FROM THE DEPARTMENT OF DIAGNOSTIC RADIOLOGY, INSTITUTE OF PATHOLOGY AND THE WALLENBERG LABORATORY, UNIVERSITY OF LUND, S-221 85 LUND, SWEDEN.
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS AND EXPERIMENTAL HEPATIC TUMOURS IN THE RAT L. EKELUND, L. STIGSSON, N. JONSSON and H.-O. SJOGREN Primary as well as secondary malignant hepatic neoplasms receive their main blood supply from the hepatic artery (BREEDIS & YOUNG 1954). Experimentally as well as clinically it has been demonstrated that ligation of the hepatic artery is followed by necrosis of tumour tissue (NILSSON & ZETTERGREN 1967, ALMERSJO et colI. 1972). However, collateral circulation promptly develops following hepatic artery ligation (BENGMARK & ROSENGREN 1970, WIRTANEN & KAUDE 1973) and thus reduces the effect. In order to find out if necrosis of tumour tissue may also be induced by hepatic artery embolization and to what extent this procedure might injure normal liver parenchyma the present investigation was carried out using the rat as an experimental model.
Material and Methods Rats of the Wistar strain with a weight of about 200 g at the beginning of the experiment were used. The technique for transfemoral selective catheterization of the coeliac artery in the rat has been described previously (EKELUND & OLIN 1970). Occasionally it was also possible to catheterize the hepatic artery. Ether anaesthesia was employed and a small film changer for industrial film (ANGANTYR & OLIN 1973) was used for the angiography with a series of 1 filmjs for 6 s (FFD 45 em, focus 1.0 mm. Film: Agfa Gevaert D4; no intensifying screens. Exposure data: 80 kV, 20 mAs, 0.04 s). For the coeliac artery 0.3 to 0.5 ml of Isopaque Cerebral (Nyegaard, Norway), injected by hand, was used and 0.2 to 0.3 ml for the hepatic artery. FolSubmitted for publication 7 September 1976. Acta Radiologica Diagnosis 18 (1977) Fasc. 6 November 41 - 775847
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
641
642
L.
EKELUND, L. STIGSSON, N. JONSSON AND H.-O. SJOGREN
Fig. 1. Coeliac angiography (lateral projection) 30 min following ligation of the hepatic artery. The hepatic artery with branches is filled indicating collaterals.
lowing a primary angiography (in the lateral projection) 0.04 to 0.08 ml of a dispension of 1 mg Spongostan powder (99.3 % gelatin, Ferrosan International, Denmark) in 4 ml of contrast medium was injected into the coeliac or hepatic artery under fluoroscopy, immediately followed by a flush of saline to prevent occlusion of the catheter. Five min later repeat angiography was performed, usually followed by another angiography 45 to 60 min later. After these procedures the catheter was withdrawn and the femoral artery ligated. In 19 normal rats Spongostan was injected in the coeliac or hepatic artery. Eight rats died within 12 to 48 hours following embolization. Most of these rats had not received any prophylactic antibiotic therapy. In the later half of the series antibiotics (penicillin) were routinely administered following embolization with improved survival rates. Repeat angiography was performed in 11 normal rats at different time intervals following embolization (l to 8 days). The second catheterization was performed via the contralateral femoral artery; the third from the left carotid artery and eventually a fourth catheterization from the right carotid artery. Eleven rats were examined by angiography 14 to 22 days after intraportal injection with 106 to 106 living cells of a syngenic colon adenocarcinoma induced with Nmethyl-Nl-nitro-N-nitrosoguanidine (STEELE & SJOGREN 1974). After subsequent embolization with Spongostan two rats were re-examined after 5 and 10 days, respectively. Eight rats died within 4 days following embolization. One rat was kept as a control and not subjected to embolization. Angiography was performed 14 and 21 days following tumour cell inoculation in this particular animal.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
a
643
b
Fig. 2. a) Primary angiography. b) 30 min following embolization. Occlusion of hepatic as well as gastroduodenal artery. c) 5 days later. Catheterization from left carotid artery. Hepatic artery still partly occluded peripherally.
c
The same procedures were performed in 3 rats with hepatic tumours induced by intraportal inoculation with 106 living sarcoma cells of a syngenic polyoma virusinduced kidney sarcoma PW 13 (BANSAL et coli. 1972). At the end of the experiments the still surviving rats were killed and autopsy performed. Sections were prepared from normal as well as tumour livers and stained with hematoxylin-erythrosin, and according to van Gieson. Results
The hepatic artery was ligated near its origin in one normal rat. Thirty min after ligation angiography of the coeliac artery was performed, demonstrating faint con-
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
644
L. EKELUND, L. STIGSSON, N. JONSSON AND H.-O. SJOGREN
a
b
Fig. 3. a) Coeliac angiography 14 days following intraportal inoculation with 108 living cells from a colon carcinoma. Multiple small tumours peripherally. b) 45 min following peripheral embolization. No contrast medium reaches tumours. Inferior phrenic artery filled and contrast medium refluxed into aorta and superior mesenteric artery.
