Severe hepatocellular injury after hematopoietic cell transplant: incidence, etiology and outcome.

Hepatic complications of transplant are a common cause of mortality. Although mild elevations of serum aminotransferase enzymes (aspartate and alanine (AST, ALT)) do not carry an adverse prognosis, this is not the case with severe hepatocellular injury. We reviewed 6225 consecutive recipients to determine the incidence and outcomes of severe hepatocellular injury (AST >1500 U/l) before day 100, which occurred in 88 patients. Causes were sinusoidal obstruction syndrome (SOS) (n = 46), hypoxic hepatitis (n = 33), varicella zoster virus (VZV) hepatitis (n = 4), drug-liver injury (n = 2) and unknown (n = 3). The incidence declined from 1.9% in the 1990s to 1.1% recently (owing to a fivefold decline in SOS and disappearance of VZV hepatitis). In hypoxic hepatitis, peak serum AST was 3545 U/l (range, 1380-25 246) within days of shock or prolonged hypoxemia; case fatality rate was 88%. In SOS, AST increases occurred 2-6 weeks after diagnosis; peak AST was 2252 U/l (range, 1437-8281); case fatality rate was 76%, with only serum bilirubin able to distinguish survivors (2.7 vs 11.3 mg/100 ml, P=0.0009). We conclude that circulatory insults (sinusoidal injury, hypotension and hypoxemia), and not infection, are the most common cause of severe hepatocellular injury, the frequency of which has declined because of a falling incidence of SOS and VZV hepatitis.

Chronic graft-versus-host disease of the liver: presentation as an acute hepatitis.

Chronic graft-versus-host disease (GVHD) of the liver usually presents as an indolent cholestatic disease in patients with skin, mouth, and eye involvement. We observed 14 patients in whom chronic GVHD of the liver presented with marked elevations of serum aminotransferases, clinically resembling acute viral hepatitis. Onset of liver dysfunction was at 294 days (range, 74-747 days) after allogeneic hematopoietic cell transplantation and coincided with a recent cessation or taper of immunosuppressive drugs. Median peak serum alanine transaminase (ALT) was 1,640 U/L (698-2,565 U/L), and median bilirubin was 12.3 mg/dL (0.9-55.9 mg/dL). All biopsies showed characteristic features of GVHD with damaged and degenerative small bile ducts. Other features included a marked lobular hepatitis, moderate to marked amounts of hepatocyte unrest, sinusoidal inflammation with perivenular necroinflammatory foci, and many acidophilic bodies scattered throughout the lobule. When high-dose immunosuppressive therapy was instituted soon after presentation, progressive improvement and eventual normalization of liver enzymes and bilirubin levels were observed. However, in cases in which the diagnosis was not made and therapy was delayed, a progressive cholestatic picture emerged with histologic evidence of loss of small bile ducts and portal fibrosis. We conclude that a distinct syndrome of chronic liver GVHD presenting as an acute hepatitis can be recognized in a patient at risk who is receiving no, or minimal, immunosuppressive medications. Liver biopsy is necessary to exclude viral causes of liver dysfunction and to confirm characteristic abnormalities of small bile ducts. Institution of high-dose immunosuppression can prevent progressive bile duct destruction and effect resolution of jaundice if given early.

Bile duct apoptosis and cholestasis resembling acute graft-versus-host disease after autologous hematopoietic cell transplantation.

BACKGROUND: Acute graft-versus-host disease (GVHD) of the liver is a frequent complication of allogeneic hematopoietic cell transplantation. This report describes hepatic GVHD following autologous transplantation. METHODS: We reviewed 116 consecutive autologous transplant recipients. A diagnosis of GVHD was based on histology (segmental to subtotal destruction of bile ductal epithelial cells with apoptosis and lymphocytic infiltrates), clinical criteria (elevated serum alkaline phosphatase), a response to immunosuppressive therapy, and finding no other cause for cholestatic liver disease. RESULTS: Two patients developed cholestatic liver disease (alkaline phosphatase levels over five times the normal upper limit) and had liver biopsies showing apoptotic and dysmorphic ductular epithelial cells typical of GVHD. Three additional patients developed cholestasis and intestinal symptoms but had gastric biopsies only, showing apoptotic crypt epithelial cells and crypt cell drop-out typical of GVHD. CONCLUSION: Two recipients of autologous hematopoietic cells developed histologic abnormalities of small bile ducts and cholestatic liver disease resembling GVHD of the liver after allogeneic transplant. The mechanisms of bile duct damage in this setting may involve immune dysregulation related to reconstitution of immunity with peripheral blood stem cells.

