Azacitidine maintenance after HSCT for AML/MDS provides no improvement in relapse-free survival in phase III study

Allogeneic hematopoietic stem cell transplant (HSCT) plays a pivotal role in the treatment of fit patients with acute myeloid leukemia (AML) or myelodysplastic syndromes (MDS), however, disease relapse after transplant occurs in around 40% of cases.¹ Whilst the hypomethylating agent azacitidine is the standard first-line treatment option for patients who are ineligible for intensive chemotherapy or HSCT, its ability to reduce relapse risk in the post-transplant setting is yet to be proven.

Here, we summarize the results of a phase III trial of azacitidine maintenance following transplant in patients with high-risk AML/MDS (NCT00887068) published by Betül Oran and colleagues in Blood Advances.¹

Study design and methods

A phase III, open-label, randomized study of azacitidine maintenance in a total of 187 patients, 18–75 years of age, with high-risk AML/MDS achieving a complete remission (CR) following HSCT.

Eligibility criteria

• For patients with AML: high-risk features, induction failure, relapsed disease, or in ≥ second remission at HSCT
• For patients with AML in first CR: high-risk features (including chromosome 5 or 7 abnormalities, complex karyotype, or FLT3 mutations)
• For patients with MDS: ≥ intermediate-1 disease risk features according to the International Prognostic Scoring System (IPSS)

Dosing

Patients were randomly assigned 1:1 to either:

• Treatment arm (n = 93): 12 cycles of azacitidine, 32 mg/m²/day for 5 days every 28 days
• Control arm (n = 94): no intervention, observation only

Primary endpoint

• Relapse-free survival (RFS), defined as time from HSCT to relapse or death

Secondary endpoints

• Overall survival (OS)
• Incidence of acute graft-versus-host disease (aGvHD)
• Toxicity

Results

Selected patient characteristics are shown in Table 1. Of the 93 patients enrolled in the treatment arm, 87 started azacitidine therapy. Although most characteristics were well balanced between arms, the azacitidine arm had more patients with a hematopoietic comorbidity index ≥ 4 (Table 1).

Patients received a median of four treatment cycles. A total of 63 patients were withdrawn from the study due to relapse (n = 29), toxicity (n = 11), patient preference (n = 9), infection (n = 7), logistics (n = 5), or GvHD (n = 2).

Table 1. Selected patient characteristics¹

Characteristic	Observation (n = 93)	Azacitidine (n = 87)
AML, acute myeloid leukemia; HCT-CI, hematopoietic comorbidity index; HSCT, hematopoietic stem cell transplant; CR, complete remission; CR1/CR2, first/second CR; MDS, myelodysplastic syndromes.
Median age, years (range)	57.5 (20–75)	57 (19–72)
Sex, male, %	60.6	58.6
MDS, %	26.6	25.3
Cytogenetics, %    Good risk    Intermediate risk    Poor risk	16.0 44.7 39.3	9.2 37.9 52.9
AML disease status at HSCT, %    CR1/CR2 with count recovery    CR without count recovery    Active disease	39.1 17.4 43.5	52.3 26.2 27.5
MDS disease status at HSCT, %    CR    Active disease	24.0 76.0	18.2 81.8
HCT-CI, %    0–1    2–3    ≥ 4	39.4 39.4 21.3	32.2 25.3 42.5

Efficacy

Patient outcomes are summarized in Table 2.

Table 2. Outcome data¹

Outcome	Observation	Azacitidine	p value
OS, overall survival; RFS, relapse-free survival.
Median follow-up (n = 88), years	4.06	4.60	–
Median RFS, years	1.28	2.07	0.19
Median OS, years	3.56	2.52	0.43

• Azacitidine maintenance offered no improvement in median RFS (HR, 0.77; 95% CI, 0.51–1.14; p = 0.19) or median OS (HR, 0.84; 95% CI, 0.56–1.28; p = 0.43), compared with the control arm
• The cumulative incidence of relapse was similar between the azacitidine maintenance and control arms, with 1-year incidences of 41% (95% CI, 31–51) and 39% (95% CI, 29–49), respectively
• One-year transplant-related mortality was low for both groups (4.3% in the azacitidine arm and 5.3% in the control arm)
• The incidence of aGvHD was also similar between arms, with low incidence of aGvHD Grade 3–4 (4.3% and 2.1% for azacitidine and control arms, respectively)

Toxicity

• The median number of adverse events (AEs) for the azacitidine arm was three, and for the control arm was two; there were no unexpected AEs in either arm
• Overall, 87.4% vs 72.3% of patients experienced ≥ 1 AE in the azacitidine and control groups, respectively
• Grade ≥ 3 AEs are summarized in Table 3. Of the 91 AEs in the azacitidine arm, 79 (87%) were attributed to azacitidine treatment
• The substantial number of AEs observed in the observation arm was attributed to post-transplant complications
• One Grade 5 aspiration pneumonia AE was reported in the azacitidine arm

Table 3. Reported Grade ≥ 3 AEs¹

Grade ≥ 3 AE, no. of events	Observation n = 215	Azacitidine* n = 302
*No Grade 5 hematological AEs; only one Grade 5 toxicity, which was aspiration pneumonia. AE, adverse event.
Hematological	5	58
Thrombocytopenia	1	29
Poor graft function	2	29
Non-hematological	56	33
Infection	19	13
Gastrointestinal	12	0
Hepatic	5	9
Pulmonary	6	4
Skin	5	2

Conclusion

Despite promising results from phase II studies reporting improved outcomes after transplant for patients with MDS/AML treated with hypomethylating agents, in this study, azacitidine maintenance failed to improve RFS for patients with AML/MDS in the post-transplant setting, compared to observation alone. When considering data interpretation, the investigators made several noteworthy observations:

• Median RFS for the study was longer than predicted, and although there was a higher absolute difference between the azacitidine and control arms (2.07 years vs 1.28 years) than the 3-month goal set at design, the difference was not statistically significant due to fewer-than-expected control arm events.
• Many patients failed to complete the 12 cycles of therapy; median cycle number was just four.
• The dose of azacitidine was lower than that given in a pretransplant setting; adapted dosage/scheduling may give better results.