trast filling of the hepatic artery and its tributaries indicating collateralization (Fig. 1) Repeat angiography was performed in 11 normal rats 1 to 8 days following embolization with Spongostan. In two rats complete recanalization of the hepatic artery was found after 6 and 8 days, respectively. In 9 rats persistent occlusion to a varying degree could be demonstrated at angiography 1 to 8 days following embolization (Fig. 2). It thus appears that embolic occlusion by means of Spongostan powder lasts for a minimum of 1 to 8 days in most rats. Microscopy of the liver of one rat (2 hours after Spongostan injection) showed that the main portal veins were moderately dilated while the sinusoids and central veins of the lobules were slightly dilated. In another rat (4 days after injection) these changes were more evident with a marked distension of the portal veins. In 2 other rats (examined 5 and 6 days, respectively, after injection) the changes were again somewhat less marked. No necrosis of liver tissue was observed. All these rats had a persistent total or partial occlusion of the hepatic artery at angiography. Angiography was performed in 11 rats with liver tumours induced by intraportal inoculation of living cells from a colon carcinoma. The vascularization of these tumours varied somewhat at angiography, but abnormal vessels were always demonstrated and so was often pooling of contrast medium in vascular lakes. Tumour nodules with an avascular center and a hypervascularized surrounding rim were sometimes observed, similar to certain human liver metastases. At angiography following embolization with Spongostan it could be demonstrated that no contrast
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
a
645
b
Fig. 4. a) Coeliac angiography 21 days following intraportal inoculation with 108 colon carcinoma cells. Large liver and abundant tortuous tumour vessels. b) 5 min following embolization. Embolic material in main hepatic artery (-+). Markedly reduced tumour vascularization. c) 60 min following embolization. Central propagation of occlusion. No contrast medium reaches tumour. c
medium reached the neoplasms (Figs 3, 4). Eight rats died within 4 days following embolization and thus no repeat angiography could be performed. In one animal multiple tumour nodules with a diameter up to 5 mm were found at angiography. After embolization the hepatic as well as the gastroduodenal arteries were occluded. Five days later the liver had increased considerably in size as estimated from palpation but unfortunately the rat died during the second catheterization procedure and therefore no angiography could be performed in vivo. However, barium mixture was injected intraarterially and films of the removed liver exposed. The hepatic artery
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
646
L. EKELUND, L. STlGSSON, N. JONSSON AND H.-O. SJOGREN
a
b
c Fig. 5. a) Coeliac angiography (late phase, lateral projection) 14 days following intraportal inoculation with 108 colon carcinoma cells. Multiple small tumour nodules (-,)-). b) After embolization the hepatic as well as the gastroduodenal artery are occluded. c) 5 days later. Angiography of liver specimen. Hepatic arteries again open. Three small areas with neovascularity (-,)-).
was again open but only three smaller areas with neovascularity could be demonstrated (Fig. 5). Macroscopically multiple tumour nodules appeared on the liver surface of this animal. In another rat repeat angiography 10 days following embolization demonstrated recanalization of the hepatic artery and some increase in size of several poorly vascularized hepatic tumours.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
647
b
a
Fig. 6. a) 14 days following intraportal inoculation with 10 carcinoma cells. Pooling of contrast medium in one small tumour in cranial part of the liver. No embolization. b) One week later. The liver has increased in size and multiple areas with neovascularity are evident. Hepatic artery wider than at previous examination indicating increased flow. 8
One rat with hepatic tumours induced by intraportal inoculation with 106 colon carcinoma cells was kept as a control. Angiography 14 days after inoculation demonstrated pooling of contrast medium in one small tumour. No embolization was carried out. One week later repeat angiography demonstrated remarkable increase in liver size and multiple areas with tumour vascularity (Fig. 6). Three rats with liver tumours induced by intraportal inoculation with sarcoma cells were subjected to embolization. In one of these animals a poorly vascularized hepatic mass was found at angiography. After embolization the hepatic artery was totally occluded. As could be demonstrated 2 days later at repeat angiography the hepatic artery was partly recanalized and the mass had increased considerably in size, but was completely avascular (Fig. 7). In one rat with colonic adenocarcinoma deposits, microscopically examined 5 days after Spongostan injection, the larger tumour infiltrates contained extensive necroses, sometimes without any demonstrable remaining vital tumour tissue, sometimes with a narrow peripheral rim of vital tumour tissue. Smaller tumour deposits had only small central necroses and consisted mainly of vital tumour tissue. The liver tissue also contained fresh coagulation necroses of varying degree, especially in strands of liver tissue intervening between tumour deposits. In one rat, examined 10 days after Spongostan injection, the necroses of tumour deposits were less marked with comparatively more vital tumour tissue. Also in this animal small necroses in normal liver tissue were apparent. In the control rat (not subjected to embolization) necroses in tumour nodules were also evident although less extensive than in the experimental rat 5 days after Spongostan injection, and many tumour nodules consisted of completely vital tumour tissue. No necroses appeared in normal liver tissue.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
