Post‐transplant maintenance therapy with azacitidine and gemtuzumab ozogamicin for high‐risk acute myeloid leukaemia
Year: 2014
Type: article
Abstract: Relapse is a major cause of treatment failure in allogeneic haematopoietic stem cell transplantation (allo-HSCT) for acute myeloid leukaemia (AML), and the clinical outcomes are very poor. Recently, the MD Anderson Cancer Centre group reported promising data on maintenance therapy with low-dose azacitidine (AZA) after allo-HSCT for AML or myelodysplastic syndrome (MDS) (de Lima et al, 2010). AZA has been reported to augment a graft-versus-leukaemia effect without a concomitant increase in graft-versus-host disease (GVHD) (Choi et al, 2010; Sanchez-Abarca et al, 2010) and to increase the anti-leukaemic effect of gemtuzumab ozogamicin (GO) (Pinto et al, 1984; Linenberger et al, 2001; Balaian & Ball, 2006; Nand et al, 2013). Therefore, combination therapy with AZA and GO seems to be a theoretically valid strategy for improving therapeutic response. We conducted a pilot study to evaluate the safety and efficacy of AZA-GO as a maintenance therapy after allo-HSCT. Ten patients with high-risk AML, eight males and two females, were enrolled in this study at a median age of 49·5 years (range: 17–65 years) (Table 1). AZA at 30 mg/m2 was administered intravenously on days 1–7, followed by GO at 3 mg/m2 on day 8. AZA-GO was repeated every 4 weeks or as soon as possible after haematological recovery. The maximum number of treatment cycles was 4 (for detailed information regarding eligibility, treatment and transplant procedure, please see 'Supporting information'). 20 F 31 M PIF MLL-AF6 Flu-BU-ATG- 4 Gy TBI 59 M Refractory relapse AML with MRC Monosomal karyotype 44 F Flu-Mel-AraC- 4 Gy TBI 65 M 57 M 44 M Flu-AraC-CY-ATG- 8 Gy TBI 62 M PIF Leukaemic evolution from CMML 55 M Flu-AraC-CY-ATG- 8 Gy TBI 17 M Flu-Mel-AraC-ATG- 4 Gy TBI The median complete blood results at the start of the first cycle of AZA-GO were: haemoglobin 108·5 (84–115) g/l, neutrophil count 2·835 (1·01–10·42) × 109/l and platelet count 88 (44–138) × 109/l. Eight patients (80%) received allo-HSCT without achieving complete remission (CR). The median duration between allo-HSCT and first cycle of AZA-GO was 78·5 d (range: 24–251 d) (Table 1). All patients were still receiving immunosuppressant(s) at the start of AZA-GO. The patient clinical course of each patient is shown in Fig 1. The median follow-up period was 474 d (range: 264–888 d) from allo-HSCT and 323 d (range: 198–858 d) after the first course of AZA-GO. Four patients (Cases 1–4) eventually relapsed. Among them, three patients (Cases 1, 3 and 4) relapsed after the first cycle of AZA-GO (Fig 1). The median number of treatment cycles was 1·5 in all patients, whereas the median number of treatment cycles was 3 in those patients who did not relapse during treatment. Eight patients had to discontinue AZA-GO: due to relapse in three patients (Cases 1, 3 and 4), infectious episode in one (Case 5), encephalopathy of uncertain aetiology in one (Case 7), non-relapse death in one (Case 9), and patient request in two (Cases 8 and 10) (Fig 1). Most patients could not receive AZA-GO every 4 weeks because of myelosuppression and were treated at a median interval of 53 d (range: 25–91 d). With regard to GVHD, all patients had already developed acute GVHD before the start of AZA-GO (Table 1). In all but one patient, GVHD did not exacerbate during the treatment. In one patient (Case 10), GVHD worsened a few weeks after rapid taper/discontinuation of FK and steroid because of suspicion of relapse. No patient developed severe chronic GVHD. With regard to toxicities, all patients developed grade 4 haematological toxicity and required platelet transfusions (Table 1). However, haematological toxicities were transient. AZA-GO was started in one patient whose platelet count was below 50 × 109/l (44 × 109/l), but the platelets recovered to the pretreatment level in 2 weeks. Because early relapse after HSCT demonstrates poor outcome (Mielcarek et al, 2007; Schmid et al, 2012), maintenance therapy should be started as soon as possible after allo-HSCT. Our result implies that AZA-GO can be safely administered even in patients with a relatively low platelet count. Febrile neutropenia (FN) occurred in three patients, but only one patient (Case 10) received granulocyte colony-stimulating factor support. Seven infectious episodes occurred in five patients, resulting in one patient (Case 5) discontinuing AZA-GO and one patient (Case 1) having to postpone AZA-GO. Furthermore, one patient died of Pneumocystis pneumonia (Table 1). However, all these infectious episodes occurred after haematological recovery. Therefore, we considered that these episodes were mainly due to the patients' immunosuppressive status, not to their myelosuppressive status by AZA-GO. One patient (Case 7) developed encephalopathy of uncertain aetiology, from which he did not fully recover. However, this episode occurred more than 80 d after AZA-GO, and it was unlikely that these drugs induced the encephalopathy (Watanabe et al, 2014). In our cohort, no patients developed sinusoidal obstruction syndrome, although three patients underwent allo-HSCT after relapse following AZA-GO therapy. These favourable results may be associated with the relatively low dose of GO (3 mg/m2). To evaluate the impact of AZA-GO on survival rates, overall survival (OS) and disease-free survival (DFS) were compared with control (no-maintenace) patients, which were randomly selected from the clinical database at our hospital (Table SI). In this retrospective matched-cohort study, although not statistically significant, the patients who received AZA-GO showed higher OS (70·0% in AZA-GO group vs. 59·8% in controls at 1 year, P = 0·138) and DFS (60·0% vs. 42·8% at 1 year, P = 0·222) compared with the control group (Fig S1A, B). However, some important factors that would affect the post-transplant course could not be included in these analyses. For example, the donor source could not be matched because a small number of patients received haploidentical HSCT. In addition, two patients (Cases 5 and 7) maintained CR for a long period despite only a single course of AZA-GO (Fig 1). Such patients might have maintained CR without any post-transplant treatment. The treatment course in most patients could not be repeated every 4 weeks. Thus, strict indication, the optimal treatment cycle and interval of AZA-GO are still matters for discussion. In summary, maintenance therapy with AZA-GO seems to be tolerable in the post-transplant setting. Furthermore, our results suggest that AZA-GO might prolong survival in patients with high-risk AML. Further study is warranted to clarify the efficacy of AZA-GO as a post-transplant maintenance therapy. We are very grateful to the patients who participated in this study. The authors would like to thank the nursing staff at Tokyo Metropolitan Cancer and Infectious Diseases Centre Komagome Hospital, for their excellent patient care. JA, TK, ND, HS, and KO designed the study. MS, GO and KK analysed the data. GO, KK and YN wrote the manuscript. The authors declare no competing financial interests. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Source: British Journal of Haematology
DOI: 10.1111/bjh.13248
Authors: Gaku Oshikawa, Kazuhiko Kakihana, Makoto Saito, Jun Aoki, Yuho Najima, Takeshi Kobayashi, Noriko Doki, Hisashi Sakamaki, Kazuteru Ohashi
Institution: Tokyo Metropolitan Komagome Hospital
Language: English
Subfield: Hematology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2051983014