| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Scientific Institute S. Raffaele, Vita-Salute
University, Olgettina 60, Milan 20132, Italy.
2 Department of Oncology, Scientific Institute S. Raffaele, Vita-Salute
University, Milan 20132, Italy.
3 Department of General Surgery, First Division, Scientific Institute S.
Raffaele, Vita-Salute University, Olgettina 60, Milan 20132, Italy.
4 Department of General Surgery, Second Division, Scientific Institute S.
Raffaele, Vita-Salute University, Milan 20132, Italy.
5 Department of Emergency Surgery, Scientific Institute S. Raffaele, Vita-Salute
University, Milan 20132, Italy.
Received March 10, 2003;
accepted after revision December 3, 2003.
Address correspondence to M. Venturini.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Catheter-port systems were placed in 204 patients with liver tumors (86.7% from colorectal metastases). Under sonographic and fluoroscopic guidance, a 5.8-French catheter was placed in the hepatic artery and connected to a subcutaneous reservoir after embolization of the gastroduodenal and right gastric arteries. Floxuridine plus dexamethasone and systemic low-dose heparin were administered. During the follow-up period, complications were classified as clinically not significant (type 1), clinically significant not requiring interruption of intrahepatic chemotherapy (type 2), clinically significant needing temporary suppression of intrahepatic chemotherapy (type 3), and clinically significant causing permanent suppression of intrahepatic chemotherapy (type 4).
RESULTS. No complications occurred during the implantation procedures. The mean number of intrahepatic chemotherapy cycles was 8.1. The mean follow-up period was 270 days. Primary and secondary patency rates of the system were 71.6% and 91.2%, respectively. Temporary suppression of intrahepatic chemotherapy was necessary in 19.6% of the patients and definitive suppression, in 8.8%. Hepatic artery thrombosis, not recanalized by local thrombolysis, was the main cause of permanent intrahepatic chemotherapy interruption (4.4%). Catheter occlusions and cerebral complications were not observed. In 91.2% of the patients, intrahepatic chemotherapy could be completed.
CONCLUSION. Percutaneous implantation of a removable and reimplantable catheter-port system for intrahepatic chemotherapy can be a safe procedure to treat unresectable liver metastases from colorectal cancer. Technical and pharmacologic complications with variable clinical relevance occurred, and various specific management strategies were necessary to reduce their incidence.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Hepatic artery infusion of floxuridine (or 5-fluorouracil) is an effective treatment for unresectable metastases from colorectal cancer [5, 6]. However, despite its pharmacologic advantage of higher tumor drug concentration with minimal systemic toxicity, local infusion of floxuridine can cause hepatobiliary and gastroduodenal toxicity [7, 8].
Surgically or percutaneously placed catheter-port systems were developed for long-term administration of intrahepatic chemotherapy. Surgically implanted intraarterial devices [9] permit the catheter to be stabilized, but these devices cannot be removed and reimplanted in case of complications. Catheter-port systems percutaneously implanted with a transaxillary [10, 11] or transfemoral [12, 13] approach can overcome the surgical stress and the irreversibility of surgical placement.
Different complication rates have been described using both implantation techniques in several reports [14, 15]. In most of them, both technical and pharmacologic complications were not completely specified and distinguished in terms of their clinical impact.
The purposes of our study were to retrospectively evaluate the main technical and pharmacologic complications occurring in the follow-up of 204 patients who underwent percutaneous transaxillary implantation of a catheter for intraarterial hepatic chemotherapy and determine their clinical relevance and management.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Technical Procedure
The implantation procedure was performed in a radiology suite with an
angiography unit. After sterilization and administration of local anesthesia,
the left subclavian–axillary artery was punctured under the clavicle
2–3 cm lateral to the insertion of the first rib, using the Seldinger
technique, under sonographic guidance. After placement of a 5-French
introducer sheath in the left axillary artery under fluoroscopic guidance,
selective catheterization of the superior mesenteric artery and celiac trunk
was preliminarily performed to study the arterial anatomy of the liver and
evaluate the patency of the portal vein. Furthermore, course, caliber, and
length of the hepatic artery and its branches, gastroduodenal artery, and
gastric arteries were evaluated: standard angiographic catheters (Cobra C2 and
Headhunter
[4–5
French]) and 0.038- or 0.035-inch guidewires with hydrophilic coating (J Stiff
guidewire, Terumo) were preferentially used.