648
L. EKELUND, L. STIGSSON, N. JONSSON AND H.-O. SJOGREN
a
b
Fig. 7. a) 21 days following intraportal inoculation with 10· living sarcoma cells. Poorly vascularized mass in ventral part of the liver. b) The hepatic and gastroduodenal arteries are occluded. c) 8 days later. Partial recanalization. The mass has increased in size with compression of normal liver tissue but is completely avascular. c
In one rat with sarcoma deposits examined 2 days after Spongostan injection tumour nodules were nearly totally necrotic with only sparse vital cells in the periphery. On examination of another rat 3 days after injection small satellite nodules of vital tumour were present in the periphery of the large necrotic nodules. In both animals normal liver tissue showed fairly extensive necrosis. Discussion
Transcatheter selective arterial embolization has recently been introduced in the management of gastrointestinal bleeding and bleeding in the tumour patient (ROSCH et colI. 1972, GOLDSTEIN et colI. 1975). Embolic occlusion of renal carcinoma with autologous muscle tissue and gelfoam has been tried preoperatively in order to
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
649
decrease blood loss during surgery and for palliation of symptoms such as flank pain or hematuria (ALMGARD et colI. 1973, GOLDSTEIN et coll. 1975). In 1967 NILSSON & ZETTERGREN reported that ligation of the hepatic artery resulted in necrosis of induced liver carcinoma in rats. Tumour necrosis has also been reported following hepatic dearterialization in man, even if no effect with regard to survival was obtained (ALMERSJO et coll. 1972). The effect of dearterialization is based upon the fact that hepatic tumours receive their main blood supply from the hepatic artery, which has been demonstrated in animals as well as in man (BREEDIS & YOUNG 1954, NILSSON & ZETTERGREN 1967). It is also well known that ligation of the hepatic artery in man is promptly followed by the development of collateral circulation, e.g. from the inferior phrenic artery with ensuing resupply of the liver (BENGMARK & ROSENGREN 1970, WIRTANEN & KAUDE 1973). This was also demonstrated in one of the present rats that had been exposed to ligation of the hepatic artery. Previously a technique for catheterization of arteries in the rats was developed (EKELUND & OLIN 1970) which has proved suitable for the angiography of experimental hepatic tumours (EKELUND et coIl. 1974). Thus, it was found natural to apply the transcatheter embolization technique with rats as an experimental model. By establishing a more peripheral occlusion of the hepatic artery the risk of collateralization might be reduced. When comparing the angiograms of the rat with ligated hepatic artery and those of the rats subjected to embolization it was evident that collateral blood flow had been reduced in the latter group. From the experiments with normal rats it was also shown that vascular occlusion with Spongostan powder (99.3 % ge1atinthus similar to gelfoam) has a temporary effect, usually lasting for at least 1 to 8 days. The hepatic arterial circulation is then again restituted, which perhaps is of clinical importance as the risk of injury to the normal liver is reduced. This means that embolization may be repeated at different time intervals. Microscopy has confirmed that the portal circulation is sufficiently effective to prevent serious ischemic injury of the normal liver after hepatic artery occlusion. Even if the tumour material is small angiography clearly demonstrated that feeding of the tumours from the hepatic artery was initially arrested by embolization. Microscopy of some of these livers revealed extensive tumour necroses with small vital tumour residues. Rather wide-spread, obviously spontaneous necroses were, however, also evident in the rat kept as control. The microscopy also indicates a rather rapid appearance of small vital satellite nodules in the periphery of largely necrotic nodules. Obviously it is impossible to draw any conclusions from this small material upon the effect of Spongostan embolization, but the results could indicate that the procedure might be of value in inducing tumour necrosis. One drawback with the method is indicated by the microscopic findings, viz. the apparently greater sensitivity to anoxic injury of preserved liver tissue in tumour-infiltrated livers. Further experiments with larger series and control groups are necessary to evaluate the value and disadvantages of the method. Another drawback with this experimental model is the high mortality resulting
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
650
L. EKELUND,
L.
STIGSSON, N. JONSSON AND H.-O. SJOGREN
from embolization of the coeliac artery. It was found that in the fatal cases extensive embolization of the gastroduodenal artery had also occurred, leading to bowel necrosis. This is due to technical difficulties to catheterize selectively the hepatic artery in the rat, which is only occasionally possible. Most of the fatal cases also occurred before prophylactic antibiotic treatment was given as a routine following embolization. These technical catheterization difficulties do not exist in man, where a selective catheterization of the hepatic artery is quite possible in most cases. This suggests that transcatheter embolization in selected cases may be an alternative procedure to hepatic artery ligation, thus saving the patient from laparotomy. As demonstrated clinically, hepatic dearterialization as a sole procedure is not enough to control tumour growth within the liver (ALMERSJO et coil. 1972) and this is probably the case also with embolization. Therefore further experiments with the present model in combination with cytotoxic drugs and immunologic manipulations are necessary in order to evaluate the potential value of transcatheter arterial embolization of hepatic neoplasms.
