Long-term liver dysfunction after allogeneic bone marrow transplantation

Bone Marrow Transplantation (2000) 26, 649–655  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt

Long-term liver dysfunction after allogeneic bone marrow transplantation: clinical features and course in 61 patients JF Toma´s1, I Pinilla1, ML Garcıá-Buey2, A Garcıá3, A Figuera1, V Go´mez-Garcıá de Soria1, R Moreno2 and JM Fernańdez-Ranãda3 Departments of 1Hematology, 2Gastroenterology and 3Pathology, Hospital Universitario La Princesa, Universidad Autońoma de Madrid, Madrid, Spain

Summary: This retrospective study has aimed at determining the prevalence, aetiology and clinical evolution of chronic liver disease (CLD) after allogeneic bone marrow transplantation (BMT). A total of 106 patients who had been transplanted in a single institution and who had survived for at least 2 years after BMT were studied. The prevalence of CLD was 57.5% (61/106). In 47.3% of cases more than one aetiopathogenic agent coexisted. The causes of CLD were iron overload (52.4%), chronic hepatitis C (47.5%), chronic graft-versus-host disease (C-GVHD) (37.7%), hepatitis B (6.5%), non-alcoholic steatohepatitis (NASH) (4.9%), autoimmune hepatitis (AIH) (4.9%) and unknown two (3.3%). Twenty-three patients with iron overload underwent venesections which were well tolerated. An improvement in liver function tests (LFTs) was observed in 21 (91%) patients. All six patients with siderosis as the only cause of CLD normalized LFT as well as three patients with HCV infection. Clinical evolution was satisfactory for patients with GVHD, AIH, NASH and hepatitis B. At the last visit 23 patients continued with abnormal LFTs, and 19 of them were infected by the HCV. A sustained biochemical and virologic response was achieved in only one case out of six patients with CHC who received interferon . We have found that CLD is a common complication in long-term BMT survivors. The aetiology is often multifactorial, iron overload, CHC and C-GVHD being the main causes. The CLD followed a rather ‘benign’ and slow course in our patients as none of them developed symptoms or signs of liver failure and we did not observe an increase in morbidity or mortality in these patients, but a longer follow-up is necessary in HCV infected patients based on the natural history of this infection in other populations. Bone Marrow Transplantation (2000) 26, 649–655. Keywords: allogeneic BMT; liver; iron overload; hepatic virus

Correspondence: Dr JF Toma´s, Servicio de Hematologıá, Fundacioń Jı´menez Dıáz, Avda Reyes Cato´licos 2, Ciudad Universitaria, 28040 Madrid, Spain Received 9 August 1999, accepted 6 May 2000

Liver disease is a well-known cause of early morbidity and mortality following an allogeneic bone marrow transplantation (BMT) which affects 80% of patients and is responsible for up to 5–10% of toxic-related deaths.1,2 Drug toxicity, veno-occlusive disease, acute graft-versus-host disease and fungal, bacterial and viral infections are the most frequent diagnoses in this period. With the growing increase in survival rate of BMT recipients, it is becoming apparent that liver disease remains a problem in many. Few studies focus on this long-term liver dysfunction,3,4 and the frequency and outcome of this complication remain unknown. Chronic viral hepatitis, chronic graft-versus-host disease (C-GVHD) and iron overload have been the main reported causes. However, these studies have been heterogeneous and often assess one specific agent. In addition, more than one aetiopathogenic agent frequently coexists, which, together with the lack of specificity of the clinical and biochemical findings, complicates the diagnosis and management of these patients. This retrospective study has aimed at determining the prevalence, aetiology and clinical evolution of chronic liver disease (CLD) in a population of long-term BMT survivors who were studied and followed-up in a single institution.

Patients and methods Patients Between October 1982 and December 1995, 298 patients underwent an allogeneic BMT in our Institution. For this study, we selected patients who had been completely followed up at our center and survived for at least 2 years after transplantation. We excluded patients who died from BMT-related complications (110 patients), patients who relapsed from their original disease (76 patients) or those who were lost to follow-up (six patients). Thus, 106 patients who had survived for at least 2 years after transplantation and who were followed-up at our center were included in the present analysis. The median follow-up period for the 106 patients has been 6 years (range 2.2– 14.3). Clinical characteristics are shown in Table 1. Transplantation procedure Methods used for BMT have been previously described.5 Briefly, conditioning regimens included busulfan (16

Long-term liver dysfunction after allogeneic BMT JF Toma´s et al

650

Table 1

Liver histopathology

Clinical characteristics of the 106 patients

Age median (range) Sex male/female Underlying disease Acute leukaemia Chronic myeloid leukaemia Severe aplastic anaemia Myelodysplastic syndrome Chronic lymphoid leukaemia Non-Hodgkin’s lymphoma Conditioning regimen BU-CY CY-TBI CY-TLI CY CY-ATG BU-VP16 CY-TBI-VP16 GVHD prophylaxis CsA + MTX CsA MTX Follow-up (years) median (range)

