Classification of Hepatic Venous Outflow Obstruction: Ambiguous Terminology of the Budd-Chiari Syndrome
JURGEN LUDWIG, M.D., ETSUKO HASHIMOTO, M.D., Division of Pathology; DOUGLAS B. McGILL, M.D., Division of Gastroenterology and Internal Medicine; JON A. van HEERDEN, M.B.,Ch.B., Section of Gastroenterologic and General Surgery
Severe hepatic venous outflow obstruction and its manifestations often are recorded under the label "Budd-Chiari syndrome." Unfortunately, this label is ambiguous; it doesnot clearly identify the site of the lesion (hepatic veins versus inferior vena cava), its morphologic features (thrombotic versus nonthrombotic), or its cause. In the literature, implied or expressed definitions vary. Use of a standardized topographic and pathogenetic classification of hepatic venous outflow obstruction would enable investigators to group patients with comparable conditions, as required for therapeutic trials, prognostic evaluations, and studies of pathogenetic pathways. Review of our own cases revealed that hepatic venous outflow obstruction involving large hepatic veins is usually thrombotic and that isolated obstruction of the inferior vena cava or of small hepatic veins is usually nonthrombotic. Application of such a classification seems feasible and may yield useful results.
The "Budd-Chiari syndrome" (BCS) is a loosely defined designation for hepatic venous outflow obstruction (HVOO). Some authors equate BCS with hepatic large-vein thrombosis.l" but most writers use the term BCS for both thrombotic and nonthrombotic obstruction of the large hepatic veins and the inferior vena cava (IVC) between the liver and the right atrium. Some authors diagnose BCS even if the obstruction is within the right atrium," whereas others specifically exclude conditions at this level-for instance, obstruction of the thoracic IVC by constrictive pericarditis.P-" Thus, BCS is an ambiguous term because it does not clearly identify the morphologic features of the obstruction or
the site involved. The ambiguity is compounded by the controversial relationship between BCS and hepatic veno-occlusive disease (HVOD). Traditionally, the latter is used to designate nonthrombotic obliteration of small hepatic veins as observed after pyrrolizidine poisoning and, more recently, after high-dose antineoplastic chemotherapy." No specific term has been coined for small-vein thrombosis, although a few authors have considered small-vein obstruction of both types as manifestations of HVOD and part of the BCS.2,8,9 In their original publications, Budd 10 and Chiari!' described inflammation of major hepatic veins or their ostia, which the authors ascribed to sepsis and alcoholism'? or syphilis.!' prevailing conditions at the time. Their observations no Address reprint requests to Dr. Jurgen Ludwig, Division of longer can be used as yardsticks for current concepts of the syndrome. Indeed, not all of Pathology, Mayo Clinic, Rochester, MN 55905. Mayo Clin Proc 65:51-55, 1990
51
52
HEPATIC VENOUS OUTFLOW OBSTRUCTION
Mayo Clin Proc, January 1990, Vol 65
Budd's three cases may have had the syndrome- DIAGNOSIS OF BCS a common experience in the study of classic In our opinion, the diagnosis of BCS should be used sparingly; this label is appropriate primardescriptions. In recent articles concerning the management ily for cases ofHVOO before the type and location of BCS, 12-15 some information is provided about ofthe obstruction are known. For the treatment topographic, morphologic, and etiologic features of individual patients, for therapeutic trials, and of the lesions; nevertheless, results still are for other retrospective and prospective studies difficult to interpret because findings usually ofHVOO, a more detailed pathogenetic classifiare discussed collectively under the label BCS. cation is needed. For any such classification, the This situation is further complicated because components of the hepatic venous outflow tract readers often are not told which definition of must be well defined by radiologic and other BCS has been applied to selection and exclusion clinical-diagnostic findings such as the results of of cases. This confusion underscores the limited color Doppler ultrasonography. The anatomic usefulness of the term BCS, but the name un- components are shown in Table 1. doubtedly is here to stay and, therefore, needs to Each case of HVOO should be defined by (1) the morphologic characteristics of the obstrucbe defined. tion-specifically, whether the lesions were DEFINITION OF BCS thrombotic or nonthrombotic; (2) the location of Review of the literature and our own experi- the obstructed venous segment, based on the ences suggests that the following definition of terms defined in Table 1; and (3) the substantiated or putative etiologic factors. Use of this BCS currently would be the most useful: approach results in a classification that is shown BCS consists of hepatic venous outflow obstruction and its in Table 2 and applied to an actual patient cohort manifestations, regardless of cause, the obstruction being in Table 3. either within the liver or in the lVC between the liver and the right atrium. Functional hepatic venous outflow obstruction caused by congestive heart failure is not considered BCS.
