What were the key findings of the RTOG 1114 trial?

← Back

RTOG 1114

Multicenter Randomized Phase II Study of R-MPV-A Chemoimmunotherapy With or Without Low-Dose Whole-Brain Radiotherapy for Newly-Diagnosed Primary CNS Lymphoma

Share Share Copied! Compare

Year of Publication: 2026

Authors: Omuro A et al.

Journal: Neuro-Oncology

Citation: Neuro Oncol. 2026 Feb 1;28(2):371-382

Link: https://doi.org/10.1093/neuonc/noaf221

Bottom Line

Adding low-dose WBRT (23.4 Gy) to R-MPV-A chemoimmunotherapy substantially improved 2-year PFS (78.7% vs 54%, HR 0.47, P=0.007) and complete response rates without producing measurable excess neurocognitive decline; OS difference was numerical only.

        Major Points
        First US multicenter randomized trial supporting a survival benefit for consolidation LD-WBRT after high-dose methotrexate-based induction in PCNSL.
23.4 Gy LD-WBRT preserved neurocognitive function: HVLT-R delayed recall (P=0.001) and recognition (P<0.001) actually improved in the ChemoRT arm, likely reflecting better disease control.
Complete response increased from 76.3% to 92.3% with the addition of LD-WBRT (absolute +16%).
OS curves favored ChemoRT (5-yr 65.8% vs 58.7%) but the difference was not significant, likely due to small sample size and effective salvage therapy.
Provides an effective consolidation option, particularly for elderly patients ineligible for autologous stem-cell transplantation.

      

Design

Study Type: Randomized Controlled Trial

Phase: II

Randomization: 1

Blinding: Open-label (non-blinded)

Enrollment Period: Sep 2011 - Jan 2016

Follow-up Duration: Median 4.6 years (PFS); 5.6 years (OS)

Centers: 26

Countries: United States, Israel

Sample Size: 91

Analysis: Intention-to-treat

Inclusion Criteria

Newly-diagnosed B-cell non-Hodgkin's lymphoma involving the brain (histologically or cytologically confirmed)
No history of immunodeficiency
Adequate organ function
Creatinine clearance >50 mL/min/1.73m²
Karnofsky performance status (KPS) ≥50 (exception: KPS 30-50 eligible if poor status attributable to a neurologic deficit expected to resolve quickly with treatment)
No evidence of systemic non-Hodgkin lymphoma on CT chest/abdomen/pelvis
Bone marrow biopsy prior to treatment with no lymphoma involvement

Exclusion Criteria

Prior immunodeficiency diagnosis
Inadequate organ function
Creatinine clearance ≤50 mL/min/1.73m²
KPS <30
Evidence of systemic NHL on staging imaging
Bone marrow involvement by lymphoma on biopsy

Baseline Characteristics

Characteristic	Control	Active
Arm Label	Chemo (R-MPV-A alone, no WBRT)	ChemoRT (R-MPV-A + 23.4 Gy LD-WBRT)
N	46	44
Median Age	60 (range 21-84)	66 (range 40-78)
Age ≥70	19.6%	31.8%
Sex - Male	50.0%	45.5%
Race - White	76.1%	86.4%
Race - Asian	13.0%	4.5%
Median KPS	80 (range 30-100)	75 (range 40-100)
KPS 30-50	13.0%	13.6%
KPS 90-100	37.0%	34.1%
Neurologic Function - Moderate/Severe	47.8%	47.7%
Diagnosis by Brain Biopsy	82.6%	93.2%
Histology - DLBCL	95.7%	95.5%
MSKCC RPA Class 3	26.1%	29.5%
Ocular Involvement	8.7%	0.0%
CSF Involvement	13.0%	13.6%

Arms

Field	ChemoRT	Control
Intervention	R-MPV induction × 4 cycles (rituximab 500 mg/m² IV d1,15; methotrexate 3.5 g/m² IV d2,16; procarbazine 100 mg/m² PO d2-8; vincristine 1.4 mg/m² IV d2,16, capped at 2.4 mg, stopped after cycle 2) + 23.4 Gy whole-brain radiotherapy in 1.8 Gy fractions (German-helmet field, sparing anterior 2/3 of orbits unless ocular involvement) + high-dose cytarabine 3 g/m²/day d1-2 × 2 cycles consolidation	Same R-MPV induction × 4 cycles + same high-dose cytarabine consolidation × 2 cycles, with NO whole-brain radiotherapy
Duration	Approximately 6-7 months	Approximately 5-6 months

