Phase 1 Results of Anti-PD-Ligand 1 (Durvalumab) & Lenalidomide in Patients with Cutaneous T Cell Lymphoma and Correlation with Programmed Death Ligand 1 Expression and Gene Expression Profile

Background:T cells in CTCL are functionally exhausted and are characterized by the expression of immune inhibitory molecules such as PD1 and PD-L1 (Cancer Immunol Res 6; 2018). These findings justify the evaluation of immune checkpoint inhibition to reverse T cell exhaustion in CTCL. We initiated a phase 1/2 clinical trial of lenalidomide and durvalumab (NCT03011814) to determine the safety and efficacy of this regimen. Durvalumab is a human monoclonal antibody with high affinity and selectivity for PD-L1, targeting exhausted T cells and distinct cells within their environment. Lenalidomide, an oral immunomodulatory drug (IMiD) and analog of thalidomide, has previously shown activity in CTCL (Blood 123; 2014). Durvalumab may restore an anti-tumor immune response, and the combination of durvalumab and lenalidomide may enhance immune checkpoint blockade-induced immune responses. Associations between immune checkpoints, gene expression profile and the clinical efficacy of durvalumab/ lenalidomide combination were evaluated. The primary objectives were to determine the recommended phase 2 dose of lenalidomide in combination with durvalumab and safety with primary endpoint of toxicity (using CTCAE 4.03). Secondary end points included objective response rate (ORR) and median duration. Relationships between gene expression profile (GEP), PD-L1 expression, and antitumor activity were exploratory end points.

Methods:Phase 1 portion to evaluate the safety and tolerability of the durvalumab and lenalidomide combination has completed enrollment. Pts were enrolled in sequential cohorts to receive durvalumab (fixed dose at 1500 mg) and dose escalation of lenalidomide (dose level 1 = 10 mg for all cycles; dose level 2 = 10 mg for cycle 1, 15 mg for all subsequent cycles; dose level 3 =10 mg for cycle 1, 15 mg for cycle 2, and 20 mg for all subsequent cycles) to characterize safety, efficacy and antitumor activity. Serial skin samples were collected to assess the impact on the tumor microenvironment and anti-tumor activity.

Results:Thirteen pts. were evaluable for toxicities and 12 patients evaluable for response. 10 males/3 females, median age 36 (29-72 y), with aggressive or refractory/advanced CTCL, clinical stages IB (2), IIA (3), IIB (6), IIIA (1), and aggressive epidermotropic CD8+ CTCL (1) and a median of prior systemic treatments of 3 (range, 2-8) have been enrolled. Dose level 1 (n=3), dose level 2 (n= 3), dose level 3 (n=4). Median follow up time was 16.1 months (range, 3.7-36.1) months. No serious AEs or DLTs were observed during the DLT evaluation period (cycles 1-3). The most frequently reported AEs were fatigue (n=8), skin pain (n=4), chills (n=3), anemia (n=3), and leukopenia (4). One grade 3 maculopapular rash (possibly due to lenalidomide) was observed, all other treatment-related AEs were grade 1/2 in severity. One patient discontinued treatment due to fatigue. Three patients developed grade 1/2 autoimmune thyroiditis that resolved with treatment. Median cycles of treatment were 7 (range, 1-28). Median duration of response was 6 (range, 1- 28+) months. Nine pts achieved PR, 2 pts maintained stable disease and 2 pts developed PD. Five pts remain on treatment. Gene expression levels for several checkpoints (PD1, PD-L1 & ICOS) (Cycle1 Day1 vs Cycle 2 Day15) were analyzed. Gene expression profile highlights downregulation of TNF-alpha signaling via NFkB, IFN-gamma, and PI3-AKT-mTOR signaling pathways in responders; up-regulation of MYC targets, IL2-, JAK-STAT and pro-inflammatory pathways were seen in PD compared to responders.

Conclusions:Anti-PD-L1/lenalidomide has significant clinical activity in refractory/advanced CTCL, which will be formally evaluated in the Phase 2 portion. Responses were durable and ongoing, and treatment was well tolerated. Dose escalation is up to maximum dose of 20 mg lenalidomide daily. Our preliminary results reveal adaptive and innate immune signatures that may be predictive of response to checkpoint blockade and yield insights into mechanisms of therapeutic resistance.