25.5 (5–54) 57/49 38 37 23 6 1 1 45 39 16 3 1 1 1 97 7 2 6 (2.2–14.3)

Bu = busulfan; Cy = cyclophosphamide; TBI = total body irradiation; TLI = total lymphoid irradiation; ATG = antithymocyte globulin; CsA = cyclosporin A; MTX = methotrexate.

mg/kg) and cyclophosphamide (120 mg/kg) (BU-CY) in 45 patients, CY (120 mg/kg) and fractioned total body irradiation (TBI, 1200 cGy divided in six doses) (CY-TBI) in 39, CY (200 mg/kg) and total lymphoid irradiation (700 cGy divided in two doses) (CY-TLI) in 16, CY (200 mg/kg) in three, CY (200 mg/kg) plus antithymocyte globulin (ATG, 90 mg/kg) in one, VP-16 (30 mg/kg) and CY-TBI in one, and BU and VP-16 (60 mg/kg) in one patient. All patients received HLA-identical non-T cell-depleted bone marrow from a sibling donor on day 0. Prophylaxis against GVHD consisted of cyclosporin A (CsA) (n = 7), methotrexate (MTX) (n = 2) or CsA plus MTX (n = 97), according to the Seattle protocol.6 Blood components were routinely HCV screened by ELISA from October 1990.

Assessment of liver dysfunction Long-term follow-up evaluation included alanine aminotransferase (ALT, measured using a standard autoanalyser; upper normal limit: 40 IU/l), gammaglutamyl transpeptidase (GGT; upper normal limit: 45 IU/l), alkaline phosphatase (ALP; upper normal limit: 279 IU/l), total bilirubin (upper normal limit: 1.2 mg/dl), albumin and prothrombin time, with determinationss at least every 3 months. All patients were tested for autoantibodies, hepatitis B virus (HBV) markers (HBsAg, HBsAb and HBcAb using commerically available enzyme immunoassays from Abbott, North Chicago, IL, USA) and hepatitis C virus (HCV) antibodies (second or third generation enzyme immunoassays: Abbott). Ninety-seven patients were evaluated for serum HCV-RNA by reverse transcription and nested polymerase chain reaction (PCR, HCV Amplicor from Roche Diagnostic System, Basel, Switzerland). Bone Marrow Transplantation

Liver specimens were available from 39/61 patients with chronic liver dysfunction, performed for diagnostic purposes. Liver biopsies were obtained percutaneously with a Menghini type disposable needle or, in some cases, by laparoscopy with a Tru-cut type needle. Samples were fixed in 10% neutral formalin and embedded in paraffin for routine stains with hematoxylin and eosin, Masson’s trichrome, Wilder’s reticulin, orcein and Perls’. Iron status This was assessed by measuring serum ferritin with an enzyme immunoassay (Access Immunoassay, Beckman, Zurich, Switzerland; reference range: 30–284 ng/ml) and by grading the stainable iron on liver sections with a Perls’ stain. The transfused iron load was assessed by recording the total number of units of red cell concentrates administered before, during and after BMT. Definitions Chronic liver dysfunction was defined biochemically as an increase in liver serum enzymes for at least 6 months. To diagnose hepatic GVHD, a liver biopsy, or typical alteration of LFT plus histologically proven involvement of other organs was required. In patients with cytolysis for at least 6 months, chronic hepatitis C was diagnosed in those with HCV antibodies and/or HCV-RNA positivity, and chronic hepatitis B (CHB) in those cases with presence of HBsAg seroconversion. Iron overload was diagnosed when elevated serum ferritin and/or liver iron stores were found. Chronic hepatitis was histologically classified according to internationally accepted criteria.7 Non-alcoholic steatohepatitis (NASH) was diagnosed according to the criteria proposed by Powell et al,8 and autoimmune hepatitis (AIH) according to the recommendations of the International Autoimmune Hepatitis Group.9 Iron chelation Twenty-three patients with chronic liver dysfunction due to iron overload were enrolled in a post-transplant iron depletion protocol. Venesections of 350–400 cm3 were performed at 2 or 3 week intervals until ferritin and/or LFT normalization. Hemoglobin and blood pressure were measured before each venesection. Volume replacement was performed with normal saline solution. Patients with unstable erythroid engraftment were treated with deferoxamine (Desferin, Novartis Pharmaceutica, Basel, Switzerland, 20–40 mg/kg/day in a 12 h subcutaneous infusion) and switched to the venesection program once a stable engraftment was achieved. Treatment regimens for chronic hepatitis C Six patients were treated with subcutaneous interferon-2b (IFN) (Intron A; Schering-Plough, Lucerne, Switzerland) at a dose of 3 million units three times per week, which was combined with venesections for iron depletion in five. One


