Haematopoeitic cell transplantation for Fanconi anaemia – when and how?


  • Margaret L. MacMillan,

    1. Fanconi Anaemia Comprehensive Care Program, Department of Pediatrics, University of Minnesota Medical School, Minneapolis, MN, USA
    Search for more papers by this author
  • John E. Wagner

    1. Fanconi Anaemia Comprehensive Care Program, Department of Pediatrics, University of Minnesota Medical School, Minneapolis, MN, USA
    Search for more papers by this author

Margaret L. MacMillan, Department of Pediatrics, University of Minnesota, MMC 484, 420 Delaware Street, S.E., Minneapolis, MN 55455, USA. E-mail: macmi002@umn.edu


Allogeneic haematopoietic cell transplantation (HCT) remains the only treatment that can correct the haematological manifestations in patients with Fanconi anaemia. Over the last two decades, sequential changes to the approach to HCT have resulted in reduced regimen-related toxicity, superior engraftment and less graft-versus-host disease (GVHD), resulting in improved survival. The two pivotal changes that most influenced these improvements were the addition of fludarabine to the preparative regimen to augment engraftment, and the use of T cell depletion to reduce GVHD. With these improved HCT outcomes, indications for HCT are quite consistent regardless of donor source. Emphasis is now being placed on developing HCT regimens that will improve quality of life by reducing late effects, particularly the risk of malignancy, sterility and endocrinopathies. This paper will review the unique challenges of HCT in FA patients, with particular emphasis on the timing and approach to HCT.

Fanconi anaemia (FA) is a genetically and phenotypically heterogeneous autosomal recessive (and rarely × linked) disorder characterized by congenital malformations, progressive bone marrow (BM) failure and predisposition to malignancy (Fanconi, 1967; Auerbach & Wolman, 1976; Schroeder et al, 1976; Alter, 1996; Alter & Young, 1998). Haematological abnormalities occur in patients with FA at a median age of 7 years (range, birth to 31 years) (Butturini et al, 1994). Based on clinical data from the International Fanconi Anaemia Registry (IFAR, n = 754 patients), the actuarial risk of developing BM failure, haematological and non-haematological malignancies by 40 years of age is 90%, 33% and 28%, respectively (Kutler et al, 2003). Similarly, an extraordinarily high incidence of malignancy was reported in a survey of 149 FA patients with risks increasing with age, particularly for solid tumours (Rosenberg et al, 2003). As of 2010, allogeneic haematopoietic cell transplantation (HCT) still remains the only treatment modality with the potential of correcting the haematological manifestations of FA. This paper will review the unique challenges of HCT in FA patients, with particular emphasis on the optimal timing and therapeutic approach to HCT.

Historical background

Early attempts at HCT for FA patients were largely unsuccessful. The conditioning regimens were based upon those administered to patients with acquired aplastic anaemia, most often consisting of cyclophosphamide (CY) 200 mg/kg with or without irradiation and resulted in excessive regimen-related toxicity (RRT), severe acute graft-versus-host disease (GVHD) and poor survival (Gluckman et al, 1980, 1983). This early clinical experience prompted in vitro laboratory studies that confirmed the hypersensitivity of FA cells to CY (Berger et al, 1980; Auerbach et al, 1983) and irradiation (Gluckman et al, 1984). As a result, Gluckman et al proposed the use of low-dose CY (20 mg/kg) and a single fraction of thoraco-abdominal irradiation [(TAI), 500 cGy] conditioning regimen, leading to markedly reduced RRT with enhanced survival after human leucocyte antigen (HLA)-matched sibling donor HCT (Gluckman et al, 1984). Over the last two decades, efforts have focused on ways to improve upon these early experiences with sequential changes to the preparative therapy. Additionally, long-term follow-up studies indicate that approximately 40% of FA patients develop a malignancy within 15–20 years after HCT (Deeg et al, 1996). Proposed risk factors include the use of irradiation and the development of chronic GVHD (Deeg et al, 1996; Socie et al, 1998; Kutler et al, 2003; Guardiola et al, 2004; Rosenberg et al, 2005) providing further impetus to develop innovative approaches to HCT.