Negative regulators of hemopoiesis and stroma function in patients with myelodysplastic syndrome.

The mechanism that leads to hemopoietic failure in patients with myelodysplastic syndrome (MDS) is not well understood. There is evidence, however, that regulatory molecules such as tumor necrosis factor (TNF)-alpha, Fas (CD95), and Fas-ligand, which negatively affect hemopoiesis by way of apoptosis are upregulated. Here we analyzed marrow samples from 80 patients with MDS in regard to TNF-alpha and Fas-ligand levels and a possible correlation with various disease parameters and risk factors. TNF-alpha levels were elevated in comparison to samples from normal marrow donors, however, no significant correlation with FAB subtype, cytogenetic risk group or score by the International Prognostic Scoring System (IPSS) was observed. However, there was an inverse correlation between the cytogenetic risk category (low, intermediate, high) and levels of soluble Fas-ligand. The major source of TNF-alpha were mononuclear (non-stromal) cells which appeared to produce TNF-alpha at maximum levels. Limiting dilution analysis of CD34+ precursor cells showed that individually assayed cells, removed from companion cells that presumably provided negative signals such as TNF-alpha or Fas-ligand, were able to generate progressively increasing numbers of colonies. Stromal layers derived from MDS marrow did not have an inhibitory effect. In fact, higher colony numbers were obtained from both normal and MDS marrow derived hemopoietic precursors propagated on irradiated stromal layers from MDS marrow than on stromal layers from normal marrow. These results show that substantial numbers of normal hemopoietic precursors persist in MDS marrow. However, differentiation into mature cells is inhibited by negative signals originating from accessory or abnormal hemopoietic precursors in the non-adherent marrow fraction.

Allogeneic and syngeneic marrow transplantation for myelodysplastic syndrome in patients 55 to 66 years of age.

We carried out bone marrow transplantation (BMT) in 50 patients with myelodysplastic syndrome (MDS) who were 55.3 to 66.2 years of age (median, 58.8 years). According to the criteria of the French-American-British (FAB) classification, 13 patients had refractory anemia (RA), 19 had RA with excess blasts (RAEB), 16 had RAEB in transformation or acute myelogenous leukemia (RAEB-T/AML), and 2 had chronic myelomonocytic leukemia (CMML). According to the recently established International Prognostic Scoring System (IPSS), available for 45 patients, 2 patients were considered low risk; 14, intermediate 1 risk; 19, intermediate 2 risk; and 10, high risk. Conditioning regimens were cyclophosphamide (CY) (120 mg/kg of body weight) plus 12-Gy fractionated total-body irradiation (FTBI) (n = 15), CY plus FTBI with lung and liver shielding (n = 4), busulfan (7 mg/kg) plus FTBI (n = 4), or busulfan (16 mg/kg) plus CY (n = 27). The busulfan-plus-CY group included 16 patients in whom busulfan was targeted to plasma levels of 600 to 900 ng/mL. In these 16 patients, steady-state levels of busulfan actually achieved were 714 to 961 ng/mL (mean +/- SD, 845 +/- 64 ng/mL; median, 838 ng/mL). The donors were HLA-identical siblings for 34 patients, HLA-nonidentical family members for 4, identical twins for 4, and unrelated volunteers for 6. All 46 patients surviving > 21 days had engraftment, and 22 patients (44%) are surviving 9 to 80 months after BMT. Specifically, among 13 patients with RA, 1 had relapse (cumulative incidence [CI] at 3 years, 8%) and 8 are surviving, for a Kaplan-Meier (KM) estimate of survival at 3 years of 59% (disease-free survival [DSF], 53%). Among 19 patients with RAEB, 3 had relapse (CI at 3 years, 16%), and 8 are surviving disease free (KM estimate at 3 years, 46%). Among 18 patients with RAEB-T/AML or CMML, 6 had relapse (CI at 3 years, 28%), and the KM estimate of DSF at 3 years is 33%. Relapse-free survival had an inverse correlation with cytogenetic risk classification and with the risk score according to the IPSS. Survival in all FAB categories was highest among patients enrolled in a protocol in which busulfan plasma levels were targeted to 600 to 900 ng/mL. These data indicate that BMT can be carried out successfully in patients with MDS who are older than 55 years of age. (Blood. 2000;95:1188-1194)

The biologic significance of rare peripheral blasts after hematopoietic cell transplantation is predicted by multidimensional flow cytometry.

