- Email: [email protected]
Partial splenic embolization in patients with idiopathic portal hypertension
European Journal of Radiology 49 (2004) 268–273
Partial splenic embolization in patients with idiopathic portal hypertension Maurizio Romano a,b,∗ , Angela Giojelli a , Gaetano Capuano c , Domenico Pomponi c , Marco Salvatore a a b
Department of Radiology and Radiotherapy, University ‘Federico II’, 5, Via Pansini, 80131 Naples, Italy Institute of Biostructures and Bioimages, National Research Council, 5, Via Pansini, 80131 Naples, Italy c Department of Hepatology, University ‘Federico II’, 5, Via Pansini, 80131 Naples, Italy Received 18 November 2002; received in revised form 7 April 2003; accepted 9 April 2003
Abstract Purpose: To evaluate the effectiveness of partial splenic embolization (PSE) in patients with idiopathic portal hypertension (IPH) in reducing variceal bleeding episodes, splenomegaly and thrombocytopenia. Materials and methods: Six patients (2M, 4F, mean age 30.3 years) with IPH presenting with splenomegaly, thrombocytopenia and recurrent variceal bleeding were treated with PSE using gelatin sponge (four patients) or Contour particles (two patients) as embolization material. Results: PSE was performed successfully in all cases; 3F coaxial microcatheters were necessary in two patients due to extreme splenic artery tortuosity. The average amount of devascularized parenchyma at CT 1 week after PSE was 71%. Splenomegaly and thrombocytopenia improved in all cases, with a mean platelet count increase of 120 000/mm3 and an average 68% reduction of spleen volume at follow up. Variceal bleeding did not recur after PSE. Esophageal or gastroesophageal varices disappeared (one patient) or significantly reduced (five patients) at endoscopic controls. No significant complications were noted. The follow up was of at least 18 months in all patients; mean follow up was 28.2 months. Conclusion: In patients with IPH PSE can be effective in preventing variceal bleedings, in reducing spleen volume and in significantly increasing platelet count; therapeutic results were durable in our population. © 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Portal hypertension; Embolization; Therapeutic; Radiology; Interventional
1. Introduction Idiopathic portal hypertension (IPH) is a disorder of unknown etiology characterized by the presence of splenomegaly, anemia and elevated values of portal pressure in the absence of cirrhosis, blood disease, parasites in the hepatobiliary system and hepatic and portal vein thrombosis; diagnosis of IPH can be made by excluding these related diseases [1–3]. Other diagnostic criteria include normal liver function tests, the presence of gastric or oesophageal varices, liver imaging not suggestive of cirrhosis [3]. Its incidence in western countries is relatively rare, being more frequent in asiatic populations. Even if etiology is actually unknown, various reports associate its pathogenesis with high serum levels of immunoglobulins, ∗ Corresponding author. Address: 3, Via Luca da Penne, 80122 Naples, Italy. Tel.: +39-081-746-3560; fax: +39-081-545-7081. E-mail address: [email protected] (M. Romano).
suggesting IPH to be an immunological disorder caused by continuous antigenic stimulation [3]. In the Japanese population women are reported to have a three times higher probability than men of having IPH [3]. Gastroesophageal varices with repeated episodes of bleeding are commonly observed in these patients, together with splenomegaly and anemia. Variceal massive bleeding is a frequent cause of death, and bleeding management is the most important therapeutic goal in IPH. IPH has been treated surgically with splenectomy and/or esophageal transection, with a reported survival rate up to 71% at 10 years [4]; however, a fairly high incidence of portal thrombosis has been reported after splenectomy [5], as well as the persistence of uncontrollable varices, ascites and hypertensive gastropathy [5,6]. Endoscopic variceal ligation is often performed many times in the lifetime of patients with IPH; however, ligation does not eliminate the primary cause of the portal hypertension, mainly due in these patients, in the initial
0720-048X/$ – see front matter © 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0720-048X(03)00134-7
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
phases of IPH, to increased splenic venous flow secondary to splenomegaly [7], and carries a significant rate of rebleeding despite treatment [8]. A number of interventional radiologic treatments are available for treating portal hypertension, namely partial splenic embolization (PSE), transjugular intrahepatic portosystemic shunt (TIPS), percutaneous trans-hepatic embolization of gastric varices. However, few reports are present in the literature of interventional radiological treatment in patients with IPH, and even fewer focusing on PSE in patients with early IPH. We have used PSE as the first treatment in six relatively young patients with IPH presenting with repeated episodes of variceal bleeding and without ascites and report here the results.