HLA matched sibling donor HCT

For patients with HLA-matched sibling donors, the Gluckman approach of low dose CY and limited field irradiation (LFI) has been the standard HCT preparative regimen. In the past decade, in experienced centres, haematopoietic recovery is expected to occur in >85% patients with survival rates >75% (Dufour et al, 2001; Farzin et al, 2007), as shown in Table I. Multivariate analysis has identified age at transplant, pre-transplant platelet count, conditioning regimen and GVHD prophylaxis as factors that correlated with survival after HLA-identical sibling donor HCT (Gluckman et al, 1995). More recent results from the International Bone Marrow Transplant Registry (IBMTR) showed that, among 209 patients transplanted between 1994 and 1999 from matched siblings, the 3-year survival was 81% [95% confidence interval (CI) 72–90%] in patients <10 years of age (n = 109) and 69% (95% CI 59–79%) in older patients (n = 100) (IBMTR/ABMTR, 2002).

Table I.   HLA-identical sibling donor HCT: clinical results with radiation-based regimens.
ReferenceConditioning regimenGVHD prophylaxisNo. of casesMedian recipient age (range)Sustained engraftmentAcute GVHD (grades III–IV)Outcome
  1. ATG, antithymocyte globulin; CSA, cyclosporine A; CY, cyclophosphamide; GVHD, graft-versus-host disease; MTX, methotrexate; TAI, thoraco-abdominal irradiation; TBI, total body irradiation.

  2. *Includes two patients who received phenotypic identical parental donor HCT.

Dufour et al (2001)CY 20 mg/kg + TAI 5−6 Gy (n = 12)
CY 20–80 mg/kg, TBI 3–6 Gy (n = 10)*
CY 100–200 (n = 5)
CSA (n = 18)
CSA + MTX (n = 8)
MTX alone (n = 1)
279 years (2·5–19·5 years)23/252/2581% alive at 3 years
Farzin et al (2007)CY 20 mg/kg + TAI 4 GyCSA + MP + ATG307·6 years27/302/3090% alive at 10 years

Over the last decade, there has been growing interest in the use of non-irradiation based regimens for FA patients with the goal of potentially reducing the risk of late malignancies (Table II) (Ebell, 2002; Tan et al, 2006; Torjemane et al, 2006; Bonfim et al, 2007; Ayas et al, 2008; Ertem et al, 2009) with the largest single institutional experience being at the Federal University of Parana, Curibata, Brazil. As part of a CY de-escalation study, Bonfim et al (2007) reported on the results of CY 15 mg/kg × 4 d (60 mg/kg total dose) in 67 patients. Patients received bone marrow (BM) from HLA-identical siblings (n = 58) or other HLA-identical relatives (n = 9). GVHD prophylaxis consisted of methotrexate (MTX) and cyclosporine A (CSA) (Bonfim et al, 2007) (C. M. Bonfim, Federal University of Parana, Curibata, Brazil, personal communication). Primary graft failure occurred in one patient who received two further transplants and is alive and well more than 5 years after the first transplant. Secondary graft failure occurred in six patients at a median of 202 (range, 138–350) days after bone marrow transplantation (BMT). Of these six patients, two had myelodysplastic syndrome (MDS) before BMT, two received HLA-identical sibling donor cells and two other HLA-identical related donor cells. At 2 years, the cumulative incidence of rejection was 11% ± 4%. Overall survival at 3 years was 87% ± 4%. Patients transplanted <10 years of age (n = 35) had a 3-year overall survival of 93% ± 7% vs. 73%± 8% for patients ≥10 years of age (n = 32) (P = 0·008). Transplant-related mortality at day 100 and 1 year was 6% and 9%, respectively. Two patients with extensive chronic GVHD developed oral squamous cell carcinoma at age 11 and 12 years. One is alive and well after surgery and the second died from disease.

Table II.   HLA-identical sibling donor HCT: clinical results with chemotherapy only-based regimens.
ReferenceConditioning regimenNo. of casesEngraftedOutcome (range)
  1. ATG, antithymocyte globulin; BU, busulfan; OKT3, Muromonab-CD3; ALG, anti-lymphoglobulin: busulfan; CY, cyclophosphamide; FLU, fludarabine.