A vexing problem after hematopoietic cell transplantation (HCT) for leukemia is assessing the biologic significance of low numbers of cells "suspicious" for relapse seen in morphologic review of peripheral blood smears (PBSs). In 27 patients, in apparent hematologic remission after HCT for leukemia, we studied the nature of such cells in PBSs to the endpoint of leukemic relapse by using multidimensional flow cytometry (MDF) on blood or bone marrow aspirates. Based on abnormal cytometric maturational patterns, +/- cell sorting of blasts with fluorescence in situ hybridization with informative markers, we differentiated benign recovering myeloid and lymphoid precursors from leukemic cells. In 17 patients, blasts were characterized by MDF as normal early hematopoietic precursors, lymphoblasts, or NK cells. Of these patients, 16 remained in remission for at least 42 days. In 10 patients, blasts were characterized by MDF as a malignant immunophenotype; 9 relapsed within 10 days and 1 relapsed 280 days after a graft-vs-leukemia effect. MDF status was strongly associated with a 90 x probability of relapse even after adjusting for other potential variables. Morphologic triggered MDF characterization of peripheral blasts is a powerful and rapid tool for distinguishing immature regenerative forms from early leukemic relapse.

Cholestasis after hematopoietic cell transplantation.

Cholestasis and jaundice are common after hematopoietic cell transplantation and may have multiple causes. Specific disorders that may contribute to cholestasis in this setting include sepsis, hemolysis, cyclosporine administration, drug toxicity, parenteral nutrition, graft versus host disease, viral infection, and extrahepatic obstruction.

Occult B cell malignancies can be detected by three-color flow cytometry in patients with cytopenias.

Patients with unexplained cytopenias often present a diagnostic dilemma with minimal morphologic or cytogenetic changes to identify the underlying disease process. We have used multidimensional flow cytometry in a study of patients with cytopenias and found that this technology established, changed, or refined the diagnosis in 17/121 patients. Using the flow cytometric technique of CD45 and right angle light scatter (SSC) gating with two additional markers in a three-color analysis, eight of 121 patients were found to have hairy cell leukemia (HCL), in the absence of definitive morphologic findings of HCL. Two additional patients were found to have non-Hodgkin's lymphoma (NHL). Myeloid abnormalities, myelodysplasia (MDS) or acute leukemia was detected in seven of 56 patients with unexplained pancytopenia. Six of 65 patients identified with cytopenias resulting from lymphoid neoplasms had been referred for bone marrow transplantation (BMT) with a presumptive diagnosis of MDS, with subsequent deferral of BMT upon correct diagnosis. The screening technique is incorporated into an extensive immunophenotyping scheme to identify hematopoietic abnormalities using multidimensional flow cytometry (MDF). HCL cells (detected as low as 1.3%) reside in the same position as normal monocytes in the CD45 and SSC plots but could be distinguished from monocytes based on the expression of HLA-DR without CD11b, and expression of CD19. Further phenotyping of the abnormal population confirmed immunoglobulin light chain restriction, CD11c, and CD25 expression. Non-Hodgkin's lymphoma was detected as aberrant mature lymphocytes expressing B lymphoid markers, CD5 and light chain restriction. Myeloid abnormalities were identified in the myeloblast or maturing myeloid cell fractions. The flow cytometric scheme described can be used in primary diagnosis. The technique is definitive, sensitive, and stresses the importance of distinguishing lymphoid from myeloid etiology of cytopenias.

Intravenous immunoglobulin and the risk of hepatic veno-occlusive disease after bone marrow transplantation.