2. Materials and methods From April 1999 through June 2001 six patients were referred to the Department of Radiology of our Institution for the evaluation and therapy of IPH (2M, 4F, mean age 36.3 years, range 30–43). Patients were referred to us from the Gastroenterology and Hepatology Departments; all had been previously hospitalized for repeated episodes of bleeding from esophageal or gastroesophageal varices and all refused to undergo surgery or any further endoscopic procedure for their repeating episodes of variceal bleeding. Two patients had esophageal varices and four gastroesophageal varices at endoscopy, while two had a previous endoscopic ligation of the esophageal varices, with recurrent bleeding. Three patients had grade 3 varices and three patients grade 2 varices before PSE. In all six cases the diagnosis of IPH was posed relatively early, with a mean duration of disease of 23 months when PSE was performed. All patients had a liver biopsy performed between 4 and 14 months prior to PSE (average 10 months), and the presence of cirrhosis was excluded in all cases; regular liver parenchyma without fibrosis was reported in 4 biopsies. Mild periportal fibrosis was reported in two patients. All patients had huge splenomegaly and thrombocytopenia; blood diseases, presence of parasites, occlusion of hepatic or portal veins were excluded by serology, hematology, sonography and contrast enhanced spiral CT of the abdomen. Portal hypertension was confirmed in all patients by means of wedged hepatic vein pressure measurement; the corrected sinusoidal pressure averaged 18 mmHg (range 15–23). No patient presented with ascites. Liver function tests were within normal limits in all cases. All patients gave informed written consense to the procedure. Antibiotic profilaxis was started 24 h before the PSE using cefoperazone and gentamicin, and continued for 5 days after the procedure. The patients’ inguinal regions were thoroughly scrubbed with iodo-povidone solution the day before and 1 h before the PSE and strict aseptic technique was used throughout the procedure.
269
Under digital subtraction angiography (DSA) the celiac trunk was catheterized through a femoral approach and imaged for anatomic and flow assessment. PSE was performed by injecting embolizing material in the distal splenic artery after its catheterization to the splenic hilum using 4F hydrophilic catheters; in two cases the extreme tortuosity of the splenic artery required the use of a 3F hydrophilic coaxial microcatheter. The embolizing material consisted in gelatin sponge cubes or powder soaked in a solution of gentamycin; in two patients Contour particles with a mean diameter of 250 were used, dispersed in contrast medium and gentamycin solution. The splenic artery flow reduction was controlled by digital video dilution technique [9]. Briefly, the decrease in blood flow in the splenic artery as a percentage of the initially measured flow was calculated from the DSA images as follows: using a field of view of 18 cm, 6 frames per second and a bolus of 2 ml of contrast medium, DSA images were acquired centered on the splenic artery before and an average of two to three times during the embolization: images were digitally transferred to a nearby workstation, where two regions of interest (ROIs) were placed, respectively on the splenic artery and nearby but outside any opacified vessel for background correction. The passage of the bolus of the contrast medium was then recorded as a time–density curve; the area under the density curve has been demonstrated to be inversely proportional to the blood flow [10] through a proportionality constant. The ratio between the pre-embolization value of the area under the density curve and the value at any given point of the embolization gave the percentage flow reduction achieved at that moment. The embolizations were then continued so to achieve approximately 75–80% reduction of the splenic arterial flow. Patients were followed up with a contrast-enhanced spiral CT of the abdomen 1 week after PSE, and then with serologic, hematologic, physical examination and spiral CT of the abdomen at 1, 6, 12, 18 months after PSE. Endoscopy was performed in the follow up every 6 months. Total splenic volumes and non-infarcted splenic volumes were calculated from spiral CT datasets on a 3D workstation (Vitrea 2, Vital Images); ROIs were manually drawn around the whole spleen and around the vascularized portions of the spleen on a few slices and were then automatically interpolated on the rest of the datasets with little need for further editing. Total splenic, infarcted and vascularized splenic volumes were then automatically calculated by the software.
3. Results PSE was successfully performed in all patients. The achieved volume of devascularized spleen parenchyma measured by contrast enhanced spiral CT 1 week after the procedure ranged from 60–85%, with an average of 71%. No significant complication was observed following PSE; all patients had fever and mild left flank pain in the
270
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
Fig. 1. (a) Sterile colliquation of embolized parenchyma 1 week after PSE (white arrow), with spleen enlargement. Note significant reduction of splenic vein size after PSE (curved arrows) with respect to pre-PSE CT scan (b).
