Phase II clinical trial of CEOP/IVE/GDP alternating therapy for peripheral T-cell lymphoma

Current standard of care regimens to treat peripheral T-cell lymphoma (PTCL) fall short in terms of complete response rates and long-term overall survival. A phase II clinical trial recently published by Cai et al.¹ in Genome Medicine investigated the efficacy and safety of alternating regimens of cyclophosphamide, epirubicin, vincristine, prednisolone (CEOP), ifosfamide, epirubicin, etoposide (IVE), and gemcitabine, cisplatin, dexamethasone (GDP; CEOP/IVE/GDP) compared with CEOP alone, in newly diagnosed PTCL patients. Tumors and peripheral blood from this cohort of patients were further analyzed for the presence of mutations that could be used as prognostic biomarkers of responsiveness to treatment, and progression free survival (PFS) using whole-exosome sequencing.

Study design and patient characteristics

• Multicenter, randomized, controlled phase II clinical trial that included 106 patients aged ≥ 16 years, with newly diagnosed, histologically confirmed PTCL, and no prior malignancies, World Health Organization (WHO) performance status ≤ 2, and radiologically measurable disease
• PTCL subtype distribution amongst the total cohort was as follows:
  • PTCL-not otherwise specified (PTCL-NOS): 46.1%
  • Angioimmunoblastic T-cell lymphoma (AITL): 33.3%
  • Anaplastic large cell lymphoma-anaplastic lymphoma kinase negative (ALCL-ALK−): 15.7%
  • Enteropathy-associated T-cell lymphoma: 3.9%
  • Hepatosplenic T-cell lymphoma: 1.0%
• Patients were randomized equally into the CEOP monotherapy group (n = 53) or the study group receiving the combination treatment of CEOP/IVE/GDP (n = 53; Table 1). Patients received therapies in 21-day cycles for a total of six cycles, or until they were withdrawn from the study
• The primary endpoint of the study was complete response rate (CRR) at end of treatment as measured by imaging
• Baseline patient characteristics were well balanced between the two groups

Table 1. Randomization and dosing schedules (21-day treatment cycles)¹

CEOP, cyclophosphamide, epirubicin, vincristine, prednisolone; GDP, gemcitabine, cisplatin, dexamethasone; IVE, ifosfamide, epirubicin, etoposide
CEOP	Day 1 Cyclophosphamide 750 mg/m2 Epirubicin 70 mg/m2 Vincristine 1.4 mg/m2 (maximum 2 mg)	Day 1–5 (Cycles 1 and 4): Prednisone 60 mg/m2 (maximum of 100 mg)	—
IVE	Days 1–3 Ifosfamide 2,000 mg/m2 Epirubicin 70 mg/m2	Days 1–3 (Cycles 2 and 5):  Etoposide 100 mg/m2	—
GDP	Days 1–8 Gemcitabine 1,000 mg/m2	Days 1–3:  Cisplatin 25 mg/m2	Days 1–4 (Cycles 3 and 6):  Dexamethasone 40 mg

 Results

Overall response to treatment and incidence of adverse events

• Seventy patients completed the full six cycles of treatment, 42 in the CEOP/IVE/GDP arm, and 28 in the CEOP monotherapy group
• Addition of IVE/GDP did not increase the efficacy of the CEOP regimen
• Overall CRR, PFS, and overall survival were similar between the CEOP and the CEOP/IVE/GDP groups (Table 2)
• The incidence of Grade 3 and 4 adverse events was similar between the two arms (Table 3)

Table 2. Treatment response outcomes¹

CEOP, cyclophosphamide, epirubicin, vincristine, prednisolone; GDP, gemcitabine, cisplatin, dexamethasone; IVE, ifosfamide, epirubicin, etoposide; OS, overall survival; PFS, progression-free survival Statistically significant p values of < 0.05 are shown in bold
Outcome	CEOP (N = 51)	CEOP/IVE/GDP (N = 51)	p value
Interim response rates, %
Overall response	60.8	84.3	0.008
Complete response	35.3	33.3	0.834
Partial response	25.5	51.0	0.008
End of treatment response rates, %
Overall response	49.0	72.5	0.015
Complete response	31.4	37.3	0.532
Partial response	17.6	35.3	0.042
Median PFS, months	9.2	15.4	0.122
2-year PFS, %	25.4	25.0	—
Median OS, months	21.9	24.3	0.178
2-year OS, %	47.1	49.2	—

