- Email: [email protected]
Intraductal cooling of the main bile ducts during radiofrequency ablation prevents biliary stenosis1
Intraductal Cooling of the Main Bile Ducts During Radiofrequency Ablation Prevents Biliary Stenosis Dominique Elias, MD, PhD, Lucas Sideris, MD, FRCSC, Marc Pocard, MD, PhD, Clarisse Dromain, MD, Thierry De Baere, MD Small-sized tumors sited closer than 15 mm to the central bile ducts cannot be destroyed by radiofrequency (RF) because the resulting heating causes destruction and stenosis of the bile ducts. STUDY DESIGN: Thirteen patients underwent intraoperative RF to treat tumors closer than 6 mm to a central bile duct. Stenosis prevention was achieved by cooling the main bile duct(s) (right, left, or both) with a 4°C saline solution quickly infused by a catheter introduced inside the bile duct through a choledochotomy. Median followup after cooling was 19.7 months (range 6 to 42 months). RESULTS: There was no mortality and no RF-related morbidity. Twelve patients (92.3%) did not present any biliary stenosis. In one patient (7.7%) biliary stenosis was detected 6 months after RF. The patient was cirrhotic but did not require any treatment. One of the 13 patients had a local recurrence. CONCLUSIONS: The intraoperative cooling of central bile ducts prevents biliary stenosis efficiently when using RF to destroy tumors closer than 5 mm to the bile duct. This new technique decreases the contraindications for the use of RF and allows more patients to be treated with RF with curative intent. ( J Am Coll Surg 2004;198:717–721. © 2004 by the American College of Surgeons) BACKGROUND:
Radiofrequency (RF) eradicates liver tumors by heat. When used for hepatectomy, it is useful to treat with curative intent (R0 result) those many patients in whom the disease is usually considered unresectable.1-4 Heat damages the bile ducts, causing destruction, stenosis, and then upstream bile duct dilatation.5,6 The scientific community does not recommend RF for tumors closer than 15 to 20 mm to the main branches of the biliary tree.1-8 In 2001, we described the intraductal cooling of the main bile duct(s) during intraoperative RF to prevent biliary stenosis when the tumor was badly located.9 There were only 3 patients and the minimal followup was 3 months. In this article, we report a series of 13 patients whose tumors, closer than 6 mm to the main biliary ducts, were treated with this new procedure. Followup ranges from 6 to 42 months.
METHODS We used bile duct cooling in humans for the first time in December 1999. From then until April 2003, this technique was applied to 13 patients. During this period, 268 patients underwent a hepatectomy for a malignant lesion, and 83 of them also underwent intraoperative RF. Eleven patients were treated with intraoperative RF without hepatectomy. Intraoperative bile duct cooling was associated with 13.8% of all RF and 4.6% of all patients undergoing hepatectomy or RF during the same period. The only selection criterion was the presence of a tumor potentially treated by RF for which one edge was closer than or equal to 5 mm from a central bile duct. All these patients were treated with curative intent. Central bile ducts were defined as follows: on the left side, only the primary duct (up to the origin of the duct for segment III); on the right side, the primary duct and its two secondary branches (right anterior trunk for segments V–VIII, and right posterior trunk for segments VI–VII). The characteristics of the 13 patients are reported in Table 1. The mean age was 57 years (range 30 to 75 years), and there were 6 men and 7 women. The mean distance between the central bile duct and the tumor was
No competing interests declared.
Received October 29, 2003; Accepted December 3, 2003. From the Departments of Surgical Oncology (Elias, Sideris, Pocard) and Radiology (Dromain, De Baere), Institut Gustave Roussy, Cancer Center, Villejuif, Cedex, France. Correspondence address: D Elias, MD, Institut Gustave Roussy, Cancer Center, 39 rue Camille Desmoulins, Villejuif, Cedex 94805, France.
© 2004 by the American College of Surgeons Published by Elsevier Inc.
717
ISSN 1072-7515/04/$30.00 doi:10.1016/j.jamcollsurg.2003.12.026
718
Hepatic Intraductal Cooling
Elias et al
J Am Coll Surg
Table 1. Patient Characteristics Distance to bile duct (mm)
Site of cooling
Pt. No.