Commonly, heart disease other than constrictive pericarditis is excluded from the list of causes of BCS. At the same time, occlusion of small hepatic veins, including classic HVOD, has become part of our definition of the syndrome. This schema may offend traditionalists because, as mentioned earlier, the men credited with discovery of the syndrome 10 ,1l described neither obstruction of the IVC nor HVOD. The difficult and controversial distinction between thrombotic and nonthrombotic (classic HVOD) occlusion of small hepatic veins and the common secondary involvement of small hepatic veins in large-vein disease, however, make it impractical to consider small-vein disease as a separate entity. Whatever the final wording may be, a definition of BCS can find wide acceptance only if it is rather broad and therefore similar to the foregoing proposed definition.
APPLICATION OF CLASSIFICATION For preparation of Table 3 (and thus testing of the proposed classification), we retrieved all Mayo Clinic cases of BCS and HVOO that had been recorded during the 12-year period from 1976 to 1987. We identified 62 patients who fulfilled the previously stated definition ofBCS; some patients from this group had already been described in a previous publication from this institution.!? In 21 instances, morphologic data were unobtainable or insufficient; therefore, these patients were excluded from the initial analysis. Thus, data from 41 patients could be reviewed in detail for the classification shown in Table 3. The results can be summarized as follows. The most common morphologic feature was thrombosis, unrelated to another structural abnormality, and the most common location was the large hepatic veins. Of the 21 patients with this condition, 15 had thrombosis of hepatic veins only, and 6 had thrombosis both of hepatic veins and of the adjacent IVC. In an additional
Mayo Clin Proc, January 1990, Vol 65
HEPATIC VENOUS OUTFLOW OBSTRUCTION
53
Table I.-Clinical Anatomy of Hepatic Venous Outflow System Designation*
Definition
Small hepatic veins
Veins that cannot be shown clearly in hepatic venograms or by ultrasound studies; they include terminal hepatic veins (central veins), intercalated veins, and interlobular veins (collecting veins)'] Veins that are regularly demonstrable in hepatic venograms and ultrasound studies; segmental hepatic vein branches generally are included A segment of the IVe that involves the entire area that is in contact with the right lobe and the caudate lobe of the liver up to the entry level of the right, middle, and left hepatic veins A segment of the IVe that extends from the entry level of the right, middle, and left hepatic veins to the junction between the Ive and the right atrium
Large hepatic veins Hepatic inferior vena cava (hepatic IVe) Suprahepatic IVe
*These designations weredefined for the purpose ofthis review; names and definitions in other publications may differ. tUniform morphologic definitions for the names of the vessels between terminal hepatic veins and veins draining hepatic segments have not been provided in the medical literature.
five patients, thrombosis of the large and small hepatic veins and the abdominal IVe had been a complication of carcinoma in the vicinity (three cases) or kinking or stricture of the suprahepatic IVe (two cases). Isolated thrombosis of the hepatic and suprahepatic IVe was not observed, although the thrombosis in the two cases with stricture or kinking ofthe IVe undoubtedly had begun in this manner. For instance, one ofthese two patients had had a thrombosis at the bifurcation of the IVe that had developed after orthotopic liver transplantation. An embolus had
broken off that clot, but it was arrested by the tight upper anastomosis of the IVe and subsequently resulted in clotting of the adjacent large hepatic veins. This observation notwithstanding, isolated thrombosis of the subdiaphragmatic IVe seems to be a rare event. Nonthrombotic HVOO was found in 15 patients (Table 3). The distribution of most lesions differed strikingly from that of the thrombotic occlusions. Thus, five patients had nonthrombotic obstruction of the Ive only, and seven patients had nonthrombotic obstruction of the
Table 2.-Topographic and Pathogenetic Classification of Hepatic Venous Outflow Obstruction: Principal Categories Site of obstruction Small hepatic veins Large hepatic veins Hepatic and suprahepatic inferior vena cava
Thrombotic
Nonthrombotic
Etiologic factors: idiopathic thrombosis or thrombosis resulting from coagulopathy of known or suspected cause or thrombosis complicating a nonthrombotic abnormality of the hepatic venous outflow tract
Etiologic factors: intravenous webs or membranes, tumor, or other obstructive abnormalities of the hepatic venous outflow tract; usually well-described morphologic features, but cause may be unknown. Some webs or membranes in the inferior vena cava might be postthrombotic!"
