Clinical and pharmacodynamic analysis of pomalidomide dosing strategies in myeloma: impact of immune activation and cereblon targets

Abstract

In preclinical studies, pomalidomide mediated both direct antitumor effects and immune activation by binding cereblon. However, the impact of drug-induced immune activation and cereblon/ikaros in antitumor effects of pomalidomide in vivo is unknown. Here we evaluated the clinical and pharmacodynamic effects of continuous or intermittent dosing strategies of pomalidomide/dexamethasone in lenalidomide-refractory myeloma in a randomized trial. Intermittent dosing led to greater tumor reduction at the cost of more frequent adverse events. Both cohorts experienced similar event-free and overall survival. Both regimens led to a distinct pattern but similar degree of mid-cycle immune activation, manifested as increased expression of cytokines and lytic genes in T and natural killer (NK) cells. Pomalidomide induced poly-functional T-cell activation, with increased proportion of coinhibitory receptor BTLA(+) T cells and Tim-3(+) NK cells. Baseline levels of ikaros and aiolos protein in tumor cells did not correlate with response or survival. Pomalidomide led to rapid decline in Ikaros in T and NK cells in vivo, and therapy-induced activation of CD8(+) T cells correlated with clinical response. These data demonstrate that pomalidomide leads to strong and rapid immunomodulatory effects involving both innate and adaptive immunity, even in heavily pretreated multiple myeloma, which correlates with clinical antitumor effects. This trial was registered at www.clinicaltrials.gov as #NCT01319422.

© 2015 by The American Society of Hematology.

Figures

Figure 1

Clinical response and survival following pomalidomide/dexamethasone in Len-refractory myeloma. Patients were randomized to therapy with pomalidomide at 2 mg/day on a 28/28-day continuous schedule (n = 19; cohort 1), or 4 mg/day on a 21/28-day intermittent schedule (n = 20; cohort 2). All patients received Pom alone for cycle 1 and with weekly dexamethasone with cycle 2 and beyond. (A) Waterfall plot of maximal reduction in measurable disease. (B) Kaplan-Meier plot comparing event-free survival in the 2 cohorts. (C) Kaplan-Meier plot comparing overall survival in the 2 cohorts.

Figure 2

Mid-cycle changes in treatment-induced immune profile following Pom therapy. (A-C) Phenotypic and numeric changes in immune cells. (A) Changes in T, NK, B, and monocytes at baseline and 1 week after therapy. (B) Changes in the expression of NKG2D, CD16, and CD56 on NK cells at baseline and after 1 week of therapy. (C) Fold change (compared with baseline) in T cells (i) and NK cells (ii) at 1 week after initiation of therapy in patients with a 28/28-day schedule (cohort 1) or a 21/28-day schedule (cohort 2) at 7 days following initiation of therapy in cycle 1 (Pom alone) or with dexamethasone (cycle 2). (D-G) Changes in functional properties of T and NK cells. (D) Changes in cytokine profile of CD4+ or CD8+ T cells. Intracellular cytokine production by CD4/CD8+ T cells was analyzed by flow cytometry at baseline vs 7 days after initiation of therapy. (E) Changes in cytokine profile of NK cells. Intracellular cytokine production by NK cells was analyzed by flow cytometry at baseline vs 7 days after initiation of therapy. (F) Changes in cytolysis proteins in T and NK cells. Expression of granzymeB and perforin by NK and CD8+ T cells was analyzed by flow cytometry at baseline vs 7 days after initiation of therapy. (G) Changes in polyfunctional T cells. Intracellular cytokine production by CD4/CD8+ T cells was analyzed by flow cytometry at baseline vs 7 days after initiation of therapy. (H) Changes in CD25hi FOXP3hi Tregs. Changes in circulating CD25hiFOXP3hi Tregs following Pom therapy were analyzed by flow cytometry. (Left) Representative FACS plot. (Center) Summary of changes before and after therapy. (Right) Comparison of changes following cycles 1 and 2. (I) Changes in coinhibitory receptors. Changes in the expression of BTLA, PD1, and Tim3 on CD4/CD8+ T or NK cells at baseline and after 1 week of therapy.

Figure 3

Correlation of immunologic effects following Pom and clinical response to therapy. (A) Changes in T and NK cells vs objective clinical response (≥PR). (B) Changes in cytokine production by T cells vs objective clinical response (≥PR). (C) Changes in cytokine production by polyfunctional CD8+ T cells vs objective clinical response (≥PR).

Figure 4

Expression of cereblon, ikaros, and aiolos protein in tumor cells and its correlation with clinical response. Expression of these proteins in bone marrow biopsies at baseline at study entry was quantified by immunohistochemistry as an H-score. (A) Correlation between immunohistochemical H-scores for ikaros, cereblon, and aiolos protein expression in tumor cells. (B) Expression of ikaros, cereblon, and aiolos protein vs objective clinical response (≥PR).

Figure 5

Changes in ikaros levels in immune cells following Pom therapy. Pre- and posttherapy samples were thawed and analyzed together for the expression of intranuclear ikaros levels in T and NK cells by flow cytometry. Data are representative of 2 patients each in cohorts 1 and 2.