SUMMARY Transcatheter hepatic arterial embolization with Spongostan (99.3 % gelatin) was performed in a group of normal rats. By repeat angiography could be demonstrated that arterial occlusions lasted for at least 1 to 8 days in most rats. Microscopy of these normal livers gave no evidence of parenchymatous liver injury. At postembolization angiography in a group of rats with experimental liver tumours it could be demonstrated that arterial supply of these tumours was temporarily completely arrested. Microscopy of some of these neoplasms revealed extensive necroses. The eventual future clinical applications of this procedure as an alternative to hepatic artery ligation is discussed.
ZUSAMMENFASSUNG Bei einer Gruppe von Ratten wurde eine Embolisierung der Leberarterie mit Spongostan (99,3 % Gelatine) durch ein Katheter durchgeflihrt. Durch wiederholte Angiographie konnte nachgewiesen werden, dass die Arterienverschltisse wenigstens 1 bis 8 Tagen bei den meisten Ratten anhielten. Die Mikroskopie dieser normalen Lebern gab keinen Anhalt flir eine parenchyrnatose Leberschadigung, Bei der Angiographie im Anschluss an eine Embolisierung bei einer Gruppe von Ratten mit experimentellen Lebertumoren konnte gezeigt werden, dass die arterielle Versorgung dieser Tumoren zeitweise vollstandig unterbrochen war. Die Mikroskopie einiger dieser Tumoren liess eine umfassende Nekrose erkennen. Die mogliche zuklinftige klinische Anwendung dieses Verfahrens urn die Unterbindung der Arteria hepatica zu ersatzen, wird diskutiert.
RESUME Une embolisation de l'artere hepatique par catheter avec Spongostan (99,3 % de gelatine) a ete effectuee sur un groupe de rats normaux. Des angiographies repetees ont montre que les occlusions arterielles durent au moins de 1 it 8 jours chez la plupart des rats. L'examen microscopique de ces foies normaux n'a pas montre de signes de lesion du parenchyme hepatique, L'angiographie apres embolisation sur un groupe de rats ayant des tumeurs
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
651
hepatiques experimentales a montre que la vascularisation arterielle de ces tumeurs est temporairement completement arretee, L'examen microscopique de certains de ces neoplasmes a montre des necroses etendues. Les auteurs examinent les applications cliniques futures eventuelles de cette technique pour remplacer la ligature de l'artere hepatique,
REFERENCES ALMERSJO 0., BENGMARK S., RUDENSTAM C. M., HAFSTROM L. O. and NILSSON L. A. V.: Evaluation of hepatic dearterialization in primary and secondary cancer of the liver. Amer. J. Surg. 124 (1972), 5. ALMGARD L. E., FERNSTROM I., HAVERLING M. and LJUNGQVIST A.: Treatment of renal adenocarcinoma by embolic occlusion of renal circulation. Brit. J. Urol. 45 (1973), 474. ANGANTYR L. G. and OLIN T.: Description of a film changer for small animals. Acta radiol. Diagnosis 14 (1973), 337. BANSAL S. c., HARGREAVES R. and SJOGREN H. 0.: Facilitation of polyoma tumor growth in rats by blocking sera and tumor eluate. Int. J. Cancer 9 (1972), 97. BENGMARK S. and ROSENGREN L.: Angiographic study of the collateral circulation to the liver after ligature of the hepatic arteries in man. Amer. J. Surg. 119 (1970), 620. BREEDIS C. and YOUNG G.: The blood supply of neoplasms in the liver. Amer. J. Pathol. 30 (1954),969. EKELUND L. and OLIN T.: Catheterization of arteries in rats. Invest. Radiol. 5 (1970), 69. - HENRIKSON H., OLIN T. and SJOGREN H. 0.: Angiography in hepatic cysts and tumors in the rat. Invest. Radiol. 9 (1974), 396. GOLDSTEIN H. M., MEDELLIN H., BEN-MENACHEM Y. and WALLACE S.: Transcatheter arterial embolization in the management of bleeding in the cancer patient. Radiology 115 (1975), 603. - - BEYDOUN M. T., WALLACE S., BEN-MENACHEM Y., BRACKEN R. B. and JOHNSON D. E.: Transcatheter embolization of renal cell carcinoma. Amer. J. RoentgenoI. 123 (1975), 557. NILSSON L. A. V. and ZETTERGREN K.: Blood supply and vascular pattern of induced primary hepatic carcinoma in rats. A microangiographic and histologic investigation. Acta path. microbiol. scand. 71 (1967), 179. - - Effect of hepatic artery ligation on induced primary liver carcinoma in rats. Acta path. microbiol. scand. 71 (1967), 187. ROSCH J., DOTTER C. T. and BROWN M. J.: Selective arterial embolization. New method for control of acute gastrointestinal bleeding. Radiology 102 (1972), 303. STEELE JR G. and SJOGREN H. 0.: Crossreacting tumor associated angigen(s) among chemically induced rat colon carcinomas. Cancer Res. 34 (1974), 1801. WIRTANEN G. W. and KAUDE J. V.: Inferior phrenic artery collateralization in hepatic artery occlusion. Amer. J. Roentgenol. 117 (1973), 615.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS AND EXPERIMENTAL HEPATIC TUMOURS IN THE RAT L. EKELUND, L. STIGSSON, N. JONSSON and H.-O. SJOGREN Primary as well as secondary malignant hepatic neoplasms receive their main blood supply from the hepatic artery (BREEDIS & YOUNG 1954). Experimentally as well as clinically it has been demonstrated that ligation of the hepatic artery is followed by necrosis of tumour tissue (NILSSON & ZETTERGREN 1967, ALMERSJO et colI. 1972). However, collateral circulation promptly develops following hepatic artery ligation (BENGMARK & ROSENGREN 1970, WIRTANEN & KAUDE 1973) and thus reduces the effect. In order to find out if necrosis of tumour tissue may also be induced by hepatic artery embolization and to what extent this procedure might injure normal liver parenchyma the present investigation was carried out using the rat as an experimental model.