54
HEPATIC VENOUS OUTFLOW OBSTRUCTION
Mayo CUn Proc, January 1990, Vol 65
Table 3.-Topographic and Pathogenetic Classification of Hepatic Venous Outflow Obstruction Applied to the Study of 41 Patients* Morphologic features of obstruction and etiologic factor Thrombotic (N Site of obstruction
Thrombosis only (N = 21)
= 26) Complicating a nonthrombotic condition (N = 5)
Thoracic IVC Abdominal IVC only Abdominal IVC, large HV Abdominal IVC, large HV, small HV Large HV only
Large HV, small HV
Small HV only
Idiopathic (2); oral contraceptive use (1); lupus anticoagulant (1) Oral contraceptive use (2) Idiopathic (4); oral contraceptive use (2); paroxysmal nocturnal hemoglobinuria (2) Idiopathic (1); polycythemia vera (2); paroxysmal nocturnal hemoglobinuria (2); agnogenic myeloid metaplasia (1); protein C deficiency (1)
Stricture or kinking of IVC after liver transplantation (2); carcinoma in the vicinity (2) Carcinoma in the vicinity (1)
Nonthrombotic (N = 15) Constrictive pericarditis (3); sclerosing mediastinitis (1) Renal cell carcinoma (1) Fibrous membrane of IVC and large HV (1)
Fibrous web at juncture with IVC (2)
Antineoplastic agents for malignant lesion (2); bone marrow or renal transplantation (2); idiopathic (oral contraceptive use [?], alcoholic liver disease [?]) (2); herbal tea (1)
*Figures in parentheses indicate numbers of patients in each etiologic category. HV = hepatic veins; IVC = inferior vena cava.
small hepatic veins only, identical or comparable to classic HVOD. Two patients had membranes at the IVC orifices of the large hepatic veins, and one patient had fibrous membranes both at these orifices and within the IVC. Among the patients with HVOO, the most common etiologic category was idiopathic thrombosis (seven cases or 17%). In five of these seven cases, only the liver was involved; the IVC apparently had remained patent. The next most common group was represented by five women who had been taking oral contraceptives; only the hepatic veins were involved in two of them, but in three patients, the thrombosis was also found in the IVC. Paroxysmal nocturnal hemo-
globinuria (four cases), polycythemia vera (two cases), protein C deficiency (one case), and agnogenic myeloid metaplasia (one case) were associated with hepatic vein thrombosis alone. One other patient had thrombosis of large hepatic veins and of the IVC; in this instance, a lupus anticoagulant was found. The etiologic factors involved in the nonthrombotic cases of HVOO are listed in Table 3.