Origin
1 2 3 4 5 6
Hepatocellular Endocrine Colon Endocrine Colon Colon
No No No No No No
4 1 5 3 4 1
Left Left ⫹ right Right Right Left Left
7 8
Colon Rectum
Central (IV, V, VIII) No
0 0
Left ⫹ right Left ⫹ right
9 10 11
Colon Colon Colon
Left hepatectomy (II–IV) Right hepatectomy (V–VIII) No, but PC treated before
2 3 0
Right Left Right
12 13
Endocrine Colon
No Right hepatectomy ⫹ 1 left wedge then percutaneous left RF
4 5
Left Left
Preliminary hepatectomy
Associated surgery
No: severe cirrhosis Whipple, 4 metastasectomies, 4 other RF Left hepatectomy (II–IV), 2 RF in right liver Left hepatectomy (II–IV), 4 RF in right liver Right hepatectomy (V–VIII) ⫹ Segment I Right hepatectomy (V–VIII) ⫹ 3 left metastasectomies Segmentectomies VI ⫹ I Anterior resection rectum ⫹ left lobectomy (II–III) 2 right metastasectomies 1 left metastasectomy No. Severe adhesions after the treatment of the PC. Right hepatectomy (V–VIII) No
Bile duct stenosis
Yes No No No No No No No No No No No No
PC, peritoneal carcinomatosis treated by resection and intraperitoneal chemohyperthermia 21 months before; RF, radiofrequency.
2.2 mm (range 0 to 5 mm). Ten of these patients underwent a hepatectomy during the same session as RF with bile duct cooling. The three other patients presented only one liver tumor close to a central bile duct and did not undergo a hepatectomy for the following reasons: 1) severe cirrhosis, 2) severe peritoneal adhesions 21 months after the complete resection of a peritoneal carcinomatosis (from a colon cancer) associated with an intra-peritoneal chemohyperthermia, and 3) history of right hepatectomy plus one left metastasectomy 16 months earlier followed by two percutaneous RF in the left liver 8 months later. Finally, two patients were treated with intraarterial oxaliplatin allowing consideration of a curative procedure when it was initially not possible. Six coolings involved the left bile duct, four the right bile duct, and three both the right and left bile ducts because the tumor was located just above the biliary confluence. Technique of bile duct cooling
Ultrasonography and hepatectomy were done first. The common bile duct was isolated and opened longitudinally, as close as possible to the hepatic hilum. The common bile duct dissection was very difficult in the two cases of preliminary intraarterial hepatic chemotherapy
because of hard sclerosis of the whole hepatic pedicle. In four cases, a second incision was necessary, higher than the first one, at the level of the main bifurcation, to successfully install the catheter at the right place. This second incision was necessary in the two patients in whom a first catheter was placed in the right anterosuperior branch and a second one in the right posteroinferior branch. The 60-cm long, 2 mm-diameter catheters (Nutricath 2181; Vygon) were inserted at the right place under ultrasonographic guidance (Fig. 1). A delicate but rapid movement of the catheter allows vizualization of its tip inside the bile ducts. Care must be taken not to inject gas inside the biliary tree, because the tip will become invisible. When placed in the appropriate location, the catheter was temporarily fixed to the edge of the bile duct incision with a thin suture. A 4°C Ringer’s lactate perfusion was initiated to wash and cool the main bile duct(s) continuously during RF, and the liquid was expelled through the choledochal incision. The flow rate was as high as possible; the bag of Ringer’s lactate was pressured with an inflatable infusor (C-Infusor 500; Medex Inc) at 300 mmHg. In this series, we did not use concomitant vascular clamping during RF to keep the physiologic heat sink effect of the blood flow. At the end of the procedure, the incisions on the bile ducts were
Vol. 198, No. 5, May 2004
Elias et al
Hepatic Intraductal Cooling
719
was 6 months, and the median was 19.7 months (range 6 to 42 months). Twelve patients (92.3%) did not present any stenosis of the bile duct. Examples are reported in Figures 2 and 3. Only one patient (7.7%) presented segmental stenosis of the bile ducts above the site of RF 6 months after treatment. This was a cirrhotic patient presenting a 33-mm hepatocellular carcinoma in segment IV (Fig. 4). No treatment was done and there was no further complication. One patient (7.7%) presented a local recurrence on the RF site 4 months after the treatment. Figure 1. Diagram showing cooling of the main bile ducts, with one or two catheters introduced in a retrograde fashion after opening the common bile duct. C, common bile duct; I, perfusion with 4°C Ringer’s lactate; T, tumor.