  2. *C. M. Bonfim, personal communication updating Bonfim et al (2007); includes nine patients who received HLA-related (non-sibling) donor transplants.

Ebell (2002)FLU 75–180 mg/m2 + ATG + OKT3755/7 alive
Tan et al (2006)FLU 125 mg/m2 + CY 20 mg/kg + ATG1110100% alive at 2 years
Torjemane et al (2006)CY 40 mg/kg + BU 6 mg/kg ± ALG171472% alive (median 16 months)
Ayas et al (2008)CY 60 mg/kg + ATG343433 (median 33·7 months)
Ertem et al (2009)BU 6 mg/kg + CY 40 mg/kg (= 2)
FLU 150 mg/kg + CY 20 mg/kg + ATG (= 2)
887 (median 2·5 years)
Bonfim, personal communication*CY 60 mg/kg676687% alive

As an alternative to higher dose CY, MacMillan et al at the University of Minnesota Medical Center, Minneapolis, United States, explored low-dose CY with fludarabine (FLU), based on promising results by Aker et al (1999). As FLU is an anti-metabolite and immunosuppressive agent that is not a DNA cross-linking agent, it was hypothesized to be a relatively safe agent in patients with FA and yet to be sufficiently immunosuppressive to allow a high probability of engraftment without the co-administration of irradiation. To date, 22 patients [median age 8·5 (range 3·2–43·3) years] have received CY (20 mg/kg), FLU (175 mg/m2) and antithymocyte globulin (ATG; 150 mg/kg) followed by an infusion of HLA-genotypic identical T cell-depleted BM (n = 15) or umbilical cord blood (UCB, n = 7) at the University of Minnesota (Tan et al, 2006). Neutrophil engraftment was observed in all patients. No patient experienced grade 3–5 RRT or developed either acute or chronic GVHD. Of the 20 FA patients with aplastic anaemia and an HLA genotypic identical donor, 18 are alive and well at a median of 3 years (range, 0·1–8·3 years).

Pasquini et al (2008) compared the early HCT outcomes using non-irradiation containing regimens (n = 71) to outcomes of regimens with irradiation (n = 77) for FA patients transplanted with HLA-matched sibling donors on behalf of the Center for International Blood and Marrow Transplant Research (CIBMTR). Haematopoietic recovery, acute and chronic GVHD and mortality were similar after the two regimens. In both cohorts, higher mortality was associated with recipients aged >10 years, prior use of androgens and cytomegalovirus (CMV) seropositivity in either the donor or recipient. With median follow-up of >5 years, the 5-year probability of overall survival was 78% after irradiation regimens and 81% after non-irradiation regimens (P = 0·61) (Pasquini et al, 2008).

Summary of HCT from HLA-matched sibling donors

HCT from HLA-identical sibling donors is generally associated with an excellent outcome if performed early prior to the development of MDS or leukaemia, particularly within the first decade of life. While promising, longer follow-up is required before determining whether non-irradiation based regimens will reduce the risk of late effects classically associated with radiation, (such as infertility, cataracts, endocrinopathies), or risk of epitheloid cancers, which is particularly common in FA adults, even without HCT. The optimal therapy would probably be one that minimizes exposure to DNA cross-linking agents such as high dose CY or irradiation, and risk of GVHD. FLU-based regimens in combination with T cell-depleted marrow have been shown to achieve these goals with excellent survival after HCT.

HCT with alternative donors

The majority of FA patients however do not have an unaffected (i.e. FA negative) HLA-identical sibling donor. Until recently, HCT for FA using alternate donors has been markedly less successful compared to that after HLA-identical sibling due to high rates of graft failure, RRT, GVHD, and opportunistic infection.

