Thalassaemia Haematopoietic stem cell transplantation for ... - Nature

Bone Marrow Transplantation (2002) 29, 101–105  2002 Nature Publishing Group All rights reserved 0268–3369/02 $25.00 www.nature.com/bmt

Thalassaemia Haematopoietic stem cell transplantation for thalassaemia major in Hong Kong: prognostic factors and outcome CK Li1, MMK Shing1, KW Chik1, V Lee1, TF Leung1, AYK Cheung2 and MP Yuen1 1

Department of Paediatrics, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong; and 2Centre for Clinical Trials and Epidemiological Research, Prince of Wales Hospital, The Chinese Universtiy of Hong Kong, Hong Kong

Summary: From August 1992 to August 1999, 44 patients received allogeneic haematopoietic stem cell transplantation in a single institution. The donors were HLA-identical siblings except for one who was a phenotypically matched father. Thirty-eight patients received bone marrow stem cells and the others received peripheral blood stem cells or umbilical cord blood (UCB). The mean age at transplant was 10.7 ± 5.1 years, ranging from 1.8 to 21 years. Patients received busulphan (16 mg/kg) and cyclophosphamide (150 to 200 mg/kg) as conditioning, and antithymocyte globulin was given to 42 patients to prevent graft rejection. All had engraftment except a patient who received a UCB transplant. Four patients died from early treatment-related mortality, and one died from interstitial pneumonitis 3 months after transplant. Two patients developed secondary graft rejection and both received a second transplant. Thirty-eight patients survived and all except one were transfusion independent. The 5-year overall and event-free survival rates were 86% and 82%, respectively. By multivariate stepwise Cox proportional hazard analyses, severe veno-occlusive disease (VOD) of liver and Pesaro class 3 features were the significant factors associated with survival. Patients aged more than 11 years were more inclined to develop VOD. In conclusion, haematopoietic stem cell transplantation should be performed early if an HLA identical sibling is available. Bone Marrow Transplantation (2002) 29, 101–105. DOI: 10.1038/sj/bmt/1703340 Keywords: transplant; thalassaemia; Chinese; prognostic factor

Thalassaemia major (TM), both alpha and beta types, is common in the southern Chinese population.1 Conventional treatment is by life-long regular blood transfusion. Patients usually die from complications of iron overload if iron chelation is not provided.2 Iron chelation therapy is expensive Correspondence: Dr CK Li, Room G15, Children’s Cancer Centre, Prince of Wales Hospital, Shatin, Hong Kong Received 3 July 2001; accepted 19 October 2001

and some patients may not be compliant. Haematopoietic stem cell transplantation (HSCT) is the only curative treatment for TM and has been reported by various centres.3–5 However, data about HSCT in the Chinese are scarce but the results of an earlier report indicate an inferior outcome as compared to Western experiences.6 This study aimed at analysis of the factors affecting outcome of TM patients after HSCT and reported a high cure rate in Chinese TM patients. Patients and methods All TM patients who received HSCT in the Prince of Wales Hospital, Hong Kong, were included in the study. From August 1992 to August 1999, 44 TM patients received allogeneic HSCT. Except for one child of Indian origin, all the other patients were of southern Chinese origin. The patients were all transfusion dependent and required regular blood transfusion once every 3 to 4 weeks. The median age of diagnosis of TM was 0.5 years, ranging from newborn in a haemoglobin Barts patient to 4.3 years. Desferrioxamine was started for iron chelation after the patients had received regular transfusion for 1 to 2 years. Three patients had not been started on desferrioxamine because of their young age at the time of HSCT. Age, and other pre-transplant characteristics are shown in Table 1. Liver biopsies were performed on 37 patients. The remaining seven patients did not have a biopsy, either because of their very young age or because they had been transplanted at an earlier stage of the programme. Grading of risk factors according to the Pesaro classification was performed for patients for whom complete information was available.7 Exclusion criteria for HSCT were presence of cirrhosis of the liver, HIV positivity or cardiomyopathy. Informed consent from parents, or patients if appropriate, was sought. The donors were all HLA-identical siblings except for one, who was a phenotypically matched father. Thirty-eight patients received bone marrow, while three received peripheral blood stem cells, two received umbilical cord blood and one had a combination of cord blood and bone marrow. Conditioning was busulphan 1 mg/kg p.o. four times per day for 4 days (total 16 mg/kg), cyclophosphamide 50 mg/kg i.v. for 4 days if the patient was less than 11 years of age, or 50 mg/kg i.v. for 3 days for older patients

