Outcome of liver transplantation for haemophilia

Abstract

Background

Prior to routine screening of blood products many patients with haemophilia were infected with hepatitis C virus (HCV) and have subsequently gone on to develop end-stage liver disease (ESLD).

Patients and Methods

We report our experience of liver transplantation (LT) in patients with haemophilia that developed ESLD secondary to HCV. Patients transplanted from 1994 to 2008 were identified retrospectively. Patient demographics pre-, intra- and post-operative details and outcome were documented.

Results

A total of 3800 LT were performed of which 13 had haemophilia A, 4 haemophilia B and one factor (F)X deficiency. All patients were male with a median age of 52 years (range 26–59), all were HCV antibody positive, 5 (28%) were human immunodeficiency virus (HIV) positive and 4 (22%) had hepatocellular carcinoma. Median intra-operative blood loss was 4.2 l (range 0.8–12) and all received coagulation factor support peri-operatively. Coagulation was unsupported by 72 h post-operatively in all recipients. Two patients developed complications as a result of post-operative bleeding. At a median follow-up of 90 months, 8 patients have died, including 4 of the 5 patients that were HIV positive. The median survival of patients with and without HIV co-infection was 26 and 118 months, respectively.

Conclusion

LT in patients with haemophilia cures the coagulation disorder and in the absence of HIV/HCV co-infection is associated with long-term patient survival.

Introduction

Up until the mid 1980s patients with haemophilia were treated by replacing low levels of clotting factors with plasma pooled from blood donors. Although effective for treating haemophilia the majority of patients were exposed to hepatitis C virus (HCV) and human immunodeficiency virus (HIV). By the time the health risk was discovered an estimated 70% of patients treated with pooled plasma had been infected with HIV and almost 100% were positive for HCV.

Haemophilia patients infected with HCV appear to have a higher risk of developing end-stage liver disease (ESLD) with a cumulative risk of hepatic decompensation of 1.7% at 10 years and 10.8% at 20 years. In the presence of HIV co-infection the chance is even higher, with an estimated risk of 17% by 10 years.^1–4 Consequently a large number of haemophilia patients have gone on to develop ESLD. Liver transplantation (LT) is established as the treatment of choice for patients with ESLD secondary to HCV-related cirrhosis. In patients with haemophilia it also corrects the underlying haemostatic defect as both factor (F)VIII and FIX are synthesized in the liver. Lewis et al. reported the first successful LT in a patient with haemophilia and ESLD.⁵ Despite initial concerns regarding LT in patients with haemophilia, a number of small case series followed and reported acceptable outcomes with patient survival at 1 and 3 years of 83% and 68%, respectively,.⁶ The introduction of highly active anti-retroviral treatment (HAART) has significantly improved the long-term outcome of HIV infection, although the survival of HCV positive patients with haemophilia co-infected with HIV appears to be significantly less.⁷ Recurrence of HCV after LT is a problem common to all transplant recipients, however, the presence of HIV co-infection is thought to enhance the replication of HCV leading to a more aggressive course post-LT.⁸

The purpose of the present study was to review the outcome of all patients with haemophilia, including those with HCV and HIV co-infection, who underwent LT at King's College Hospital.

Patients and methods

Patients

A prospective database of all patients undergoing LT at King's College Hospital has been maintained. The present study is based on an analysis of the prospectively collected data and supplemented by review of clinical notes. From January 1994 to December 2008, a total of 3800 patients underwent LT, of which 18 (0.47%) had haemophilia and formed the basis of the present study. All were male patients with a median age of 52 years (range 26–59). The severity of haemophilia was determined according to the level of clotting factor. Severe was defined as a level <1%, moderate 1–5% and mild 5–40% of normal. Thirteen patients had FVIII deficiency (haemophilia A), of which seven were severe and five mild (data not available for one patient). Four patients had FIX deficiency (haemophilia B or Christmas disease), of which two were severe and one was mild (data not available for one patient). One patient had mild FX deficiency.