5–12 days following the procedure, well controlled with non-steroidal antiinflammatory drugs. One small left sided pleural effusion was noted 1 week after PSE, spontaneously regressed after 1 additional week, without the need for
thoracentesis. No cases of portal thrombosis have been observed. No patient developed a splenic abscess while only one patient experienced a fever lasting for 22 days. In this patient, enlargement of the spleen with sterile colliquation
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
271
Fig. 2. Same patient of Fig. 1: 1 year follow-up CT. Splenic volume significantly decreased. The spleen is irregular in shape due to the scarring of the embolized parenchyma.
of the embolized parenchyma was observed at control CT 1 week after PSE (Fig. 1); sterility of the colliquation was confirmed by percutaneous aspiration and cultures. The patient was followed up closely with ultrasound; antibiotic therapy was continued for additional 10 days and the spleen showed size reduction and no fluid collections at ultrasound 1 month post PSE. The 1 year control CT scan of this patient showing splenic volume reduction is shown in Fig. 2. Spleen volume reduction ranged from 59 to 78% 18 months after PSE as measured by contrast enhanced spiral CT (Fig. 3). A significant and progressive increase in platelet count has been observed in all patients; platelet count increased on average by 120 000/mm3 (range 95 000–145 000/mm3 ) at 18 months follow-up.
Esophageal varices disappeared in one patient; in the remaining five patients varices passed from grade 3 to grade 1 (two patients) and from grade 2 to grade 1 (three patients). No repeated episodes of variceal bleeding has been reported after the procedure in our patients. No patient required so far a repeated procedure. Results are summarized in Table 1.
4. Discussion Banti first described IPH as a distinct nosological entity from liver cirrhosis with splenomegaly [11]. Disease progression, histology and portal hemodynamics have been described to be different between IPH and cirrhosis and the pathogenesis of the former has been recently associated with
Fig. 3. Variation of splenic volumes after PSE measured at CT, expressed as a percentage of the pre-PSE volumes.
272
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
Table 1 Summary of results Patient
Gastric varices before PSE
Gastric varices 18 months follow-up
Oesophageal varices grade before PSE
Oesophageal varices grade 18 months follow-up
Platelets, before PSE (103 /mm3 )
Platelets 18 months follow-up (103 /mm3 )
Variceal bleeding episodes after PSE
1 2 3 4 5 6
+++ − +++ ++ − +++
+ − + + − ++
3 2 3 3 2 2
1 1 1 0 1 1
10.8 25.2 52.2 66.7 12.5 49.9
106.1 170.8 152.4 189.6 147.8 179.9
− − − − − −
immunological disorders. Okudaira et al. [3] reported high levels of immunoglobulins in the majority of patients with IPH, suggesting it to be an immunological disorder mediated by a continuous antigenic or superantigenic stimulation, indicating immunologic abnormalities as an etiologic factor in Japanese patients; they also reported that IPH incidence has markedly declined in recent years in Japan, indirectly suggesting a role of infection. In the early stages of the disease the increase in portal venous pressure is thought to be secondary mainly to the splenomegaly and relative increase of splenic venous flow; sclerotic changes in the intrahepatic proximal portal branches have been observed in the Japanese patients with long-lasting IPH [7,12–14] giving a possible late pathogenetic role to an increment in portal vascular resistance. Surgical splenectomy alone or in association with esophageal transection has been used in Japan in patients with IPH, with a reported 10 year survival rate of approximately 71% [4]; however, a high incidence of portal thrombosis has been reported following surgical splenectomy, up to 25% in patients with IPH [5]; the authors suggested that the total removal of the splenic venous contribution to the portal flow with subsequent reduction of blood flow in the portal vein could explain the relatively high thrombosis rate, while others highlighted the relationship between portal thrombosis and thrombocytosis after total removal of the spleen [15]. In early stages of IPH, when the increase of splenic venous flow rather than the increase of portal resistance is considered the leading cause of portal hypertension, PSE can be an extremely effective therapeutic modality [16], as confirmed by our results. Our patients’ age was relatively low when compared to the average onset age reported by the Japan IPH Research Committee [12] and none was in a late stage of disease; two of our patients underwent liver biopsy 4 and 8 months before PSE, resulting in mild periportal fibrosis, and the remaining four had a liver biopsy performed between 10 and 14 months before PSE, without evidence of sclerotic changes in the intrahepatic proximal portal branches. The principal cause of hypertension in the portal circulation was therefore removed or greatly diminished by PSE, resulting in an elevated therapeutic effect. Gastroesophageal varices disappeared in one subject and significantly reduced in size
in the rest of the population, with no rebleeding, together with significant spleen volume reduction and rise in platelets count at 18 months follow-up. Hirota et al. [16] described the results of different interventional radiologic procedures in patients with IPH, including PSE. They performed PSE, aiming to infarct 60% of spleen volume, as the only treatment in three patients 20, 23 and 60 years old, with excellent results and with a follow up duration comparable to our population’s. PSE was performed by these authors, respectively two times, three times and once in the three patients. Hoping to reduce the number of repeated procedures we aimed to embolize and infarct an higher percentage of splenic parenchyma (75–80%) during the first PSE. All of our patients showed to well tolerate the procedure even when more than 80% of splenic parenchyma resulted devascularized at control CT 1 week after the procedure, with complication rates comparable to previous literature reports. No patient experienced portal thrombosis due to excessive reduction of splenic venous flow. Until present none of our patients required a repeated PSE. The use of a quantitative or semi-quantitative mean of monitoring the blood flow reduction during the embolization, like the video-dilution technique used in our patients [9], appears to be a crucial step in PSE in order to prevent overembolization of splenic parenchyma that could lead to an increased rate of complications. In conclusion, PSE in patients with IPH can be an effective therapeutic modality, particularly in young patients and in patients in the early stages of the disease. In patients with more advanced disease, in which the increase of portal vascular resistance plays a significant role, PSE can be a good alternative when patients are considered inoperable, particularly in association with other interventional techniques like percutaneous variceal embolization and/or TIPS.