DISCUSSION Although most clinicians implicitly use in their publications a classification similar to the one described herein, that information rarely is communicated. We believe that use of a stan-
Mayo Clin Proc, January 1990, Vol 65
HEPATIC VENOUS OUTFLOW OBSTRUCTION
55
dardized classification would facilitate thera- REFERENCES 1. Brinson RR, Curtis WD, Schuman BM, Mills LR: peutic trials, prognostic evaluations, and study Recovery from hepatic vein thrombosis (Budd-Chiari of pathogenetic pathways. Obviously, for pasyndrome) complicating ulcerative colitis. Dig Dis Sci 33:1615-1620, 1988 tient management, many other factors should be 2. McClure S, Dincsoy HP, Glueck H: Budd-Chiari considered, such as the degree of liver dysfuncsyndrome and antithrombin III deficiency. Am J Clin tion and the duration of the disease. For diagPathol 78:236-241, 1982 3. Disney TF, Sullivan SN, Haddad RG, Lowe D, nostic considerations, the data shown in Table 3 Goldbach MM: Budd-Chiari syndrome with inferior are important; they clearly support the experivena cava obstruction associated with systemic lupus ence that thrombotic HVOO is located primarily erythematosus. J Clin Gastroenterol 6:253-256, in the large hepatic veins and that thrombosis of 1984 4. Feingold ML, Litwak RL, Geller SS, Baron MM: the IVC or the small hepatic veins usually occurs Budd-Chiari syndrome caused by a right atrial tuby extension from the large hepatic veins. mor. Arch Intern Med 127:292-295,1971 Conversely, nonthrombotic HVOO involves 5, Solano FX Jr, Young E, Talamo TS, Dekker A: Constrictive pericarditis mimicking Budd-Chiari synalmost exclusively the IVC, and the small hedrome. Am J Med 80:113-115,1986 patic veins in the few instances of classic HVOD. 6. Dahms BB: Case 5: hepatic veno-occlusive disease We considered the orifices of the large hepatic with clinical Budd-Chiari syndrome. Pediatr Pathol veins as part of the IVC system. Thus, on the 4:173-179, 1985 7. Rollins BJ: Hepatic veno-occlusive disease. Am J basis of our data, occlusion oflarge hepatic veins Med 81:297-306,1986 should be considered thrombotic even if no 8', Alpert LI: Veno-occlusive disease of the liver associmorphologic confirmation is obtainable. Whether ated with oral contraceptives: case report and review of literature. Hum Pathol 7:709-718,1976 such a thrombosis is primary or whether it 9, Nagashima H, Itoshima T, Yamamoto K: Buddcomplicates other lesions such as a tumor, howChiari syndrome. Geriatr Med (Jpn) 24:1209-1213, ever, cannot be judged by the location of the 1986 thrombosis. Unfortunately, no explanation ex- 10. Budd G: On Diseases of the Liver. Philadelphia, Lea and Blanchard, 1846 ists for the topographic predilection of primary 11. Chiari H: Ueber die selbstandige Phlebitis obliterans hepatic vein thrombosis. der Hauptstamme der Venae hepaticae als Todesur-
sache. Beitr Pathol Anat Allg Pathol 26:1-18, 1899 Langer B, Stone RM, Colapinto RF, Meindok H, Phillips MJ, Fisher MM: Clinical spectrum of the Budd-Chiari syndrome and its surgical management. Am J Surg 129:137-145, 1975 13. OrloffMJ, Johansen KH: Treatment of Budd-Chiari syndrome by side-to-side portacaval shunt: experimental and clinical results. Ann Surg 188:494-510, 1978 14. Millikan WJ Jr, Henderson JM, Sewell CW, Guyton RA, Potts JR III, Cranford CAJr, Cramer AR, Galambos JT, Warren WD: Approach to the spectrum of Budd-Chiari syndrome: which patients require portal decompression? Am J Surg 149:167-175, 1985 15. Mahony MJ, Littlewood JM, Losowsky MS, Robinson PJ, Giles GR: Budd-Chiari syndrome treated by Senning operation. Arch Dis Child 63:669-671, 1988 16. Sevenet F, Deramond H, Hadengue A, Casadevall N, Delamarre J, Capron J-P: Membranous obstruction ofthe inferior vena cava associated with a myeloproliferative disorder: a clue to membrane formation? Gastroenterology 97:1019-1021, 1989 17. McCarthy PM, van Heerden JA, Adson MA, Schafer LW, Wiesner RH: The Budd-Chiari syndrome: medical and surgical management of 30 patients. Arch Surg 120:657-661, 1985
12.
CONCLUSION We prefer not to use the term BCS or, if that cannot be achieved, to apply it only to the symptom complex ofnoncardiogenic HVOO either (1) before complete diagnostic workup of a patient or (2) as a collective term for all cases in a study. For patient management and all other purposes-in particular, publication of therapeutic trials and prognostic evaluations-specific etiologic and pathogenetic designations should be used. For example, a diagnostic group could be described as "HVOO, IVC, nonthrombotic, due to fibrous web" or "HVOO, small hepatic veins, thrombotic, due to antineoplastic agents." Implementation of such a system would clarify many publications, facilitate interpretation of data, and ultimately enhance our understanding of the many facets ofHVOO.