closed with a continuous suture of a thin absorbable material without any specific drainage. RESULTS There was no mortality, and the morbidity rate was 31%, mainly because of lung atelectasis and right pleural effusion. There was no complication related to RF or cooling. All 13 patients underwent a CT scan of the liver before leaving the hospital to assess the result of the RF and to rule out any bile duct stenosis. Ultrasonography was performed between 3 and 4 weeks after operation to detect a dilatation of the biliary tree above the treated site. These two early explorations never detected any dilatation; followup was done every 3 months, as usual. The minimal followup after cooling for the entire series
DISCUSSION This series is the first to report successful late results for the cooling of main bile ducts to prevent biliary stenosis after heat damage after RF. Heat damages bile ducts and results in stenosis in pigs10,11 and in humans.5,6 This was studied later in pigs in which a specific experimental model of post-RF biliary stenosis was developed.12 This stenosis had no real impact when located peripherally, but could be dangerous when located centrally in the main bile ducts. These damages led the scientific community to contraindicate RF to destroy tumors closer than 15 to 20 mm to the central bile ducts. Two procedures eventually prevented biliary stenosis: the bile duct cooling we described in three patients in 2001,9 and the endoscopically placed internal stent, as reported for one patient by Bilchik and colleagues the same year.13 In this article, we report on 13 patients who presented with tumors located 0 to 5 mm from the main bile duct
Figure 2. Patient No. 7. (A) Aspect on CT scan of the recurrence after a central hepatectomy. (B) Result 6 months after radiofrequency plus cooling.
720
Elias et al
Hepatic Intraductal Cooling
J Am Coll Surg
Figure 3. Patient No. 2. (A) Preoperative magnetic resonance imaging. (B) Magnetic resonance imaging 3 months after radiofrequency with bile duct cooling. This patient with an endocrine tumor underwent a Whipple resection plus multiple wedge liver resections associated with radiofrequency in the left part of the liver. Three months later, he underwent a right hepatectomy.
with a strict definition of them. The preventive cooling allowed avoidance of biliary stenosis in 92% of our patients. Retrospectively, we did not find any explanation for the unique failure we observed, except that it occurred in a cirrhotic parenchyma. Also, the biliary dilatation was not clearly detectable 3 months after RF, but only 6 months later. This stenosis was well tolerated and did not necessitate any treatment. An experimental study performed on pigs about the preventive effect of bile duct cooling has recently been published and clearly proved its efficacy.12 Despite the fact that the catheter could be placed only in the common bile duct
(and not inside the right or left duct), only 5 of the 35 animals presented with histologic lesions 3 weeks after RF performed in close contact to the biliary bifurcation. This study showed that necrosis of tissue was complete around the biliary duct when vascular clamping was associated with RF, but was not complete around the bile duct without clamping. Despite this experimental finding, we observed in this series only one local recurrence after RF in the cooling site. Technically, introducing the catheter in the right duct can be difficult, and if the incision on the main bile duct is too low, a second one can be made as close as possible
Figure 4. Patient No. 1. (A) Magnetic resonance imaging aspect 3 months after radiofrequency. (B) Magnetic resonance imaging aspect 6 months after radiofrequency: note the bile duct dilatation upstream of the radiofrequency.