References [1] Boyer JL, Sen Gupta KP, Biswas SK, Pal NC, Basu Mallik KC, Iber FL, et al. Idiopathic portal hypertension. Comparison with the portal hypertension of cirrhosis and extrahepatic portal vein obstruction. Ann Intern Med 1967;66:41–68.
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273 [2] Okuda K, Nakashima T, Okudaira M, et al. Anatomic basis of hepatic venographic alterations in idiopathic portal hypertension. Liver 1981;1:255–63. [3] Okudaira M, Ohbu M, Okuda K. Idiopathic portal hypertension and its pathology. Semin Liver 2002;22(1):59–72. [4] Fukazawa M, Sugiura M. Portal hypertension. Oka H Sugiura M (Eds.), Rinsho kanzobyogaku 1989;:221–2Asakura ShotenTokyo. [5] Eguchi A, Hashizume M, Kitano S, Tanoue K, Wada H, Sugimachi K. High rate of portal thrombosis after splenectomy in patients with esophageal varices and idiopathic portal hypertension. Arch Surg 1991;126:752–5. [6] Ohashi K, Kojima K, Fukazawa M, Beppu T, Futagawa S. Long term prognosis of nonshunt operation for idiopathic portal hypertension. J Gastroenterol 1998;33:241–6. [7] Ohnishi K, Saito M, Sato S, Terabayashi H, Iida S, Nomura F, et al. Portal hemodynamics in idiopathic portal hypertension (Banti’s syndrome): comparison with chronic persistent hepatitis and normal subjects. Gastroenterology 1987;92:751–8. [8] Singh P, Pooran N, Indaram A, Bank S. Combined ligation and sclerotherapy versus ligation alone for secondary prophylaxis of esophageal variceal bleeding: a meta-analysis. Am J Gastroenterol 2002;97(3):623–9.
273
[9] Link DP, Seibert JA, Gould J, Lantz BMT. On-line monitoring of sequential blood flow reduction during splenic embolization. Acta Radiologica 1988;30:101–3. [10] Link DP, Foerster JM, Lantz BMT, Holcroft JW. Assessment of peripheral blood flow in man by video dilution technique. A preliminary report. Invest Radiol 1981;16(4):298–304. [11] Banti BC. Ueber morbus Banti. Folia Haematol 1910;10:33–74. [12] Okuda K, Nakajima T, Kameda H, Sugiura M, Ohnishi K, Kobayashi M. Idiopathic noncirrhotic portal hypertension: a national study. In: Popper H, Brunner H, Thaler H, editors. Hepatology. New York: Raven Press; 1984, p. 95–108. [13] Ohnishi K, Sato S, Nomura F, Iida S. Splanchnic hemodynamics in idiopathic portal hypertension: comparison with chronic persistent hepatitis. Am J Gastroenterol 1989;84:403–8. [14] Okuda K, Kono K, Ohnishi K, Kimura K, Omata M, Koen H, et al. Clinical study of eighty-six cases of idiopathic portal hypertension and comparison with cirrhosis with splenomegaly. Gastroenterology 1984;86:600–10. [15] Belli L, Rommani F, Riolo F. Thrombosis of the portal vein in the absence of hepatic disease. Surg Gynecol Obstet 1989;169:46–9. [16] Hirota S, Ichikawa S, Matsumoto S, Motohara T, Fukuda T, Yoshikawa T. Interventional radiologic treatment for idiopathic portal hypertension. Cardiovasc Intervent Radiol 1999;22:311–4.