All patients were HCV RNA positive (Table 1). Genotype details were available for 11 patients; 4 were genotype 3a, 3 were genotype 1, 2 were genotype 1a and 1 was genotype 1b. Eleven (61%) patients received anti-HCV therapy pre-LT. Eight received non-pegylated interferon monotherapy, and all were non-responders. Three were treated with a combination of pegylated interferon and ribavirin of which one was a non-responder and two responded and relapsed within 6-months of completing therapy.

Table 1.

Pre-liver transplant clinical details

Patient No	Age (Years)	Haemophilia type	Severity	Co-aetiology	HCV Genotype	Albumin (g/l)	Bilirubin (μ mol/l)	Creatinine (μ mol/l)	INR	MELD score
1	46	A	Mild	HCV	1a	27	45	86	1.55	15
2	26	A	Sever	HCV/HIV	–	24	28	52	1.5	13
3	54	Factor X deficiency	Mild	HCV	–	36	60	85	1.4	15
4	48	B	Sever	HCV	–	17	56	114	1.9	21
5	36	A	Sever	HCV/HIV	1a	28	90	67	1.35	16
6	59	B	unknown	HCV	–	36	25	105	1.24	12
7	34	A	Mild	HCV/HIV	1b	29	30	63	1.49	13
8	41	A	Sever	HCV	–	22	597	238	1.8	36
9	57	B	Sever	HCV	–	31	28	92	1.21	11
10	38	A	unknown	HCV	–	26	31	63	1.7	15
11	41	A	Sever	HCV/HIV	1	38	13	75	1.02	7
12	56	B	Mild	HCV/HCC	1	39	22	72	1.28	10
13	58	A	Mild	HCV/HCC	3a	27	19	379	1.26	23
14	38	A	Sever	HCV/HIV	1	29	54	73	1.47	15
15	47	A	Sever	HCV/HCC	3	36	35	137	1.04	12
16	41	A	Mild	HCV	3a	25	39	107	1.25	14
17	51	A	Mild	HCV	3a	28	44	69	1.41	14
18	58	A	Sever	HCV/HCC	3a	27	53	70	1.25	13

HCV, hepatitis C virus; HCC, hepatocellular carcinoma; HIV, human immunodeficiency virus; INR, international normalized ratio; MELD, model of end-stage liver disease.

Five of the 18 (28%) patients had HIV co-infection (Table 2). The mean duration of HIV infection prior to LT was 9.5 years (range 5–14). The mean CD4+ T-lymphocyte count pre-LT was 228 cells/µl (range 160–297). HAART became available in 1995 and three patients were treated pre-LT and all recipients were treated post-LT.

Table 2.

Demographic details of HIV-infected liver transplant recipients

Patient No	Diagnosis of HIV prior to LT (years)	CD4a count	HIV viral loadb	HAART
		Pre- LT/Latest	Pre-LT/Latest	Pre-LT	After LT
2	5	160/NA	NA/NA	No	AZT
5	10	280/NA	NA/NA	No	3TC,AZT, SQV
7	14	260/47	965/NA	3TC,AZT, IDV	3TC,AZT, IDV
11	10	229/NA	<50/NA	d4T,TEN,3TC,EFV	ddI,TEN,3TC,EFV
14	14	297/NA	<50/<50	ABC,3TC,KAL,ddI	SQV,KAL,truvada

^aCD4 × 10⁶/l.

^bViral load copies/ml.

LT, liver transplant; HAART, highly active anti-retroviral treatment; NA, not available; HIV, human immunodeficiency vírus; Drug abbreviations: 3TC, lamivudine; AZT, zidovudine; IDV, indinavir; d4T, stavudine; TEN, tenofovir; EFV, efavirenz; SQV, saquinavir; KAL, kaletra; ddI, didanosine.

The pre-operative clinical details of all patients with haemophilia that underwent LT are summarized in Table 1. Four (22%) had hepatocellular carcinoma (HCC), all of which were within Milan criteria on pre-operative imaging and were treated with doxorubicin-based (40 mg/m²) transarterial chemoembolization prior to LT.

Protocol of clotting factors replacement

Haematologists were involved in the peri-operative management. Pre-operative coagulation studies were performed, including international normalized ratio (INR), activated partial thromboplastin time (APTT) and fibrinogen level. Thromboelastography (TEG) was used to monitor the coagulation profile intra-operatively. Depending upon the type of haemophilia, FVIII, IX or X levels were measured immediately prior to the LT, 12 and 24 h post-LT, and thereafter daily until normal. Patients with FVIII deficiency were administered FVIII pre-operatively, maintained on an infusion intra-operatively (2 IU/kg/h) and post-operatively (1 IU/kg/h) until the INR was <1.5. Patients with FIX deficiency were administered FIX pre-operatively, intra-operatively and discontinued post-operatively once the level was >50 IU/l. Patients with FX deficiency were managed according to our standard LT coagulation protocol of fresh-frozen plasma (FFP) or cryoprecipitate for fibrinogen deficiency and platelet concentrates to support coagulation parameters.

Recurrent HCV

Recurrence of HCV in the allograft was diagnosed by the presence of biochemical dysfunction in association with histological changes on liver biopsy. Liver biopsies were performed in HCV-positive patients to investigate graft dysfunction and since 1999 routinely at 12 months post-LT in all recipients.

Immunosuppression protocol

All patients received prednisolone tapered over 3 months and a calcineurin inhibitor, either cyclosporine (n = 5) prior to 1997 or tacrolimus (n = 13). The trough levels were maintained within the range of 100–150 µg/l and 5–10 ng/ml, respectively. Rejection episodes were treated with three 1-g boluses of intravenous methylprednisolone.

Statistical analysis

Data were entered into an electronic database and statistical analysis performed using SAS version 9.1 (SAS Institute, Cary, NC, USA). Descriptive statistics have been used to characterize the study population. Time was measured from the date of LT to graft loss, death or last follow-up. Patient and graft survival was calculated at 3 months, 1, 3 and 5 years post-LT using the Kaplan–Meier method and compared with all adult (>16 years of age) recipients transplanted over the same time period (n = 1593).

Results

Pre-LT patient demographics are summarized in Table 1. Sixteen patients were transplanted with a whole liver, of which 14 were from a deceased after brain death (DBD) donor and 2 were after cardiac death. Two patients received a split right lobe DBD graft. The median donor age was 33 years (range 14–73) with a median intensive care unit (ICU) stay of 2 days (range 1–9). Median cold and warm ischaemic times were 690 min (range 480–1050) and 38 min (30–58), respectively. Two donors were hepatitis B anti-core antibody positive. The recipients of these grafts were hepatitis B surface antigen negative and received long-term hepatitis B immunoglobulin and lamivudine therapy post-transplant.

The median operative time was 300 min (range 270–450) (Table 3). The median intra-operative blood loss was 4.2 l (range 0.8–12) with a median transfusion requirement of 6 units of packed red cells (range 1–11) and 13 units of fresh frozen plasma (FFP) (4–24). Fourteen patients required platelet transfusion with a median of three bags (range 1–4) transfused per patient. Two patients received two units of cryoprecipitate each.

Table 3.

Post-liver transplant clinical details

	Median (range)	Number (%)
ICU stay (days)	2 (1–8)
Packed cell transfusion (units per patient)	6 (4–24)
FFP transfusion (units per patient)	13 (1–11)
Platelet transfusion (units per patient)	3 (1–4)
Postoperative bleeding		2 (11%)
Infectious complications		4 (22%)
Renal failure		3 (16%)
Acute cellular rejection		4 (22%)
HCV recurrence		11 (61%)

ICU, intensive care unit; FFP, fresh frozen plasma; HCV, hepatitis C virus.

All recipients had normal factor levels by 72 h post-LT. In patients with haemophilia A the median FVIII level at 72 h post-LT was 150 IU/dl (range 97–215). In patients with haemophilia B the median FIX level at 72 h post-LT was 148 IU/dl (range 104–236 IU/dl). The one patient with FX deficiency had a level of 125 IU/dl at 72 h post LT.