BMT for thalassaemia in Hong Kong CK Li et al

Table 1

Number Agea at diagnosis (year) Sex: male/female Diagnosis ␣-thalassaemia ␤-thalassaemia Agea started desferrioxamine (year) Agea at transplant (year) Splenectomy done Pre-transplant characteristics Number of transfusiona Serum ferritina (pmol/l) Elevated SGOT Liver biopsy (n = 37) Hepatic fibrosis Haemosiderosis grades I, II/III, IV Pesaro classification Class 1/2/3 Hepatitis status Hepatitis B antigen carrier Hepatitis C antibody positive a

Results

Patient characteristics 44 0.5 (0–4.3) 19/25 2 42 3.7 (1.5–18) 11.4 (1.8–21) 14 patients 126 (11–266) 3699 (970–21000) 9 patients 11 22/15 8/25/7 1 1

Presented as median (range).

(150 to 200 mg/kg). Antithymocyte globulin (ATG) was given to all except the first two patients, to prevent graft rejection. The first 20 patients received ATG at 10 mg/kg i.v. daily from day −5 to day +5 (total 11 days), whereas the subsequent 20 patients received 30 mg/kg i.v. daily for 3 days from day −3 to day −1. The median transplanted nucleated cell dose of BM or PBSC was 4.6 × 108/kg, ranging from 2.5 to 18.6 × 108/kg. Graft-versus-host disease (GVHD) prophylaxis was cyclosporin 3 mg/kg/day i.v. from day −1 and methotrexate on days +1, 3, 6 and 11. Cyclosporin was administered orally when the patient could tolerate oral feeding, and gradual tapering was started from day 50 and discontinued by day 180. Monitoring of CMV reactivation was initially done by shell vial culture and later by CMV pp65 antigen assessment. Ganciclovir prophylaxis was initially given to all patients but later on, according to CMV antigenaemia. Pre-transplant patient characteristics were collected. Outcome parameters included treatment-related complications, namely veno-occlusive disease of the liver (VOD), GVHD, transplant-related mortality (TRM), survival and event-free survival. Severe VOD was defined as a serum bilirubin level of more than 200 ␮mol/l which required diuretic and analgesic treatment. Events were defined as death or rejection of the donor cells. Continuous variables were expressed as median and range. Overall and event-free survival (EFS) were estimated by the Kaplan–Meier method. An event was defined as death or the recurrence of thalassaemia. Differences in survival were assessed by the log-rank test. Since conditioning was according to age and patients over 16 years were regarded as adults, patients were divided into three groups: less than 11 years, 11 to 15.9 years, 16 years or above. The potential factors were analysed with outcome by univariate and multivariate Cox proportional hazards and logistic regression models. Statistical significance was considered to be reached at P ⬍ 0.05. Data were analysed by SPSS for Windows (Release 10.1). Bone Marrow Transplantation