Partial splenic embolization in patients with idiopathic portal hypertension Maurizio Romano a,b,∗ , Angela Giojelli a , Gaetano Capuano c , Domenico Pomponi c , Marco Salvatore a a b
Department of Radiology and Radiotherapy, University ‘Federico II’, 5, Via Pansini, 80131 Naples, Italy Institute of Biostructures and Bioimages, National Research Council, 5, Via Pansini, 80131 Naples, Italy c Department of Hepatology, University ‘Federico II’, 5, Via Pansini, 80131 Naples, Italy Received 18 November 2002; received in revised form 7 April 2003; accepted 9 April 2003
Abstract Purpose: To evaluate the effectiveness of partial splenic embolization (PSE) in patients with idiopathic portal hypertension (IPH) in reducing variceal bleeding episodes, splenomegaly and thrombocytopenia. Materials and methods: Six patients (2M, 4F, mean age 30.3 years) with IPH presenting with splenomegaly, thrombocytopenia and recurrent variceal bleeding were treated with PSE using gelatin sponge (four patients) or Contour particles (two patients) as embolization material. Results: PSE was performed successfully in all cases; 3F coaxial microcatheters were necessary in two patients due to extreme splenic artery tortuosity. The average amount of devascularized parenchyma at CT 1 week after PSE was 71%. Splenomegaly and thrombocytopenia improved in all cases, with a mean platelet count increase of 120 000/mm3 and an average 68% reduction of spleen volume at follow up. Variceal bleeding did not recur after PSE. Esophageal or gastroesophageal varices disappeared (one patient) or significantly reduced (five patients) at endoscopic controls. No significant complications were noted. The follow up was of at least 18 months in all patients; mean follow up was 28.2 months. Conclusion: In patients with IPH PSE can be effective in preventing variceal bleedings, in reducing spleen volume and in significantly increasing platelet count; therapeutic results were durable in our population. © 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Portal hypertension; Embolization; Therapeutic; Radiology; Interventional
1. Introduction Idiopathic portal hypertension (IPH) is a disorder of unknown etiology characterized by the presence of splenomegaly, anemia and elevated values of portal pressure in the absence of cirrhosis, blood disease, parasites in the hepatobiliary system and hepatic and portal vein thrombosis; diagnosis of IPH can be made by excluding these related diseases [1–3]. Other diagnostic criteria include normal liver function tests, the presence of gastric or oesophageal varices, liver imaging not suggestive of cirrhosis [3]. Its incidence in western countries is relatively rare, being more frequent in asiatic populations. Even if etiology is actually unknown, various reports associate its pathogenesis with high serum levels of immunoglobulins, ∗ Corresponding author. Address: 3, Via Luca da Penne, 80122 Naples, Italy. Tel.: +39-081-746-3560; fax: +39-081-545-7081. E-mail address: [email protected] (M. Romano).
suggesting IPH to be an immunological disorder caused by continuous antigenic stimulation [3]. In the Japanese population women are reported to have a three times higher probability than men of having IPH [3]. Gastroesophageal varices with repeated episodes of bleeding are commonly observed in these patients, together with splenomegaly and anemia. Variceal massive bleeding is a frequent cause of death, and bleeding management is the most important therapeutic goal in IPH. IPH has been treated surgically with splenectomy and/or esophageal transection, with a reported survival rate up to 71% at 10 years [4]; however, a fairly high incidence of portal thrombosis has been reported after splenectomy [5], as well as the persistence of uncontrollable varices, ascites and hypertensive gastropathy [5,6]. Endoscopic variceal ligation is often performed many times in the lifetime of patients with IPH; however, ligation does not eliminate the primary cause of the portal hypertension, mainly due in these patients, in the initial
0720-048X/$ – see front matter © 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0720-048X(03)00134-7
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
phases of IPH, to increased splenic venous flow secondary to splenomegaly [7], and carries a significant rate of rebleeding despite treatment [8]. A number of interventional radiologic treatments are available for treating portal hypertension, namely partial splenic embolization (PSE), transjugular intrahepatic portosystemic shunt (TIPS), percutaneous trans-hepatic embolization of gastric varices. However, few reports are present in the literature of interventional radiological treatment in patients with IPH, and even fewer focusing on PSE in patients with early IPH. We have used PSE as the first treatment in six relatively young patients with IPH presenting with repeated episodes of variceal bleeding and without ascites and report here the results.