Outcomes

Outcome	Type	Control	Intervention	HR / OR / RR	P-value
Progression-free survival from randomization	Primary	Median 2.1 years; 2-yr PFS 54% (95% CI 39.1-69%)	Median not reached; 2-yr PFS 78.7% (95% CI 66.4-91.1%)	0.47	0.007 (one-sided)
Overall survival	Secondary	5-yr OS 58.7% (44.0-73.4%); 18 deaths	5-yr OS 65.8% (51.2-80.4%); 14 deaths	0.71	0.33 (two-sided)
Complete response rate (confirmed)	Secondary	76.3%	92.3%		Absolute difference +16%
Overall response rate (CR+CRu+PR)	Secondary	86.9%	100%		0.03 (Fisher exact)
3-year cumulative incidence of neurocognitive failure	Secondary	41.6% (21.0-61.2%)	17.9% (6.3-34.1%)	0.52	0.18 (Gray's test)
3-year cumulative clinician-assessed neurotoxicity	Secondary	9.2% (2.9-20.1%)	11.9% (4.3-23.8%)	1.45	0.53
HVLT-R Delayed Recall longitudinal change	Secondary	Less improvement	Greater improvement		0.001
HVLT-R Delayed Recognition longitudinal change	Secondary	Less improvement	Greater improvement		<0.001
EORTC QLQ-C30 Global Health Status	Secondary	Less improvement	Greater improvement over time		0.005
Treatment-related death	Adverse	1 patient ChemoRT (cause unknown) vs 1 patient Chemo (sepsis, deemed unlikely related)
Overall tolerability	Adverse	Grade 3-5 toxicities described as 'in general transient and reversible, consistent with the previously described profile of R-MPV'; the immunochemotherapy regimen was relatively well tolerated in both arms
Detailed Grade 3-5 toxicity	Adverse	Full breakdown in study Appendix 1C; no signal of excess toxicity from LD-WBRT
LD-WBRT adherence	Adverse	5/44 (11%) randomized to ChemoRT did not receive WBRT

Subgroup Analysis

Benefit appeared more pronounced in elderly patients (the population with highest relapse risk after methotrexate-only consolidation), though subgroup analysis was exploratory and underpowered. Adjusted analyses confirmed PFS benefit across MSKCC RPA class, neurologic function, and sex strata.

Criticisms

Phase II design with relatively small sample size (n=90 evaluable), limiting OS conclusions and subgroup power
Open-label, non-blinded design
11% of ChemoRT arm did not receive the assigned radiotherapy
Stratification was by MSKCC RPA rather than age, leading to a 6-year median age difference between arms (60 vs 66)
20% of patients declined neurocognitive testing, and disease progression/salvage confounded longitudinal cognitive analyses
No direct comparison with HDCT/ASCT consolidation, which is increasingly the standard for fit patients
Limited number of OS events with effective salvage therapy may have masked a true OS benefit

Funding

NCI/NIH grants U10CA180868 and U10CA180822 (RTOG / NRG Oncology). No industry support.

Based on: RTOG 1114 (Neuro-Oncology, 2026)

Authors: Omuro A et al.

Citation: Neuro Oncol. 2026 Feb 1;28(2):371-382

Content summarized and formatted by NeuroTrials.ai.

RTOG 1114

(2026)

Objective

Evaluate whether adding consolidation low-dose whole-brain radiotherapy (LD-WBRT, 23.4 Gy) to R-MPV induction plus high-dose cytarabine consolidation improves progression-free survival without unacceptable neurocognitive toxicity in newly-diagnosed primary CNS lymphoma.

Study Summary

• Adding LD-WBRT (23.4 Gy) to R-MPV-A markedly improved median PFS (not reached vs 2.1 years; HR 0.47, 95% CI 0.26–0.87, one-sided P=0.007)
• 2-year PFS: 78.7% (ChemoRT) vs 54% (Chemo)
• Complete response: 92.3% (ChemoRT) vs 76.3% (Chemo); overall response 100% vs 86.9% (P=0.03)
• 5-year OS numerically higher with ChemoRT (65.8% vs 58.7%; HR 0.71, P=0.33, NS)
• No excess clinical neurotoxicity (3-yr cumulative 11.9% ChemoRT vs 9.2% Chemo, P=0.53); HVLT-R delayed recall and recognition favored ChemoRT (P=0.001 and P<0.001)
• Global Health Status improved more over time in ChemoRT (P=0.005)

Intervention

R-MPV induction (rituximab, methotrexate 3.5 g/m², procarbazine, vincristine) × 4 cycles + high-dose cytarabine consolidation × 2 cycles, with or without consolidation low-dose whole-brain radiotherapy (23.4 Gy in 1.8 Gy fractions) between induction and cytarabine.