In the absence of hepatic artery anomalies, 130 gastroduodenal arteries (63.7%) were embolized with Gianturco microcoils (Cook [diameter, 3–5 mm]) to avoid extrahepatic diffusion of the chemotherapeutic drug. In 62 patients (30.3%), the right gastric artery, when enlarged [10], was also embolized, preferentially using a 3-French coaxial microcatheter. In the presence of hepatic artery anomalies (36.3%) with separate origin of the right and left hepatic arteries, such as a right hepatic artery arising from the superior mesenteric artery or the common trunk of the left hepatic artery and left gastric artery, the catheter was usually placed in the largest artery. The proximal part of the smaller artery (usually the left hepatic artery) was embolized with Gianturco microcoils [10] to cause the development of collateral vessels from the contralateral hepatic lobe and to obtain bilateral and homogeneous chemotherapeutic infusion of the liver.
After removal of the introducer sheath, a 5.8-French radiopaque polyurethane catheter with an end hole was advanced over a 0.038-inch J Stiff guidewire and inserted into the hepatic artery. The catheter tip was placed in the proper hepatic artery in 48.5% or in the proximal part of the right (or left) hepatic artery in 51.4% of the cases. In the few patients (3.9%) with normal arterial anatomy but very short common and proper hepatic arteries, the catheter tip was located deep in the right hepatic artery and a side hole corresponding to the left hepatic artery origin was made in the catheter. The final position of the catheter during the implantation procedure was chosen according to the arterial anatomy of each patient, course and caliber of the hepatic artery, and distribution of the metastatic lesions on preliminary contrast-enhanced CT or MRI.
After the definitive positioning of the catheter tip and local anesthesia administration, an incision that was approximately 3 cm long was made 3–4 cm below the puncture site in the left subclavian area. A subcutaneous pocket was formed by a blunt dissection above the pectoral fascia. Then the extravascular part of the catheter was cut off, pushed through a tunnel from the puncture site to the pocket in the subcutaneous tissue, and connected to the reservoir. The proximal part of the catheter was placed in a way to form a slight curve to allow shoulder movement and prevent too much tension in the distal part. After the final control angiogram of the catheter tip position was obtained, an absorbable running suture was made to close the pocket. Skin stitches were finally placed. The patients were mobilized immediately after the procedure. The stitches were removed after 15 days.
Intrahepatic chemotherapy was started 7 days after the percutaneous procedure with a continuous infusion of floxuridine (dose escalation, 0.15–0.30 mg/kg per day for 14 days every 28 days) plus 28 mg of dexamethasone and 500 IU of heparin. In 15 patients (7.3%), 5-fluorouracil was administered when floxuridine was temporarily unavailable: also in these patients, 5-fluorouracil (250–300 mg/day) was administered for 14 days every 28 days, plus 28 mg of dexamethasone and 500 IU/day of heparin through the port. Subcutaneous low-molecular-weight heparin (enoxaparin, 2,000 IU/day for 14 days every 28 days) was administered in all the patients during the period free from intrahepatic chemotherapy (at home).
During the follow-up period, catheter patency and hepatic artery patency were assessed before each treatment cycle through review of a control angiogram performed by injection of contrast material into the catheter-port system. A whole-body contrast-enhanced CT examination was performed before intrahepatic chemotherapy and every third cycle. Thoracic CT without contrast administration and abdominal MRI were performed in four patients with chronic renal insufficiency.
The durability of the system was analyzed in terms of the number of cycles administered, mean follow-up period, and primary and secondary patency rates of the system. The complications occurring after the procedure were classified and evaluated considering time from onset, clinical relevance, and specific management.