2. Materials and methods From April 1999 through June 2001 six patients were referred to the Department of Radiology of our Institution for the evaluation and therapy of IPH (2M, 4F, mean age 36.3 years, range 30–43). Patients were referred to us from the Gastroenterology and Hepatology Departments; all had been previously hospitalized for repeated episodes of bleeding from esophageal or gastroesophageal varices and all refused to undergo surgery or any further endoscopic procedure for their repeating episodes of variceal bleeding. Two patients had esophageal varices and four gastroesophageal varices at endoscopy, while two had a previous endoscopic ligation of the esophageal varices, with recurrent bleeding. Three patients had grade 3 varices and three patients grade 2 varices before PSE. In all six cases the diagnosis of IPH was posed relatively early, with a mean duration of disease of 23 months when PSE was performed. All patients had a liver biopsy performed between 4 and 14 months prior to PSE (average 10 months), and the presence of cirrhosis was excluded in all cases; regular liver parenchyma without fibrosis was reported in 4 biopsies. Mild periportal fibrosis was reported in two patients. All patients had huge splenomegaly and thrombocytopenia; blood diseases, presence of parasites, occlusion of hepatic or portal veins were excluded by serology, hematology, sonography and contrast enhanced spiral CT of the abdomen. Portal hypertension was confirmed in all patients by means of wedged hepatic vein pressure measurement; the corrected sinusoidal pressure averaged 18 mmHg (range 15–23). No patient presented with ascites. Liver function tests were within normal limits in all cases. All patients gave informed written consense to the procedure. Antibiotic profilaxis was started 24 h before the PSE using cefoperazone and gentamicin, and continued for 5 days after the procedure. The patients’ inguinal regions were thoroughly scrubbed with iodo-povidone solution the day before and 1 h before the PSE and strict aseptic technique was used throughout the procedure.
269
Under digital subtraction angiography (DSA) the celiac trunk was catheterized through a femoral approach and imaged for anatomic and flow assessment. PSE was performed by injecting embolizing material in the distal splenic artery after its catheterization to the splenic hilum using 4F hydrophilic catheters; in two cases the extreme tortuosity of the splenic artery required the use of a 3F hydrophilic coaxial microcatheter. The embolizing material consisted in gelatin sponge cubes or powder soaked in a solution of gentamycin; in two patients Contour particles with a mean diameter of 250 were used, dispersed in contrast medium and gentamycin solution. The splenic artery flow reduction was controlled by digital video dilution technique [9]. Briefly, the decrease in blood flow in the splenic artery as a percentage of the initially measured flow was calculated from the DSA images as follows: using a field of view of 18 cm, 6 frames per second and a bolus of 2 ml of contrast medium, DSA images were acquired centered on the splenic artery before and an average of two to three times during the embolization: images were digitally transferred to a nearby workstation, where two regions of interest (ROIs) were placed, respectively on the splenic artery and nearby but outside any opacified vessel for background correction. The passage of the bolus of the contrast medium was then recorded as a time–density curve; the area under the density curve has been demonstrated to be inversely proportional to the blood flow [10] through a proportionality constant. The ratio between the pre-embolization value of the area under the density curve and the value at any given point of the embolization gave the percentage flow reduction achieved at that moment. The embolizations were then continued so to achieve approximately 75–80% reduction of the splenic arterial flow. Patients were followed up with a contrast-enhanced spiral CT of the abdomen 1 week after PSE, and then with serologic, hematologic, physical examination and spiral CT of the abdomen at 1, 6, 12, 18 months after PSE. Endoscopy was performed in the follow up every 6 months. Total splenic volumes and non-infarcted splenic volumes were calculated from spiral CT datasets on a 3D workstation (Vitrea 2, Vital Images); ROIs were manually drawn around the whole spleen and around the vascularized portions of the spleen on a few slices and were then automatically interpolated on the rest of the datasets with little need for further editing. Total splenic, infarcted and vascularized splenic volumes were then automatically calculated by the software.
3. Results PSE was successfully performed in all patients. The achieved volume of devascularized spleen parenchyma measured by contrast enhanced spiral CT 1 week after the procedure ranged from 60–85%, with an average of 71%. No significant complication was observed following PSE; all patients had fever and mild left flank pain in the
270
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
Fig. 1. (a) Sterile colliquation of embolized parenchyma 1 week after PSE (white arrow), with spleen enlargement. Note significant reduction of splenic vein size after PSE (curved arrows) with respect to pre-PSE CT scan (b).