 Table 3. Incidence of adverse events¹

ALT, alanine aminotransferase; AST, aspartate aminotransferase; CEOP, cyclophosphamide, epirubicin, vincristine, prednisolone; GDP, gemcitabine, cisplatin, dexamethasone; IVE, ifosfamide, epirubicin, etoposide
Adverse event	CEOP (N = 51)			CEOP/IVE/GDP (N = 51)
	Grade 1–2	Grade 3	Grade 4	Grade 1–2	Grade 3	Grade 4
Hematological events (%)
Neutropenia	24	18	45	22	18	43
Thrombocytopenia	20	2	6	24	4	4
Anemia	61	10	0	67	8	2
Non-hematological events (%)
Infection	20	12	2	18	8	4
Nausea or vomiting	37	2	0	29	4	0
Fatigue	67	2	0	73	2	0
Headache	10	2	0	25	2	0
Mucositis	29	2	0	25	0	0
ALT or AST increase	16	2	0	18	4	0

 Identification of mutations associated with poor PTCL prognosis

• To identify mutations associated with clinical prognosis of PTCL, genomic DNA was isolated from tumor tissue of 27 patients and peripheral blood collected from 21 patients and analyzed using whole-exosome sequencing
• Thirty-three genes were analyzed (Table 4), and the most commonly identified mutations were in genes associated with
  • epigenetics (histone modification, DNA methylation, chromatin remodeling)
  • tumor suppression
  • inflammatory pathways (Janus kinase/signal transducer and activator of transcription)
  • transcription regulators
• Notably, mutations in the histone modification genes KMT2D, CREBBP, and IDH2 correlated with fewer end of treatment responses, and inferior to poor PFS (Table 4)
•  According to histological subtypes, the most frequent mutated gene was KMT2D (7/34, 20.6%) in PTCL-NOS, TET2 (9/18, 50.0%) in AITL, and DNMT3A (2/7, 28.6%) in ALCL-ALK−. EP300, TET2, and ARID1B mutations had a significantly higher proportion in AITL than in PTCL-NOS (22.2% vs 2.9%, p = 0.043; 50.0% vs 14.7%, p = 0.010; 33.3% vs 2.9%, p = 0.003, respectively)

Table 4. Most common mutations identified in PTCL tumors¹

EOT, end-of-treatment; JAK, Janus kinase; OS, overall survival; PFS, progression-free survival; PTCL, peripheral T-cell lymphoma; STAT, signal transducer and activator of transcription
Class of genes	Most common mutations (% incidence)	Mutations associated with fewer EOT responses	Mutations associated with poor/inferior PFS	Mutations associated with poor/inferior OS
Histone modification	KMT2D (22.6) EP300 (8.1) CREBBP (4.8) KMT2A (3.2) SETD2 (1.6)	KMT2D CREBBP	KMT2D CREBBP	CREBBP
DNA methylation	TET2 (22.6) DNMT3A (11.3) IDH2 (4.8) TET1 (1.6)	IDH2	IDH2	IDH2
Chromatin remodeling	ARID1B (14.5) ARID1A (9.7) ARID2 (1.8) CHD8 (1.8)	—	ARID1B	—
Tumor suppressors	TP53 (6.5) ATM (4.8) MGA (3.2) NF1 (1.6)	—	—	—
JAK-STAT pathway	SOCS1 (4.8) JAK3 (1.6) STAT3 (1.6)	—	—	—
Transcription regulation	ASXL3 (1.6) PRDM1 (1.6)	—	—	—
Others	RHOA (11.3) NOTCH1 (1.6)	—	—	—

 Conclusions

The results from this phase II clinical trial indicate that there is no survival or CRR advantage to the addition of IVE/GDP to CEOP regimens in newly diagnosed PTCL patients. However, this study did identify several mutations associated with PTCL that predicted clinical outcomes. In particular, KMT2D, CREBBP, IDH2, and ARID1B, which are associated with histone modification, DNA methylation, and chromatin remodeling, were found in patients with fewer end of treatment responses and inferior to poor PFS. Mutations in this subset of genes have also been correlated with poor prognosis in other lymphomas. Further validation of these results might facilitate the use of these mutations as biomarkers for clinical prognosis of PTCL in newly diagnosed patients.