Vol. 198, No. 5, May 2004
to the main bifurcation. A good rule is to open the common bile duct as high as possible. Visualization of the catheter tip inside the biliary ducts under ultrasonography is difficult and requires experience. It would be very useful to develop specifically marked catheter tips. No complication was observed in relation to closure of the duct or with the expelled liquid. The other means of preventing biliary stenosis seems to be stent placement in the biliary tree through endoscopic retrograde pancreatography before RF. Bilchik recently reported 10 cases with no postablation injury; the stent was removed 4 to 6 weeks after the procedure.14 There is no information concerning the site of tumors or followup. This preventive maneuver necessitates two endoscopic procedures in addition to the intraoperative RF. It is impossible to conclude whether bile duct cooling is more efficient, less invasive, and less expensive than stenting. Only a prospective randomized study comparing the two approaches could answer this question. In conclusion, intraoperative cooling of the main bile ducts to avoid a stenosis is efficient and safe in humans. This procedure allows using RF to treat small-sized tumors sited less than 5 mm from the main bile ducts, increasing the number of patients in whom it is possible to propose an association of hepatectomy plus RF with curative intent. Author Contributions
Study conception and design: Elias Acquisition of data: Elias, Dromain, De Baere Analysis and interpretation of data: Elias, Sideris Drafting of manuscript: Elias, Sideris Critical revision: Sideris, Pocard Supervision: Elias
Elias et al
Hepatic Intraductal Cooling
721
REFERENCES 1. Curley SA, Izzo F, Delrio P, et al. Radiofrequency ablation of unresectable primary and metastatic hepatic malignancies. Results in 123 patients. Ann Surg 1999;230:1–8. 2. Elias D, Debaere T, Mutillo I, et al. Intraoperative use of radiofrequency treatment allows an increase in the rate of curative liver resection. J Surg Oncol 1998;67:190–191. 3. Siperstein A, Garland A, Engle K, et al. Local recurrence after laparoscopic radiofrequency thermal ablation of hepatic tumors. Ann Surg Oncol 2000;7:106–113. 4. Bilchik AJ, Wood TF, Allegra D, et al. Cryosurgical ablation and radiofrequency ablation for unresectable hepatic malignant neoplasm. Arch Surg 2000;135:657–664. 5. 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. Am J Roentgenol 2000;175:1619–1625. 6. Bilchik AJ, Rose DM, Allegra DP, et al. Radiofrequency ablation: a minimally invasive technique with multiple applications. Cancer J Sci Am 1999;5:356–361. 7. Dodd GD, Soulen MC, Kane R A, et al. Minimally invasive treatment of malignant hepatic tumors: at the threshold of a major breakthrough. Radiographics 2000;20:9–27. 8. Goldberg SN, Gazelle GS, Mueller PR. Thermal ablation therapy for focal malignacy: a unified approach to underlying principles, techniques and diagnosis imaging guidance. Am J Roentgenol 2000;174:323–331. 9. Elias D, El Otmany A, Goharin A, et al. Intraductal cooling of the main bile duct during intraoperative radiofrequency ablation. J Surg Oncol 2001;76:297–300. 10. Patterson EJ, Scudamore CH, Nagy AG, Buczkowsky AD. Radiofrequency ablation of porcine liver: in vivo effect of blood flow and treatment on lesion size. Ann Surg 1998;227:559–565. 11. Hansen PD, Rogers S, Corless CL, et al. Radiofrequency ablation lesions in a pig liver model. J Surg Research 1999;87:114– 121. 12. Marchal F, Elias D, Rauch P, et al. Biliary lesions during radiofrequency ablation in liver: study on the pig. Eur Surg Res. 2004;36:88–94. 13. Bilchik AJ, Wood TF, Allegra DP. Radiofrequency ablation of unresectable hepatic malignancies: lessons learned. Oncologist 2001;6:24–33. 14. Bilchik AJ. Thermal ablation of hepatic colorectal metastasis: should this minimally invasive technique replace resection? Educational Book, 39th annual meeting of the Am Soc Clin Oncol; 2003:699–710.
Radiofrequency (RF) eradicates liver tumors by heat. When used for hepatectomy, it is useful to treat with curative intent (R0 result) those many patients in whom the disease is usually considered unresectable.1-4 Heat damages the bile ducts, causing destruction, stenosis, and then upstream bile duct dilatation.5,6 The scientific community does not recommend RF for tumors closer than 15 to 20 mm to the main branches of the biliary tree.1-8 In 2001, we described the intraductal cooling of the main bile duct(s) during intraoperative RF to prevent biliary stenosis when the tumor was badly located.9 There were only 3 patients and the minimal followup was 3 months. In this article, we report a series of 13 patients whose tumors, closer than 6 mm to the main biliary ducts, were treated with this new procedure. Followup ranges from 6 to 42 months.
METHODS We used bile duct cooling in humans for the first time in December 1999. From then until April 2003, this technique was applied to 13 patients. During this period, 268 patients underwent a hepatectomy for a malignant lesion, and 83 of them also underwent intraoperative RF. Eleven patients were treated with intraoperative RF without hepatectomy. Intraoperative bile duct cooling was associated with 13.8% of all RF and 4.6% of all patients undergoing hepatectomy or RF during the same period. The only selection criterion was the presence of a tumor potentially treated by RF for which one edge was closer than or equal to 5 mm from a central bile duct. All these patients were treated with curative intent. Central bile ducts were defined as follows: on the left side, only the primary duct (up to the origin of the duct for segment III); on the right side, the primary duct and its two secondary branches (right anterior trunk for segments V–VIII, and right posterior trunk for segments VI–VII). The characteristics of the 13 patients are reported in Table 1. The mean age was 57 years (range 30 to 75 years), and there were 6 men and 7 women. The mean distance between the central bile duct and the tumor was
No competing interests declared.