5–12 days following the procedure, well controlled with non-steroidal antiinflammatory drugs. One small left sided pleural effusion was noted 1 week after PSE, spontaneously regressed after 1 additional week, without the need for
thoracentesis. No cases of portal thrombosis have been observed. No patient developed a splenic abscess while only one patient experienced a fever lasting for 22 days. In this patient, enlargement of the spleen with sterile colliquation
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
271
Fig. 2. Same patient of Fig. 1: 1 year follow-up CT. Splenic volume significantly decreased. The spleen is irregular in shape due to the scarring of the embolized parenchyma.
of the embolized parenchyma was observed at control CT 1 week after PSE (Fig. 1); sterility of the colliquation was confirmed by percutaneous aspiration and cultures. The patient was followed up closely with ultrasound; antibiotic therapy was continued for additional 10 days and the spleen showed size reduction and no fluid collections at ultrasound 1 month post PSE. The 1 year control CT scan of this patient showing splenic volume reduction is shown in Fig. 2. Spleen volume reduction ranged from 59 to 78% 18 months after PSE as measured by contrast enhanced spiral CT (Fig. 3). A significant and progressive increase in platelet count has been observed in all patients; platelet count increased on average by 120 000/mm3 (range 95 000–145 000/mm3 ) at 18 months follow-up.
Esophageal varices disappeared in one patient; in the remaining five patients varices passed from grade 3 to grade 1 (two patients) and from grade 2 to grade 1 (three patients). No repeated episodes of variceal bleeding has been reported after the procedure in our patients. No patient required so far a repeated procedure. Results are summarized in Table 1.
4. Discussion Banti first described IPH as a distinct nosological entity from liver cirrhosis with splenomegaly [11]. Disease progression, histology and portal hemodynamics have been described to be different between IPH and cirrhosis and the pathogenesis of the former has been recently associated with
Fig. 3. Variation of splenic volumes after PSE measured at CT, expressed as a percentage of the pre-PSE volumes.
272
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273
Table 1 Summary of results Patient
Gastric varices before PSE
Gastric varices 18 months follow-up
Oesophageal varices grade before PSE
Oesophageal varices grade 18 months follow-up
Platelets, before PSE (103 /mm3 )
Platelets 18 months follow-up (103 /mm3 )
Variceal bleeding episodes after PSE
1 2 3 4 5 6
+++ − +++ ++ − +++
+ − + + − ++
3 2 3 3 2 2
1 1 1 0 1 1
10.8 25.2 52.2 66.7 12.5 49.9
106.1 170.8 152.4 189.6 147.8 179.9
− − − − − −
immunological disorders. Okudaira et al. [3] reported high levels of immunoglobulins in the majority of patients with IPH, suggesting it to be an immunological disorder mediated by a continuous antigenic or superantigenic stimulation, indicating immunologic abnormalities as an etiologic factor in Japanese patients; they also reported that IPH incidence has markedly declined in recent years in Japan, indirectly suggesting a role of infection. In the early stages of the disease the increase in portal venous pressure is thought to be secondary mainly to the splenomegaly and relative increase of splenic venous flow; sclerotic changes in the intrahepatic proximal portal branches have been observed in the Japanese patients with long-lasting IPH [7,12–14] giving a possible late pathogenetic role to an increment in portal vascular resistance. Surgical splenectomy alone or in association with esophageal transection has been used in Japan in patients with IPH, with a reported 10 year survival rate of approximately 71% [4]; however, a high incidence of portal thrombosis has been reported following surgical splenectomy, up to 25% in patients with IPH [5]; the authors suggested that the total removal of the splenic venous contribution to the portal flow with subsequent reduction of blood flow in the portal vein could explain the relatively high thrombosis rate, while others highlighted the relationship between portal thrombosis and thrombocytosis after total removal of the spleen [15]. In early stages of IPH, when the increase of splenic venous flow rather than the increase of portal resistance is considered the leading cause of portal hypertension, PSE can be an extremely effective therapeutic modality [16], as confirmed by our results. Our patients’ age was relatively low when compared to the average onset age reported by the Japan IPH Research Committee [12] and none was in a late stage of disease; two of our patients underwent liver biopsy 4 and 8 months before PSE, resulting in mild periportal fibrosis, and the remaining four had a liver biopsy performed between 10 and 14 months before PSE, without evidence of sclerotic changes in the intrahepatic proximal portal branches. The principal cause of hypertension in the portal circulation was therefore removed or greatly diminished by PSE, resulting in an elevated therapeutic effect. Gastroesophageal varices disappeared in one subject and significantly reduced in size
in the rest of the population, with no rebleeding, together with significant spleen volume reduction and rise in platelets count at 18 months follow-up. Hirota et al. [16] described the results of different interventional radiologic procedures in patients with IPH, including PSE. They performed PSE, aiming to infarct 60% of spleen volume, as the only treatment in three patients 20, 23 and 60 years old, with excellent results and with a follow up duration comparable to our population’s. PSE was performed by these authors, respectively two times, three times and once in the three patients. Hoping to reduce the number of repeated procedures we aimed to embolize and infarct an higher percentage of splenic parenchyma (75–80%) during the first PSE. All of our patients showed to well tolerate the procedure even when more than 80% of splenic parenchyma resulted devascularized at control CT 1 week after the procedure, with complication rates comparable to previous literature reports. No patient experienced portal thrombosis due to excessive reduction of splenic venous flow. Until present none of our patients required a repeated PSE. The use of a quantitative or semi-quantitative mean of monitoring the blood flow reduction during the embolization, like the video-dilution technique used in our patients [9], appears to be a crucial step in PSE in order to prevent overembolization of splenic parenchyma that could lead to an increased rate of complications. In conclusion, PSE in patients with IPH can be an effective therapeutic modality, particularly in young patients and in patients in the early stages of the disease. In patients with more advanced disease, in which the increase of portal vascular resistance plays a significant role, PSE can be a good alternative when patients are considered inoperable, particularly in association with other interventional techniques like percutaneous variceal embolization and/or TIPS.
