Oral Azacytidine in Patients with Relapsed/Refractory Angioimmunoblastic T-Cell Lymphoma: Final Analysis of the Oracle Phase III Study

Abstract

Background: Angioimmunoblastic T-cell Lymphoma (AITL) and nodal lymphomas of follicular helper T-cell origin have in common frequent mutations affecting genes regulating DNA methylation and hydroxymethylation such as TET2, DNMT3A and IDH2. Preliminary studies suggested efficacy of the hypomethylating agent 5-azacytidine (Blood. 2018; 132(21):2305-09) in relapsed/refractory AITL patients. We present here the final results of the ORACLE study (NCT03593018), a phase III trial comparing CC-486, an oral form of 5-azacytidine, to single agent treatment chosen by the investigator. Patients & methods: Eighty-six patients with relapsed/refractory AITL or nodal follicular helper T-cell lymphoma were randomized between CC-486 (n=42) and investigator's choice (gemcitabine, n=24, bendamustine n=16, romidepsin n=4) between November 2018 and February 2021. CC-486 was given at a daily dose of 300 mg/day (200 mg/day in Asian patients, based on previous phase I pharmacokinetics results) every day for 14 days out of 28 day-cycles, until progression or unacceptable toxicity. Gemcitabine and bendamustine were both administered for 6 cycles, while romidepsin was administered until progression. Patients were evaluated for response using 2014 international working group CT-based criteria. The primary endpoint was progression-free survival (PFS) based on local assessment of progressive disease, and overall survival was a key secondary endpoint in this study. Main secondary endpoints were PFS based on central review, overall response rate (ORR), complete response (CR) rate and safety. Statistical hypotheses were built on a PFS of 12 months in the experimental arm vs 5 months in the standard arm (HR=0.417), with a one-sided alpha risk of 0.025 and 90% power. Results: Median age was 69 years (IQR 62-76), and the male/female ratio was 1.4. Patients had received a median of 2 (IQR 1-2) previous lines of treatment, 90.6% of them had advanced stage (III-IV), 57% had elevated LDH, 23.3% had an ECOG performance status >1, 31.4% had an IPI score 4-5 and 35.8% had a bone marrow involvement. Patients received a median number of 5.5 (IQR 2;14) cycles of CC486, and respectively 2 (1-4), 3.5 (2-6) and 13.5 (5-25) cycles of gemcitabine, bendamustine and romidepsin. The primary endpoint was analysed after a follow-up of 14.4 months. Median PFS in the CC-486 arm was 5.6 (95%CI, 2.66-8.11) months vs 2.8 (95%CI, 1.87-4.83) months in the standard arm (stratified log-rank test p=0.0421), with a hazard ratio of 0.634 (95%CI, 0.38; 1.07), which did not reach the required significance of p<0.025. Median overall survival was 18.4 months (95%CI, 12.9-31.5) in the CC-486 arm vs 10.3 (95%CI, 4.2-13.5) in the standard arm, with a hazard ratio of 0.557 (95%CI, 0.323-0.961). Best overall response and complete response rates at 3 months in the CC-486 arm were 33.3% (95%CI, 19.6%-49.5%) and 11.9% (95%CI, 4%-25.6%) respectively, vs 43.2% (95%CI, 28.3%-59.0%) and 22.7% (95%CI, 11.5%-37.8%) in the standard arm. The most frequent (>40%) treatment emergent adverse events (TEAEs) for the CC-486 versus (vs) standard arm respectively were: blood and lymphatic system disorders (76.2% vs 93%) [neutropenia (42.9% vs 58.1%) and thrombocytopenia (23.8% vs 48.8%)], infections (35.7% vs 67.4%), gastrointestinal disorders (71.4% vs 55.8%). At least one grade 3/4 TEAE occurred in 76.2% patients in CC-486 arm vs 97.7% in the standard arm, and at least one serious TEAE occurred in 26.2% patients in CC-486 arm vs 44.2% patients in the standard arm. Central pathological review confirmed the diagnosis of AITL and TFH PTCL in 69 and 9 cases. We detected TET2, RHOA, DNMT3A and IDH2 mutations in 68/77 (88%), 51/75 (68%), 24/75 (32%) and 25/75 (33%) samples. None of these mutations were associated with PFS or OS of the whole population or the 5-azacytidine treated patients. Conclusion: The trial did not meet its primary endpoint, most likely due to an optimistic hypothesis of PFS improvement, resulting in a study including 86 patients which could be underpowered to detect a clinically meaningful difference. 5-azacytidine had a favourable safety profile compared to standard of care and was associated with prolonged overall survival, suggesting that this drug might have a role in the treatment of TFH PTCL and could be investigated in combination with other agents. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal