Lenalidomide Maintenance Significantly Improves Survival Figures in Patients with Relapsed Diffuse Large B-Cell Lymphoma (rDLBCL) Who Are Not Eligible for Autologous Stem Cell Transplantation (ASCT): Final Results of a Multicentre Phase II Trial

Background: Patients (pts) with rDLBCL not eligible for ASCT or experiencing relapse after ASCT have a low likelihood of cure. Single-drug maintenance after salvage therapy may be an attractive strategy to prolong survival in these pts. Lenalidomide (LEN) is a suitable candidate for long-lasting maintenance as it is an oral drug, active against DLBCL, that can be taken for years with an acceptable toxicity profile. Accordingly, we designed a multicentre phase II trial addressing LEN maintenance in pts with chemosensitive relapse of DLBCL not eligible for ASCT or experiencing relapse after ASCT (clinicaltrials.gov NCT00799513).

Methods: HIV-neg pts (age ≥18 ys) with histologically-proven de novo or transformed DLBCL and relapsed disease responsive to conventional rituximab-containing salvage therapy were registered and treated with LEN 25 mg/day for 21 days out of 28, until lymphoma progression or unacceptable toxicity. Primary endpoint was 1-year progression-free survival (PFS). Simon's two-stage optimal design was used. To demonstrate a 1-yr PFS improvement from 30% (P0) to 50% (P1), 47 pts (one-sided; α 5%; β 80%) were needed. Maintenance would be considered effective if ≥19 pts were progression-free survivors at 1 yr. Cell of origin was assessed by NanoString Technology and Hans algorithm, and cereblon expression was assessed by immunohistochemistry.

Results: 46 of 48 enrolled pts were assessable (median age 72 ys; range 34-86); 36 pts had de novo DLBCL, 10 had transformed DLBCL. All pts were previously treated with anthracycline- and rituximab-based combination, plus ASCT in 6 pts; the median TTP after the prior therapy was 16 months (range 3-121). Thirty-three pts were enrolled at 1st relapse, 13 at 2^nd relapse; salvage therapy contained high doses of cytarabine or ifosfamide in two-thirds of cases, and response was complete in 26 pts and partial in 20. Most pts had unfavourable features: IPI ≥2 in 38 (83%) pts, advanced stage in 35 (76%), extranodal disease in 29 (63%), high LDH level in 21 (46%); 28 (61%) pts were older than 70 ys.

At a median follow-up of 25 (range 6-87) months, 556 LEN courses were delivered, with an average of 12 courses/pt (range 3-41); 19 pts are still in treatment. LEN was well tolerated: with the exception of neutropenia, grade 3-4 toxicities were uncommon, occurring in ≤3% of delivered courses. Infections were rare, and well controlled with oral antibiotics (grade 1-2 in 8 courses; grade 3 in 3). LEN dose reduction was indicated in 23 pts (transient in 19), and was due to neutropenia (12), rash (7), diarrhoea (2), and neurotoxicity (2); LEN was discontinued in 6 of them. One (2%) pt died of acute toxicity (intestinal infarction) and one due to secondary myelodysplastic syndrome at 56 months of follow-up. Pts with HBV/HCV seropositivity (n=12) or prior ASCT (n=6) did not experience unexpected toxicity after >1 yr of treatment.

At one year from trial registration, 28 pts were still progression free, which was significantly higher than the pre-determined efficacy threshold (n≥19). During the whole observation period, there were 21 events: progressive disease in 19 pts, death of toxicity in one, death while off therapy in one, with a 1-yr PFS (primary endpoint) of 70 ± 7%. The duration of response to LEN was longer than response duration after the prior treatment line in 27 (59%) pts, and was twice as long in 15 of them.

The benefit of LEN maintenance was observed both in pts with de novo or transformed DLBCL. According to the Hans' algorithm, the 1-yr PFS was 64 ± 11% for GCB-DLBCL and 67 ± 11% for nonGCB-DLBCL (p= 0.67). Results using the Nanostring technique were consistent with the Hans' algorithm, with a concordance rate of 86%. There was no significant association between cereblon expression and PFS. Multivariate analysis confirmed that treatment at first relapse and a prior TTP ≥12 months were independently associated with better PFS. Overall, 33 (72%) pts are alive, with a 1- and 3-yr OS of 81 ± 6% and 71 ± 8%, respectively.

Conclusions: With the limitations of a non-randomized design, this trial soundly promotes the use of LEN maintenance in pts with chemosensitive relapse of DLBCL not eligible for ASCT or experiencing relapse after ASCT. LEN was well tolerated in this elderly population, with survival benefit both in pts with de novo or transformed DLBCL, and both in pts with GCB- or nonGCB-DLBCL. These results warrant further investigation of immunomodulatory drugs as maintenance in these high-risk pts.