Received October 29, 2003; Accepted December 3, 2003. From the Departments of Surgical Oncology (Elias, Sideris, Pocard) and Radiology (Dromain, De Baere), Institut Gustave Roussy, Cancer Center, Villejuif, Cedex, France. Correspondence address: D Elias, MD, Institut Gustave Roussy, Cancer Center, 39 rue Camille Desmoulins, Villejuif, Cedex 94805, France.
© 2004 by the American College of Surgeons Published by Elsevier Inc.
717
ISSN 1072-7515/04/$30.00 doi:10.1016/j.jamcollsurg.2003.12.026
718
Hepatic Intraductal Cooling
Elias et al
J Am Coll Surg
Table 1. Patient Characteristics Distance to bile duct (mm)
Site of cooling
Pt. No.
Origin
1 2 3 4 5 6
Hepatocellular Endocrine Colon Endocrine Colon Colon
No No No No No No
4 1 5 3 4 1
Left Left ⫹ right Right Right Left Left
7 8
Colon Rectum
Central (IV, V, VIII) No
0 0
Left ⫹ right Left ⫹ right
9 10 11
Colon Colon Colon
Left hepatectomy (II–IV) Right hepatectomy (V–VIII) No, but PC treated before
2 3 0
Right Left Right
12 13
Endocrine Colon
No Right hepatectomy ⫹ 1 left wedge then percutaneous left RF
4 5
Left Left
Preliminary hepatectomy
Associated surgery
No: severe cirrhosis Whipple, 4 metastasectomies, 4 other RF Left hepatectomy (II–IV), 2 RF in right liver Left hepatectomy (II–IV), 4 RF in right liver Right hepatectomy (V–VIII) ⫹ Segment I Right hepatectomy (V–VIII) ⫹ 3 left metastasectomies Segmentectomies VI ⫹ I Anterior resection rectum ⫹ left lobectomy (II–III) 2 right metastasectomies 1 left metastasectomy No. Severe adhesions after the treatment of the PC. Right hepatectomy (V–VIII) No
Bile duct stenosis
Yes No No No No No No No No No No No No
PC, peritoneal carcinomatosis treated by resection and intraperitoneal chemohyperthermia 21 months before; RF, radiofrequency.
2.2 mm (range 0 to 5 mm). Ten of these patients underwent a hepatectomy during the same session as RF with bile duct cooling. The three other patients presented only one liver tumor close to a central bile duct and did not undergo a hepatectomy for the following reasons: 1) severe cirrhosis, 2) severe peritoneal adhesions 21 months after the complete resection of a peritoneal carcinomatosis (from a colon cancer) associated with an intra-peritoneal chemohyperthermia, and 3) history of right hepatectomy plus one left metastasectomy 16 months earlier followed by two percutaneous RF in the left liver 8 months later. Finally, two patients were treated with intraarterial oxaliplatin allowing consideration of a curative procedure when it was initially not possible. Six coolings involved the left bile duct, four the right bile duct, and three both the right and left bile ducts because the tumor was located just above the biliary confluence. Technique of bile duct cooling
Ultrasonography and hepatectomy were done first. The common bile duct was isolated and opened longitudinally, as close as possible to the hepatic hilum. The common bile duct dissection was very difficult in the two cases of preliminary intraarterial hepatic chemotherapy
because of hard sclerosis of the whole hepatic pedicle. In four cases, a second incision was necessary, higher than the first one, at the level of the main bifurcation, to successfully install the catheter at the right place. This second incision was necessary in the two patients in whom a first catheter was placed in the right anterosuperior branch and a second one in the right posteroinferior branch. The 60-cm long, 2 mm-diameter catheters (Nutricath 2181; Vygon) were inserted at the right place under ultrasonographic guidance (Fig. 1). A delicate but rapid movement of the catheter allows vizualization of its tip inside the bile ducts. Care must be taken not to inject gas inside the biliary tree, because the tip will become invisible. When placed in the appropriate location, the catheter was temporarily fixed to the edge of the bile duct incision with a thin suture. A 4°C Ringer’s lactate perfusion was initiated to wash and cool the main bile duct(s) continuously during RF, and the liquid was expelled through the choledochal incision. The flow rate was as high as possible; the bag of Ringer’s lactate was pressured with an inflatable infusor (C-Infusor 500; Medex Inc) at 300 mmHg. In this series, we did not use concomitant vascular clamping during RF to keep the physiologic heat sink effect of the blood flow. At the end of the procedure, the incisions on the bile ducts were
Vol. 198, No. 5, May 2004
Elias et al
Hepatic Intraductal Cooling
719