A 10-year-old girl failed to engraft after UCB transplant. The nucleated cell dose transplanted was 2.9 × 107/kg. She experienced autologous bone marrow recovery and again became transfusion dependent. Four patients died before day 50 because of TRM. They all had severe VOD and interstitial pneumonia. No organisms were identified, including CMV. Post-mortem examination showed diffuse alveolar damage of the lung. These four patients were aged between 12 and 17 years. One patient had stable engraftment but developed interstitial pneumonitis on day 103. He died from respiratory failure and post-mortem examination showed diffuse alveolar damage. Two patients, both aged 12 years, had initial engraftment and became transfusion independent. However, secondary graft rejection occurred 5 and 8 months post transplant and marrow aplasia developed. Second transplants were performed in these two patients with the same BM donors. The first one did not engraft and subsequently died from complications of marrow aplasia. The second one had successful engraftment and remained in complete donor chimerism up to 30 months after the second transplant. CMV disease was not documented in any of the patients. Thirty-eight patients survived and 37 remained transfusion independent. The median follow up was 65 months, ranging from 20 months to 105 months. The 5-year overall survival and EFS of the whole group were 86% and 82%, respectively. The overall survival according to age group was 96%, 86% and 63% for age less than 11 years, 11 to less than 16 years, and 16 years or above, respectively (P = 0.03) (Figure 1). The EFS were 91%, 79% and 63% for the three groups respectively, and there was no significant difference (P = 0.13). According to the Pesaro classification, the overall survival was 100%, 88% and 57% for class 1, 2 and 3, respectively (P = 0.036). The EFS was 1.0

16 years

0.6 0.5 0.4 0.3 0.2

P = 0.03

0.1 0.0

0

20

40

60

80

100

120

Follow-up in months Figure 1 Overall survival of thalassaemia patients after BMT according to age.


100%, 80% and 57% for class 1, 2 and 3, respectively (P = 0.1). Age at transplant was an important factor for survival. The other pre-transplant factors did not have an impact on survival or EFS (Table 2). Patients aged 16 years or older at time of HSCT had 5.9 times risk of mortality as compared to younger patients (P = 0.03). Presence of VOD did not affect survival, but severe VOD was associated with 20 times the risk of mortality. Pesaro class 3 was associated with a higher risk of mortality, risk ratio of 5.768. Hepatic fibrosis also had a borderline adverse effect on survival (P = 0.054). There were four deaths among the 11 patients with hepatic fibrosis, but there was only one death among patients without hepatic fibrosis. By multivariate Cox regression, Pesaro class 3 and severe VOD were the two important factors associated with mortality. Severe VOD was the only significant factor associated with EFS survival (P = 0.001). Since severe VOD was a significant factor associated with survival, further analysis of predisposing factors for VOD was carried out. Patients aged less than 11 years had a decreased risk of developing VOD, odds ratio of 0.132. The greater the number of transfusions the patient had received before BMT the higher the risk of developing VOD. The other factors, including pre-transplant serum ferritin, liver enzymes and bilirubin, hepatic fibrosis and haemosiderosis did not affect the incidence of VOD. By multivariate analysis, number of transfusions before BMT was the most significant factor for VOD. GVHD occurred in 23 of 40 (57.5%) evaluable patients. Most were grade I, and eight patients had grade II GVHD (20%). None of the patients had grade III or more. Acute GVHD was treated with methylprednisolone 2 mg/kg/day and all were controlled. By univariate logistic regression analysis, haemosiderosis of liver was the only significant factor predisposing to GVHD, with odds ratio of 5.9 for grade 3 or 4 vs 1

Table 2

or 2. None of the long-term survivors developed chronic GVHD. Chimerism studies were performed by fluorescent in situ hybridisation for X and Y chromosomes if the donor– recipient pairs were sex-mismatched, or by DNA variable number tandem repeat in a sex-matched setting. Four patients did not have an informative marker for DNA chimerism, five patients died before day 50 and one had primary graft rejection. Finally, 35 patients had complete follow-up data on chimerism. Complete donor chimerism occurred in 24 patients (69%), and mixed chimerism in 11 patients (31%). One patient with mixed chimerism had graft rejection at 8 months after BMT. Of the other 10 patients, six had persistent mixed chimerism up to 9 years after transplant, ranging from 13% to 61% of recipient cells. None of these patients required blood transfusion. Quality of life of survivors was excellent. One 16-year-old patient developed diabetes mellitus 2 months after BMT and required insulin treatment. However he had had poor compliance with desferrioxamine before BMT. No patient developed cardiac disease after BMT, and body iron gradually returned to normal after BMT, either with venesection or desferrioxamine.8 Discussion TM is the commonest hereditary transfusion-dependent anaemia in the southern Chinese. Repeated blood transfusion carries the risks of transfusion-transmitted infection, iron overload and immunomodulation.9,10 Successful BMT results in a normal haematopoietic system and patients are thus freed from the need for further transfusion. However, BMT also has risks of transplant-related mortality and graft rejection. The factors leading to failure have been analysed