JURGEN LUDWIG, M.D., ETSUKO HASHIMOTO, M.D., Division of Pathology; DOUGLAS B. McGILL, M.D., Division of Gastroenterology and Internal Medicine; JON A. van HEERDEN, M.B.,Ch.B., Section of Gastroenterologic and General Surgery
Severe hepatic venous outflow obstruction and its manifestations often are recorded under the label "Budd-Chiari syndrome." Unfortunately, this label is ambiguous; it doesnot clearly identify the site of the lesion (hepatic veins versus inferior vena cava), its morphologic features (thrombotic versus nonthrombotic), or its cause. In the literature, implied or expressed definitions vary. Use of a standardized topographic and pathogenetic classification of hepatic venous outflow obstruction would enable investigators to group patients with comparable conditions, as required for therapeutic trials, prognostic evaluations, and studies of pathogenetic pathways. Review of our own cases revealed that hepatic venous outflow obstruction involving large hepatic veins is usually thrombotic and that isolated obstruction of the inferior vena cava or of small hepatic veins is usually nonthrombotic. Application of such a classification seems feasible and may yield useful results.
The "Budd-Chiari syndrome" (BCS) is a loosely defined designation for hepatic venous outflow obstruction (HVOO). Some authors equate BCS with hepatic large-vein thrombosis.l" but most writers use the term BCS for both thrombotic and nonthrombotic obstruction of the large hepatic veins and the inferior vena cava (IVC) between the liver and the right atrium. Some authors diagnose BCS even if the obstruction is within the right atrium," whereas others specifically exclude conditions at this level-for instance, obstruction of the thoracic IVC by constrictive pericarditis.P-" Thus, BCS is an ambiguous term because it does not clearly identify the morphologic features of the obstruction or
the site involved. The ambiguity is compounded by the controversial relationship between BCS and hepatic veno-occlusive disease (HVOD). Traditionally, the latter is used to designate nonthrombotic obliteration of small hepatic veins as observed after pyrrolizidine poisoning and, more recently, after high-dose antineoplastic chemotherapy." No specific term has been coined for small-vein thrombosis, although a few authors have considered small-vein obstruction of both types as manifestations of HVOD and part of the BCS.2,8,9 In their original publications, Budd 10 and Chiari!' described inflammation of major hepatic veins or their ostia, which the authors ascribed to sepsis and alcoholism'? or syphilis.!' prevailing conditions at the time. Their observations no Address reprint requests to Dr. Jurgen Ludwig, Division of longer can be used as yardsticks for current concepts of the syndrome. Indeed, not all of Pathology, Mayo Clinic, Rochester, MN 55905. Mayo Clin Proc 65:51-55, 1990
51
52
HEPATIC VENOUS OUTFLOW OBSTRUCTION
Mayo Clin Proc, January 1990, Vol 65
Budd's three cases may have had the syndrome- DIAGNOSIS OF BCS a common experience in the study of classic In our opinion, the diagnosis of BCS should be used sparingly; this label is appropriate primardescriptions. In recent articles concerning the management ily for cases ofHVOO before the type and location of BCS, 12-15 some information is provided about ofthe obstruction are known. For the treatment topographic, morphologic, and etiologic features of individual patients, for therapeutic trials, and of the lesions; nevertheless, results still are for other retrospective and prospective studies difficult to interpret because findings usually ofHVOO, a more detailed pathogenetic classifiare discussed collectively under the label BCS. cation is needed. For any such classification, the This situation is further complicated because components of the hepatic venous outflow tract readers often are not told which definition of must be well defined by radiologic and other BCS has been applied to selection and exclusion clinical-diagnostic findings such as the results of of cases. This confusion underscores the limited color Doppler ultrasonography. The anatomic usefulness of the term BCS, but the name un- components are shown in Table 1. doubtedly is here to stay and, therefore, needs to Each case of HVOO should be defined by (1) the morphologic characteristics of the obstrucbe defined. tion-specifically, whether the lesions were DEFINITION OF BCS thrombotic or nonthrombotic; (2) the location of Review of the literature and our own experi- the obstructed venous segment, based on the ences suggests that the following definition of terms defined in Table 1; and (3) the substantiated or putative etiologic factors. Use of this BCS currently would be the most useful: approach results in a classification that is shown BCS consists of hepatic venous outflow obstruction and its in Table 2 and applied to an actual patient cohort manifestations, regardless of cause, the obstruction being in Table 3. either within the liver or in the lVC between the liver and the right atrium. Functional hepatic venous outflow obstruction caused by congestive heart failure is not considered BCS.