Material and Methods Rats of the Wistar strain with a weight of about 200 g at the beginning of the experiment were used. The technique for transfemoral selective catheterization of the coeliac artery in the rat has been described previously (EKELUND & OLIN 1970). Occasionally it was also possible to catheterize the hepatic artery. Ether anaesthesia was employed and a small film changer for industrial film (ANGANTYR & OLIN 1973) was used for the angiography with a series of 1 filmjs for 6 s (FFD 45 em, focus 1.0 mm. Film: Agfa Gevaert D4; no intensifying screens. Exposure data: 80 kV, 20 mAs, 0.04 s). For the coeliac artery 0.3 to 0.5 ml of Isopaque Cerebral (Nyegaard, Norway), injected by hand, was used and 0.2 to 0.3 ml for the hepatic artery. FolSubmitted for publication 7 September 1976. Acta Radiologica Diagnosis 18 (1977) Fasc. 6 November 41 - 775847
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
641
642
L.
EKELUND, L. STIGSSON, N. JONSSON AND H.-O. SJOGREN
Fig. 1. Coeliac angiography (lateral projection) 30 min following ligation of the hepatic artery. The hepatic artery with branches is filled indicating collaterals.
lowing a primary angiography (in the lateral projection) 0.04 to 0.08 ml of a dispension of 1 mg Spongostan powder (99.3 % gelatin, Ferrosan International, Denmark) in 4 ml of contrast medium was injected into the coeliac or hepatic artery under fluoroscopy, immediately followed by a flush of saline to prevent occlusion of the catheter. Five min later repeat angiography was performed, usually followed by another angiography 45 to 60 min later. After these procedures the catheter was withdrawn and the femoral artery ligated. In 19 normal rats Spongostan was injected in the coeliac or hepatic artery. Eight rats died within 12 to 48 hours following embolization. Most of these rats had not received any prophylactic antibiotic therapy. In the later half of the series antibiotics (penicillin) were routinely administered following embolization with improved survival rates. Repeat angiography was performed in 11 normal rats at different time intervals following embolization (l to 8 days). The second catheterization was performed via the contralateral femoral artery; the third from the left carotid artery and eventually a fourth catheterization from the right carotid artery. Eleven rats were examined by angiography 14 to 22 days after intraportal injection with 106 to 106 living cells of a syngenic colon adenocarcinoma induced with Nmethyl-Nl-nitro-N-nitrosoguanidine (STEELE & SJOGREN 1974). After subsequent embolization with Spongostan two rats were re-examined after 5 and 10 days, respectively. Eight rats died within 4 days following embolization. One rat was kept as a control and not subjected to embolization. Angiography was performed 14 and 21 days following tumour cell inoculation in this particular animal.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
a
643
b
Fig. 2. a) Primary angiography. b) 30 min following embolization. Occlusion of hepatic as well as gastroduodenal artery. c) 5 days later. Catheterization from left carotid artery. Hepatic artery still partly occluded peripherally.
c
The same procedures were performed in 3 rats with hepatic tumours induced by intraportal inoculation with 106 living sarcoma cells of a syngenic polyoma virusinduced kidney sarcoma PW 13 (BANSAL et coli. 1972). At the end of the experiments the still surviving rats were killed and autopsy performed. Sections were prepared from normal as well as tumour livers and stained with hematoxylin-erythrosin, and according to van Gieson. Results
The hepatic artery was ligated near its origin in one normal rat. Thirty min after ligation angiography of the coeliac artery was performed, demonstrating faint con-
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
644
L. EKELUND, L. STIGSSON, N. JONSSON AND H.-O. SJOGREN
a
b
Fig. 3. a) Coeliac angiography 14 days following intraportal inoculation with 108 living cells from a colon carcinoma. Multiple small tumours peripherally. b) 45 min following peripheral embolization. No contrast medium reaches tumours. Inferior phrenic artery filled and contrast medium refluxed into aorta and superior mesenteric artery.
