Results from the phase II P-2001 study on pevonedistat plus azacytidine for high-risk MDS

Pevonedistat (pevo) is a first-in-class inhibitor of the neural precursor cell expressed, developmentally downregulated 8 (NEDD8)-activating enzyme, and acts by inhibiting the ubiquitination of specific proteasome-related proteins.¹ This action leads to tumor cell death and apoptosis. In vivo studies have shown that pevo works synergistically with the hypomethylating agent (HMA) azacytidine (aza), to kill leukemic clones in mouse models of acute myeloid leukemia (AML).¹

To further explore the clinical efficacy and safety of pevo in patients with high-risk myelodysplastic syndromes (MDS), chronic myelomonocytic leukemia (CMML) or low-blast AML, Mikkael Sekeres and colleagues designed the phase II P-2001 trial (NCT02610777). Initial results from the intention-to-treat population showed a nonsignificant trend for longer event-free survival (EFS) in patients receiving pevo plus aza than aza alone (p = 0.076).¹ Further clinical and cytogenetic risk analyses for the high-risk MDS cohort were presented by Mikkael Sekeres, during the 62nd American Society of Hematology (ASH) Annual Meeting and Exposition¹ and are summarized below.

Study design

• Phase II, randomized, open-label, global, multicenter trial that included patients with high-risk MDS (n = 67), high-risk CMML (n = 17) or low-blast AML (n = 36), who were ineligible for allogeneic transplantation and had not previously been treated with HMAs
• In this analysis, only patients with high-risk MDS were considered (n = 67)
• Patients were randomized 1:1 to one of the following arms:
  • Pevo + aza: aza (75 mg/m² intravenously [iv] or subcutaneously [sc] on Days 1–5, 8, 9, in combination with pevo (20 mg/m² iv on Days 1, 3, 5)
  • Aza: aza monotherapy (75 mg/m² iv or sc on Days 1–5, 8, 9)
• The primary endpoint of this analysis is overall survival (OS), while secondary endpoints include EFS, 6- and 12-month survival rates
• Patient baseline characteristics were well balanced between the two arms as shown in Table 1
• Analyses of patients included clinical and cytogenetic risk factors including:
  • Revised International Prognostic Scoring System (IPSS-R)
  • Genetic risk factors
  • The Cleveland Clinic model formula (defines four patient risk groups [low, intermediate-1, intermediate-2, high] based on age, IPSS-R score, and mutations in EZH2, SF3B1 and TP53 

Table 1. Key patient baseline characteristics¹

High-risk MDS patients	Pevo + Aza (n = 32)	Aza (n = 35)
Aza, azacytidine; IPSS-R, revised International Prognostic Scoring System; MDS, myelodysplastic syndrome; pevo, pevonedistat; RAEB, refractory anemia with excess blasts; WHO, World Health Organization.
Males, %	75	71
Median age, years (range)	75 (47–91)	70 (44–84)
Disease type, %   De novo   Secondary	100 0	97 3
WHO tumor classification, %   RAEB-1   RAEB-2	38 56	43 51
IPSS-R risk category, %   Intermediate   High   Very high	34 34 31	26 29 46
Median time from diagnosis, months (range)	2.30 (0.2–58.4)	1.74 (0.6–79.1)

Results

• When classifying patients based on the IPSS-R:
  • Pevo + aza led to significantly prolonged median EFS than aza alone (20.2 versus 14.8 months; HR = 0.539; 95% confidence interval [CI], 0.292–0.995; p = 0.045)
  • The longer EFS was more evident in patients with high-risk (n = 21; HR = 0.53; 95% CI, 0.17–1.72) or very high-risk IPSS-R MDS (n = 26; HR = 0.47; 95% CI, 0.19–1.18) when treating with pevo + aza than aza alone
  • OS was not significantly different between the two arms with pevo + aza resulting in a median OS of 23.9 months and aza alone of 19.1 months (HR = 0.701; 95% CI, 0.386–1.273; p = 0.240)
• When using the Cleveland Clinic model formula that incorporates both clinical and genetic factors:
  • Similarly, pevo + aza led to a longer EFS in patients with high-risk MDS than aza alone (20.2 vs 11.7 months; HR = 0.388; 95% CI, 0.166–0.902; p = 0.023)
  • OS was not significantly different between the two arms but was trending close to significance (pevo + aza, 24.2 vs aza, 14.2 months; HR = 0.447; 95% CI, 0.190–1.050; p = 0.056)
• Response outcomes between the two arms are shown in Table 2. Overall, complete response (CR) rate was almost double in the pevo + aza arm when compared with aza monotherapy (p = 0.05) but overall response rate (ORR) was not significantly different between the arms (p = 0.065; Table 2)
• Patients with high-risk MDS had higher duration, but not rate, of red blood cell and platelet transfusion independence after receiving pevo + aza than aza alone (23.3% vs 11.6%; HR = 0.11; 95% CI, 0.01–0.94; p = 0.016)

Table 2. Response outcomes¹

Response outcomes (n = 59)	Pevo + Aza	Aza	p value
Aza, azacytidine; CI, confidence interval; CR, complete response; DoR, duration of response; HI, hematological improvement; NE, not evaluable; pevo, pevonedistat; PR, partial response; ORR, overall response rate.
ORR, %	79	57	0.065
CR, %	52	27	0.050
PR, %	3	13	—
HI, %	24	17	—
Median DoR, months (95% CI)	34.6 (11.53–34.60)	13.1 (12.02–NE)	—
Median time to first CR or PR among responders, months (range)	3.83 (1.8–25.8) (n = 16)	4.29 (2.0–13.2) (n = 12)	—

• Adverse event (AE) rates were similar or lower for patients receiving pevo + aza. Exposure-normalized AE rates were lower in the pevo + aza arm and are shown in Table 3

Table 3. Exposure-normalized AEs¹

Exposure-normalized AE rates*	Pevo + Aza (n = 32)	Aza (n = 35)
AE, adverse event; aza, azacytidine; pevo, pevonedistat *AE rates normalized by mean number of aza cycles.
Any AE (n)	1.96 (32)	3.27 (35)
Treatment-related AE (n)	1.35 (22)	2.52 (27)
Serious AE (n)	1.47 (24)	1.87 (20)
Treatment-related serious AE (n)	0.25 (4)	0.28 (3)
Grade ≥ 3 AE (n)	1.84 (30)	2.71 (29)

• The presence of poor-risk or adverse-risk (such as TP53 mutations) cytogenetics did not affect the observed pevo + aza clinical efficacy when compared with aza alone

Conclusions

The results of this phase II trial indicate that pevo + aza combination is more efficacious than aza monotherapy in patients with high-risk MDS, leading to more prolonged EFS, higher CR rate, and longer response durability. Nevertheless, the OS was not significantly different between the arms. It should be noted that the study was originally powered on EFS as the primary endpoint and was not meant to detect changes in OS. However, this changed later based on regulatory feedback, with OS being later considered as the primary endpoint. This treatment combination seems to be efficacious even in patients with poor-risk or adverse-risk cytogenetics (TP53 mutations). Based on these promising preliminary results, the phase III trial PANTHER (NCT03268954) has been designed and is currently at the patient accrual stage.