Recent reports using historical controls or registry cohorts suggest, respectively, either an increase in the mortality or a decrease in the incidence of hepatic veno-occlusive disease (VOD) with the administration of intravenous immunoglobulin (i.v.Ig) after bone marrow transplantation. These divergent results prompted us to conduct a retrospective analysis of two randomized clinical trials conducted at our center to determine the effect of i.v.Ig infusions on the development and severity of VOD. Patients were randomized to receive (n=318) or not to receive (n=315) i.v.Ig prophylaxis after human leukocyte antigen-identical sibling (n=414), mismatched or unrelated (n=178), or autologous or syngeneic (n=41) marrow transplantation. To determine the relationship of i.v.Ig to the development and severity of VOD, a single observer reviewed data displays created for each patient for grading VOD without knowledge of patient i.v.Ig use. In this analysis, VOD was defined as hyperbilirubinemia > or =2.0 mg/dL before day 20 and abrupt weight gain > or =2% before day 14 posttransplant in the absence of other causes of liver disease. Hepatic VOD developed in 235 (37%) of the 633 randomized patients. No evidence for VOD was found in 230 (36%) patients. The remaining 168 (27%) patients were classified as having liver disease of uncertain etiology. Hepatic VOD was judged to be severe in 63 (10%) and mild or moderate in 172 (27%) patients. The number of patients developing any VOD or severe VOD was similar between those randomized to i.v.Ig prophylaxis and untreated controls (115 vs. 120 and 32 vs. 31, respectively). Logistic regression models identified several covariates as significant (p < 0.01) factors associated with the development of severe VOD. Increased risk occurred with elevated pretransplant serum aspartate aminotransferase (odds ratio [OR] = 2.64) and earlier year of transplant (OR = 3.73); decreased risk occurred with autologous or twin donors (OR = 0.09) and acute myeloid leukemia (OR = 0.39). The development of any VOD was associated with an elevated pretransplant alkaline phosphatase (OR = 4.1), pretransplant use of vancomycin (OR = 1.6) or amphotericin (OR = 3.0), posttransplant use of cyclosporine (OR = 2.5), older patient age (OR = 1.03), and obesity (OR = 0.78). We concluded from the controlled trials of 633 patients that the administration of i.v.Ig did not influence the development or severity of VOD after bone marrow transplantation.

The effect of prophylactic fluconazole on the clinical spectrum of fungal diseases in bone marrow transplant recipients with special attention to hepatic candidiasis. An autopsy study of 355 patients.

We reviewed 355 autopsies performed between 1990 and 1994 at a major marrow transplant center to determine whether fluconazole prophylaxis prevented visceral fungal infection. Fluconazole prophylaxis was defined by a minimum of 5 prophylactic doses. Fungal infection (any site) was found in 40% of patients transplanted and autopsied at the center. Overall, the proportion of autopsies with any fungal infection was not different for those patients receiving no fluconazole prophylaxis versus those with prophylactic fluconazole. With fluconazole prophylaxis, candidal infections were less frequent, decreasing from 27% to 8%, while Aspergillus infections were more frequent, increasing from 18% to 29%. No increase in deaths related to non-albicans Candida infections was seen. Of the 329 patients with livers examined, hepatic infection caused by Candida species was significantly less common in patients who had received fluconazole. Fungal liver infection was found in 31 patients (9%), 16% of those who were not treated with fluconazole and 3% of those who were treated with fluconazole. Since patients with candidal infections died earlier after marrow transplant than patients with mold infections, we speculate that a longer length of survival may dispose toward acquisition of mold infections. Fluconazole prophylaxis in this cohort of marrow transplant patients undergoing autopsy resulted in a significant reduction in infection caused by Candida species and an increase in mold infections.

Multidimensional flow cytometry of marrow can differentiate leukemic from normal lymphoblasts and myeloblasts after chemotherapy and bone marrow transplantation.

Serial bone marrow aspirates from patients previously given a diagnosis of acute lymphoblastic leukemia (ALL) who had undergone chemotherapy, bone marrow transplantation (BMT), or both were analyzed by multidimensional flow cytometry (MDF) to detect residual disease (lower limit of detection 0.3%). Correlation between the results of morphologic examination and MDF showed concordant results on 100 of 118 specimens. The MDF-positive, morphologic examination-negative specimens were positive by cytogenetic examination or were obtained from patients in whom the ALL eventually relapsed. Similar correlations between MDF and the results of cytogenetic examination were obtained. Leukemic cells were detected in 29 of 62 patients before BMT and 12 of 52 after BMT Normal regenerating lymphoblasts were identified and quantified by MDF in patients without detectable leukemic lymphoblasts. Patients with leukemic lymphoblasts found by MDF in specimens obtained immediately before BMT were 3.28 times more likely to experience relapse after BMT compared with MDF-negative patients, even when leukemic lymphoblasts were undetectable by histopathologic examination, cytogenetic examination, or both. All patients who had undergone BMT with leukemic lymphoblasts found by MDF, with or without morphologic or cytogenetic confirmation, experienced relapse according to conventional criteria within 42 days of the MDF analysis. The detection of residual disease before overt relapse may provide information for early intervention, while definitive recognition of normal recovering blasts may prevent unnecessary treatment.

