Bortezomib plus R-CHOP for distinct molecular DLBCL subtypes: Results from the phase III trial REMoDL-B

On 5 April 2019, Andrew Davies from the University of Southampton, Southampton, UK, and colleagues, published in the Lancet Oncology results from the phase III clinical trial REMoDL-B (randomised evaluation of molecular guided therapy for diffuse large B-cell lymphoma with bortezomib; NCT01324596).

In this multicenter, randomized, open-label, adaptive, superiority, phase III trial, the efficacy of bortezomib addition to standard chemoimmunotherapy for diffuse large B-cell lymphoma (DLBCL) was investigated. Initially, all patients received one cycle of standard rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone (R-CHOP) treatment. Following that, gene-expression analysis was undertaken, patients were stratified by molecular DLBCL subtype (germinal center B-cell [GCB], activated B-cell [ABC], or unclassified) and further randomized to continue R-CHOP alone or with the addition of bortezomib. The primary endpoint of the study was 30-month progression-free survival (PFS) for the GCB and ABC populations. Secondary endpoints included, 30-month PFS according to cell-of-origin subgroup, overall survival (OS), disease-free survival (DFS), response rates and duration, and toxicity.

Study design & baseline characteristics

• N = 918 patients with previously untreated, histologically-confirmed DLBCL, with an Eastern Cooperative Oncology Group (ECOG) performance status ≤ 2, and bulky stage I or disease stage II−IV,
• Molecular DLBCL subtypes:
  • ABC: 26.6% of patients (n = 244)
  • GCB: 51.7% of patients (n = 475)
  • Unclassified: 21.7% of patients (n = 199)
• All patients initially received one 21-day R-CHOP cycle and then were randomly 1:1 assigned to R-CHOP or R-CHOP plus bortezomib (RB-CHOP) as below:
  • R-CHOP (n = 459; five additional 21-day cycles):
    • R: 375 mg/m² intravenously (IV) on Day 1
    • C: 750 mg/m² IV on Day 1
    • H: 50 mg/m² IV on Day 1
    • O: 1.4 mg/m² (maximum 2 mg) IV on Day 1
    • P: 100 mg orally once daily on Days 1−5
  • RB-CHOP (n = 459; five 21-day cycles):
    • R-CHOP as above
    • Bortezomib: 1.3 mg/m² IV or 1.6 mg/m² subcutaneously on Days 1 and 8 for cycles 2−6
  • Further treatment cycles were given when patient neutrophils had recovered to 1.0 × 10⁹/L and platelets to 100 × 10⁹/L
• Primary analysis on the  modified intention-to-treat (mITT) populations comprising the GCB and ABC subgroups
• Baseline characteristics were well balanced between the R-CHOP and RB-CHOP arms:

DLBCL molecular subtype	R-CHOP arm, n (%)	RB-CHOP arm, n (%)
Activated B cell	121 (26·4%)	123 (26.8%)
Germinal centre B cell	240 (52·3%)	235 (51.2%)
Unclassified	98 (21.4%)	101 (22.0%)

 Key findings

• The following statistically significant clinical differences between molecular DLBCL subgroups were observed:

	ABC subgroup (n = 244)	GCB subgroup (n = 475)	Unclassified subgroup (n = 199)	P value (ABC vs GCB)
Age, years	67 (22–86)	63 (20–82)	63 (20–84)	0.0045
Bone marrow involvement	33/240 (13.8%)	66/465 (14.2%)	42/191 (22.0%)	0.017
Bulky disease >10 cm	50/241 (20.7%)	158/467 (33.8%)	55/198 (27.8%)	< 0.0001

• Nineteen patients were identified with primary mediastinal lymphoma. Of those, 74% (n = 14) had been allocated to the GCB group and 26% (n = 5) to the unclassified group
• After a median follow-up of 30 months:
  • 30-month PFS rates:
    • R-CHOP: 70.1% (95% CI, 65.0−7)
    • RB-CHOP: 74.3% (95% CI, 69.3−7)
    • Comparison: HR = 0.86; (95% CI, 0.65−1.13); P = 0.28
      • Adjusted HR = 0.84; (95% CI, 0.64−1.11); P = 0.23
    • OS events (i.e. deaths):
      • R-CHOP: 62 events
      • RB-CHOP: 54 events
    • 30-month OS rates:
      • R-CHOP: 82.7% (95% CI, 78.2−3)
      • RB-CHOP: 83.6% (95% CI, 79.0−3)
      • Comparison: HR = 0.89; (95% CI, 0.62−1.28); P = 0.52
        • Adjusted HR = 0.85; (95% CI, 0.59−1.23); P = 0.40
      • After a median follow-up of 42.3 months:
        • 30-month PFS rates:
          • R-CHOP: 70.6% (95% CI, 65.5−0)
          • RB-CHOP: 75.2% (95% CI, 70.3−4)
          • Comparison: adjusted HR = 0.82; (95% CI, 0.63−1.08); P = 0.16
        • Secondary analysis by DLBCL molecular subtype revealed that bortezomib did not significantly affect PFS in the ABC (adjusted HR = 0.78, [95% CI, 0.51–1.21]; P = 0.27), GCB (HR = 0.85, 95% CI, [0·60–1.20]; P = 0.35), or unclassified patient subgroup (HR = 1.29, [95% CI, 0.77–2.16]; P = 0.34)

Safety

• Bortezomib was generally well tolerated, with the most common Grade ≥ 3 adverse events (AEs) being (R-CHOP vs RB-CHOP):
  • Hematological toxicity (39.8% vs 1%)
• No significant increase in the proportion of patients with Grade ≥ 3 neutropenia, febrile neutropenia, thrombocytopenia or anemia was observed in the RB-CHOP group as compared to the R-CHOP arm
• Any grade neuropathy was more frequently observed in the RB-CHOP (56.8%) versus the R-CHOP arm (41.6%; P< 0.0001)
• Serious adverse events:
  • R-CHOP: 42.5% (n = 190/447)
  • RB-CHOP: 50.2% (n = 223/444)
• Deaths in the safety population:
  • R-CHOP: 16.3% (n = 73/447)
  • RB-CHOP: 15·3% (n = 68/444)
• Reason of death:
  • Progressive lymphoma:
    • R-CHOP: 68·5% (n = 50/73)
    • RB-CHOP: 79·4% (n = 54/68)
  • Treatment-related:
    • R-CHOP: 6.8% (n = 5/73)
    • RB-CHOP: 9% (n = 4/68)
  • In the ITT population, dose reductions of any treatment drug were less frequent in the R-CHOP arm (34.5%), when compared to the RB-CHOP group (42.9%)

Conclusion

This multicenter phase III trial with real-time DLBCL molecular characterisation showed that bortezomib addition to R-CHOP did not improve PFS in patients with DLBCL, despite their molecular subtype (ABC, GCB or unclassified).