trast filling of the hepatic artery and its tributaries indicating collateralization (Fig. 1) Repeat angiography was performed in 11 normal rats 1 to 8 days following embolization with Spongostan. In two rats complete recanalization of the hepatic artery was found after 6 and 8 days, respectively. In 9 rats persistent occlusion to a varying degree could be demonstrated at angiography 1 to 8 days following embolization (Fig. 2). It thus appears that embolic occlusion by means of Spongostan powder lasts for a minimum of 1 to 8 days in most rats. Microscopy of the liver of one rat (2 hours after Spongostan injection) showed that the main portal veins were moderately dilated while the sinusoids and central veins of the lobules were slightly dilated. In another rat (4 days after injection) these changes were more evident with a marked distension of the portal veins. In 2 other rats (examined 5 and 6 days, respectively, after injection) the changes were again somewhat less marked. No necrosis of liver tissue was observed. All these rats had a persistent total or partial occlusion of the hepatic artery at angiography. Angiography was performed in 11 rats with liver tumours induced by intraportal inoculation of living cells from a colon carcinoma. The vascularization of these tumours varied somewhat at angiography, but abnormal vessels were always demonstrated and so was often pooling of contrast medium in vascular lakes. Tumour nodules with an avascular center and a hypervascularized surrounding rim were sometimes observed, similar to certain human liver metastases. At angiography following embolization with Spongostan it could be demonstrated that no contrast
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
a
645
b
Fig. 4. a) Coeliac angiography 21 days following intraportal inoculation with 108 colon carcinoma cells. Large liver and abundant tortuous tumour vessels. b) 5 min following embolization. Embolic material in main hepatic artery (-+). Markedly reduced tumour vascularization. c) 60 min following embolization. Central propagation of occlusion. No contrast medium reaches tumour. c
medium reached the neoplasms (Figs 3, 4). Eight rats died within 4 days following embolization and thus no repeat angiography could be performed. In one animal multiple tumour nodules with a diameter up to 5 mm were found at angiography. After embolization the hepatic as well as the gastroduodenal arteries were occluded. Five days later the liver had increased considerably in size as estimated from palpation but unfortunately the rat died during the second catheterization procedure and therefore no angiography could be performed in vivo. However, barium mixture was injected intraarterially and films of the removed liver exposed. The hepatic artery
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
646
L. EKELUND, L. STlGSSON, N. JONSSON AND H.-O. SJOGREN
a
b
c Fig. 5. a) Coeliac angiography (late phase, lateral projection) 14 days following intraportal inoculation with 108 colon carcinoma cells. Multiple small tumour nodules (-,)-). b) After embolization the hepatic as well as the gastroduodenal artery are occluded. c) 5 days later. Angiography of liver specimen. Hepatic arteries again open. Three small areas with neovascularity (-,)-).
was again open but only three smaller areas with neovascularity could be demonstrated (Fig. 5). Macroscopically multiple tumour nodules appeared on the liver surface of this animal. In another rat repeat angiography 10 days following embolization demonstrated recanalization of the hepatic artery and some increase in size of several poorly vascularized hepatic tumours.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
647
b
a
Fig. 6. a) 14 days following intraportal inoculation with 10 carcinoma cells. Pooling of contrast medium in one small tumour in cranial part of the liver. No embolization. b) One week later. The liver has increased in size and multiple areas with neovascularity are evident. Hepatic artery wider than at previous examination indicating increased flow. 8
One rat with hepatic tumours induced by intraportal inoculation with 106 colon carcinoma cells was kept as a control. Angiography 14 days after inoculation demonstrated pooling of contrast medium in one small tumour. No embolization was carried out. One week later repeat angiography demonstrated remarkable increase in liver size and multiple areas with tumour vascularity (Fig. 6). Three rats with liver tumours induced by intraportal inoculation with sarcoma cells were subjected to embolization. In one of these animals a poorly vascularized hepatic mass was found at angiography. After embolization the hepatic artery was totally occluded. As could be demonstrated 2 days later at repeat angiography the hepatic artery was partly recanalized and the mass had increased considerably in size, but was completely avascular (Fig. 7). In one rat with colonic adenocarcinoma deposits, microscopically examined 5 days after Spongostan injection, the larger tumour infiltrates contained extensive necroses, sometimes without any demonstrable remaining vital tumour tissue, sometimes with a narrow peripheral rim of vital tumour tissue. Smaller tumour deposits had only small central necroses and consisted mainly of vital tumour tissue. The liver tissue also contained fresh coagulation necroses of varying degree, especially in strands of liver tissue intervening between tumour deposits. In one rat, examined 10 days after Spongostan injection, the necroses of tumour deposits were less marked with comparatively more vital tumour tissue. Also in this animal small necroses in normal liver tissue were apparent. In the control rat (not subjected to embolization) necroses in tumour nodules were also evident although less extensive than in the experimental rat 5 days after Spongostan injection, and many tumour nodules consisted of completely vital tumour tissue. No necroses appeared in normal liver tissue.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
648
L. EKELUND, L. STIGSSON, N. JONSSON AND H.-O. SJOGREN
a
b
Fig. 7. a) 21 days following intraportal inoculation with 10· living sarcoma cells. Poorly vascularized mass in ventral part of the liver. b) The hepatic and gastroduodenal arteries are occluded. c) 8 days later. Partial recanalization. The mass has increased in size with compression of normal liver tissue but is completely avascular. c