was 6 months, and the median was 19.7 months (range 6 to 42 months). Twelve patients (92.3%) did not present any stenosis of the bile duct. Examples are reported in Figures 2 and 3. Only one patient (7.7%) presented segmental stenosis of the bile ducts above the site of RF 6 months after treatment. This was a cirrhotic patient presenting a 33-mm hepatocellular carcinoma in segment IV (Fig. 4). No treatment was done and there was no further complication. One patient (7.7%) presented a local recurrence on the RF site 4 months after the treatment. Figure 1. Diagram showing cooling of the main bile ducts, with one or two catheters introduced in a retrograde fashion after opening the common bile duct. C, common bile duct; I, perfusion with 4°C Ringer’s lactate; T, tumor.
closed with a continuous suture of a thin absorbable material without any specific drainage. RESULTS There was no mortality, and the morbidity rate was 31%, mainly because of lung atelectasis and right pleural effusion. There was no complication related to RF or cooling. All 13 patients underwent a CT scan of the liver before leaving the hospital to assess the result of the RF and to rule out any bile duct stenosis. Ultrasonography was performed between 3 and 4 weeks after operation to detect a dilatation of the biliary tree above the treated site. These two early explorations never detected any dilatation; followup was done every 3 months, as usual. The minimal followup after cooling for the entire series
DISCUSSION This series is the first to report successful late results for the cooling of main bile ducts to prevent biliary stenosis after heat damage after RF. Heat damages bile ducts and results in stenosis in pigs10,11 and in humans.5,6 This was studied later in pigs in which a specific experimental model of post-RF biliary stenosis was developed.12 This stenosis had no real impact when located peripherally, but could be dangerous when located centrally in the main bile ducts. These damages led the scientific community to contraindicate RF to destroy tumors closer than 15 to 20 mm to the central bile ducts. Two procedures eventually prevented biliary stenosis: the bile duct cooling we described in three patients in 2001,9 and the endoscopically placed internal stent, as reported for one patient by Bilchik and colleagues the same year.13 In this article, we report on 13 patients who presented with tumors located 0 to 5 mm from the main bile duct
Figure 2. Patient No. 7. (A) Aspect on CT scan of the recurrence after a central hepatectomy. (B) Result 6 months after radiofrequency plus cooling.
720
Elias et al
Hepatic Intraductal Cooling
J Am Coll Surg
Figure 3. Patient No. 2. (A) Preoperative magnetic resonance imaging. (B) Magnetic resonance imaging 3 months after radiofrequency with bile duct cooling. This patient with an endocrine tumor underwent a Whipple resection plus multiple wedge liver resections associated with radiofrequency in the left part of the liver. Three months later, he underwent a right hepatectomy.
with a strict definition of them. The preventive cooling allowed avoidance of biliary stenosis in 92% of our patients. Retrospectively, we did not find any explanation for the unique failure we observed, except that it occurred in a cirrhotic parenchyma. Also, the biliary dilatation was not clearly detectable 3 months after RF, but only 6 months later. This stenosis was well tolerated and did not necessitate any treatment. An experimental study performed on pigs about the preventive effect of bile duct cooling has recently been published and clearly proved its efficacy.12 Despite the fact that the catheter could be placed only in the common bile duct
(and not inside the right or left duct), only 5 of the 35 animals presented with histologic lesions 3 weeks after RF performed in close contact to the biliary bifurcation. This study showed that necrosis of tissue was complete around the biliary duct when vascular clamping was associated with RF, but was not complete around the bile duct without clamping. Despite this experimental finding, we observed in this series only one local recurrence after RF in the cooling site. Technically, introducing the catheter in the right duct can be difficult, and if the incision on the main bile duct is too low, a second one can be made as close as possible
Figure 4. Patient No. 1. (A) Magnetic resonance imaging aspect 3 months after radiofrequency. (B) Magnetic resonance imaging aspect 6 months after radiofrequency: note the bile duct dilatation upstream of the radiofrequency.