Conditioning regimens

Gluckman et al (1995) were the first to report the outcomes of alternative donor HCT in 48 FA patients from multiple institutions, using the database of the IBMTR. Graft failure occurred in 24% patients, and overall survival was 29% at 2 years. Similar rates of graft failure were observed in other published series of FA patients undergoing alternative donor HCTs, suggesting that the reduced intensity conditioning used in these patients was insufficient to achieve reliable engraftment. MacMillan et al (2000) observed a particularly high rate of graft failure in FA patients with T cell mosaicism, suggesting that the presence of diepoxybutane (DEB)-resistant T cells increased the risk of graft rejection. It was hypothesized that CY/TBI 450–600 cGy/ATG may be insufficient for their complete eradication of DEB resistant T cells (MacMillan et al, 2000). Based on these observations, most centres in the last decade added FLU to the conditioning regimens with improved results.

The outcomes of unrelated donor HCT in patients with FA (= 98) were reported by Wagner et al (2007) on behalf of the CIBMTR. Of 83 patients surviving at least 21 d, the overall incidence of neutrophil recovery was 78% at a median of 11 d after HCT. In multivariate analysis, the use of a FLU-containing regimen was associated with a higher probability of neutrophil engraftment. Among recipients of non-FLU-containing regimens, engraftment tended to be poorer in those with evidence of DEB T cell mosaicism (52% vs. 89%; odds ratio 0·14, 95% CI 0·02–1·23, P = 0·076). The day 100 mortality rate was significantly lower in recipients of a FLU-containing regimen (24% vs. 65% respectively, P < 0·001). Corresponding 3-year adjusted overall survival rates were 52% vs. 13%, P < 0·001). Factors associated with higher survival were younger recipient age (<10 years), CMV seronegativity in the recipient, history of fewer than 20 blood product transfusions and use of FLU in the preparative regimen.

Published data on the larger series of HCT outcomes in FA patients transplanted with alternative donors in the past decade are summarized in Table III (Guardiola et al, 2000; Gluckman et al, 2007; Wagner et al, 2007; Chaudhury et al, 2008; MacMillan et al, 2009). More recently efforts have been focused on reducing or eliminating irradiation in the conditioning regimen even for patients requiring alternative donor HCT.

Table III.   Alternative donor HCT: clinical results.
ReferenceConditioning regimenGVHD prophylaxisNo. of casesRecipient age (range)Sustained engraftmentAcute GVHDSurvival
  1. CY, cyclophosphamide; TBI, total body irradiation; rATG, rabbit anti-thymocyte globulin; FLU, fludarabine; TCD, T cell depletion; TCD, CSA, cyclosporine A.

  2. *Probability of secondary graft failure was 19%.

Guardiola et al (2000)VariedVaried6910·8 years (4·0–37·4 years)83%*34%33% alive at 3 years
Wagner et al (2007)VariedVaried9812 years (0·8–33 years)89% (FLU)
69% (no FLU)
21% (TCD)
70% (no TCD)
52% (FLU) at 3 years
13% (no FLU) at 3 years
Gluckman et al (2007)VariedVaried938·6 (1–45 years)60% ± 5%32% ± 5%74% ± 13% (HLA 6/6, n = 12)
48% ± 9% (HLA 5/6, n = 35)
25% ± 7% (HLA 3–4/6, n = 45)
Chaudhury et al (2008)CY 40 mg/kg
TBI 450 cGy
rATG 10 mg/kg
TCD2121/21  14/21 alive
MacMillan et al (2009)CY 40 mg/kg
TBI 300/150 cGy with thymic shielding
FLU 140 mg/m2
ATG 150 mg/kg
TCD + CSA248·8 years (4·0–21·2 years)22/22 (TBI 300 cGy)
0/2 (TBI 150 cGy)
19/22 alive
2/2 alive