Sinusoidal endothelial cells as a target for acetaminophen toxicity. Direct action versus requirement for hepatocyte activation in different mouse strains.

Hepatic congestion occurs early in acetaminophen poisoning. This study examines whether acetaminophen is toxic to sinusoidal endothelial cells (SEC), which might lead to microcirculatory disruption. Acetaminophen toxicity was examined in vivo and in vitro in SEC and hepatocytes from C3H-HEN and Swiss Webster mice. In both strains, there was significantly more toxicity to SEC than to hepatocytes; in SEC from C3H-HEN mice, acetaminophen was directly toxic, but the presence of hepatocytes was required for toxicity to Swiss SEC. Acetaminophen, 750 mg/kg, by gavage caused toxicity with variability within and between strains, but all animals died between 3.5 and 6 hr with zone 3 hemorrhagic necrosis. Pretreatment of C3H-HEN SEC with aminobenzotriazole, a suicide inhibitor of P450, abolished toxicity. Baseline glutathione (GSH) levels were comparable, but a 12-hr incubation with acetaminophen decreased GSH by 60 and 8%, respectively, in C3H-HEN and Swiss SEC in single cell type culture. In co-culture, under conditions where Swiss SEC viability declined by 73%, hepatocyte viability and GSH only decreased by 21 and 20%, respectively. In conclusion, acetaminophen was toxic to SEC. It was directly toxic to SEC in one mouse strain and required hepatocyte activation in another strain. The lack of direct toxicity to Swiss SEC may be due to the lack of an activating P450 isozyme. Zone 3 hemorrhagic necrosis in vivo was comparable in both strains, despite differences in the pathways leading to SEC toxicity in vitro. We propose that toxicity to SEC may contribute to hepatic congestion in acetaminophen intoxication.

Acute pancreatitis in marrow transplant patients: prevalence at autopsy and risk factor analysis.

Pancreatitis has been described as an infrequent complication of marrow transplantation. This study investigated the prevalence of pancreatitis at autopsy in marrow transplant patients and determined risk factors for its development. We reviewed consecutive autopsy reports from 1991 to 1993. Medical records and laboratory reports were reviewed for analysis of clinical variables. Autopsy findings and clinical variables were correlated with the autopsy diagnosis of pancreatitis. Pancreatitis was found in 51 of 184 (28%) patients at autopsy. Of those with pancreatitis, 35% had abdominal pain, 10% had measurements of serum pancreatic enzymes, and 20% had abdominal imaging studies in the week prior to death. By univariable analysis, risk factors associated with development of pancreatitis included clinical grades 3 and 4 GVHD, GVHD at autopsy, liver GVHD at autopsy, major infection at autopsy, and increasing days of survival. By multivariable analysis, independent risk factors for its development included any GVHD at autopsy, increasing length of survival after transplantation, and major infection at autopsy. We conclude that pancreatitis is a common but often subclinical complication of marrow transplantation. Its development may be associated with a high prevalence of biliary sludge and prolonged treatment of GVHD with cyclosporine and prednisone.

Hepatitis C virus infection and liver failure in patients undergoing allogeneic bone marrow transplantation.

The role of hepatitis C virus (HCV) infection in severe liver failure (LF) following bone marrow transplantation is still uncertain. We therefore decided to determine the presence of HCV-RNA in 31 patients who died of severe LF after BMT and in 26 matched BMT controls who did not develop LF. HCV-RNA was identified by polymerase chain reaction and anti-HCV by second generation enzyme-linked immunoassay and by 4-band recombinant immunoblotting assay in serum samples obtained before and after BMT. Biochemical and clinical parameters of liver disease were obtained by reviewing clinical records. LF developed at a median interval of 80 days (20-570) from transplantation and was clinically assessed as VOD (n = 7), liver GVHD (n = 5) or hepatitis (n = 19). HCV-RNA was detected, respectively, in 15/31 (48%) and in 12/26 (46%) of LF patients and controls (P = 0.9). Conversely, the risk of dying of LF was 62% and 53% (P = 0.5) respectively, for HCV-RNA positive and negative patients. Anti-HCV profile did not correlate with viremia, nor with type of liver disease. These findings indicate that, despite a 47% prevalence of HCV infection in our series, HCV-RNA positivity was neither a predictor of VOD nor a marker for life-threatening liver disease.

