A Case-Matched Analysis Comparing Lenalidomide After Autologous or After Allogeneic Stem Cell Transplantation Demonstrates a Survival Advantage in Allografted Myeloma Patients

Lenalidomide (Len) is a highly effective drug against multiple myeloma (MM). It acts through several mechanisms, such as a direct cytotoxic effect, anti-angiogenesis, microenvironment modifications, and immunomodulation. The latter property is particularly interesting in the setting of allogeneic stem cell transplantation (AlloSCT), since Len may interact favourably with the graft-versus-myeloma (GVM) effect. On the other side, the possibility of an over activation of the donor immune system raises some concerns about the safety of this treatment. In order to verify if Len is more effective when given after AlloSCT, we conducted among 7 Italian transplant centers a case-matched analysis comparing Len after autologous SCT (AutoSCT) vs. Len after AlloSCT. The hypothesis is that Len is more active when administered after AlloSCT. A secondary end-point of the study was the safety profile.

In this retrospective study, the main matching criterion was represented by the number of treatment lines received before Len. In an attempt to uniform the treatment regimens, an intra-center matching was carried out. We collected data from 40 patients in each group. Baseline characteristics between Auto and Allo patients (pts) were similar, except for age at diagnosis (55 years, range 39–70, in Auto pts; 47 years, range 29 – 62, in Allo pts). The median number of previous lines of treatment was 3 (range 1–6) for both groups. Thirty-one (77%) Auto and 36 (90%) Allo pts received bortezomib. Similarly, 35 (87%) Auto and 21 (52%) Allo pts were previously treated with thalidomide. Twenty-one out of 40 (52%) Allo pts received SCT as a second or subsequent line of therapy. Before Len treatment, 14 (35%) Allo pts had acute graft-versus-host disease (aGVHD) (G1, 3 pts; G2, 8 pts; G3, 2 pts; G4, 1 pts), and 15 (37%) pts had chronic GVHD (cGVHD) (limited, 6 pts; extensive 9 pts). Median time between diagnosis and Len start was 65 months (range 14–162) in Auto, and 72 months(range 19–246) in Allo pts. Median time from transplant to Len start was 50 months (range 7–159) in Auto, and 21 months (range 6–134) in Allo pts. In all cases Len dosage was 25 mg, and it was combined with high or intermediate dose dexamethasone.

Best responses for Auto and Allo patients were as follows: 5 vs. 3 CR, 6 vs. 8 VGPR, 12 vs. 13 PR, 9 vs. 8 SD, 8 vs. 8 PD. Time from Len start to the best response was 4 months for both groups. With a median follow-up from Len start of 22 months (range 2–60+), 3-years progression-free survival (PFS) was 13% in the Auto, and 43% in the Allo group (p=0.03). Median PFS was 9 months in Auto, and 16 months in Allo pts. Median OS was 22 months in the Auto group, and was not reached in the Allo group (p=0.05). Hematological toxicities were similar among the 2 groups. Neutropenia was observed in 16 (40%) patients in both the Auto and Allo group, thrombocytopenia in 10 (25%) Auto and 8 (20%) Allo pts, and anemia in 5 (12%) Auto and 2 (5%) Allo pts. Non-hematological toxicities were mild in both groups. In the Allo group one patients developed aGVHD after donor lymphocyte infusion, and 3 pts had a worsening of a pre-existing extensive cGVHD.

The comparison between Auto and Allo pts has shown a benefit in terms of PFS and OS in the Allo group. This observation supports the hypothesis that Len may be synergic with the GVM effect. Interestingly, despite the same response rate between the 2 groups, clinical responses in Allo pts were more durable, reinforcing the hypothesis of the immunological control. On the other side, since Len has a potent immunomodulatory effect, this can raise concerns about its use after AlloSCT. The Dutch HOVON 76 trial has shown that the early administration of Len 10 mg after AlloSCT induces aGVHD in a substantial proportion of patients, causing unacceptable toxicity and the premature discontinuation of the study. However, in our retrospective study the median time between AlloSCT and Len administration was 21 months, and we did not observed any increase of GVHD. This suggests that the mechanisms of immune tolerance can mitigate the activation induced by Len, allowing a safe administration. Moreover, since in all cases dexamethasone was given with Len, its immunosuppressive effects may have harnessed the Len-induced immune activation, contributing to the treatment tolerability. In conclusion our case-matched study suggests that Len is more active when given after AlloSCT, still retaining a favorable toxicity profile.

No relevant conflicts of interest to declare.