Complications: Classification and Management
Complications observed during the follow-up period were retrospectively
classified on the basis of cause and clinical relevance. We distinguished
local complications due to the arterial puncture (group A), technical or
mechanical complications due to the permanent catheter in the hepatic artery
(group B), and pharmacologic complications due to the prolonged
chemotherapeutic drug infusion (group C).
On the basis of their clinical relevance, we distinguished clinically not significant complications that can be easily corrected without intrahepatic chemotherapy interruption (type 1), clinically significant complications not requiring intrahepatic chemotherapy interruption (type 2), clinically significant complications needing temporary suppression of intrahepatic chemotherapy (type 3), and clinically significant complications causing permanent suppression of intrahepatic chemotherapy (type 4). Different management was used for the complications of groups A, B, and C.
Group A (local complications).—Complications in this group included local hematoma and pneumothorax. Local hematoma was treated by manual compression, and pneumothorax was treated by pleural drainage (8-French catheter) if clinically significant.
Group B (technical or mechanical complications).—Complications included catheter malposition, hepatic artery thrombosis, and sepsis. Catheter malposition was caused by catheter dislodgment in the celiac trunk or abdominal aorta or catheter tip in too distal of a position (right or left hepatic artery). In cases of catheter dislodgment in either the celiac trunk or the abdominal aorta, surgical opening of the pocket in the subclavian area and temporary disconnection of the proximal part of the catheter from its reservoir were necessary. In cases of dislodgment in the celiac trunk, a hydrophilic guidewire was inserted in the catheter and pushed in the hepatic artery. Then the same catheter was advanced in the hepatic artery over the guidewire. In cases of dislodgment in the abdominal aorta, a new catheter-port system was placed using the same transaxillary approach, and by maintaining a hydrophilic guidewire in the aorta, removing the catheter, and then performing a new catheterization of the hepatic artery with an angiographic catheter. In both cases, the catheter of the port was more distally positioned than the previous position. In cases of the catheter tip distally placed in the right or left hepatic artery, retraction of the catheter tip in a more proximal position (proper hepatic artery) was achieved with a 5-French Simmons catheter using a transfemoral approach. The catheter of the port was hooked in the abdominal aorta at the level of the celiac trunk entrance and pulled down, thus allowing retraction of the catheter tip.
In cases of hepatic artery thrombosis, local thrombolysis was attempted through the catheter-port system with the administration of recombinant tissue plasminogen activator. The total drug dose (maximum dose for an 80-kg patient, 50 mg) was adjusted on the basis of patient weight and was administered as follows: 10 mg as bolus, 10 mg in slow infusion in 10 min, and 10 mg/hr for 3 hr. During follow-up, in cases of thrombosis successfully treated by local thrombolysis, subcutaneous low-molecular-weight heparin (enoxaparin) was increased from 2,000 to 6,000 IU/12 hr during the free period from local chemotherapy.
In cases of sepsis, antibiotic therapy, catheter-port system removal, and prolonged manual compression of the left subclavian area were performed.
Group C (pharmacologic complications).— Complications in this group included gastroduodenitis, gastroduodenal ulcer, cholecystitis, hepatobiliary toxicity, and arteriobiliary fistula.
In cases of gastroduodenitis or gastroduodenal ulcer, medical therapy (omeprazole) was administered, and, when necessary, transfemoral embolization of the right gastric artery (or other small gastric branches), arising distally to the catheter tip, was performed with a 3-French microcatheter and Gianturco microcoils.
In cases of cholecystitis, medical therapy was administered and cholecystectomy, if necessary, was performed.
In cases of hepatobiliary toxicity, according to Miller's classification of hepatobiliary toxicity [16], clinically manifested as elevations of aspartate aminotransferase, alkaline phosphatase, and bilirubin levels, we distinguished grades 2 and 3 from grade 4 causing temporary or permanent interruption of intrahepatic chemotherapy, respectively. After temporary suppression, a variable reduction of chemotherapeutic dose was considered, whereas chemotherapy was definitively interrupted when serum bilirubin increased to more than 3 mg/dL.