Fungal liver infection in marrow transplant recipients: prevalence at autopsy, predisposing factors, and clinical features.

To determine the prevalence of fungal liver infection at autopsy in marrow transplant recipients, we reviewed autopsy results for the period 1980-1989. Cases were compared to randomly chosen autopsied controls without fungal infection. Fungal liver infection was found in 67 (9%) of 731 patients. Fungal cultures of liver lesions were positive for 34 of 67 patients, most of whom had been culture-positive for the same fungal species (largely Candida) during life. Multivariate analysis revealed that independent predictors of fungal liver infection were deep fungal infection after transplantation (RR, 35), colonization or superficial infection after transplantation (RR, 13), and severe liver dysfunction caused by veno-occlusive disease of the liver and/or graft-versus-host disease (RR, 7). Clinical and laboratory findings during the last month of life revealed no differences between cases and controls. Liver imaging studies performed during the last 15 days of life had a sensitivity of only 18% for detecting fungal liver lesions.

Utility of transvenous liver biopsies and wedged hepatic venous pressure measurements in sixty marrow transplant recipients.

Sixty marrow transplant recipients with liver dysfunction underwent transvenous liver biopsy and measurement of the hepatic venous pressure gradient. Biopsies were done on 29 patients using a Cook needle inserted through the jugular vein, on 30 patients through the femoral vein with a Mansfield biopsy forceps, and on 1 patient using both instruments. The average number of evaluable portal spaces was 4.0 for aggregated Cook needle specimens and 5.2 for Mansfield forceps specimens. The average number of central venules was 2.6 for Cook needle specimens and 3.5 for Mansfield specimens. Tissue obtained with the Mansfield forceps had crush artifact, especially along the edges, making assessment of bile ducts more difficult than in Cook needle specimens. Liver histology aided management in 53/60 patients by confirming the clinical diagnosis in 24 (40%) and by providing additional diagnoses in 29 (48%). Hepatic venous pressure gradient > 10 mmHg correlated with a histologic diagnosis of veno-occlusive disease (P = 0.001); this gradient value provided 91% specificity and 86% positive predictive value. Eleven patients had bleeding complications, 9 after Cook needle biopsy and 2 after Mansfield forceps biopsy. There were 3 procedure-related deaths, 2 from intraperitoneal bleeding after Cook needle biopsy and 1 from femoral vein bleeding after Mansfield biopsy. We conclude that transvenous liver biopsy and pressure measurements provide useful diagnostic information in marrow transplant patients with liver disease. In our hands, the Mansfield forceps was associated with a lower risk of intra-abdominal bleeding and capsular perforation of the liver while providing adequate histology for diagnosis. Hepatic venous pressure gradients > 10 mmHg were highly specific for a diagnosis of veno-occlusive disease in this patient population.

Veno-occlusive disease of the liver after marrow transplantation: histological correlates of clinical signs and symptoms.

We evaluated the relationship between the signs and symptoms of the clinical syndrome called veno-occlusive disease of the liver after bone marrow transplantation and the histological findings in 76 patients who later came to autopsy. Coded necropsy liver was scored for individual histological features that were correlated with prospectively assessed clinical features that the patients had exhibited during life. Patients were stratified into two groups: those with severe clinical veno-occlusive disease (n = 32) and those without. Clinically severe veno-occlusive disease was statistically correlated with several zone 3 acinar changes: occluded hepatic venules, the frequency of occluded hepatic venules x degree of occlusion, eccentric luminal narrowing/phlebosclerosis, zone 3 sinusoidal fibrosis and zone 3 hepatocyte necrosis (all p < or = 0.03). There was a significant relationship between the number of these histological abnormalities in zone 3 of the liver acinus and a clinical diagnosis of severe veno-occlusive disease (p = 0.003). The presence of ascites was significantly correlated with occluded venules, zone 3 sinusoidal fibrosis and zone 3 hepatocyte necrosis (p = 0.001). Maximum serum bilirubin in the first 20 days after transplant was significantly correlated with sinusoidal fibrosis, hepatocyte necrosis and eccentric luminal sclerosis/phlebosclerosis (p < 0.01) but not with venular occlusion. The clinical syndrome of liver toxicity (commonly called veno-occlusive disease) that results from cytoreductive therapy before bone marrow transplant is strongly correlated with a constellation of histological lesions involving structures in zone 3 of the liver acinus and the hepatic venules into which sinusoidal blood flows. This study suggests that there is no single diagnostic histological feature. The severity of clinical veno-occlusive disease appears to be proportional to the number of such histological changes and is not due solely to occlusion of small hepatic venules.