Factors affecting outcome Overall survival Risk ratio

Age at BMT: ⬍11 years vs ⭓11 years 0.181 ⭓16 years vs ⬍16 years 5.932 Splenectomy: yes vs no 0.863 Pesaro class: 3 vs 2 and 1 5.768 Number of transfusion 1.010 Serum ferritin 0.999 SGOT 0.984 Bilirubin level 0.959 Presence of VOD 4.612 severe vs others 20.356 Presence of GVHD 0.877 grade II vs none or grade I 0.833 ATG dose: 90 mg vs 110 mg 1.581 Presence of hepatic fibrosis 5.305 Haemosiderosis of liver: severe vs 0.727 mild Nucleated cell of bone marrow 1.090 transplanted *

103

P = 0.01–⬍0.05;

P = 0.001–⬍0.01;

**

95% CI

0.021–1.547 1.191–29.544 0.158–4.713 1.161–28.647 0.998–1.023 0.999–0.999 0.945–1.024 0.855–1.076 0.843–25.244 3.925–105.577 0.177–4.344 0.087–8.006 0.318–7.846 0.971–28.984 0.133–3.974 0.919–1.293

Event-free survival

VOD

P value

Risk ratio

95% CI

P value

Risk ratio

95% CI

P value

0.118 0.030* 0.865 0.032* 0.113 0.611 0.427 0.476 0.078 ⬍0.001*** 0.872 0.874 0.575 0.054 0.713

0.372 4.375 1.090 4.323 1.009 0.999 0.989 0.962 2.994 14.796 0.640 0.605 1.147 3.492 0.567

0.072–1.917 0.976–19.620 0.211–5.623 0.966–19.359 0.997–1.020 0.999–0.999 0.956–1.023 0.867–1.067 0.669–13.398 3.201–68.406 0.143–2.861 0.068–5.413 0.256–5.131 0.781–15.615 0.110–2.923

0.237 0.054 0.918 0.056 0.144 0.913 0.510 0.458 0.152 0.001** 0.559 0.653 0.858 0.102 0.498

0.132 2.273 0.762 2.250 1.012 0.999 0.988 1.033 — — 0.471 0.444 1.488 0.779 0.963

0.030–0.577 0.479–10.781 0.198–2.931 0.251–20.130 1.001–1.022 0.999–0.999 0.962–1.014 0.966–1.104 — — 0.132–1.676 0.072–2.740 0.413–5.351 0.182–3.336 0.253–3.669

0.007** 0.301 0.692 0.468 0.034* 0.490 0.363 0.342 — — 0.245 0.382 0.543 0.737 0.956

1.032

0.856–1.246

0.739

0.975

0.813–1.170

0.787

0.321

P ⬍ 0.001.

