Tumor-Preferential Induction of Immune Responses and Epidermal Cell Death in Actinic Keratoses by Ingenol Mebutate

Abstract

The rapid and strong clinical efficacy of the first-in-class, ingenol mebutate, against actinic keratosis (AK) has resulted in its recent approval. We conducted the first comprehensive analysis of the cellular and molecular mode of action of topical ingenol mebutate 0.05% gel in both AK and uninvolved skin of 26 patients in a phase I, single-center, open-label, within-patient comparison. As early as 1 day after application, ingenol mebutate induced profound epidermal cell death, along with a strong infiltrate of CD4+ and CD8+ T-cells, neutrophils, and macrophages. Endothelial ICAM-1 activation became evident after 2 days. The reaction pattern was significantly more pronounced in AK compared with uninvolved skin, suggesting a tumor-preferential mode of action. Extensive molecular analyses and transcriptomic profiling of mRNAs and microRNAs demonstrated alterations in gene clusters functionally associated with epidermal development, inflammation, innate immunity, and response to wounding. Ingenol mebutate reveals a unique mode of action linking directly to anti-tumoral effects.

Trial registration: ClinicalTrials.gov NCT01387711.

Conflict of interest statement

S.E., H.A.H., M.S., and M.P.S. declare no competing financial interests. J.R.Z., T.L., A.H., and M.H.H. are employees of LEO Pharma A/S, the manufacturer of ingenol mebutate gel. This does not alter our adherence to PLOS ONE policies on sharing data and materials.

Figures

Fig 1. Inflammatory skin reactions induced by ingenol mebutate 0.05% gel are more pronounced in actinic keratosis (AK)-lesioned areas compared with uninvolved skin.

(A) CONSORT study diagram. This was a phase I, single-center, open-label, within-patient comparison trial to explore the biological effects of ingenol mebutate gel applied once daily for 2 consecutive days in patients with actinic keratosis (AK) in a 25-cm2 area on the extremities and a 25-cm2 area of uninvolved-skin (US) on the inner upper arm. (B) Typical skin reactions during the course of the trial in AK treatment areas of three representative patients at the dorsum of the hand (upper panel) (n = 26) and uninvolved skin of the inner upper arm (lower panel) (n = 26) at day 0 (baseline), day 1 (after one treatment application), and day 2 (after two treatment applications) as well as during follow-up at days 8 and 29. (B) The composite local skin response (LSR) score is the sum of six individual LSR scores including erythema, flaking/scaling, crusting, swelling, vesiculation/pustulation, and erosion/ulceration, which range from 0 to 4, with higher numbers indicating more severe reactions. This was calculated at each study visit for each patient, with a theoretical maximum composite score of 24. Patients were assessed on days 0, 1, 2, 8, and 29. No unexpected signs of local or systemic toxicity were noted. Illustrated are averages of the LSR; error bars indicate standard error of the mean.

Fig 2. Ingenol mebutate 0.05% gel treatment causes rapid infiltration of T-cells, macrophages, and neutrophilic granulocytes.

Biopsy specimens from all patients (n = 26) were subjected to histopathological evaluation. Skin tissues from the two treatment areas and all time points were assessed, and representative images are depicted. The panels depict hematoxylin & eosin staining as well as immunohistochemical staining for CD4+ T-lymphocytes, CD8+ T-lymphocytes, CD68+ macrophages/histiocytes, and myeloperoxidase (MPO+) neutrophils as indicated. The scale bar represents 100 μm.

Fig 3. Ingenol mebutate 0.05% gel activates cutaneous blood vessels and induces epidermal cell death.

Biopsy specimens from both treatment areas of all patients (n = 26) and all time points were assessed immunohistochemically for expression of CD20+ B-lymphocytes, ICAM-1 (CD54), and CD1a+ cells. Cleaved caspase 3 and TdT-mediated dUTP-biotin nick end labeling (TUNEL) reactivity indicating apoptotic responses were detected in five of these patients (arrows indicate examples of TUNEL positive epidermal cells). The figure depicts representative slides from all stainings as indicated. The scale bar represents 100 μm.

Fig 4. Cells of the adaptive and innate immune system increase strongly upon topical treatment with ingenol mebutate 0.05% gel.