A total body irradiation (TBI) dose de-escalation trial was conducted at the University of Minnesota Medical Center in an attempt to reduce the risks of opportunistic infections and late effects and to determine the lowest possible dose of TBI required for engraftment in FA patients undergoing alternative donor HCT. All patients received CY 40 mg/kg, FLU 140 mg/m2, ATG 150 mg/kg, and a single fraction of TBI with CT guided thymic shielding (hypothesized to minimize thymic damage and improve immune reconstitution), shown to reduce the risk of opportunistic infections (MacMillan et al, 2006). TBI dose de-escalation strata were: TBI 300 cGy (cohort 1); TBI 150 cGy (cohort 2); no TBI (cohort 3). Between July 2006 and September 2009, 22 FA patients were enrolled in cohort 1; two in cohort 2 (MacMillan et al, 2009). Median age was 8·8 (range 4–21·1) years. All patients achieved primary engraftment at a median of 10·5 d (9–30). However two of two patients who received TBI 150 developed secondary graft failure at 76 and 114 d after HCT. They both received subsequent HCT and are alive and well. However, as a result of this observation, the dose de-escalation was stopped with TBI 300 cGy identified as the lowest possible dose in the context of FLU/CY given concomitantly. Of the entire cohort, three of 24 patients developed grade II acute GVHD with none having extensive chronic GVHD to date. With a median follow up of 1·3 years, 21 of the 24 are alive with good engraftment and without GVHD. These results demonstrate that TBI 300 cGy is sufficient for consistent engraftment in recipients of CY-FLU-ATG and HLA matched or mismatched T cell-depleted alternative donor BM or unmanipulated UCB (MacMillan et al, 2009). Longer follow up is needed to quantify the impact of lower dose radiation on immune recovery and risk for malignancy.

In addition, it is important to note that there has been some success in eliminating irradiation in the conditioning regimens even in recipients of alternative donor HCT. At the Federal University of Parana (C. M. Bonfim, personal communication), 49 patients (median age 8 years) received CY 60 mg/kg, FLU 125 mg/m2, and rabbit ATG 4 mg/kg, followed by alternative donor HCT with BM (16 HLA matched, 8 HLA-mismatched) or UCB [HLA matched (n = 3); one (n = 7) or two (n = 17) HLA-mismatched units]. Overall survival at 3 years was 49% ± 7% with a median follow up of 50 (range, 7–72) months. Of the patients transplanted with haematopoietic stem cells from 8/8 HLA-matched BM (antigen level class 1 and allele level class II) overall survival at 3 years was 81%. Details on risks of acute and chronic GVHD were not reported.

At the Birmingham Children’s Hospital (Birmingham, UK), Motwani et al (2005) also used a non-irradiation based preparative therapy consisting of FLU 125 mg/m2, CY 20–30 mg/kg and ATG to transplant seven children with FA between 2000 and 2004. Donor sources included HLA-matched unrelated UCB (= 4), HLA matched unrelated (= 2) and haploidentical (= 1) peripheral blood grafts. Five of seven patients had sustained engraftment with the two remaining patients who rejected their grafts later achieving engraftment after a second transplant. All patients developed grade II–III acute GVHD. All patients were alive 1 year after HCT.

HCT for FA patients with advanced MDS or leukaemia

Experience in the treatment of FA patients who have developed advanced MDS (≥5% blasts) or leukaemia is limited. Untreated patients generally die of disease progression within months to a year of diagnosis. Clinical experience with chemotherapy in these patients is scant, although there are anecdotal reports of responses to chemotherapy and HCT. These data indicate that long-term survival is possible in these patients and they should be considered candidates for HCT. The Cincinnati group reported on the outcomes of four patients with FA and MDS or acute myeloid leukaemia (AML) who were treated with a reduced intensity mini-FLAG regimen prior to HCT, consisting of FLU 30 mg/m2, cytosine arabinoside 300 mg/m2 each on days 2–4 and granulocyte-colony stimulating factor (G-CSF) 5 μg/kg on days 1–5 (Mehta et al, 2007). Although the chemotherapy was well tolerated, it was unclear whether chemotherapy prior to HCT improved the survival rate after HCT for these high risk patients (Mehta et al, 2007). Experience at the University of Minnesota does not clearly support the use of cytoreduction prior to HCT as these patients still frequently die of RRT (Wagner, unpublished observations).

At the University of Minnesota we developed a pilot study using a busulfan (BU)-based regimen for FA patients with leukaemia (MacMillan et al, 2004). Between December 2002 and July 2007, 6 FA patients with acute leukaemia (four AML, two acute lymphoblastic leukaemia) received BU 3·2 mg/kg, FLU 140 mg/m2, CY 40 mg/kg and ATG. Stem cell sources included 5/6 HLA-matched related T cell-depleted marrow (n = 1), 5/6 HLA-matched unrelated donor T cell-depleted (TCD) marrow (n = 2), 6/6 HLA-matched unrelated UCB (n = 1) and two double UCBs (4/6 + 5/6 HLA matched UCBs; and 2 5/6 HLA matched UCBs). All patients achieved neutrophil engraftment at a median of 16 d (range 11–20 d). One patient developed grade I acute GVHD and no patient developed chronic GVHD. Relapse occurred in two patients. The overall probability of survival at 1 year was 52% with a median follow up of 5·3 years. However, as RRT was high and particularly severe, we have elected to not further pursue this approach. At this time, at the University of Minnesota, all FA patients with or without advanced MDS/leukaemia are treated with our CY/FLU/ATG/TBI 300 cGy with thymic shielding. Although there is no standard practice for FA patients with leukaemia, low dose CY/FLU alone is probably not sufficient conditioning therapy as there is a substantial number of residual host cells early after transplant with this approach and the risk of relapse would probably be higher than after more aggressive conditioning therapy. Therefore patients with MDS/AML and an HLA-matched sibling donor still receive TBI 300 cGy.

Graft-versus-host disease

In an attempt to eliminate or reduce the risk of GVHD, we have utilized T cell depletion routinely. Historically, acute GVHD was a major cause of morbidity and mortality after HCT in FA patients. In a retrospective analysis of 37 FA patients who underwent HLA-identical bone marrow transplantation at the Hopital Saint Louis, grade II–IV acute GVHD occurred in 62% (95% CI, 43–75%) patients (Guardiola et al, 2004). Compared to a cohort of 73 patients with aplastic anaemia treated at the same institution, FA patients were more likely to have steroid-refractory acute GVHD and require second-line therapy. In univariate analysis, factors associated with grade II–IV acute GVHD in the FA patients included urogenital malformations (RR, 4·67; 95% CI, 1·46–14·9; < 0·01), donor/recipient ABO blood group incompatibility (RR, 4·40; 95% CI, 1·47–13·1; < 0·01), and donor/recipient sex mismatch (RR, 3·31; 95% CI, 1·47–7·47; < 0·01) (Guardiola et al, 2004). The only factor associated with increased risk of grade III–IV acute GVHD in FA patients was donor/recipient major ABO blood group incompatibility (RR, 10·5, 95% CI, 3·09–36·0, < 0·001). In addition, donor/recipient major ABO blood group incompatibility (RR, 6·02; 95% CI, 1·66–21·80; < 0·01) was associated with an increased risk of steroid-resistant acute GVHD. Importantly, by 10 years after transplantation, grades II–IV acute GVHD was a significant predictor of head and neck malignancy, which was only observed in FA patients experiencing one of these two complications. The 10-year cumulative incidence of head and neck malignancy in FA patients experiencing grades II–IV acute GVHD was 28% vs. 0% for patients without GVHD (Guardiola et al, 2004).

The outcomes of unrelated donor HCT in patients with FA (= 69) were reported by Guardiola et al (2000) on behalf of the European Group for Blood and Marrow Transplantation (EMBT) and the European Fanconi Anaemia Registry. Overall survival was 33% at 3 years. The primary causes of death were acute GVHD (= 18), primary or secondary graft failure (= 13), chronic GVHD (= 4), infections (= 11) and sinusoidal obstructive syndrome (SOS) of the liver (= 1). The probability of developing grade II–IV and severe (grade III–IV) acute GVHD was 43% and 34%, respectively. In multivariate analysis, factors associated with a high risk of severe GVHD included absence of T cell depletion (RR, 20·00; 95% CI, 2·54–142·86; = 0·004), malformations of the urogenital tract and/or the kidneys (RR after day 14, 6·60; 95% CI, 1·38–31·64; < 0·02), elevated serum alanine/aspartate transaminases value before starting the conditioning regimen (RR, 2·52; 95% CI, 1·06–6·01; < 0·04), and limb malformations (RR, 2·55; 95% CI, 1·05–6·17; < 0·04).

Similarly, a reduced risk for acute GVHD in recipients of a TCD graft was observed in the recent CIMBTR analysis by Wagner et al (2007). Grade II–IV acute GVHD was statistically significantly higher in recipients of a non-FLU-containing regimen who received a non–TCD graft (RR, 4·29; 95% CI, 1·82–10·10; < 0·001). The probabilities of grade II–IV acute GVHD at day 100 after HCT were 70% and 21% for non–TCD and TCD bone marrow recipients after a non-FLU-containing regimen, respectively. The corresponding probability of acute GVHD after a FLU-containing regimen was 16%.

Hematopoietic stem cell sources

Given the high historical rate of graft failure in FA patients, the preference was to use BM as the cell source because subsequent cells might be required in the event of graft failure. Over time, the use of UCB evolved. In fact the first successful UCBT in the world was performed by Gluckman et al (1989) in a patient with FA. With the development of in vitro fertilization and pre-implantation genetic diagnosis (IVF-PGD), sibling donor UCB is being used more often (Grewal et al, 2004).

On behalf of Eurocord-Netcord and EBMT Gluckman et al (2007) reported the outcomes of unrelated donor HCT in patients with FA (= 93), restricting the analysis to recipients of a HLA-matched (6/6 match, = 12) or partially HLA-matched (5/6 match, = 35; 3-4/6 match, = 45) unrelated UCB unit). The incidence of neutrophil recovery was 60% ± 5% with a higher incidence of recovery in recipients of FLU and a unit containing ≥4·9 × 107 nucleated cells (NC)/kg recipient body weight. The incidence of acute grade II–IV and chronic GVHD was 32% ± 5% and 16% ± 4%, respectively. The overall survival was 40% + 5% with higher survival rates in CMV-seronegative patients and in recipients of FLU and a unit containing ≥4·9 × 107 NC/kg recipient body weight. Survival by HLA match was 74% ± 13% (6/6 match, = 12), 48% ± 9% (5/6 match, = 35) and 25 ± 7% (3–4/6 match, = 45), suggesting that more mismatched grafts should be avoided when possible. Anecdotally, TCD peripheral blood stem cells provided similar rates of neutrophil recovery, however there are no published data in the literature to support this finding.

Summary of HCT from alternative donors in the treatment of FA

With improved outcomes, the indications for alternative donor HCT are increasingly similar to those described for sibling donor HCT. To date there have been a number of observations regarding risk factors for survival in FA patients transplanted with alternative donor haematopoietic stem cells. As seen with HCT in general, age, organ function, CMV serostatus, stem cell dose and Karnofsky performance status are predictive of survival (Guardiola et al, 2000; MacMillan et al, 2000; Wagner et al, 2007). More recent studies have demonstrated the number of malformations (≥3 sites), prior exposure to androgens and absence of FLU in the preparative therapy predicts poorer survival (Guardiola et al, 2000; MacMillan et al, 2000; Wagner et al, 2007). With markedly improved rates of survival in the current era of HCT in FA, emphasis is now being placed on improving quality of life by reducing late effects, particularly the risks of malignancy, sterility and endocrinopathies. Whether reducing irradiation will reduce late effects remains to be seen. Nevertheless, many groups have shown that successful HCT can be achieved with less than the standard 450 cGy TBI. Improvements in therapy are likely to proceed more rapidly if patients are treated according to common protocols and data are shared between institutions.

When to transplant

With improved HCT outcomes after alternative donor HCT, indications for HCT are quite consistent regardless of donor source. The clinical data suggest that the optimal time for HCT is prior to 20 red cell and/or single donor platelet transfusions, and androgen therapy, and before the development of advanced MDS or leukaemia. However, transplant might be delayed or not considered if the patient has high risk features, such as older age, organ dysfunction, uncontrolled infection, or poorly HLA-matched donor (e.g. >7/8 BM or >5/6 UCB). All patients should undergo a bone marrow aspirate and biopsy with cytogenetic evaluation annually and more frequently if dysplastic changes or cytogenetic abnormalities are noted.