***

Bone Marrow Transplantation


104

by the Pesaro Centre.7 Hepatomegaly, hepatic fibrosis and poor compliance with iron chelation were identified as being independent prognostic factors, and a classification system according to these factors was developed and adopted by most centres.11 Age is also an important factor when predicting outcome, with adult patients usually having a poorer outcome.12 However, the analysis was based on experience from a single institution and the patients were mainly Caucasians. The results of BMT from centres outside Italy were in general inferior.4,5 The largest thalassaemia population in the world is Chinese but the data from the Chinese are scanty.6 We have attempted to analyse transplant outcome in a homogeneous Chinese population treated in a single institution in Hong Kong. We identified age as one of the important pre-transplant factors leading to mortality. Older patients receive many more transfusions and are thus exposed to prolonged iron toxicity of various organs. Liver biopsy for quantitative assessment of body iron is a reliable indicator.13 However, the iron content of liver biopsy at the time of transplant may not reflect long-term toxicity. These patients may have received more intensive chelation just before transplantation. Hepatic fibrosis may be a better marker of tissue damage due to long-term iron toxicity. In the Pesaro analysis, hepatic fibrosis is an independent prognostic factor. Our study also showed a trend towards higher mortality in patients with hepatic fibrosis. By applying multivariate analysis, we found that Pesaro class 3 was associated with a higher risk of mortality that is in concordance with the Pesaro experience. The Pesaro classification is the most useful predictor for failure based on pre-transplant factors. Four of our patients experienced early TRM and they all had severe VOD and idiopathic interstitial pneumonitis. Autopsy showed severe VOD and diffuse pulmonary damage. Pulmonary damage due to iron overload cannot be diagnosed accurately pre-transplant as most patients have normal pulmonary function. Busulphan is toxic to the lungs and may cause pulmonary fibrosis. Busulphan in conditioning probably aggravates any underlying pulmonary toxicity and contributes to mortality. Younger patients have more rapid clearance of busulphan and thus TRM is lower in young patients.14 In our previous study, busulphan levels (area under curve) greater than 908 ␮mol × min/l were associated with higher mortality.15 The occurrence of VOD is related to number of transfusions and age. Thus, older patients are at risk of severe VOD and other treatmentrelated complications. Once severe VOD occurs, there is a marked increase in mortality. Special precautions should be taken to reduce TRM. We are now measuring busulphan levels in order to make adjustments if necessary. Graft rejection seems to be more common in TM transplants. It may be related to alloimmunisation to HLA antigens due to repeated blood transfusions. Rejection rate ranges from 5% to 35%, and is higher in class 3 and adult patients.4,16,17 The lower dose of cyclophosphamide used in class 3 and adult patients may contribute to the higher rejection rate in these patients.12,16 However, a higher dose of cyclophosphamide is associated with a higher TRM. High rejection rates are reported from centres outside Italy. The exact reason is unknown, but the widespread use of leuko-


cyte depletion filters in Italy may reduce alloimmunisation and consequently lead to decreased rejection.18 In our study, we used ATG routinely. There were two late graft rejections. Primary rejection after cord blood transplantation is likely to be due to the low cell dose transplanted. The 5% rejection rate for the whole group of our patients is low. However, we cannot establish whether ATG is useful in reducing graft rejection. A randomised study of ATG on graft rejection in TM is required. Grade II GVHD occurred in only 20% of patients. None of the patients had grade III GVHD or more. This low GVHD incidence may be related to the relatively young age of our patients or to the homogeneity of the population in southern China. Severe haemosiderosis is the only significant factor associated with GVHD. Iron deposits in the skin and other tissues may cause a predisposition to GVHD. The overall and event-free survivals of the whole group were 86% and 82%, respectively. Patients younger than 11 years had excellent event-free survival. Thus, patients should be transplanted early, before severe iron toxicity develops. We observed mixed chimerism in 31% of patients after BMT, and one patient finally rejected the graft. Six patients (18%) had persistent mixed chimerism but none of them required blood transfusions even when the recipient cells were as high as 61%. Mixed chimerism is common in TM after BMT,19 and some patients may finally develop graft rejection. However, some patients may run a more benign course and may not require blood transfusions for many years.20 It has also been reported that disparate lympho–erythroid mixed chimerism may also occur.21 The use of donor leukocyte infusions should be carefully considered in patients with stable chimerism since the treatment may be associated with significant risks.22 In conclusion, BMT for Chinese TM patients has a similar outcome as for Caucasian patients. Older age is associated with treatment failure, and transplantation should be performed early if HLA-identical siblings are available. Severe VOD is the most important cause of TRM in this group of patients.

Acknowledgements This work was supported in part by The Hong Kong Paediatric Bone Marrow Transplant Fund of the Chinese University of Hong Kong. We thank the nursing staff for providing dedicated care to the BMT patients.