Commonly, heart disease other than constrictive pericarditis is excluded from the list of causes of BCS. At the same time, occlusion of small hepatic veins, including classic HVOD, has become part of our definition of the syndrome. This schema may offend traditionalists because, as mentioned earlier, the men credited with discovery of the syndrome 10 ,1l described neither obstruction of the IVC nor HVOD. The difficult and controversial distinction between thrombotic and nonthrombotic (classic HVOD) occlusion of small hepatic veins and the common secondary involvement of small hepatic veins in large-vein disease, however, make it impractical to consider small-vein disease as a separate entity. Whatever the final wording may be, a definition of BCS can find wide acceptance only if it is rather broad and therefore similar to the foregoing proposed definition.
APPLICATION OF CLASSIFICATION For preparation of Table 3 (and thus testing of the proposed classification), we retrieved all Mayo Clinic cases of BCS and HVOO that had been recorded during the 12-year period from 1976 to 1987. We identified 62 patients who fulfilled the previously stated definition ofBCS; some patients from this group had already been described in a previous publication from this institution.!? In 21 instances, morphologic data were unobtainable or insufficient; therefore, these patients were excluded from the initial analysis. Thus, data from 41 patients could be reviewed in detail for the classification shown in Table 3. The results can be summarized as follows. The most common morphologic feature was thrombosis, unrelated to another structural abnormality, and the most common location was the large hepatic veins. Of the 21 patients with this condition, 15 had thrombosis of hepatic veins only, and 6 had thrombosis both of hepatic veins and of the adjacent IVC. In an additional
Mayo Clin Proc, January 1990, Vol 65
HEPATIC VENOUS OUTFLOW OBSTRUCTION
53
Table I.-Clinical Anatomy of Hepatic Venous Outflow System Designation*
Definition
Small hepatic veins
Veins that cannot be shown clearly in hepatic venograms or by ultrasound studies; they include terminal hepatic veins (central veins), intercalated veins, and interlobular veins (collecting veins)'] Veins that are regularly demonstrable in hepatic venograms and ultrasound studies; segmental hepatic vein branches generally are included A segment of the IVe that involves the entire area that is in contact with the right lobe and the caudate lobe of the liver up to the entry level of the right, middle, and left hepatic veins A segment of the IVe that extends from the entry level of the right, middle, and left hepatic veins to the junction between the Ive and the right atrium
Large hepatic veins Hepatic inferior vena cava (hepatic IVe) Suprahepatic IVe
*These designations weredefined for the purpose ofthis review; names and definitions in other publications may differ. tUniform morphologic definitions for the names of the vessels between terminal hepatic veins and veins draining hepatic segments have not been provided in the medical literature.
five patients, thrombosis of the large and small hepatic veins and the abdominal IVe had been a complication of carcinoma in the vicinity (three cases) or kinking or stricture of the suprahepatic IVe (two cases). Isolated thrombosis of the hepatic and suprahepatic IVe was not observed, although the thrombosis in the two cases with stricture or kinking ofthe IVe undoubtedly had begun in this manner. For instance, one ofthese two patients had had a thrombosis at the bifurcation of the IVe that had developed after orthotopic liver transplantation. An embolus had
broken off that clot, but it was arrested by the tight upper anastomosis of the IVe and subsequently resulted in clotting of the adjacent large hepatic veins. This observation notwithstanding, isolated thrombosis of the subdiaphragmatic IVe seems to be a rare event. Nonthrombotic HVOO was found in 15 patients (Table 3). The distribution of most lesions differed strikingly from that of the thrombotic occlusions. Thus, five patients had nonthrombotic obstruction of the Ive only, and seven patients had nonthrombotic obstruction of the
Table 2.-Topographic and Pathogenetic Classification of Hepatic Venous Outflow Obstruction: Principal Categories Site of obstruction Small hepatic veins Large hepatic veins Hepatic and suprahepatic inferior vena cava
Thrombotic
Nonthrombotic
Etiologic factors: idiopathic thrombosis or thrombosis resulting from coagulopathy of known or suspected cause or thrombosis complicating a nonthrombotic abnormality of the hepatic venous outflow tract
Etiologic factors: intravenous webs or membranes, tumor, or other obstructive abnormalities of the hepatic venous outflow tract; usually well-described morphologic features, but cause may be unknown. Some webs or membranes in the inferior vena cava might be postthrombotic!"