patient received the same dose of IFN plus 1000 mg of oral ribavirin as salvage therapy. Treatment regimens for C-GVHD Immunosuppressive therapies were prednisone 1–2 mg/kg/day p.o., alone (five patients) or in combination with cyclosporine (Sandimmun; Novartis), at a dose of 12 mg/kg/day p.o. divided in two doses, with further adjustments to maintain serum levels of 100–300 ng/ml (14 patients); azathioprine (Imurel; Medeva) 1.5 mg/kg/day p.o. plus prednisone (four patients). Ursodeoxycholic acid (UDCA) (Ursochol, Zambon Group, Bresso, Milan, Italy) at a dose of 8–10 mg/kg/day p.o. was used in six patients, with or without immunosuppression. Results Prevalence and etiology of chronic liver dysfunction Sixty-one of 106 (57.5%) patients developed CLD. A single cause of liver dysfunction was found in 49.2% of cases and more than one aetiopathogenic agent coexisted in 47.3% (Table 2). Causes of CLD identified were iron overload (52.4%), chronic hepatitis C (CHC) (47.5%), C-GVHD (37.7%), hepatitis B (6.5%), NASH (4.9%) and AIH (4.9%). In two patients the aetiopathogenic agent could not be identified. Chronic hepatitis C HCV contributed to CLD in 29/61 (47.5%) patients. It was the only cause of CLD in 11 patients and was associated with other factors (iron overload in 17 and C-GVHD in Table 2 BMT

Frequency and aetiology of chronic liver disease following

five) in 18 patients. The most profound alteration in LFT occurred after withdrawal of immunosuppressive drugs: between days +132 and +1515 post BMT (median +209), with a median peak ALT level of 552 IU/l (range 204– 1600) in the group of patients with only CHC and between days +152 and 1284 (median +351), with median peak ALT of 656 IU/l (range 109–1935) in the group of patients with other causes of CLD associated with the HCV infection. A significant finding was that among the 29 patients with CHC, seven (24%) were constantly anti-HCV negative but HCV-RNA positive. Liver biopsy was performed in 21/29 patients after day +100 post BMT. Moderate chronic active hepatitis (CAH) was found in 12/21 patients while the rest showed mild chronic hepatitis or minimal, nonspecific changes. None of the specimens showed cirrhosis. Four of 11 (36.4%) patients with CHC alone normalized their LFT during follow-up without any therapy and LFTs remained normal at the last visit, while none of the patients in the group with other causes of CLD associated with HCV normalized their ALT without treatment. Six patients were treated with IFN, combined with venesection for iron depletion in five. Four patients were non-responders. Two patients achieved biochemical and virologic response, which was sustained in only one case. The other patient relapsed shortly after therapy was discontinued and subsequently responded to the combination of IFN with ribavirin. Iron depletion improved the LFTs in 66.6% (6/9) of patients with CHC and siderosis and three experienced sustained normalization of LFTs. Seven of 11 (63.6%) patients with CHC alone retained their abnormal LFTs, with the ALT fluctuating between normal and three times the ULN. Moreover, in 14/18 (61%) patients with CHC and other associated factors, the CLD remained and the ALT levels were between normal and eight times the ULN. None of them developed any signs or symptoms of liver failure.

651

Iron overload

Chronic liver disease Yes No Aetiology of chronic liver disease Single cause C-GVHD Chronic hepatitis C Iron overload Hepatitis B Autoimmune hepatitis (AIH) Multifactorial cause Chronic hepatitis C + iron overload C-GVHD + iron overload Chronic hepatitis C + iron overload + C-GVHD C-GVHD + NASH Hepatitis B + iron overload C-GVHD +iron overload + hepatitis B C-GVHD + chronic hepatitis C Iron overload + NASH Iron overload + AIH Hepatitis B + AIH C-GVHD + iron overload + NASH C-GVHD + iron overload + drug hepatotoxicity Unknown aetiology

No. of patients

(%)

61 45

(57.5) (42.5)

30 11 11 6 1 1 29 13 3 4 1 1 1 1 1 1 1 1 1 2

(49.4%) (18.1) (18.1) (10) (1.6) (1.6) (47.3%) (21.3) (5) (6.6) (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (3.3)