In one rat with sarcoma deposits examined 2 days after Spongostan injection tumour nodules were nearly totally necrotic with only sparse vital cells in the periphery. On examination of another rat 3 days after injection small satellite nodules of vital tumour were present in the periphery of the large necrotic nodules. In both animals normal liver tissue showed fairly extensive necrosis. Discussion
Transcatheter selective arterial embolization has recently been introduced in the management of gastrointestinal bleeding and bleeding in the tumour patient (ROSCH et colI. 1972, GOLDSTEIN et colI. 1975). Embolic occlusion of renal carcinoma with autologous muscle tissue and gelfoam has been tried preoperatively in order to
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
649
decrease blood loss during surgery and for palliation of symptoms such as flank pain or hematuria (ALMGARD et colI. 1973, GOLDSTEIN et coll. 1975). In 1967 NILSSON & ZETTERGREN reported that ligation of the hepatic artery resulted in necrosis of induced liver carcinoma in rats. Tumour necrosis has also been reported following hepatic dearterialization in man, even if no effect with regard to survival was obtained (ALMERSJO et coll. 1972). The effect of dearterialization is based upon the fact that hepatic tumours receive their main blood supply from the hepatic artery, which has been demonstrated in animals as well as in man (BREEDIS & YOUNG 1954, NILSSON & ZETTERGREN 1967). It is also well known that ligation of the hepatic artery in man is promptly followed by the development of collateral circulation, e.g. from the inferior phrenic artery with ensuing resupply of the liver (BENGMARK & ROSENGREN 1970, WIRTANEN & KAUDE 1973). This was also demonstrated in one of the present rats that had been exposed to ligation of the hepatic artery. Previously a technique for catheterization of arteries in the rats was developed (EKELUND & OLIN 1970) which has proved suitable for the angiography of experimental hepatic tumours (EKELUND et coIl. 1974). Thus, it was found natural to apply the transcatheter embolization technique with rats as an experimental model. By establishing a more peripheral occlusion of the hepatic artery the risk of collateralization might be reduced. When comparing the angiograms of the rat with ligated hepatic artery and those of the rats subjected to embolization it was evident that collateral blood flow had been reduced in the latter group. From the experiments with normal rats it was also shown that vascular occlusion with Spongostan powder (99.3 % ge1atinthus similar to gelfoam) has a temporary effect, usually lasting for at least 1 to 8 days. The hepatic arterial circulation is then again restituted, which perhaps is of clinical importance as the risk of injury to the normal liver is reduced. This means that embolization may be repeated at different time intervals. Microscopy has confirmed that the portal circulation is sufficiently effective to prevent serious ischemic injury of the normal liver after hepatic artery occlusion. Even if the tumour material is small angiography clearly demonstrated that feeding of the tumours from the hepatic artery was initially arrested by embolization. Microscopy of some of these livers revealed extensive tumour necroses with small vital tumour residues. Rather wide-spread, obviously spontaneous necroses were, however, also evident in the rat kept as control. The microscopy also indicates a rather rapid appearance of small vital satellite nodules in the periphery of largely necrotic nodules. Obviously it is impossible to draw any conclusions from this small material upon the effect of Spongostan embolization, but the results could indicate that the procedure might be of value in inducing tumour necrosis. One drawback with the method is indicated by the microscopic findings, viz. the apparently greater sensitivity to anoxic injury of preserved liver tissue in tumour-infiltrated livers. Further experiments with larger series and control groups are necessary to evaluate the value and disadvantages of the method. Another drawback with this experimental model is the high mortality resulting
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
650
L. EKELUND,
L.
STIGSSON, N. JONSSON AND H.-O. SJOGREN
from embolization of the coeliac artery. It was found that in the fatal cases extensive embolization of the gastroduodenal artery had also occurred, leading to bowel necrosis. This is due to technical difficulties to catheterize selectively the hepatic artery in the rat, which is only occasionally possible. Most of the fatal cases also occurred before prophylactic antibiotic treatment was given as a routine following embolization. These technical catheterization difficulties do not exist in man, where a selective catheterization of the hepatic artery is quite possible in most cases. This suggests that transcatheter embolization in selected cases may be an alternative procedure to hepatic artery ligation, thus saving the patient from laparotomy. As demonstrated clinically, hepatic dearterialization as a sole procedure is not enough to control tumour growth within the liver (ALMERSJO et coil. 1972) and this is probably the case also with embolization. Therefore further experiments with the present model in combination with cytotoxic drugs and immunologic manipulations are necessary in order to evaluate the potential value of transcatheter arterial embolization of hepatic neoplasms.
SUMMARY Transcatheter hepatic arterial embolization with Spongostan (99.3 % gelatin) was performed in a group of normal rats. By repeat angiography could be demonstrated that arterial occlusions lasted for at least 1 to 8 days in most rats. Microscopy of these normal livers gave no evidence of parenchymatous liver injury. At postembolization angiography in a group of rats with experimental liver tumours it could be demonstrated that arterial supply of these tumours was temporarily completely arrested. Microscopy of some of these neoplasms revealed extensive necroses. The eventual future clinical applications of this procedure as an alternative to hepatic artery ligation is discussed.