In cases of arteriobiliary fistula, permanent interruption of local chemotherapy was always necessary. Transaxillary embolization with Gianturco microcoils was performed when necessary: a 3-French coaxial microcatheter was pushed in the catheter of the temporarily disconnected port.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Group A Complications
These complications occurred immediately after the procedure in 1.4% of the
patients and were due to a local hematoma and two cases of pneumothorax.
The local hematoma and one pneumothorax resolved after prolonged manual compression and spontaneously without major complications (type 1 complication), respectively. In the second case of pneumothorax, pleural drainage was necessary for 4 days, but intrahepatic chemotherapy could be started again regularly after 7 days (type 2 complication).
Group B Complications
These complications occurred in 17.4% of the patients and were due to 18
cases of catheter malposition, 14 cases of hepatic artery thrombosis, and four
cases of sepsis.
Of the 18 cases of catheter malposition (type 3 complications), 14 cases were due to catheter dislocation: nine in the celiac trunk (Fig. 1A, 1B) and five in the abdominal aorta. In all cases, the catheter was successfully repositioned after making a pocket surgical opening, and a new dislocation did not occur during the follow-up period. In the nine cases of dislocation in the celiac trunk, the same catheter-port system was used. In the five cases of dislocation in the abdominal aorta, a new catheter-port system was used to replace the original one. In the remaining four cases of catheter malposition, the catheter tip had been too distally placed in the right hepatic artery (Fig. 2A, 2B, 2C, 2D, 2E, 2F, 2G). Catheter retraction via a transfemoral approach was easily achieved in all cases, and no secondary complications due to the loop of the catheter in the abdominal aorta were observed during the follow-up period.
|
|
|
|
|
|
|
|
|
All the cases of catheter malposition occurred during the first 8 weeks of follow-up (mean, 4.56 weeks; range, 1–8 weeks).
In five of the 14 patients with hepatic artery thrombosis, thrombolysis was successfully performed with satisfactory recanalization of the hepatic artery and its main distal branches (Fig. 3A, 3B). During the follow-up period, hepatic artery patency allowed regular administration of intrahepatic chemotherapy in four cases (type 3 complications), and in one case, a recurrent thrombosis caused permanent interruption of intrahepatic chemotherapy (type 4 complication). In the other eight cases of thrombosis, recanalization was unsuccessfully attempted (type 4 complications). No complications due to the thrombolytic treatment (i.e., hemorrhage) were observed. In one case, hepatic artery thrombosis was an incidental finding on a control angiogram: the presence of rich collateral vessels allowed regular distribution of the drug to the distal branches of the hepatic artery (type 2 complication).
|
|
The time from onset of hepatic artery thrombosis was variable (mean, 15.45 weeks; range, 3–33 weeks).
In four cases of sepsis, permanent interruption of intrahepatic chemotherapy was necessary (type 4 complications). No major complications were observed after removal of the catheter-port system and manual compression of the subclavian area. Sepsis did not occur in those patients who had a pocket surgical opening due to catheter dislocation.
All cases of sepsis occurred 15 weeks or more after the implantation procedure (mean, 24.72 weeks; range, 15–31 weeks).
Group C Complications
These complications occurred in 18.6% of the cases and were due to 15 cases
of gastroduodenitis, nine cases of gastroduodenal ulcer, two cases of
cholecystitis, 10 cases of hepatobiliary toxicity, and two cases of
arteriobiliary fistula. In 24 patients, epigastric pain, nausea, and
gastroduodenal acidity of variable grade and intensity occurred during
intrahepatic chemotherapy infusion.
In 13 cases of gastroduodenitis, medical therapy was sufficient, and no temporary interruption of intrahepatic chemotherapy was necessary (type 1 complications). In the other two more severe cases of gastroduodenitis, transfemoral embolization of two right gastric arteries was necessary and was successfully performed. In one case, intrahepatic chemotherapy interruption was not necessary (type 2 complication), whereas in the second case, it was necessary (type 3 complication).