Iron overload contributed to chronic liver dysfunction in 32/61 (52.4%) patients. In six patients, siderosis was the only cause of liver disease and in 26, more than one factor was identified. The other aetiologic agents associated with the iron overload were: chronic hepatitis C in 17 patients, C-GVHD in 10, hepatitis B in two, NASH in two, AIH in one, and drug-related hepatotoxicity in one. The severest degree of liver dysfunction after day +100 post BMT of the 32 patients with iron overload was observed between days +111 and 1284 (median day +308). The median peak ALT level was 425 IU/l (range 52–1935 IU/l), median peak GGT 380 IU/l (64–2958), median peak ALP 682 IU/l (183– 2126) and median peak total bilirubin 1 mg/dl (0.3–15.8). Albumin and prothrombin times remained normal. Serum ferritin values ranged from 476 to 9100 ng/ml, with a median of 2260 ng/ml. Twenty-three patients underwent liver biopsy for diagnostic purposes. In 22/23 (95%) patients the hepatic iron storage was more prominent in Kupffer cells and macrophages of the parenchyma and portal tracts, and 21 of them also had mild to severe deposits of iron in hepatocytes. In one patient, the pattern of iron accumulation was exclusively in periportal parenchymal cells, resembling a hereditary haemochromatosis. Fibrosis Bone Marrow Transplantation


652

was observed in 19/23 (82.6%) of liver biopsies which also revealed other causes of liver injury associated with siderosis in the 23 cases. Cirrhosis was not found in any case. Total red cell concentrates (from the diagnoses of the underlaying disease to BMT) transfused to the 32 patients with iron overload ranged from 6 to 151, with a median of 33 units. Twenty-three patients underwent venesection, including all six patients with iron overload as the only cause of liver disease and 17 with one or more factors associated with siderosis (chronic hepatitis C in 11, C-GVHD in five cases, hepatitis B in one, NASH in one, AIH in one and toxicity due to oral anticonceptives in one) (Table 3). Three patients with poor erythroid engraftment were treated with DFX for 2, 10 and 11 months, and were switched to the venesection program once stable engraftment was achieved. In one patient the DFX was interrupted for 3 months due to subcutaneous abscesses at the infusion site. The venesections were initiated between day +188 and +2077 post BMT (median day +539). A median of 12 (4– 30) phlebotomies was performed. The median serum ferritin levels prior to, and at the end of, the iron depletion therapy were 2260 ng/ml (range 729–8750) and 441 ng/ml (50–1275), respectively. In 21/23 (91.3%) of the patients, an improvement in LFT was observed at the end of the treatment. The median ALT values before starting and after iron depletion therapy were 90 IU/l (26–561) and 32 IU/l (12–177), respectively. LFTs normalized in the six patients in whom iron overload was the only cause of the chronic liver disease, in five of them immediately after phlebotomies. One had a reduction in ALT levels from 225 IU/l to 95 IU/l after the last venesection, and the LFTs normalized 2 months later. The two patients with siderosis and C-GVHD achieved a sustained response after iron depletion therapy without immunosuppressive therapy. Nine of the 23 patients venesected had chronic hepatitis C and iron overload. In 6/9 (66.6%) of these patients, a reduction in Table 3

ALT levels after the iron depletion was observed; 3/9 normalized their LFTs and had maintained normal ALT levels at the last follow-up without further therapy. Four of the nine patients received IFN at the same time or after venesections, one of them achieving a sustained biochemical and virologic response, and another relapsed after the withdrawal of IFN, achieving a sustained response on combined therapy with IFN and ribavirin. One patient with siderosis, chronic hepatitis C and C-GVHD showed an improvement on venesection and IFN, but without normalization of ALT levels or virological response, and the other one, with extensive C-GVHD, experienced a striking improvement in LFTs with UDCA and venesection, and is still on the venesection program. The patient with AIH and siderosis normalized his LFTs with venesection UDCA therapy and immunosuppression. The patient with siderosis, C-GVHD and hepatotoxicity due to contraceptives had an improvement after withdrawal of these agents and treatment with steroids and UDCA, finally normalizing her LFTs after iron depletion. One patient with hepatitis B and one with NASH associated with siderosis normalized his LFTs after venesection without further therapy. Venesections were well tolerated and no anaemia or other complications were observed. Chronic graft-versus-host disease C-GVHD contributed to CLD in 23 patients (37.7%). In 11 patients, this was the only factor and in the remaining 12 patients, it was combined with other causes of CLD: iron overload in 10, CHC in five, NASH in two, hepatitis B in one and drug-related hepatotoxicity in one. The median peak ALT, GGT, ALP and total bilirubin levels for the patients with C-GVHD alone were 268 IU/l (range 41– 943), 489 IU/l (242–1175), 676 IU/l (435–1599) and 0.9 mg/dl (0.3–5.4), respectively, and were observed between