54
HEPATIC VENOUS OUTFLOW OBSTRUCTION
Mayo CUn Proc, January 1990, Vol 65
Table 3.-Topographic and Pathogenetic Classification of Hepatic Venous Outflow Obstruction Applied to the Study of 41 Patients* Morphologic features of obstruction and etiologic factor Thrombotic (N Site of obstruction
Thrombosis only (N = 21)
= 26) Complicating a nonthrombotic condition (N = 5)
Thoracic IVC Abdominal IVC only Abdominal IVC, large HV Abdominal IVC, large HV, small HV Large HV only
Large HV, small HV
Small HV only
Idiopathic (2); oral contraceptive use (1); lupus anticoagulant (1) Oral contraceptive use (2) Idiopathic (4); oral contraceptive use (2); paroxysmal nocturnal hemoglobinuria (2) Idiopathic (1); polycythemia vera (2); paroxysmal nocturnal hemoglobinuria (2); agnogenic myeloid metaplasia (1); protein C deficiency (1)
Stricture or kinking of IVC after liver transplantation (2); carcinoma in the vicinity (2) Carcinoma in the vicinity (1)
Nonthrombotic (N = 15) Constrictive pericarditis (3); sclerosing mediastinitis (1) Renal cell carcinoma (1) Fibrous membrane of IVC and large HV (1)
Fibrous web at juncture with IVC (2)
Antineoplastic agents for malignant lesion (2); bone marrow or renal transplantation (2); idiopathic (oral contraceptive use [?], alcoholic liver disease [?]) (2); herbal tea (1)
*Figures in parentheses indicate numbers of patients in each etiologic category. HV = hepatic veins; IVC = inferior vena cava.
small hepatic veins only, identical or comparable to classic HVOD. Two patients had membranes at the IVC orifices of the large hepatic veins, and one patient had fibrous membranes both at these orifices and within the IVC. Among the patients with HVOO, the most common etiologic category was idiopathic thrombosis (seven cases or 17%). In five of these seven cases, only the liver was involved; the IVC apparently had remained patent. The next most common group was represented by five women who had been taking oral contraceptives; only the hepatic veins were involved in two of them, but in three patients, the thrombosis was also found in the IVC. Paroxysmal nocturnal hemo-
globinuria (four cases), polycythemia vera (two cases), protein C deficiency (one case), and agnogenic myeloid metaplasia (one case) were associated with hepatic vein thrombosis alone. One other patient had thrombosis of large hepatic veins and of the IVC; in this instance, a lupus anticoagulant was found. The etiologic factors involved in the nonthrombotic cases of HVOO are listed in Table 3.
