/TELV in patients who underwent right PVE in our series was 9%. Our results show slightly less contralateral hypertrophy, however, than that achieved by others using ethanol and cyanoacrylate [7-9, 15].
As reported by Madoff et al. [16], we found that PVE with PVA showed no sustained elevation in serum transaminases or bilirubin or any deterioration of liver synthetic function after embolization, suggesting that little to no hepatic necrosis occurred as a result of PVE. In addition, despite the relatively high contrast volume occasionally needed for PVE with PVA, no patient experienced postembolization or postoperative renal insufficiency. Although some patients in our series had recanalization of embolized portal vein branches on the first postembolization CT, the hypertrophy induced in these patients was not significantly different from that in those whose portal vein remained occluded, suggesting that recanalization visible radiographically involves primarily the segmental branches, whereas the small terminal branch vessels remain occluded.
Fifty-one PVEs were initially approached percutaneously from an ipsilateral portal vein branch, of which 50 were performed successfully. The single unsuccessful attempt at ipsilateral portal vein access was in a patient to undergo right PVE. The patient returned a second time, and the right portal vein was successfully embolized from a contralateral approach. Seven embolizations in this series were left PVEs performed from a right portal vein puncture site.
Unlike the transileocolic approach, percutaneous access to the portal venous system can be performed in the interventional suite with the patient under conscious sedation. This eliminates the risk of general anesthesia and should reduce the cost, procedure-related complications, and in-patient recovery time.
Various approaches to the portal venous system have been used for PVE, including transhepatic [7-9] and transileocolic [7, 11], the latter of which is relatively invasive and requires general anesthesia. Of the authors using a transhepatic approach, de Baere et al. [8] specifically used the left portal vein for access, whereas Nagino et al. [9] described a technique in which the target portal vein is used for access. Nagino's technique, however, for ipsilateral percutaneous PVE required the use of specially constructed catheters. We use standard angiographic catheters that are both readily available (eliminating the need for special-order equipment) and familiar to all interventional radiologists. More importantly, the ipsilateral approach limits the potential for complications involving the FLR. Azoulay et al. [6] reported a case in which ipsilateral access to the right portal vein was complicated by injury to the right hepatic artery. Although the patient developed hepatic necrosis, this did not preclude a potential curative operation (although the patient ultimately refused the operation). If, however, Azoulay et al. had chosen a contralateral approach, injury to the future remnant hepatic artery could have precluded potentially curative surgery.
The mean FLR/TELV ratio of 38.8% in our series before right PVE is somewhat higher than reported in other series. This is due to the fact that 54 patients (93%) received cytotoxic chemotherapy before embolization. We believe that patients who receive cytotoxic chemotherapy before PVE may have decreased hepatic reserve, despite adequate liver volume, and that may limit the ability of the liver remnant to hypertrophy after resection. Therefore, we perform this procedure even though the volume of the FLR is greater than 30% in such cases. The high FLR/TELV ratio in our patient population also reflects the fact that most of the patients in this series had small, poorly positioned metastases that would require extended resection of normal liver parenchyma for relatively low-volume disease.
In conclusion, PVE with PVA using an ipsilateral approach and standard angiographic catheters is safe and successful in achieving adequate left hemi-liver hypertrophy in patients with small predicted FLR. In contrast, left PVE using a contralateral approach did not induce significant right hemi-liver hypertrophy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. L. Denys, T. De Baere, and F. Doenz Portal vein embolization: a plea for strict patient selection. Am. J. Roentgenol., July 1, 2006; 187(1): W125 - W125. [Full Text] [PDF]
|
|
|
|
|
|
A. M. Covey and L. A. Brody Reply Am. J. Roentgenol., July 1, 2006; 187(1): W126 - W126. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