Venesection therapy in 23 patients with post-BMT chronic liver disease Onset day

Pre-therapy ferritin

Post-therapy ferritin

Pre-therapy ALT

Post-therapy ALT

No. of venesections

Siderosis (n = 6) Siderosis + GVHD (n = 2)

+498 (255–866) +546

2612 (1250–4180) 2725

446 (50–586)

76 (61–225)

27 (21–95)

11 (7–30)

507

52

23

18

Siderosis + CHC (n = 9) Siderosis + CHC + GVHD (n = 2)

+973 (706–1974) +948

3390 (729–8750) 3430

178 (61–1275) 794

197 (90–561)

60 (22–143)

14 (4–29)

163

52

10

Siderosis + hepatitis B (n = 1)

+532

1710

237

26

12

16

Siderosis + AIH (n = 1) Siderosis + GVHD + drug toxicity (n = 1) Siderosis + NASH (n = 1)

+188

973

261

62

177

9

+2077

1450

305

69

32

6

+455

1300

461

56

23

5

CHC = chronic hepatitis C; UDCA = ursodeoxycolic acid; IFN = interferon; LFT = liver function tests; AIH = autoimmune hepatitis; NASH = nonalcoholic steatohepatitis. Bone Marrow Transplantation


days +107 and +341 (median +230); and for the patients with other associated causes of CLD, the median peaks were ALT 206 IU/l (range 77–1900), GGT 491 IU/l (97– 2958), ALP 842 IU/l (255–2120) and bilirubin 4.6 mg/dl (0.4–15.8), observed between days +136 and +566 (median +298). Liver biopsy material was available from 17 patients. Bile duct epithelial cell damage was observed in all liver specimens, bile duct obliteration in 8/17 and fibrosis in 13/17; none of the biopsies displayed cirrhosis. Fourteen patients received immunosuppressive therapy, nine of them requiring more than one treatment. Six patients received UDCA, associated with immunosuppressive drugs in three cases. All patients with C-GVHD alone normalized their LFTs, and at last follow-up were not receiving immunosuppressive therapy and had no CLD, while 5/12 patients with C-GVHD associated with other aetiopathogenic agents had no CLD and were receiving treatment. The other seven patients retained liver function values between 1.5 and seven times the upper level of normal (ULN), on no immunosuppressive therapy and without clinical signs of liver failure. Hepatitis B This was found in 4/61 (6.5%) of the patients with CLD and was the only cause of liver disease in one and associated with siderosis, siderosis and C-GVHD, and AIH in the others. All patients were HBsAg negative pretransplant and received BMT from HBsAg and HBsAb-negative donors. In three patients, who were HBsAb positive pretransplant, hepatitis B appeared when the protective antibodies were lost after BMT. Hepatitis episodes occurred when the immunosuppression was decreased or after its withdrawal, with a median peak ALT value of 1614 IU/l (range 225– 1900). None of the four patients developed liver failure or received antiviral therapy. All patients cleared the HBsAg between 3 and 6 months after the onset of the hepatitis. Autoimmune hepatitis (AIH) Three out of 61 (4.9%) of the patients with CLD were diagnosed with AIH, this being the only cause of liver disease in one patient and associated with iron overload or hepatitis B in the others. The most profound alterations in LFT were observed between days +139 and +313, with peak ALT levels between 222 and 1626 IU/l. Autoantibodies found were ANA (1/80–1/2560), anti-DNA (1/60), smooth muscle antibody (1/1024), antimicrosomal (1/102 400) and antithyroglobulin (1/6400). The three patients underwent liver biopsy which showed mild to severe necroinflammatory activity. One patient with mild AIH responded to therapy with UDCA and, at the last visit, maintained the response without treatment. One more patient was treated with prednisolone plus UDCA, and another patient with prednisolone plus azathioprine. Both have normalized their LFTs and are still on treatment. Other causes Nonalcoholic steatohepatitis (NASH) was found as the cause of CLD in 3/61 (4.9%) patients, associated with C-

GVHD and/or siderosis. The factors that were probably associated with the development of NASH in these patients were hyperlipaemia, type II diabetes, obesity, long-term steroid therapy and total parenteral nutrition. The peak ALT value was between 77 and 443 IU/l. Liver biopsy material was available in one patient and showed severe macrovesicular steatosis, moderate microvesicular steatosis, lobular inflammation and mild fibrosis. Two patients have normalized their LFTs. In two (83.3%) patients the cause of CLD was not determined. Both had mild liver dysfunction and liver biopsy was not performed. One of them normalized his LFTs without therapy.