DISCUSSION Although most clinicians implicitly use in their publications a classification similar to the one described herein, that information rarely is communicated. We believe that use of a stan-
Mayo Clin Proc, January 1990, Vol 65
HEPATIC VENOUS OUTFLOW OBSTRUCTION
55
dardized classification would facilitate thera- REFERENCES 1. Brinson RR, Curtis WD, Schuman BM, Mills LR: peutic trials, prognostic evaluations, and study Recovery from hepatic vein thrombosis (Budd-Chiari of pathogenetic pathways. Obviously, for pasyndrome) complicating ulcerative colitis. Dig Dis Sci 33:1615-1620, 1988 tient management, many other factors should be 2. McClure S, Dincsoy HP, Glueck H: Budd-Chiari considered, such as the degree of liver dysfuncsyndrome and antithrombin III deficiency. Am J Clin tion and the duration of the disease. For diagPathol 78:236-241, 1982 3. Disney TF, Sullivan SN, Haddad RG, Lowe D, nostic considerations, the data shown in Table 3 Goldbach MM: Budd-Chiari syndrome with inferior are important; they clearly support the experivena cava obstruction associated with systemic lupus ence that thrombotic HVOO is located primarily erythematosus. J Clin Gastroenterol 6:253-256, in the large hepatic veins and that thrombosis of 1984 4. Feingold ML, Litwak RL, Geller SS, Baron MM: the IVC or the small hepatic veins usually occurs Budd-Chiari syndrome caused by a right atrial tuby extension from the large hepatic veins. mor. Arch Intern Med 127:292-295,1971 Conversely, nonthrombotic HVOO involves 5, Solano FX Jr, Young E, Talamo TS, Dekker A: Constrictive pericarditis mimicking Budd-Chiari synalmost exclusively the IVC, and the small hedrome. Am J Med 80:113-115,1986 patic veins in the few instances of classic HVOD. 6. Dahms BB: Case 5: hepatic veno-occlusive disease We considered the orifices of the large hepatic with clinical Budd-Chiari syndrome. Pediatr Pathol veins as part of the IVC system. Thus, on the 4:173-179, 1985 7. Rollins BJ: Hepatic veno-occlusive disease. Am J basis of our data, occlusion oflarge hepatic veins Med 81:297-306,1986 should be considered thrombotic even if no 8', Alpert LI: Veno-occlusive disease of the liver associmorphologic confirmation is obtainable. Whether ated with oral contraceptives: case report and review of literature. Hum Pathol 7:709-718,1976 such a thrombosis is primary or whether it 9, Nagashima H, Itoshima T, Yamamoto K: Buddcomplicates other lesions such as a tumor, howChiari syndrome. Geriatr Med (Jpn) 24:1209-1213, ever, cannot be judged by the location of the 1986 thrombosis. Unfortunately, no explanation ex- 10. Budd G: On Diseases of the Liver. Philadelphia, Lea and Blanchard, 1846 ists for the topographic predilection of primary 11. Chiari H: Ueber die selbstandige Phlebitis obliterans hepatic vein thrombosis. der Hauptstamme der Venae hepaticae als Todesur-
sache. Beitr Pathol Anat Allg Pathol 26:1-18, 1899 Langer B, Stone RM, Colapinto RF, Meindok H, Phillips MJ, Fisher MM: Clinical spectrum of the Budd-Chiari syndrome and its surgical management. Am J Surg 129:137-145, 1975 13. OrloffMJ, Johansen KH: Treatment of Budd-Chiari syndrome by side-to-side portacaval shunt: experimental and clinical results. Ann Surg 188:494-510, 1978 14. Millikan WJ Jr, Henderson JM, Sewell CW, Guyton RA, Potts JR III, Cranford CAJr, Cramer AR, Galambos JT, Warren WD: Approach to the spectrum of Budd-Chiari syndrome: which patients require portal decompression? Am J Surg 149:167-175, 1985 15. Mahony MJ, Littlewood JM, Losowsky MS, Robinson PJ, Giles GR: Budd-Chiari syndrome treated by Senning operation. Arch Dis Child 63:669-671, 1988 16. Sevenet F, Deramond H, Hadengue A, Casadevall N, Delamarre J, Capron J-P: Membranous obstruction ofthe inferior vena cava associated with a myeloproliferative disorder: a clue to membrane formation? Gastroenterology 97:1019-1021, 1989 17. McCarthy PM, van Heerden JA, Adson MA, Schafer LW, Wiesner RH: The Budd-Chiari syndrome: medical and surgical management of 30 patients. Arch Surg 120:657-661, 1985
12.
CONCLUSION We prefer not to use the term BCS or, if that cannot be achieved, to apply it only to the symptom complex ofnoncardiogenic HVOO either (1) before complete diagnostic workup of a patient or (2) as a collective term for all cases in a study. For patient management and all other purposes-in particular, publication of therapeutic trials and prognostic evaluations-specific etiologic and pathogenetic designations should be used. For example, a diagnostic group could be described as "HVOO, IVC, nonthrombotic, due to fibrous web" or "HVOO, small hepatic veins, thrombotic, due to antineoplastic agents." Implementation of such a system would clarify many publications, facilitate interpretation of data, and ultimately enhance our understanding of the many facets ofHVOO.