Eight cases of gastroduodenal ulcers caused temporary interruption of intrahepatic chemotherapy (type 3 complications). In one case, permanent interruption was necessary (type 4 complication): transfemoral embolization of six right gastric arteries, arising distally to the catheter tip, was successfully performed.
Two cases of cholecystitis were treated in a different manner: in one case, medical therapy was sufficient (type 2 complication), and in the other case, surgical cholecystectomy was necessary (type 3 complication). After cholecystectomy, intrahepatic chemotherapy was restarted without other complications.
In eight cases of hepatobiliary toxicity due to floxuridine administration, temporary interruption of intrahepatic chemotherapy (type 3 complications) was necessary with a variable reduction of chemotherapeutic dose when restarting the treatment. In two cases of grade IV hepatobiliary toxicity, the significant persistence of elevated levels of aspartate aminotransferase, alkaline phosphatase, and bilirubin caused permanent interruption of intrahepatic chemotherapy (type 4 complications): bilirubin levels in both cases increased to more than 3 mg/dL, and intrahepatic chemotherapy administration was interrupted to prevent the development of sclerosing cholangitis. A sonographic examination did not show biliary dilatation, and percutaneous transhepatic cholangiography due to biliary strictures was not necessary in any case.
Finally, in two patients with melena treated with local administration of 5-fluorouracil, two arteriobiliary fistulas were shown on a contrast-enhanced CT examination and an angiographic control study (Fig. 4A, 4B). In one case, the fistula, located distal in relation to the catheter tip, was successfully embolized by pushing a few Gianturco microcoils through a coaxial 3-French microcatheter into the catheter of the port-system and obtaining prompt melena resolution. In the other case, the fistula, adjacent to the catheter tip, spontaneously resolved a few days after intrahepatic chemotherapy suppression. In both cases, permanent interruption of intrahepatic chemotherapy was necessary (type 4 complications). Gastroduodenal complications (and cholecystitis) occurred in a variable period during the follow-up (mean, 12.09 weeks; range, 1–28 weeks), whereas all cases of hepatobiliary toxicity (mean, 21.96 weeks; range, 15–27 weeks) and arteriobiliary fistulas (mean, 21 weeks; range, 18–24 weeks) occurred 15 or more weeks after the implantation procedure.
|
|
In summary, the percentages of complications based on their different clinical relevance were as follows: 7.4% (type 1), 2.0% (type 2), 19.6% (type 3), and 8.8% (type 4).
In 186 (91.2%) of 204 patients, intraarterial hepatic chemotherapy could be completed, notwithstanding all the complications that occurred. Hepatic artery thrombosis, not recanalized by local thrombolysis, was the main cause of permanent interruption of intrahepatic chemotherapy (4.4%). Neither occlusion of the catheter-port system nor cerebral complications occurred in this study.
Complications with variable clinical relevance are listed in Table 1: the percentages refer to the number of the cases and not to the patients because some complications were linked (gastroduodenitis and catheter dislodgment in the celiac trunk, for example).
|
We compared the technical complications in our study with those of other authors using a transaxillary or transfemoral percutaneous approach. These data are listed in Table 2.
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Group A Complications
As in other studies, we observed a low incidence of local complications: a
local hematoma and two cases of pneumothorax, only one of which required a
pleural drainage. Local complications due to the arterial puncture (hematomas,
hemorrhages, pseudoaneurysms, arteriovenous fistulas) also became rarer when
using a transaxillary approach after the method of Arai et al.
[21] was modified so that
arterial access was obtained directly under sonographic guidance: the direct
puncture of only the anterior wall of the left subclavian–axillary
artery—avoiding the vein that runs more cranial and ventral—can
reduce the incidence of local complications
[17,
22].