653

Serum liver enzymes at the latest visit LFTs were normal at the last evaluation in 38/61 (62%) of the patients with CLD. It is interesting to note that 19 out of the 23 patients with persistent elevations in ALT had HCV infection. In seven cases, chronic hepatitis C (CHC) was the only proven cause of CLD. In the remaining 12 patients with HCV infection C-GVHD (three cases) and siderosis (nine cases) were also presented. Three out of the four patients with persistent CLD without HCV infection suffered from C-GVHD, one with chronic hepatitis B and other with NASH. The cause of CLD in the remaining patient was unknown.

Discussion Long-term liver disease is a common complication of BMT; in the present study 57.5% (61/106) of BMT recipients surviving for at least 2 years after transplantation developed CLD. This prevalence is similar to that observed by Locasciulli et al3 in a cohort of pediatric bone marrow recipients (53%). The aetiology of CLD following BMT is complex as often more than one aetiopathogenic agent is present. In our series, 47.3% of the patients had multifactorial causes of CLD, the most frequent one being the combination of chronic hepatitis C and iron overload (21.3% of all patients with CLD). Overall, the main causes of CLD were iron overload (52.4%), chronic hepatitis C (47.5%) and C-GVHD (37.7%). Attention should also be paid to the diagnosis of other less frequent causes of liver dysfunction, namely, hepatitis B (6.5% in our study, probably due to reactivation of latent HBV infection), NASH (4.9%) and AIH (4.9%). In our study siderosis which was found in 52.4% of the cases, was the most prevalent pathological finding and was the single cause of the liver dysfunction in 18.7% of these patients. Iron overload, which can cause fibrosis and eventually cirrhosis and liver failure, is a recognized cause of CLD.10–12 The possible role of siderosis as a cause of liver dysfunction in long-term survivors of acute leukemia or BMT recipients has been reported in a few studies. McKay et al,13 in a series of long-term survivors of autologous or allogeneic bone marrow transplantation, found that in half the cases where liver dysfunction was present, this was due to transfusion iron overload. However, they did not include patients with GVHD in their study. Harrison et Bone Marrow Transplantation


654

al14 reported a similar prevalence in a series of 38 longterm survivors of acute leukaemia. Iron status was assessed by determining serum ferritin in all patients, and in 72% of the cases a grading of the stainable iron on liver biopsies by the Perls’ stain was carried out.15 This method, although somewhat subjective, provides an estimation of iron stores, including all cell types and their distribution. In addition, a non-uniform pattern of iron distribution and sampling variability in hepatic iron concentration in liver needle biopsies of thalassaemic patients has been reported.16 Although serum ferritin is an acute reactant, the long period of time between BMT and its determination rules out an elevation due to toxicity of the conditioning regimen or early complications. The pattern of hepatic iron storage in our study was consistent with a transfusional mechanism of iron loading, above all in the macrophages, although most of the patients also showed parenchymal accumulation, with preferential loading in the periportal areas. In one patient the hepatic iron storage was exclusively hepatocellular. We found fibrosis in 82.6% of the liver specimens, all of which had more than one factor contributing to the liver dysfunction. This suggests that, iron, when combined with other hepatotoxic agents, may enhance fibrogenesis. Slow and spontaneous decrease in iron stores has been reported in thalassemic children in the years following a BMT.17–19 This ‘natural iron depletion’ could normalize iron stores in individuals with mild siderosis. However, in patients with moderate to severe iron overload, this slow depletion could not prevent the development of liver dysfunction. For this reason, several transplantation teams have developed iron depletion protocols with satisfactory outcomes.13,19,20 We performed venesection in 23 patients with iron overload and chronic liver disease, achieving a striking improvement in liver serum enzymes in 86.9% (20/23) of patients, including 14 with other associated causes of liver dysfunction, and normalisation of serum transaminases was sustained in 73.9% (17/23). At last follow-up, all patients (6/6) with siderosis as the only cause of liver disease, and 70.6% (12/17) of patients with other associated agents had normalised their ALT levels, in some cases by combining therapy with IFN or UDCA. The improvement observed in serum liver enzymes with phlebotomy alone in our patients suffering from CLD due to siderosis and chronic hepatitis C, with three cases experiencing sustained normalisation of ALT levels after iron depletion, is of interest as this depletion may eliminate the need for antiviral therapy. In our study, hepatitis C virus was a major aetiological agent for CLD, involving 47.5% of the cases. The reported prevalence of HCV infection in BMT recipients is highly variable, ranging from 3% to 70%21–28 due to the geographic differences in the prevalence of HCV in the general population, the inclusion of patients transplanted before the introduction of blood donor screening and the diagnostic procedures used. In our series, which included 46 patients transplanted before routine blood component screening, the prevalence of HCV infection post BMT was 33% (data not shown), and 82.8% of the infected patients developed chronic hepatitis, as has been reported by other transplant teams21–28 or in immunocompetent individuals.29 As has been previously reported in this immunosuppressed popu-

