AndreÌs Ferreri, MD
The MATRIX regimen, comprised of a methotrexate/cytarabine backbone plus thiotepa and rituximab, improved outcomes in patients with primary CNS lymphoma, according to results from the international randomized phase II ELSG #32 trial presented at the 13th International Conference on Malignant Lymphoma, held in Lugano, Switzerland.
Overall survival (OS) and failure-free survival (FFS) were improved with the MATRIX regimen compared with either methotrexate/cytarabine alone or in combination with rituximab. More patients with CNS lymphoma receiving MATRIX also achieved a complete remission (CR).
Lead investigator AndreÌs Ferreri, MD, Ospedale San Raffaele, Milano, Italy, presented results on behalf of the International Extranodal Lymphoma Study Group. The study randomized patients with histologically-proven primary CNS lymphoma and measurable disease to receive a maximum of four 3-week cycles of methotrexate at 3.5 g/m2
on day 1 and cytarabine at 2 g/m2
twice daily on days 2 and 3, either alone (arm A; n = 75), in combination with 375 mg/m2
of rituximab on day 5 (arm B; n = 69), or combined with rituximab at the same dose plus 30 mg/m2
of thiotepa on day 4 (MATRIX arm; n = 75).
The study was conducted at 52 locations across five countries. The median patient age was 58 years (range, 18-70) and all patients were HIV-negative. Overall, patients had an ECOG PS ≤3, with patients aged 66 to 70 years having an ECOG PS ≤2. Patient characteristics were well balanced among the study arms.
Autologous stem cells were collected after the second treatment course. Response was assessed after the second and fourth courses. Histology and neuroimaging were centrally reviewed. The coprimary endpoints of the study were CR and 2-year FFS.
A total of 733 (84%) of planned treatment courses were delivered. Successful collection of autologous stem cells was possible in 152 patients (94%).
At a median follow-up of 20 months, 118 patients were failure-free: 48 patients in the MATRIX arm, 35 patients in arm B, and 35 patients in arm A. Responding patients underwent a second randomization comparing consolidation with whole-brain irradiation or autologous stem cell transplantation.
Two-year OS rates were 66% (± 6%; P
= .01) in the MATRIX arm, 58% (± 6%) in arm B, and 40% (± 6%) in arm A. At 2 years, 124 patients are still alive: 52 patients in the MATRIX arm, 41 patients in arm B, and 31 patients in arm A.
FFS rates at 2 years were 64% (±6%) with MATRIX (P
= .0006), 52% (±6%) in arm B, and 34% (±6%) in arm A. Five-year FFS rates were 54% (±11%), 43% (±8%), and 34% (±10%), respectively (P
= .0001 for arm A vs C).
In the MATRIX arm, the overall response rate was 87% (95% CI, 80-94) compared with 74% (95% CI, 64-84), and 53% (95% CI, 42-64) in arms B and A, respectively (P
= .00001 for A vs C). The CR rate was 49% (95% CI, 38-60) in the MATRIX arm, compared with 31% in arm B (95% CI, 21-42), and 23% (95% CI, 14-31) in arm A (P
= .0007 for A vs C). A partial response was achieved by 37%, 43%, and 31%, of patients in the three arms, respectively. Stable disease was reported for 1%, 4%, and 8%, whereas progressive disease occurred in 8%, 16%, and 29% of patients in the MATRIX arm, arm B, and arm A, respectively.
“Conventional chemoradiotherapy is associated with suboptimal disease control and a high risk of neurotoxicity. In the search for new active drugs, there are suitable candidates. Rituximab has changed the natural history of diffuse large B-cell lymphoma but it is doubtful that it can cross the blood-brain barrier. Tiotepa is active against aggressive lymphomas and is able to cross the blood-brain barrier,” Ferreri said.
The most frequently reported adverse events were infective complications, which occurred at a similar incidence among the three cohorts.
G4 hematological toxicity occurred more often in the MATRIX arm; G4 neutropenia (P
= .01) and thrombocytopenia (P
= .0001) were reported significantly more frequently in patients receiving the MATRIX regimen.
Thirteen patients (6%) died due to toxicity and chemotherapy was interrupted due to toxicity in 21 patients (9%). “Toxic deaths and infectious complications were not significantly different across arms,” Ferreri said.
“There is a dire need for new agents to treat primary CNS lymphoma,” according to discussant Tracy T. Batchelor MD, MPH, Dana-Farber Cancer Institute, who was not involved in the study. “This was a well-conducted study, particularly since randomization was stratified by IELSG scores, which were balanced across treatment arms; the IELSG has raised the bar for randomized trials in our field,” he said.
No industry funding was reported for this study.