The post-operative outcome for all recipients is summarized in Table 3. Of note, two patients had post-operative bleeding complications. One patient developed a retroperitoneal haematoma with compression of the renal vein associated with renal dysfunction requiring laparotomy for control of bleeding day 2 post-LT. Subsequently his renal function returned to normal and he made an uneventful recovery. The second patient had a large subdural haematoma. At the time of his subdural bleed, the FIX level and all his clotting parameters were within the normal range. Surgical evacuation of the clot was performed but the patient died day 12 post-LT. Eleven (61%) patients developed histological proven HCV recurrence on liver biopsy at a median of 6 months (range 3–80) post-LT. Five of these patients remain alive; one patient failed to respond to anti-HCV treatment and is currently being considered for re-transplantation, two patients are currently receiving anti-HCV treatment and the remaining 2 patients have early signs of HCV recurrence and at the time of writing have not been started on anti-HCV treatment.

The overall patient survival at 3 months, 1, 3 and 5 years post-LT was 88.9%, 88.9%, 64.2%, and 53.5%, respectively (Fig. 1). The 1-, 3- and 5-year survival for all other LT recipients over the same period (n = 1593) was 86.0%, 80.9% and 77.5%, respectively (Fig. 1). Of the 18 patients with haemophilia transplanted 8 died over a median follow-up period of 90 months (range 12–184). Two died within the first year post-LT; one patient with a subdural haematoma (described above), the other a patient with HIV co-infection who developed fungal sepsis. The remaining six deaths occurred after 12 months and were attributable to HCV recurrence.

Figure 1.

Recipient survival for patients with haemophilia after liver transplantation (n = 18) compared with adult recipients transplanted over the same time period (n = 1593)

Four of the 5 HIV co-infected patients died resulting in a 1- and 3-year survival for HIV co-infected patients of 81% and 60%, respectively (Fig. 2). One died at day 90 post-LT from fungal sepsis (already described). The remaining 3 died as a result of HCV recurrence at a medium of 30.5 months (range 16–46) post-LT. None of the HIV-positive patients developed HIV-related complications.

Figure 2.

Survival after liver transplant for hepatitis C virus (HCV)-positive patients with haemophilia that were HIV positive (dashed line) or negative (continuous line) (P = 0.025)

All patients (n = 4) with a pre-operative diagnosis of HCC are alive with no evidence of tumour recurrence at a median of 21 months (range 12–36) post-LT.

Discussion

The development of recombinant factor replacement therapy has revolutionized the management of haemophilia. Prior to its routine use pooled plasma was the only therapeutic option and sadly many patients were infected with HCV and HIV. These patients are now presenting with complications of chronic liver disease. The results of this study demonstrate that patients with decompensated liver disease and clotting factor deficiency can be safely transplanted without excessive transfusion requirement with acceptable long-term recipient outcomes.

The success of LT in treating patients with haemophilia raises the question whether prophylactic LT should be recommended as a treatment for haemophilia. Although slight, the risk of transfusion-related complications as a result of repeat blood product administration and the possibility of developing complications of ESLD that preclude transplantation, such as advanced HCC, are arguments that favour earlier transplantation. Against this is the inherent morbidity and mortality associated with LT, requirement for life-long immunosuppression and for many re-infection of the graft with HCV that has a more aggressive natural history. Gene therapy would be the ideal solution with correction of the defective gene to enable for endogenous expression the deficient coagulation factor protein. To date, gene therapy has proven difficult to transition into the clinic because of inefficient protein production and dependency on retrovirus vectors. Until gene therapy becomes a clinical reality LT remains the only ‘cure’ for haemophilia. As a result of the inherent risks associated with LT, haemophiliac patients with chronic liver disease should be closely monitored and offered LT only when they develop complications of chronic liver disease.