In order to avoid transfusions, standard risk patients (<18 years of age, good organ function, absence of advanced MDS or leukaemia) should undergo HCT when persistent and moderately severe cytopenia develops (i.e. haemoglobin <80 g/l; absolute neutrophil count <0·5 × 109 per litre; and/or platelet count <20 × 109 per litre). Patients with evidence of advanced MDS or leukaemia should also be considered. Earlier transplantation may be considered for patients with specific mutations deemed to be particularly high risk for rapid progression to MDS or leukaemia and poor survival (e.g. presence of biallelic breast cancer [BRCA] gene mutations).

Other considerations

As most FA patients at the time of HCT have experienced cytopenia for an extended period, a number of considerations need to be made prior to HCT to optimize the outcomes. Assessment for iron overload and necessity for chelation should be evaluated particularly for heavily transfused patients.

The use of androgens should generally be avoided as data suggests an adverse effect on HCT outcomes. Although newer formulations may be safer, this is yet unknown. Androgen use before HCT has been shown to correlate with a higher risk of mortality after HCT (Guardiola et al, 2000). While efficacy is not known, newer formulations of androgens may be an effective and safer approach for those patients deemed too high risk for HCT. If patients are receiving androgens, they should be discontinued at least a month prior to the initiation of preparative therapy and after a thorough radiological investigation for liver adenomas, which can cause life-threatening complications during HCT (Kumar et al, 2004).

Prior to HCT, all patients should have an extensive evaluation for organ function and opportunistic infections. To reduce the risk of infections, patients with neutropenia should receive G-CSF. Also, consideration should be given to prophylaxis therapy with a mould-active azole (e.g. voriconazole or posoconazole) for a month prior to HCT to help eradicate any undetectable nidus of infection.

Prevention and management of late effects has recently become more crucial for FA patients as a greater proportion are now surviving into adulthood, largely due to major advances in HCT. Long-term follow-up in FA patients is considerably more complex than for patients with acquired illnesses as late effects arise from issues related to FA (e.g. congenital anomalies, endocrinopathies, risk of malignancies, infertility) and its treatment (e.g. iron overload from transfusions, side effects of androgens, and HCT late effects). Although there are no published reports specifically addressing late effects in FA patients, studies in other populations of patients, especially those with cancer or HCT recipients are instructive, including joint recommendations recently made by the EBMT, CIBMTR and the American Society of Blood and Marrow Transplantation (Rizzo et al, 2006).

It is imperative that all FA patients undergo rigorous screening and participate in cancer prevention programs. Although the few reports in the literature cite different relative risks of malignancies in FA patients (Kutler et al, 2003; Rosenberg et al, 2003; Guardiola et al, 2004) it is clear that the risk is extraordinary high, estimated to be 48–500-fold higher than the non-FA population. Solid tumors (particularly squamous cell carcinoma of the upper aerodigestive and anogenital regions) in FA patients begin to appear at a very young age, and the risk appears nonlinear and does not reach a plateau (Rosenberg et al, 2003). There is no clear evidence that the risk for malignancies is higher after HCT but studies are needed to address this question.


Over the last two decades sequential changes to the approach to HCT has resulted in significant HCT outcomes and survival in FA patients. The two pivotal changes have been the addition of FLU to improve engraftment rates and the use of T cell depletion to reduce GVHD. Improved outcomes are also due in part to the fact that FA patients are now coming to HCT prior to the development of high risk features (e.g. advanced MDS, leukaemia, impaired organ function, opportunistic infections). With these improved HCT outcomes, indications for HCT are quite consistent regardless of donor source. Emphasis is now being placed on developing approaches to HCT, which will improve quality of life by reducing late effects, particularly the risk of malignancy, sterility and endocrinopathies. In light of the complexity of FA, it is recommended that HCT for FA be performed at select centres experienced in the care of FA patients with and without transplant, which will also provide long term care monitoring for late effects of the disease and therapeutic interventions employed.