References 1 Lau YL, Chan LC, Chan YYA et al. Prevalence and genotypes of ␣- and ␤-thalassemia carriers in Hong Kong – implications for population screening. New Engl J Med 1997; 336: 1298–1301. 2 Modell B, Khan M, Darlison M. Survival in ␤-thalassaemia major in the UK: data from the UK Thalassaemia Register. Lancet 2000; 355: 2051–2052. 3 Lucarelli G, Galimberti M, Giardini C et al. Bone marrow transplantation in thalassemia. The experience of Pesaro. Ann NY Acad Sci 1998; 850: 270–275. 4 Walters MC, Sullivan KM, O’Reilly RJ et al. Bone marrow


5 6 7 8 9 10 11 12 13 14

transplantation for thalassemia. the USA experience. Am J Pediatr Hematol Oncol 1994; 16: 11–17. Roberts IAG, Darbyshire PJ, Will AM. BMT for children with ␤-thalassemia major in the UK. Bone Marrow Transplant 1997; 19 (Suppl. 2): 60–61. Lin KH, Lin KS. Allogeneic bone marrow transplantation for thalassemia in Taiwan: factors associated with graft failure. Am J Pediatr Hematol Oncol 1989; 11: 417–423. Lucarelli G, Galimberti M, Polchi P et al. Bone marrow transplantation in patients with thalassemia. New Engl J Med 1990; 332: 417–421. Li CK, Lai DH, Shing MMK et al. Early iron reduction programme for thalassaemia patients after bone marrow transplantation. Bone Marrow Transplant 2000; 25: 653–656. Olivieri NF. The ␤-thalassemias. New Engl J Med 1999; 341: 99–109. Li K, Li CK, Wong RPO et al. Transfusion-related immunomodulation in Chinese children with thalassaemia. Vox Sang 1997; 73: 167–173. Lucarelli G, Galimberti M, Polchi P et al. Marrow transplantation in patients with thalassemia responsive to iron chelation therapy. New Engl J Med 1993; 329: 840–844. Lucarelli G, Clift RA, Galimberti M et al. Bone marrow transplantation in adult thalassemic patients. Blood 1999; 93: 1164–1167. Angelucci E, Brittenham GM, McLaren CE et al. Hepatic iron concentration and total body iron stores in thalassemia major. New Engl J Med 2000; 343: 327–331. Slattery JT, Sanders JE, Buckner CD et al. Graft-rejection and toxicity following bone marrow transplantation in relation to

15 16 17

18 19 20

21

22

busulfan pharmacokinetics. Bone Marrow Transplant 1995; 16: 31–42. Li CK, Yuen PMP, Wong R et al. Busulphan level and early mortality in thalassaemia patients after BMT. Bone Marrow Transplant 1999; 23: 307–310. Lucarelli G, Clift RA, Galimberti M et al. Marrow transplantation for patients with thalassemia: results in class 3 patients. Blood 1996; 87: 2082–2088. Solh H, Rao K, da Cunha AM et al. Engraftment failure following bone marrow transplantation in children with thalassemia major using busulfan and cyclophosphamide conditioning. Pediatr Hematol Oncol 1997; 14: 73–77. Reverberi R, Menini C. Clinical efficacy of five filters specific for leukocyte removal. Vox Sang 1990; 58: 188–191. Andreani M, Manna M, Lucarelli G et al. Persistence of mixed chimerism in patients transplanted for the treatment of thalassemia. Blood 1996; 87: 3494–3499. Andreani M, Nesci S, Lucarelli G et al. Long-term survival of ex-thalassemic patients with persistent mixed chimerism after bone marrow transplantation. Bone Marrow Transplant 2000; 25: 401–404. Weinberg RS, Vlachos A, Najfeld V et al. Disparate lympho– erythroid donor to recipient chimaerism in a beta degreesthalassaemia bone marrow transplant recipient with red cell indices indicative of apparent full engraftment. Br J Haematol 1997; 99: 61–63. Aker M, Kapelushnik J, Pugatsch T et al. Donor lymphocyte infusions to displace residual host hematopoietic cells after allogeneic bone marrow transplantation for ␤-thalassemia major. J Pediatr Hematol Oncol 1998; 20: 145–148.

105