Venoocclusive disease of the liver: development of a model for predicting fatal outcome after marrow transplantation.

PURPOSE: Hepatic venoocclusive disease (VOD) is a common complication of cytoreductive therapy for marrow transplantation. Only 25% of patients who develop VOD have severe disease. We tested the hypothesis that early clinical signs of VOD would predict which patients would recover and which would die. PATIENTS AND METHODS: We evaluated 355 consecutive patients who had transplants between August 6, 1987 and July 21, 1988 for occurrence of VOD and whether it was reversible within 100 days of transplant. Total serum bilirubin and weight gain from day -7 through day +16 posttransplant were compared among patients with no, severe, or nonsevere VOD. Logistic regression models were developed to estimate probabilities of severe VOD at each of six time intervals. The accuracy of these models was tested by applying them to 392 consecutive patients who underwent transplantation between July 22, 1988 and July 20, 1989. RESULTS: As early as day -1, bilirubin and weight gain were significantly different between patients whose VOD proved to be severe and patients with reversible VOD or no disease. Regression models were used to generate coefficients (beta 0, beta 1, beta 2) for the equation P = 1/(1 + e-z), where P is the probability of severe VOD and z = beta 0 + beta 1 (In total serum bilirubin [mg/dL]) + beta 2 (percent weight gain). Application of this equation to the next 392 patients allowed us to calculate sensitivity, specificity, and positive predictive value for a range of probabilities. CONCLUSION: The course of VOD after cytoreductive therapy can be predicted by knowing the serum bilirubin and weight gained within 1 to 2 weeks of transplantation. Probability estimates derived from patient data are highly specific and moderately sensitive. Such probability estimates may be useful when considering potentially risky interventions to treat VOD, such as recombinant human tissue plasminogen activator.

Veno-occlusive disease of the liver and multiorgan failure after bone marrow transplantation: a cohort study of 355 patients.

OBJECTIVE: To determine the incidence and clinical course of veno-occlusive disease of the liver (VOD) after bone marrow transplantation and to analyze risk factors for severe VOD. DESIGN: Cohort study of 355 consecutive patients. SETTING: A bone marrow transplantation center. MEASUREMENTS: Each patient was prospectively evaluated for VOD, and many risk factors for severe VOD were analyzed using logistic regression models. The relation of VOD to renal and cardiopulmonary failure was analyzed using time-dependent proportional hazards models. RESULTS: Veno-occlusive disease developed in 190 of 355 patients (54%; 95% CI, 48% to 59%): Fifty-four patients had severe VOD and 136 had mild or moderate VOD. Independent variables derived from a multivariate model for predicting severe VOD included elevated transaminase values before transplantation (relative risk, 4.6; P < 0.0001); vancomycin therapy during cytoreductive therapy (relative risk, 2.9; P = 0.003); cytoreductive therapy with a high-dose regimen (relative risk, 2.8; P = 0.01); acyclovir therapy before transplantation (relative risk, 4.8; P = 0.02); mismatched or unrelated donor marrow (relative risk, 2.4; P = 0.02); and previous radiation therapy to the abdomen (relative risk, 2.2; P = 0.04). Vancomycin therapy was a marker for persistent fever. Multiorgan failure was more frequent among patients with VOD and usually followed the onset of liver disease. CONCLUSIONS: Veno-occlusive disease, which developed in 54% of bone marrow transplant recipients, is frequently associated with renal and cardiopulmonary failure. Pretransplant transaminase elevations, use of high-dose cytoreductive therapy, and persistent fever during cytoreductive therapy are independent predictors of severe VOD.