For detailed quantification, two complementary quantitative analyses were conducted to ensure high-quality measurements. Automated histomorphometric analyses (A) and manual counting of positive cells (B) were performed on all samples (actinic keratosis [AK] lesions at day 0, day 1, and day 2, and uninvolved skin (US) at day 0 and day 2 from all patients (n = 26). Depicted data represent averages +/- standard error of the mean. There was a statistically significant overall concordance with the two technologies, e.g. for CD4+ epidermis (r = 0.4) and dermis (r = 0.4) and for CD8+ epidermis (r = 0.5) and dermis (r = 0.7). The data for the computer-based histomorphometrical analysis were logarithmically transformed. Statistical significance was determined by analysis of variance for: CD4+ T-lymphocytes, epidermis/dermis AK0 versus AK2 (P = 0.0007/P = 0.0006), US0 versus US2 (P < 0.0001/P < 0.0001) and for US2 versus AK2 (not significant [NS]/P < 0.0001), CD8+ T-lymphocytes, epidermis/dermis AK0 versus AK2 (P = 0.0024/P = 0.0019), US0 versus US2 (P < 0.0001/P < 0.0001), and for US2 versus AK2 (NS/P = 0.0065), CD68+ macrophages/histiocytes, dermis (epidermis was NS) AK0 versus AK2 (P < 0.0001), US0 versus US2 (P < 0.0001), and for US2 versus AK2 (P = 0.0042), and MPO+ neutrophils, epidermis/dermis AK0 versus AK2 (P < 0.0001/P < 0.0001), US0 versus US2 (P < 0.0001/P < 0.0001), and for US2 versus AK2 (NS/NS). A full statistical overview is available in S1 Table.

Fig 5. Topical ingenol mebutate 0.05% gel activates cutaneous blood vessels, attracts B-cells, and induces epidermal cell death.

Automated histomorphometric analyses on all samples from all patients (n = 26) were performed to quantitate expression of CD1a+ dendritic cells (A), CD20+ B-cells (B), and CD54 (ICAM-1) expressing cutaneous blood vessels and inflammatory cells within the dermis (C). Expression of both CD20 and CD54 increased rapidly and strongly upon treatment with ingenol mebutate, and actinic keratosis (AK) lesions showed generally higher expression levels compared with uninvolved skin. (D) Dyskeratotic (dead) keratinocytes within the epidermis of all samples (n = 26 patients; five biopsy specimens each) were assessed using a scoring system ranging from 0 (absent, not depicted), 1 (few dyskeratotic cells), 2 (several dead cells) to 3 (numerous necrotic cells). No necrosis was evident in the dermis. TdT-mediated dUTP-biotin nick end labeling (TUNEL)+ apoptotic cells (E) and cleaved caspase 3 (CC3)+ nuclei (F) were determined in biopsies from five patients. A full statistical overview is available in S1 Table.

Fig 6. Topical treatment with ingenol mebutate 0.05% gel affects expression of gene clusters relevant for inflammatory and wound healing responses, and induces a characteristic microRNA (miRNA) pattern.

The profiles illustrated are heat-maps and 2-way hierarchical clusterings of deregulated genes of (A) the 95 most variable mRNAs, and (B) the 64 most variable miRNAs across samples (n = 24) for the first six patients in the study. Differentially expressed genes and miRNAs were identified by pair-wise comparison of the following groups: actinic keratosis day 0 (AK0) versus uninvolved skin day 0 (US0), AK2 versus AK0, US2 versus US0, and AK2 versus US2. The expression analysis included variance filtering (VAR >0.2), statistical significance test by analysis of variance (P < 0.01), expression cut-off (>2-fold), and the false-discovery-rate was controlled by the Benjamini-Hochberg procedure (q < 0.05). Microarray data can be found in the GEO repository (GSE63107). Arrows indicate mRNA genes and miRNAs that were selected for validation by quantitative polymerase chain reaction. The colors above the heat map indicate: US0 (dark blue, n = 6), US2 (light blue, n = 6), AK0 (green, n = 6), and AK2 (red, n = 6) samples, respectively. The red and green shades on the heatmap represent up- and down-regulated genes, respectively.