Vol. 198, No. 5, May 2004
to the main bifurcation. A good rule is to open the common bile duct as high as possible. Visualization of the catheter tip inside the biliary ducts under ultrasonography is difficult and requires experience. It would be very useful to develop specifically marked catheter tips. No complication was observed in relation to closure of the duct or with the expelled liquid. The other means of preventing biliary stenosis seems to be stent placement in the biliary tree through endoscopic retrograde pancreatography before RF. Bilchik recently reported 10 cases with no postablation injury; the stent was removed 4 to 6 weeks after the procedure.14 There is no information concerning the site of tumors or followup. This preventive maneuver necessitates two endoscopic procedures in addition to the intraoperative RF. It is impossible to conclude whether bile duct cooling is more efficient, less invasive, and less expensive than stenting. Only a prospective randomized study comparing the two approaches could answer this question. In conclusion, intraoperative cooling of the main bile ducts to avoid a stenosis is efficient and safe in humans. This procedure allows using RF to treat small-sized tumors sited less than 5 mm from the main bile ducts, increasing the number of patients in whom it is possible to propose an association of hepatectomy plus RF with curative intent. Author Contributions
Study conception and design: Elias Acquisition of data: Elias, Dromain, De Baere Analysis and interpretation of data: Elias, Sideris Drafting of manuscript: Elias, Sideris Critical revision: Sideris, Pocard Supervision: Elias
Elias et al
Hepatic Intraductal Cooling
721
REFERENCES 1. Curley SA, Izzo F, Delrio P, et al. Radiofrequency ablation of unresectable primary and metastatic hepatic malignancies. Results in 123 patients. Ann Surg 1999;230:1–8. 2. Elias D, Debaere T, Mutillo I, et al. Intraoperative use of radiofrequency treatment allows an increase in the rate of curative liver resection. J Surg Oncol 1998;67:190–191. 3. Siperstein A, Garland A, Engle K, et al. Local recurrence after laparoscopic radiofrequency thermal ablation of hepatic tumors. Ann Surg Oncol 2000;7:106–113. 4. Bilchik AJ, Wood TF, Allegra D, et al. Cryosurgical ablation and radiofrequency ablation for unresectable hepatic malignant neoplasm. Arch Surg 2000;135:657–664. 5. 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. Am J Roentgenol 2000;175:1619–1625. 6. Bilchik AJ, Rose DM, Allegra DP, et al. Radiofrequency ablation: a minimally invasive technique with multiple applications. Cancer J Sci Am 1999;5:356–361. 7. Dodd GD, Soulen MC, Kane R A, et al. Minimally invasive treatment of malignant hepatic tumors: at the threshold of a major breakthrough. Radiographics 2000;20:9–27. 8. Goldberg SN, Gazelle GS, Mueller PR. Thermal ablation therapy for focal malignacy: a unified approach to underlying principles, techniques and diagnosis imaging guidance. Am J Roentgenol 2000;174:323–331. 9. Elias D, El Otmany A, Goharin A, et al. Intraductal cooling of the main bile duct during intraoperative radiofrequency ablation. J Surg Oncol 2001;76:297–300. 10. Patterson EJ, Scudamore CH, Nagy AG, Buczkowsky AD. Radiofrequency ablation of porcine liver: in vivo effect of blood flow and treatment on lesion size. Ann Surg 1998;227:559–565. 11. Hansen PD, Rogers S, Corless CL, et al. Radiofrequency ablation lesions in a pig liver model. J Surg Research 1999;87:114– 121. 12. Marchal F, Elias D, Rauch P, et al. Biliary lesions during radiofrequency ablation in liver: study on the pig. Eur Surg Res. 2004;36:88–94. 13. Bilchik AJ, Wood TF, Allegra DP. Radiofrequency ablation of unresectable hepatic malignancies: lessons learned. Oncologist 2001;6:24–33. 14. Bilchik AJ. Thermal ablation of hepatic colorectal metastasis: should this minimally invasive technique replace resection? Educational Book, 39th annual meeting of the Am Soc Clin Oncol; 2003:699–710.