Phase 1b/2 study evaluating safety, efficacy and immune effects of TLR9 agonist cavrotolimod with anti-PD-1 antibodies among patients with advanced solid tumors

Abstract

Background: There is an unmet need for novel immunotherapies to overcome immune evasion in patients with advanced skin cancers resistant to programmed death (PD)-1 / PD-ligand 1 (PD-L1) blockade. Cavrotolimod is a novel spherical nucleic acid configuration of a toll-like receptor 9 agonist oligonucleotide, designed to trigger innate and adaptive immune responses to tumors.

Patients and methods: The safety, pharmacokinetics, pharmacodynamics and preliminary efficacy of intratumoral cavrotolimod, first dosed alone and then in combination with anti-PD-1 antibodies (pembrolizumab or cemiplimab), were assessed in a combined Phase 1b/2 dose escalation/dose expansion study in patients with advanced skin cancers, including melanoma, Merkel cell carcinoma and cutaneous squamous cell carcinoma (www.

Clinicaltrials: gov; NCT03684785).

Results: A total of 58 patients (20 in dose-escalation and 38 in expansion cohorts) were enrolled. 55 (95%) of the 58 patients experienced progressive disease on prior anti-PD-(L)1 therapy. Cavrotolimod, in combination with anti-PD-1 therapy, produced objective responses in 6 (12%) and stable disease (SD) in 8 (16%) of 51 evaluable patients on this study, leading to a disease control rate of 27% (14/51). 5 of 6 (83%) patients with an objective response and 13 of 14 (93%) patients with disease control had progressed on prior anti-PD-(L)1 therapy. Disease control was durable, with median duration of 54 (range 24-88+) weeks for responses and 24 (range 11-35+) weeks for SD. Regression of both injected and non-injected tumors was observed. Cavrotolimod, alone and in combination with anti-PD-1 therapy, had a manageable safety profile with mostly transient adverse events (AEs). The most frequent Grade 3/4 cavrotolimod-related AEs were fatigue and injection site reactions. Cavrotolimod dosing was associated with robust chemokine/cytokine induction and lymphocyte activation in peripheral blood. Serial tumor biopsies of injected tumors suggested upregulation of genes associated with the interferon pathway, antiviral proteins, immune checkpoints, chemokines, granzymes and costimulatory proteins, along with increases in certain immune cell populations.

Conclusions: Cavrotolimod had a manageable safety profile and showed clinical activity in anti-PD-(L)1 refractory cutaneous malignancies, suggesting potential for further development as an antitumor immunotherapy in combination with other agents.

Trial registration number: NCT03684785.

Keywords: Immune Checkpoint Inhibitor; Intratumoral; Nanoparticle; Skin Cancer; Toll-like receptor - TLR.

Figure 1. AST-008–102 study overview. (A) Study design. (B) Schedule of dosing, tumor assessments and DLT periods. CSCC, cutaneous squamous cell carcinoma; DL, dose level; DLT, dose-limiting toxicity; MCC, Merkel cell carcinoma; N, number of patients enrolled; RP2D, recommended Phase 2 dose; SC, subcutaneous; IT, intratumoral.

Figure 2. Antitumor activity of cavrotolimod in combination with pembrolizumab or cemiplimab. (A) Best change from baseline of the sum of longest diameters (sLD) for target tumors by study phase (dose escalation, left panel; dose expansion, right panel). (B) Individual percent change from baseline in the sLD of target lesions over time. (C) Duration of treatment, onset, and duration of disease control. The cohort, tumor type (if not indicated by the cohort name) and dose are given. Photos: a patient in their 70s who had experienced PD on prior pembrolizumab monotherapy with melanoma metastases on the scalp. Cavrotolimod was injected into only the larger of the two pictured lesions at baseline (D) and at study week 12 (E). Regression of both the injected and non-injected tumors was observed at week 12. This patient went on to a CR at week 84. A patient in their 90s with three MCC lesions on the right arm after PD on prior pembrolizumab monotherapy. Cavrotolimod was injected into two lesions, one of which is pictured in (F). One injected lesion completely regressed by week 12 (not pictured) while the other completely regressed by week 36 (G). The non-injected lesion (pictured at baseline in (H)) completely regressed by week 24 (I). This patient experienced a CR at week 36. CEM, cemiplimab; CR, complete response; EOT, end of treatment; ESC, dose escalation cohort; MCC, Merkel cell carcinoma dose expansion cohort; Melanoma, melanoma dose expansion cohort; PD, progressive disease; PEM, pembrolizumab; PR, partial response; RECIST, Response Evaluation Criteria in Solid Tumors; SC, subcutaneous dose expansion cohort; SD, stable disease. *This patient was not anti-programmed death 1 antibody refractory prior to study enrollment.

Figure 3. Peripheral blood cytokine/chemokine levels are increased and lymphocytes are activated after intratumor or SC cavrotolimod. (A) Chemokine/cytokine concentrations and (B) lymphocyte activation by dose escalation level and cohort and in the blood before and after cavrotolimod alone or in combination with an anti-programmed death 1 antibody. The horizontal bar in each group represents the group mean. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001 versus baseline. Ordinary two-way analysis of variance and Dunnett’s multiple comparisons test of post-dose samples versus baseline were used. The number above each bar represents the number of patient samples in that group. CSCC, cutaneous squamous cell carcinoma; MCC, Merkel cell carcinoma; NK, natural killer; SC, subcutaneous.

Figure 4. Differential gene expression and immune cell composition changes elicited by cavrotolimod. (A) Volcano plot of differential gene expression comparing cavrotolimod-injected tumor RNA derived from lesions at screening and pre-dose week 3. Datapoints corresponding to differentially expressed genes (unadjusted p<0.05) are labeled with the gene name abbreviation. N=18 pairs of samples. (B) Changes in immune cell score in cavrotolimod-injected lesions. p<0.05, **p<0.01. Paired two-tailed t-test versus baseline. Cell abundance estimates from these gene expression signatures, referred to as cell type scores, are presented in log₂ scale so an increase of one unit corresponds to a doubling in abundance. N=18 and 14 paired biopsies for cavrotolimod alone, and cavrotolimod plus anti-PD-1, respectively. CR, complete response; NE, not evaluable; PD, progressive disease; PD-1, programmed death 1; PR, partial response; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease.