- Top of page
- Patients and methods
The management of patients with severe aplastic anaemia (SAA) who do not have a matched sibling donor and fail a course of horse anti-thymocyte globulin (h-ATG)/ciclosporin (CsA) is uncertain. Repeated courses of ATG-based immunosuppression are often employed; in children and increasingly in adults, alternative donor haematopoietic stem cell transplantation is an option. We analysed the success rate of retreatment with rabbit ATG (r-ATG)/CsA in 43 patients treated at our institution in the last 5 years; 22 were refractory (20 adults; two children) to h-ATG/CsA-based regimens and 21 (17 adults; four children) had relapsed after h-ATG/CsA-based regimens. The overall response rate was 30% in patients who were refractory to h-ATG and 65% in patients who had relapsed following h-ATG. The 1000-d survival in patients who responded to r-ATG was 90% compared with 65% in non-responders. Six patients developed a clonal haematological disorder; two were responders, two were non-responders and in two the evolution occurred before the response could be assessed at 3 months following r-ATG. Thirteen patients died; three were responders, six were non-responders and four patients died prior to 3 months when response was assessed. In our study, the response rate in refractory patients was inferior to what has been previously reported.
Due to age or the absence of a histocompatible donor, the majority of patients with SAA do not undergo haematopoietic stem cell transplant (HSCT) from a matched sibling and are treated instead by immunosuppression, usually the combination of a horse anti-thymocyte globulin (h-ATG) and ciclosporin (CsA) (Frickhofen & Rosenfeld, 2000). A favourable haematological response to h-ATG and CsA is observed in approximately two-thirds of patients; and the one-third who do not respond are considered refractory. Relapse among responders occurs frequently, estimated at about 35% at 5 years in one large study, with relapse defined as the requirement for further immunosuppressive drug treatment (Rosenfeld et al, 2003).When a more liberal definition was used, based on resumption of transfusions after achieving transfusion independence, relapse was observed in 12% of cases at 3 years (Bacigalupo et al, 2000). Both refractory aplastic anaemia and relapse are frequently treated with further courses of ATG. In patients who lack a histocompatible sibling, alternative donor HSCT is usually sought, with a higher success rate reported in younger patients (Georges & Storb, 2002; Bacigalupo et al, 2005).
The response rates of retreatment with h-ATG or rabbit ATG (r-ATG) in refractory patients or relapsed patients have varied significantly, from 22% to 77% (Marsh et al, 1987; Means et al, 1988; Stein et al, 1994; Tichelli et al, 1998; Di Bona et al, 1999). In three studies of patients refractory to a course of anti-lymphocyte globulin alone, retreatment resulted in a response in four of 18 (22%), three of eight (38%) and 16 of 25 (64%) cases (Marsh et al, 1987; Means et al, 1988; Tichelli et al, 1998). Retreatment with rabbit anti-lymphocyte globulin following horse anti-lymphocyte globulin failure resulted in response of two of eight (25%) cases (Stein et al, 1994). There are fewer studies assessing the success of retreatment with r-ATG in patients who fail to respond to the combination of h-ATG and CsA. In one report, 30 patients unresponsive to h-ATG were retreated with r-ATG; response was reported in 23 (77%) (Di Bona et al, 1999). The response rate with retreatment in relapsed patients following h-ATG have been more consistent, with results in the 50–60% range (Schrezenmeier et al, 1993; Tichelli et al, 1998).
Rabbit anti-thymocyte globulin has been used in conditioning regimens for solid organ and haematopoietic stem cell transplant and as a primary treatment in aplastic anaemia, mainly in Europe. Due to the few studies using r-ATG following failure to respond to h-ATG/CsA and the large variability of response with retreatment in this setting, we conducted a retrospective analysis at our institution of patients who received r-ATG after failing h-ATG/CsA. We also analysed the success of r-ATG in patients who relapsed after h-ATG/CsA. Here, we summarise the results of r-ATG retreatment in patients with SAA who were refractory to or who had relapsed following treatment with h-ATG.