Inclusion Criteria

Newly-diagnosed B-cell non-Hodgkin's lymphoma involving the brain; KPS ≥50 (or 30–50 if expected to improve with treatment); creatinine clearance >50 mL/min/1.73m²; adequate organ function; no systemic NHL on CT chest/abdomen/pelvis; bone marrow biopsy free of lymphoma.

Study Design

Arms: ChemoRT (R-MPV-A + 23.4 Gy LD-WBRT) vs Chemo alone (R-MPV-A without WBRT)

Patients per Arm: ChemoRT n=44, Chemo n=46 (90 eligible of 91 randomized)

Outcome

• Primary PFS: HR 0.47 (95% CI 0.26–0.87), one-sided P=0.007; adjusted HR 0.36 (0.19–0.70), P=0.002
• 2-yr PFS 78.7% vs 54%
• OS: 5-yr 65.8% vs 58.7%; HR 0.71 (0.35–1.42), P=0.33
• CR 92.3% vs 76.3%; ORR 100% vs 86.9% (P=0.03)
• 3-yr cumulative neurocognitive failure 17.9% (ChemoRT) vs 41.6% (Chemo) (HR 0.52, P=0.18)
• Clinical neurotoxicity 3-yr 11.9% vs 9.2% (HR 1.45, P=0.53)
• HVLT-R Delayed Recall (P=0.001) and Delayed Recognition (P<0.001) favored ChemoRT

Clinical Question

In newly-diagnosed primary CNS lymphoma, does adding consolidation low-dose whole-brain radiotherapy (23.4 Gy) to R-MPV induction plus high-dose cytarabine consolidation improve progression-free survival without causing excess neurocognitive toxicity?

Bottom Line

Major Points

First US multicenter randomized trial supporting a survival benefit for consolidation LD-WBRT after high-dose methotrexate-based induction in PCNSL.
23.4 Gy LD-WBRT preserved neurocognitive function: HVLT-R delayed recall (P=0.001) and recognition (P<0.001) actually improved in the ChemoRT arm, likely reflecting better disease control.
Complete response increased from 76.3% to 92.3% with the addition of LD-WBRT (absolute +16%).
OS curves favored ChemoRT (5-yr 65.8% vs 58.7%) but the difference was not significant, likely due to small sample size and effective salvage therapy.
Provides an effective consolidation option, particularly for elderly patients ineligible for autologous stem-cell transplantation.

Study Design

Study Type: Randomized Controlled Trial
Randomization: Yes
Blinding: Open-label (non-blinded)
Sample Size: 91
Follow-up: Median 4.6 years (PFS); 5.6 years (OS)
Centers: 26
Countries: United States, Israel

Primary Outcome

Definition: Progression-free survival from randomization

Control	Intervention	HR/OR	P-value
Median 2.1 years; 2-yr PFS 54% (95% CI 39.1-69%)	Median not reached; 2-yr PFS 78.7% (95% CI 66.4-91.1%)	0.47 (0.26-0.87)	0.007 (one-sided)

Limitations & Criticisms

Phase II design with relatively small sample size (n=90 evaluable), limiting OS conclusions and subgroup power
Open-label, non-blinded design
11% of ChemoRT arm did not receive the assigned radiotherapy
Stratification was by MSKCC RPA rather than age, leading to a 6-year median age difference between arms (60 vs 66)
20% of patients declined neurocognitive testing, and disease progression/salvage confounded longitudinal cognitive analyses
No direct comparison with HDCT/ASCT consolidation, which is increasingly the standard for fit patients
Limited number of OS events with effective salvage therapy may have masked a true OS benefit

Citation

Neuro Oncol. 2026 Feb 1;28(2):371-382

View Original Publication →

RTOG 1114

Multicenter Randomized Phase II Study of R-MPV-A Chemoimmunotherapy With or Without Low-Dose Whole-Brain Radiotherapy for Newly-Diagnosed Primary CNS Lymphoma

Bottom Line

Major Points

Design

Inclusion Criteria

Exclusion Criteria

Baseline Characteristics

Arms

Outcomes

Subgroup Analysis

Criticisms

Funding

Ask about RTOG 1114