Group B Complications
Most "catheter-induced" complications were caused by catheter
malposition (8.8%) and hepatic artery thrombosis (6.8%). Among the cases of
catheter malposition, we distinguished 14 proximal dislocations, needing a
surgical opening of the pocket, from four distal dislocations. In cases of
proximal dislocation in the celiac trunk, the same catheter-port system was
correctly repositioned by pushing the catheter over a hydrophilic guidewire
previously advanced into the hepatic artery, whereas in cases of dislocation
in the abdominal aorta, a new catheter-port system was necessary with an
evident major cost. A higher frequency of catheter dislocation in the
abdominal aorta in the early implantations than in the total implantations was
probably related to the learning curve about the procedure. Generally, when
the catheter was placed in the proximal part of the right or left hepatic
artery, the first control angiogram after the implantation procedure showed
catheter retraction in the proper or common hepatic artery, due to the
patient's physiologic movements, allowing the regular distribution of drug in
both hepatic lobes.
In four patients, contrast-enhanced CT or MRI showed a unilateral response after intrahepatic chemotherapy due to catheter permanency in the right hepatic artery (distal dislocations). Catheter repositioning in the proper hepatic artery was necessary to allow complete efficacy of intrahepatic chemotherapy. Catheter retraction with a Simmons catheter via a transfemoral approach was an easy procedure, and no complications due to the small loop of the catheter in the abdominal aorta were observed and surgical opening of the pocket was avoided. Like other authors [13, 20], we believe that the creation of a subcutaneous tunnel for the extravascular part of the catheter and a low tension in the configuration of the intravascular part of the catheter are crucial in reducing the frequency of dislocation. The catheter can be stabilized using the Arai technique [23] that was modified by other authors [24]. For this modified technique, a catheter permanently is fixed with bucrylate in the gastroduodenal artery and a side hole corresponding to the origin of the proper hepatic artery. With this technique, removal (and eventual reimplantation) in case of catheter dysfunction is not feasible.
Variable percentages of dislocation have been found by other authors, as shown in Table 2, without a significant difference between the transaxillary and transfemoral approaches. In a recent study about the complications encountered with a transfemoral placement of a catheter-port system, Kuroiwa et al. [15] compared a heparin-coated polyurethane catheter with a stiffer catheter of the same caliber (5-French) that was coated with a copolymer composed of fluorine, acryl, styrene, urethane, and silicone, and they found a lower percentage of dislocations using the stiffer catheter.
In our study, catheter placement in the proximal part of the right or left hepatic artery reduced the number of dislocations, especially in patients with a short proper hepatic artery; however, the incidence of thrombosis increased with this placement technique because the caliber of the right hepatic artery was smaller than that of the proper hepatic artery. Thus, the relation between the lumen of the target vessel and the size of catheter seems to be important in the rate of hepatic artery thrombosis. Using a catheter smaller than 5.8 French, we found that the incidence of thrombosis could be reduced but probably with an increased incidence of occluded catheters. Comparing technical complications among the studies listed in Table 2, it is evident that a relatively high percentage of thrombosis is often associated with a low percentage of catheters occluded and vice versa.
We believe that anticoagulation prophylaxis, softness (polyurethane), and coaxial position of the catheter in the hepatic artery can contribute to a reduction in the number of the cases of thrombosis. In patients with permanent infusion devices, repetitive mechanical trauma of the catheter and the toxic effect of chemotherapeutic agents on the arterial wall may be considered the main causes of thrombosis [25]. Arteritis represents another possible cause of arterial occlusion reported by a few authors [26]: histopathologic findings of arteritis have been reported by Marymont et al. [27] in five cystic arteries after cholecystectomy due to cholecystitis associated with intrahepatic chemotherapy. Involvement of the arterial wall due to arteritis could explain one arteriobiliary fistula located distal to the catheter tip.
In our study, clinical relevance of arterial thrombosis differed (types 2–4 complications), mainly in relation to the efficacy of thrombolytic treatment: permanent suppression of intrahepatic chemotherapy was necessary in nine of 14 cases because of unsuccessful thrombolysis, and arterial thrombosis was the main cause of type 4 complications (9/204 cases [4.4%]). Efficacy of local thrombolysis depends on many factors [28], but thrombolysis can probably be more effective on recent thrombosis. We therefore think that regular follow-up angiography through the catheter-port system is important to monitor not only the position but also the function of the catheter and condition of the hepatic artery. For example, in case of early diagnosis of arteritis or a hypertrophic right gastric artery, temporarily interrupting intrahepatic chemotherapy, maintaining dexamethasone infusion, or performing a transfemoral embolization, respectively, could reduce the incidence of thrombosis and ulcer.
Sepsis, the second cause of permanent interruption of intrahepatic chemotherapy (1.9%) in our study, was treated with antibiotic therapy and definitive catheter-port system removal to avoid a recurrent infection in a new system [18, 29, 30]. When comparing the reports of percutaneous and surgical implantations [5], we found no significant differences in terms of sepsis rates. Interventional radiology suites seem to provide sufficient hygienic conditions for these procedures. In our experience, all cases occurred a long time after the implantation procedure probably because of improper skin preparation and care during the puncture of the reservoir and chemotherapeutic infusion.
Group C Complications
Gastroduodenitis and gastroduodenal ulcers are generally due to the
inadvertent perfusion of the stomach and duodenum, which occurs as a result of
an incompletely occluded gastroduodenal artery or a hypertrophic right gastric
artery [30] arising distally
to the catheter tip. In our experience, no gastroduodenal complications were
due to revascularization of a previously embolized gastroduodenal artery.
However, in eight cases of gastroduodenal complications (six cases of ulcer),
transfemoral embolization of the right gastric artery was performed after the
implantation procedure. Right gastric artery embolization is regularly
performed by Japanese authors
[15,
17], differently from European
authors [11,
13,
18]. Yoshikawa et al.
[10] explained that they also
perform right gastric artery embolization when the right gastric artery is
enlarged. Routine right gastric artery embolization
[31] could further reduce the
incidence of gastroduodenitis and ulcers (4.4% in our study). However, we
recommend accurate control of position, caliber, and course of the right
gastric artery during the catheter-port system implantation after
gastroduodenal artery embolization. Often the right gastric artery or other
small arteries directed to the stomach became hypertrophic after
gastroduodenal artery embolization. Other cases of ulcers were caused by
inadvertent drug infusion in the left gastric artery due to catheter
dislocations in the celiac trunk. In all cases of gastroduodenal complications
except one, intrahepatic chemotherapy could be restarted after temporary
interruption and transfemoral embolization or catheter repositioning,
maintaining omeprazole protection.
Two cases of cholecystitis were also caused by drug perfusion of the cystic artery. In one, cholecystectomy was necessary because of severe chemical inflammation. A retrospective analysis of this case showed the cystic artery just above the catheter tip.
In addition to peptic disease, the main problem due to floxuridine intraarterial infusion was hepatobiliary toxicity. In two of 10 cases, grade IV hepatobiliary toxicity required permanent interruption of intrahepatic chemotherapy. Hepatobiliary toxicity is clinically characterized by elevation of liver blood parameters and, in more severe cases, by jaundice and sclerosing cholangitis [32]. Bile ducts are particularly susceptible to floxuridine toxicity because they receive their blood supply from the hepatic artery, whereas hepatocytes are perfused by a dual blood supply (portal vein and hepatic artery). After a surgical implantation procedure, biliary toxicity may be increased by an ischemic injury of the bile duct wall secondary to the extensive hepatic hilum dissection and by the ligation of small arterial branches of the biliary tract. This factor can be avoided using percutaneous implantation procedures [33].
In previous studies [34] and our study, dexamethasone was administered with floxuridine and 5-fluorouracil, and intrahepatic chemotherapy was performed for 2 consecutive weeks every 4 weeks to decrease hepatobiliary toxicity and arterial wall damage due to chemotherapeutic infusion. Steroids are known to reverse jaundice, particularly in cases of acute hepatitis but also in cases of obstructive jaundice [35]. The preferential use of floxuridine for intrahepatic chemotherapy was based on its elevated efficacy due to high hepatic extraction when compared with 5-fluorouracil (94–99% vs 19–55%, respectively) and a minor systemic toxicity. In some patients, 5-fluorouracil was used when floxuridine was temporarily unavailable. Complications due to hepatobiliary toxicity were not observed, but two cases of arteriobiliary fistulas (type 4 complications) were shown on a control angiogram, probably due to the high endothelial toxicity of 5-fluorouracil. In one case, the fistula located distal to the catheter tip was successfully embolized by advancing a coaxial microcatheter into the catheter. The possibility of a conservative treatment with a covered stent [36] was not considered because of the high probability of infection.