ZUSAMMENFASSUNG Bei einer Gruppe von Ratten wurde eine Embolisierung der Leberarterie mit Spongostan (99,3 % Gelatine) durch ein Katheter durchgeflihrt. Durch wiederholte Angiographie konnte nachgewiesen werden, dass die Arterienverschltisse wenigstens 1 bis 8 Tagen bei den meisten Ratten anhielten. Die Mikroskopie dieser normalen Lebern gab keinen Anhalt flir eine parenchyrnatose Leberschadigung, Bei der Angiographie im Anschluss an eine Embolisierung bei einer Gruppe von Ratten mit experimentellen Lebertumoren konnte gezeigt werden, dass die arterielle Versorgung dieser Tumoren zeitweise vollstandig unterbrochen war. Die Mikroskopie einiger dieser Tumoren liess eine umfassende Nekrose erkennen. Die mogliche zuklinftige klinische Anwendung dieses Verfahrens urn die Unterbindung der Arteria hepatica zu ersatzen, wird diskutiert.
RESUME Une embolisation de l'artere hepatique par catheter avec Spongostan (99,3 % de gelatine) a ete effectuee sur un groupe de rats normaux. Des angiographies repetees ont montre que les occlusions arterielles durent au moins de 1 it 8 jours chez la plupart des rats. L'examen microscopique de ces foies normaux n'a pas montre de signes de lesion du parenchyme hepatique, L'angiographie apres embolisation sur un groupe de rats ayant des tumeurs
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
TRANSCATHETER ARTERIAL EMBOLIZATION OF NORMAL LIVERS
651
hepatiques experimentales a montre que la vascularisation arterielle de ces tumeurs est temporairement completement arretee, L'examen microscopique de certains de ces neoplasmes a montre des necroses etendues. Les auteurs examinent les applications cliniques futures eventuelles de cette technique pour remplacer la ligature de l'artere hepatique,
REFERENCES ALMERSJO 0., BENGMARK S., RUDENSTAM C. M., HAFSTROM L. O. and NILSSON L. A. V.: Evaluation of hepatic dearterialization in primary and secondary cancer of the liver. Amer. J. Surg. 124 (1972), 5. ALMGARD L. E., FERNSTROM I., HAVERLING M. and LJUNGQVIST A.: Treatment of renal adenocarcinoma by embolic occlusion of renal circulation. Brit. J. Urol. 45 (1973), 474. ANGANTYR L. G. and OLIN T.: Description of a film changer for small animals. Acta radiol. Diagnosis 14 (1973), 337. BANSAL S. c., HARGREAVES R. and SJOGREN H. 0.: Facilitation of polyoma tumor growth in rats by blocking sera and tumor eluate. Int. J. Cancer 9 (1972), 97. BENGMARK S. and ROSENGREN L.: Angiographic study of the collateral circulation to the liver after ligature of the hepatic arteries in man. Amer. J. Surg. 119 (1970), 620. BREEDIS C. and YOUNG G.: The blood supply of neoplasms in the liver. Amer. J. Pathol. 30 (1954),969. EKELUND L. and OLIN T.: Catheterization of arteries in rats. Invest. Radiol. 5 (1970), 69. - HENRIKSON H., OLIN T. and SJOGREN H. 0.: Angiography in hepatic cysts and tumors in the rat. Invest. Radiol. 9 (1974), 396. GOLDSTEIN H. M., MEDELLIN H., BEN-MENACHEM Y. and WALLACE S.: Transcatheter arterial embolization in the management of bleeding in the cancer patient. Radiology 115 (1975), 603. - - BEYDOUN M. T., WALLACE S., BEN-MENACHEM Y., BRACKEN R. B. and JOHNSON D. E.: Transcatheter embolization of renal cell carcinoma. Amer. J. RoentgenoI. 123 (1975), 557. NILSSON L. A. V. and ZETTERGREN K.: Blood supply and vascular pattern of induced primary hepatic carcinoma in rats. A microangiographic and histologic investigation. Acta path. microbiol. scand. 71 (1967), 179. - - Effect of hepatic artery ligation on induced primary liver carcinoma in rats. Acta path. microbiol. scand. 71 (1967), 187. ROSCH J., DOTTER C. T. and BROWN M. J.: Selective arterial embolization. New method for control of acute gastrointestinal bleeding. Radiology 102 (1972), 303. STEELE JR G. and SJOGREN H. 0.: Crossreacting tumor associated angigen(s) among chemically induced rat colon carcinomas. Cancer Res. 34 (1974), 1801. WIRTANEN G. W. and KAUDE J. V.: Inferior phrenic artery collateralization in hepatic artery occlusion. Amer. J. Roentgenol. 117 (1973), 615.
Downloaded from acr.sagepub.com at University of Manitoba Libraries on September 9, 2015