Recurrent HCV was the leading cause of graft loss in our series and remains a persistent problem for all patients with HCV undergoing LT. Attempts to prevent re-infection have not been successful. In patients with compensated liver disease pre-transplant treatment with pegylated interferon and ribavirin therapy was shown in the HALT-C trial to result in a sustained viral response (SVR) of only 11%.⁹ The treatment of individuals with decompensated liver disease, who represent the majority of potential transplant recipients, is even less encouraging. In our series only 61% patients had sufficient hepatic reserve pre-transplant to receive anti-HCV therapy pre-LT. Only two responded and none developed a sustained virological response. It has been observed previously that patients with haemophilia respond poorly to interferon therapy which may account for a higher proportion requiring LT than in the non-haemophilic population.¹⁰ Both pre-emptive and directed therapies against HCV have been attempted after transplant, with varying success as a result of poor treatment tolerability. A randomized trial of pre-emptive treatment with pegylated interferon and ribavirin showed a SVR of only 8% with an early discontinuation rate of 31%.^11–13 A directed approach, only treating those with at least stage II fibrosis, resulted in a SVR of 7–26% and discontinuation rate of between 30 and 50%.^11–13 Because SVR is only achieved in a minority of patients, modulating the severity of recurrent disease to prevent graft loss remains the primary therapeutic goal. Unfortunately for patients with haemophilia the rate of progression appears to be more aggressive. Eleven (61%) of the patients in our series developed histological changes of HCV-associated chronic hepatitis at a median 6 months (3–80 months) post-LT. As a comparison, in a longitudinal study of 149 patients reported by Gane et al. from the same centre, 27% of patients transplanted for HCV developed moderate chronic hepatitis at a median of 35 months and cirrhosis developed in 8% at a median of 51 months post-LT.¹⁴ The possible viral, donor and recipient factors that account for the more aggressive clinical course in patients with haemophilia have not been identified.

It is well recognized that HIV accelerates the progression of chronic liver disease in HCV co-infected patients. HCV-related liver disease has become a major cause of death in HIV-infected patients stabilized on HAART.^15–20 The risk of hepatic decompensation has been estimated to be 21-fold higher in HIV-positive patients compared with HIV-negative patients.¹ The mechanism by which HIV results in higher rate of progression of HCV liver disease remains speculative. Given that HIV does not appear to cause hepatitis directly the effect is likely to be mediated by suppression of the immune system against HCV. Supporting this, HCV RNA levels tend to be higher in HIV-positive haemophilic patients compared with those that are HIV negative.⁶ However, it is not known whether the level of HCV RNA correlates with the degree of liver injury and in a study of 26 co-infected haemophilia patients they found no association between the stage of HIV infection and degree of fibrosis on liver biopsy.²¹ It has been postulated that pharmacological interactions between calcineurin inhibitors, the primary immunopressive used post-LT and protease inhibitor-containing HAART regimens may explain the poor outcome in co-infected patients. Pharmacokinetic interaction between HAART and calcineurin inhibitors necessitates up to a 50-fold reduction in calcineurin inhibitor dosage.^22,23 In our series tacrolimus dose reductions were not observed prior to 2001. It is possible that inadvertent over immuno-suppression in the three patients transplanted before 2001 may have contributed to HCV recurrence and poor outcome. There is also evidence that HCV has a detrimental effect on the course of HIV infection. The Swiss HIV Cohort study of over 3000 HIV positive patients on HAART found that the relative risk of developing an AIDS-defining illness or death in the co-infected group was 1.7 compared with HCV-negative patients.²⁴ None of the patients in our cohort have developed AIDS-defining illnesses.

Although the long-term outcome for patients that are HCV positive and HIV negative is better than for patients who are positive for both, all co-infected patients that have stable HIV infection on or off HAART should be considered for LT. Prior to the use of HAART the prognosis for patients with HIV infection was poor. Gordon et al. reported 1- and 3-year survival in 6 HIV-HCV co-infected hemophilic patients of 67% and 23%, respectively, which was significantly worse than for the 19 HIV negative, HCV positive patients (90% and 83%, respectively).⁷ With the improvement in anti-retroviral therapy graft survival rates after LT in HIV-positive patients has dramatically improved. A recent retrospective review of 16 HIV-positive liver allograft recipients (11 of which who were co-infected with HCV) from the USA reported a 2-year actuarial survival of 80%.²⁵ Many of the HIV-positive patients in our cohort were transplanted before effective anti-retroviral therapy was available which may explain the difference we observed in recipient survival.

In summary, LT can be safely performed in haemophilic patients with acceptable peri-operative morbidity and mortality; however, recurrence of HCV is common and our findings suggest that HIV co-infection results in a more rapid and aggressive recurrence.

Conflicts of interest

None declared.