Time of Onset of Complications
Finally, a variable time of onset in the different complications was
observed: early complications (local complications, catheter malposition),
late complications (sepsis, hepatobiliary toxicity, arteriobiliary fistulas),
and complications of variable onset (hepatic artery thrombosis, gastroduodenal
complications) were distinguished.
Knowledge of these data can be useful in preventing possible development of new complications through a more strict clinical and radiologic follow-up. For example, if a catheter dislocation does not occur in the first 2 months, then it will rarely occur during the remaining follow-up. On the contrary, the risk of thrombosis persists for the entire period, and a strict regimen of angiographic controls is still mandatory.
In conclusion, in agreement with other authors, we also believe that percutaneous transaxillary or transfemoral implantation of a permanent catheter—one that is removable and reimplantable—for intraarterial hepatic chemotherapy will be well accepted by patients and can serve as a safe and effective procedure, particularly for patients with unresectable liver metastases from colorectal cancer. In previous reports regarding complications of intrahepatic chemotherapy, both technical and pharmacologic complications were rarely analyzed, and their different clinical relevance was generally not considered, probably also because of the difficulty in monitoring all treated patients and obtaining accurate follow-up. In our opinion, for a true balance of occurred complications, careful and continuous monitoring of the condition of the catheters and the hepatic artery based on repetitive control angiograms is necessary for the entire period of treatment in all patients. Our retrospective study, based on a relatively large number of patients, has shown that intraarterial hepatic chemotherapy could be completed in 91.2% of the patients, notwithstanding the variable occurred complications.
Different management and various strategies to reduce the incidence of complications have been described. In agreement with other authors and also in our experience, catheter dislocation and hepatic artery thrombosis represented the main complications that occurred and the first cause of temporary and permanent interruption of intraarterial hepatic chemotherapy, respectively. The creation of a subcutaneous tunnel for the extravascular part of the catheter and the absence of tension in its final configuration can reduce the frequency of dislocation, and anticoagulation prophylaxis, softness of polyurethane, and coaxial position of the catheter into the hepatic artery can reduce the incidence of thrombosis. Probably the creation of a new catheter composed by a soft distal part of small caliber (4-French) into the hepatic artery and a stiffer and larger (5- or 6-French) proximal part into the abdominal aorta could contribute to further reduce the cases of dislocation, thrombosis, and catheter occlusion.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Pulitano, M. Arru, M. Catena, E. Guzzetti, G. Vitali, M. Ronzoni, M. Venturini, E. Villa, G. Ferla, and L. Aldrighetti Results of Preoperative Hepatic Arterial Infusion Chemotherapy in Patients Undergoing Liver Resection for Colorectal Liver Metastases Ann. Surg. Oncol., June 1, 2008; 15(6): 1661 - 1669. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Seki, T. Ozaki, and M. Shiina Side-Hole Catheter Placement for Hepatic Arterial Infusion Chemotherapy in Patients with Liver Metastases from Colorectal Cancer: Long-Term Treatment and Survival Benefit Am. J. Roentgenol., January 1, 2008; 190(1): 111 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Nagaoka, S. Itano, H. Nagamatsu, J. Akiyoshi, J. Kurogi, N. Tajiri, M. Kajiwara, and M. Sata Temporary Indwelling Catheter System via the Left Brachial Artery: Evaluation in 83 Patients with Hepatic Tumors Am. J. Roentgenol., March 1, 2007; 188(3): 652 - 658. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Seki and M. Shiina Placement of a long tapered side-hole catheter in the hepatic artery: technical advantages, catheter stability, and arterial patency. Am. J. Roentgenol., November 1, 2006; 187(5): 1312 - 1320. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||