References [1] Boyer JL, Sen Gupta KP, Biswas SK, Pal NC, Basu Mallik KC, Iber FL, et al. Idiopathic portal hypertension. Comparison with the portal hypertension of cirrhosis and extrahepatic portal vein obstruction. Ann Intern Med 1967;66:41–68.
M. Romano et al. / European Journal of Radiology 49 (2004) 268–273 [2] Okuda K, Nakashima T, Okudaira M, et al. Anatomic basis of hepatic venographic alterations in idiopathic portal hypertension. Liver 1981;1:255–63. [3] Okudaira M, Ohbu M, Okuda K. Idiopathic portal hypertension and its pathology. Semin Liver 2002;22(1):59–72. [4] Fukazawa M, Sugiura M. Portal hypertension. Oka H Sugiura M (Eds.), Rinsho kanzobyogaku 1989;:221–2Asakura ShotenTokyo. [5] Eguchi A, Hashizume M, Kitano S, Tanoue K, Wada H, Sugimachi K. High rate of portal thrombosis after splenectomy in patients with esophageal varices and idiopathic portal hypertension. Arch Surg 1991;126:752–5. [6] Ohashi K, Kojima K, Fukazawa M, Beppu T, Futagawa S. Long term prognosis of nonshunt operation for idiopathic portal hypertension. J Gastroenterol 1998;33:241–6. [7] Ohnishi K, Saito M, Sato S, Terabayashi H, Iida S, Nomura F, et al. Portal hemodynamics in idiopathic portal hypertension (Banti’s syndrome): comparison with chronic persistent hepatitis and normal subjects. Gastroenterology 1987;92:751–8. [8] Singh P, Pooran N, Indaram A, Bank S. Combined ligation and sclerotherapy versus ligation alone for secondary prophylaxis of esophageal variceal bleeding: a meta-analysis. Am J Gastroenterol 2002;97(3):623–9.
273
[9] Link DP, Seibert JA, Gould J, Lantz BMT. On-line monitoring of sequential blood flow reduction during splenic embolization. Acta Radiologica 1988;30:101–3. [10] Link DP, Foerster JM, Lantz BMT, Holcroft JW. Assessment of peripheral blood flow in man by video dilution technique. A preliminary report. Invest Radiol 1981;16(4):298–304. [11] Banti BC. Ueber morbus Banti. Folia Haematol 1910;10:33–74. [12] Okuda K, Nakajima T, Kameda H, Sugiura M, Ohnishi K, Kobayashi M. Idiopathic noncirrhotic portal hypertension: a national study. In: Popper H, Brunner H, Thaler H, editors. Hepatology. New York: Raven Press; 1984, p. 95–108. [13] Ohnishi K, Sato S, Nomura F, Iida S. Splanchnic hemodynamics in idiopathic portal hypertension: comparison with chronic persistent hepatitis. Am J Gastroenterol 1989;84:403–8. [14] Okuda K, Kono K, Ohnishi K, Kimura K, Omata M, Koen H, et al. Clinical study of eighty-six cases of idiopathic portal hypertension and comparison with cirrhosis with splenomegaly. Gastroenterology 1984;86:600–10. [15] Belli L, Rommani F, Riolo F. Thrombosis of the portal vein in the absence of hepatic disease. Surg Gynecol Obstet 1989;169:46–9. [16] Hirota S, Ichikawa S, Matsumoto S, Motohara T, Fukuda T, Yoshikawa T. Interventional radiologic treatment for idiopathic portal hypertension. Cardiovasc Intervent Radiol 1999;22:311–4.