Bone Marrow Transplantation

lation, some patients never developed anti-HCV antibodies and the diagnosis of HCV infection could only be made from determination of HCV-RNA in serum. Liver biopsies showed a prevalence of mild to moderate histological damage in liver tissue without cirrhosis at a median follow-up of nearly 7 years. Regarding therapy, only one out of six patients treated with interferon experienced a sustained biochemical and virologic response, although no special adverse effects regarding the onset or reactivation of GVHD or intolerance due to cytopenias were observed. Although no cases of liver failure were observed in our population it should be remembered that 19 of the 23 patients who retained abnormal serum liver enzymes at their last visit were infected with HCV and that this infection was the main cause of the persistent abnormalities in serum liver enzymes after BMT. In a recent study, cirrhosis was identified in 31 out of 3721 patients surviving for more than 1 year after a haematopoietic stem cell transplant, and HCV was the cause in 25 of cases, the median time from transplant to the diagnoses of cirrhosis being 10.1 years (1.2–24.9).30 However, the same group reported that in long-term survivors, HCV infection was not associated with excess mortality after more than 10 years of follow-up.31 C-GVHD remains a major cause of CLD after BMT. In our series, immunosuppressive therapy led to normalisation of LFTs in all patients with C-GVHC as the only cause of CLD and to improvement in liver dysfunction in patients with multifactorial disease, although most of them required more than one treatment. UDCA, with or without immunosuppression, controlled hepatic C-GVHD, as previously reported by Fried et al,32 but in contrast to their observations, our patients had a sustained response after cessation of therapy. Our study confirms that CLD is a common complication in long-term BMT survivors, and the aetiology is often multifactorial with iron overload and chronic hepatitis C being the main causes in this non-thalassaemic cohort. Iron depletion with venesection after BMT is safe, can be accomplished without causing anaemia or other significant adverse effects and constitutes an inexpensive and easily administered therapy. We have also found a ‘benign’ and slow course of post-BMT CLD in our cohort of patients and although we have not performed serial liver biopsies none of them developed symptoms or clinical signs of end-stage liver disease and no increase in morbidity and mortality in patients with CLD has been observed. Nevertheless, the evolution of cirrhosis might be slow in these patients and could take more than one decade. Therefore, whether CLD in patients after allogenic BMT will remain stable or will progress to cirrhosis needs to be determined by continued follow-up over the ensuing decade.

Acknowledgements This work was supported in part by grant number FIS 97/0157 awarded by the Fondo de Investigaciones Sanitarias de la Seguridad Social of Spain.


References 17 1 McDonald GB, Shulman HM, Wolford JL, Spencer GD. Liver disease after human marrow transplantation. Semin Liv Dis 1987; 7: 210–229. 2 Locasciulli A, Alberti A, de Bock R et al. Impact of liver disease and hepatitis infections on allogeneic bone marrow transplantation in Europe: a survey from the European Bone Marrow Transplantation (EBMT) Group-Infectious Diseases Working Party. Bone Marrow Transplant 1994; 14: 833–837. 3 Locasciulli A, Testa M, Valsecchi MG et al. Morbidity and mortality due to liver disease in children undergoing allogeneic bone marrow transplantation: a 10-year prospective study. Blood 1997; 90: 3799–3805. 4 Ljungman P, Johansson N, Aschan JM et al. Long-term effects of hepatitis C virus infection in allogeneic bone marrow transplant recipients. Blood 1995; 86: 1614–1618. 5 Toma´s JF, Go´mez-Garcia de Soria V, Lo´pez-Lorenzo JL et al. Autologous or allogeneic bone marrow transplantation for acute myeloblastic leukemia in second complete remission: importance of duration of first complete remission in final outcome. Bone Marrow Transplant 1996; 17: 979–984. 6 Storb R, Deeg HJ, Whitehead J et al. Methotrexate and cyclosporine compared with cyclosporine alone for prophylaxis of acute graft-versus-host disease after marrow transplantation for leukemia. New Engl J Med 1986; 314: 729–735. 7 Desmet VJ, Gerber M, Hoofnagle JH et al. Classification of chronic hepatitis: diagnosis, grading and staging. Hepatology 1994; 19: 1513–1520. 8 Powell EE, Cooksley WG, Hanson R et al. The natural history of nonalcoholic steatohepatitis: a follow-up study of forty-two patients for up to 21 years. Hepatology 1990; 11: 74–80. 9 Johnson PJ, McFarlane IG, Alvarez F et al. Meeting Report. International Autoimmune Hepatitis Group. Hepatology 1993; 18: 998–1005. 10 Bacon BR, Healey JF, Brittenham GM et al. Hepatic microsomal function in rats with chronic dietary iron overload. Gastroenterology 1986; 90: 1844–1853. 11 Gualdi R, Casalgrandi G, Montosi G et al. Excess iron into hepatocytes is required for activation of collagen type I during iron experimental siderosis. Gastroenterology 1994; 107: 1118–1124. 12 Friedman SL. Seminars in medicine of the Beth Israel Hospital, Boston: the cellular basis of hepatic fibrosis. New Engl J Med 1993; 24: 1828–1835. 13 McKay PJ, Murphy JA, Cameron S et al. Iron overload and liver dysfunction after allogeneic bone marrow transplantation. Bone Marrow Transplant 1996; 17: 63–66. 14 Harrison P, Neilson JR, Marwah SS et al. Role of nontransferrin bound iron in iron overload and liver dysfunction in long term survivors of acute leukaemia and bone marrow transplantation. J Clin Pathol 1996; 49: 853–856. 15 McSween R, Anthoy P, Scheuer P et al. Pathology of the Liver. Churchill Livingstone: New York, 1994. 16 Faa G,Arrbu R, Nurchi V et al. Sampling variability in hepatic