- Top of page
- Patients and methods
Although immunosuppressive therapy has decreased the morbidity and improved the survival of patients with SAA, unresponsiveness to h-ATG + CsA and relapse after successful treatment remain important problems. Proposals have been made to standardise response criteria (Camitta, 2000), but we prefer the simple definition of no longer meeting response parameters, which we and others have validated as strongly correlating to freedom from requirement for transfusions and susceptibility to infection, as well as long term (Kim et al, 2003; Rosenfeld et al, 2003). The outcome of patients who are unresponsive to immunosuppression is poor, with death from complications of pancytopenia common (Rosenfeld et al, 2003). The management of such cases is uncertain. Alternative donor HSCT has resulted in a survival rate of about 70% in patients of younger age (Deeg et al, 2001; Vassiliou et al, 2001); however, most studies have been small, follow up has been relatively short, and the optimal conditioning regimen has not been defined. In a recent report from the European Group for Blood and Marrow Transplantation on SAA Working Party, 38 patients with refractory SAA underwent an unrelated donor HSCT following a radiation-free conditioning; engraftment and survival was superior in younger patients (<14 years of age) when compared with older patients (Bacigalupo et al, 2005). Overall, outcomes in adults following alternative donor HSCT have not been as favourable, and instead of transplant, repeat courses of immunosuppression are often administered.
In an Italian multicentre study, 30 aplastic anaemia patients who had failed h-ATG + CsA received second-line treatment with r-ATG and CsA (Di Bona et al, 1999). r-ATG was given at 3·5 mg/kg, i.v. over 6–8 h from days 1 to 5; ciclosporin 5 mg/kg p.o., from days 1 to 180; 24 patients received granulocyte colony-stimulating factor at 5 μg/kg, s.c. from days 1 to 90. The median interval between the first and second courses of ATG was 151 d. Overall response, defined as transfusion-independence, was achieved in 23 of 30 (77%) patients after a median of 95 d and overall survival was 93% with a median follow up of 914 d, with no patient having relapsed at the time of publication. In a much smaller American experience, r-ATG ± ciclosporin was administered in combination with prednisone ± androgens in eight patients who had previously received h-ATG; two (25%) responded (Stein et al, 1994). In our experience, response to a course of r-ATG in patients with refractory SAA was about 30%, considerably lower than in the Italian study. There are several possible explanations for this difference. First, the median time from last course of immunosuppression to r-ATG was about 2 months longer in our study compared with that reported by Di Bona et al (1999) (205 d vs. 151 d respectively). Since responses to h-ATG can occur up to 6 months after treatment, it is possible that some patients who responded to r-ATG in the Italian study were actually late responders to h-ATG. Secondly, the definition of response was based on transfusion-independence, whereas we defined response by peripheral blood count criteria. Although cessation of transfusion requirements almost always correlates to haematopoietic recovery, it still remains a subjective criterion that can be influenced by the physician's tolerance and the patient's adaptability to chronically low blood counts.
The median survival in refractory patients to two courses of immunosuppression in our cohort was not reached; nevertheless, the impact on quality of life with supportive care can be considerable. Therefore, the probability of survival must be weighed against the burden of supportive care on an individual basis as other treatment options are considered, such as alternative donor HSCT.
The response rate of retreatment with r-ATG in patients who relapsed following h-ATG/CsA is comparable with those reported by others, in the range of 50–60%. The occurrence of relapse or lack of response to retreatment following relapse has been associated with a worse survival compared with patients who either did not relapse or who did respond to retreatment (Schrezenmeier et al, 1993). Our experience has been that relapse following h-ATG/CsA does not equate to a worse survival, compared with those who do not relapse (Rosenfeld et al, 2003). However, here the survival at 1000 d was inferior among non-responders to retreatment with r-ATG compared with responders, but this difference in our study did not reach statistical significance. The difference in response and survival between patients who were refractory and those who had relapsed following h-ATG could reflect different pathogenesis between the two groups. A prior response to h-ATG implies an immunological pathophysiology for marrow failure and responsiveness to repeated treatment may be expected. In contrast, refractoriness to h-ATG may be due to inadequacy of the initial immunosuppressive treatment regimen (r-ATG may be more potent or provide different specificities of antibodies); severe stem cell depletion (from an immunological cause) precluding haematopoietic recovery; or a non-immune basis for disease.
Systematic studies comparing the efficacy of h- and r-ATG in SAA are lacking. In solid organ transplantation, r-ATG has been shown to be superior to h-ATG in preventing graft rejection in randomised studies (Gaber et al, 1998; Brennan et al, 1999). In aplastic anaemia, the use of r-ATG in a heterogeneous population could account for the variability of the response rates observed after a repeat course of immunosuppression. We are currently conducting a prospective study of r-ATG in patients who failed h-ATG to better define the response rate of r-ATG in the salvage setting; and a randomised prospective study, to be initiated this year at our institution, will determine the response rate of r-ATG as initial treatment as well as compare the efficacy between h- and r-ATG in treatment naïve patients.