18

19

20

21 22 23 24 25

26

27

28 29 30 31 32

iron concentration in beta-thalassemia. Bone Marrow Transplant 1997; 19 (Suppl. 2): 173–175. Lucarelli G, Angelucci E, Giardini C et al. Fate of iron stores in thalassemia after bone marrow transplantation. Lancet 1993; 342: 1388–1391. Muretto P, Del Fiasco S, Angelucci E, Lucarelli G. Bone marrow transplantation in thalassemia: modification of hepatic iron overload and related pathologies after long-term engrafting. Liver 1994; 14: 14–24. Angelucci E, Muretto P, Lucarelli G et al. and the Italian Cooperative Group for Phlebotomy Treatment of Transplanted Thalassemia Patients. Phlebotomy to reduce iron overload in patients cured of thalassemia by bone marrow transplantation. Blood 1997; 90: 994–998. Toma´s JF, Pinilla I, Go´mez-Garciá de Soria V et al. Longterm liver dysfunction after allogeneic bone marrow transplantation (BMT): importance of hepatitis C virus (HCV) infection and iron overload. Blood 1995; 86 (Suppl. 1): 856a (Abstr.). Kolho E, Ruutu P, Ruutu T. Hepatitis C infection in BMT patients. Bone Marrow Transplant 1993; 11: 119–123. Chen PM, Fan S, Hsieh RK et al. Liver disease in patients with liver dysfunction prior to bone marrow transplantation. Bone Marrow Transplant 1992; 9: 415–419. Fujii Y, Kaku K, Tanaka M et al. Hepatitis C virus infection and liver disease after allogeneic bone marrow transplantation. Bone Marrow Transplant 1994; 13: 523–526. Norol F, Roche B, Satin Marc Girardin MF et al. Hepatitis C virus infection and allogeneic bone marrow transplantation. Transplantation 1994; 57: 393–397. Locasciulli A, Bacigalupo A, VantLint MT et al. Hepatitis C virus infection and liver failure in patients undergoing allogeneic bone marrow transplantation. Bone Marrow Transplant 1995; 16: 407–411. Ljungman P, Fassas A, Ferrant A et al. Long-term effects of hepatitis C virus (HCV) infection after bone marrow transplantation. Bone Marrow Transplant 1998; 21 (Suppl. 1): S152 (Abstr.). Rodri´guez-Iń˜igo E, Toma´s JF, Go´mez-Garciá de Soria V et al. Hepatitis C and G virus infection and liver dysfunction after allogeneic bone marrow transplantation: results from a prospective study. Blood 1997; 90: 1326–1331. Kuo G, Choo QL, Alter HJ et al. An assay for circulating antibodies to a major etiologic virus of human non-A, non-B hepatitis. Science 1989; 244: 362–364. Hoofnagle JH. Hepatitis C: the clinical spectrum of disease. Hepatology 1997; 26 (Suppl. 1): 15S–20S. Strasser SI, Sullivan KM, Myerson D et al. Cirrhosis of the liver in long-term marrow transplant survivors. Blood 1999; 93: 3259–3266. Strasser SI, Myerson D, Spurgeon CL et al. Hepatitis C virus infection and bone marrow transplantation: a cohort study with 10-year follow-up. Hepatology 1999; 29: 1893–1899. Fried R, Murakami CS, Willson RA et al. Ursodeoxycholic acid treatment of refractory chronic graft-versus-host disease of the liver. Ann Intern Med 1992; 116: 624–629.

655

Bone Marrow Transplantation