RF1 - Pilot Study: Identification of the tumor microenvironment of premalignant and malignant anal lesions

Purpose/Background: While rare, anal cancer incidence has risen 2.2% each year over the past decade. Ninety percent of anal squamous cell cancers (ASCC) are caused by human papillomavirus (HPV). HPV undergoes host immune evasion via DNA integration and upregulation of immune-suppressive pathways, including tumor infiltrating lymphocytes (TIL) by promoting the expression of oncoproteins E6 and E7. Currently, anal cancer screening includes anal cytology and high-resolution anoscopy (HRA). Screening is burdened with sampling error, patient discomfort and requires specialized training. Identification of less invasive, simple, and readily available diagnostic anal tumor biomarkers are essential to improve anal cancer screening.

Methods/Interventions: This is a prospective, IRB-approved human translational study, which analyzes the tumor microenvironment of premalignant [low grade squamous intraepithelial (LSIL) and high grade squamous intraepithelial lesions (HSIL)] and malignant lesions using flow cytometry to phenotype the immune infiltrates.

Peripheral blood, anal pre/malignant specimens and normal anal tissue were obtained from individuals presumed to have at least squamous intra-epithelial lesions (SIL) who underwent an operative procedure, including examination under anesthesia with anoscopy, HRA, or abdominoperineal resection. Peripheral blood mononuclear cells (PBMC) were isolated from the patient’s peripheral blood and anal specimens were digested. A 28-color antibody panel plus a viability dye was used to stain the samples and samples were acquired via the Cytek® Aurora (Fremont, CA) spectral flow cytometer. Data was analyzed with FlowJo software.

Results/Outcomes: Ten patients underwent immunophenotyping. Demographics of the study population includes average age 48.9 (27-72 years), 7 (70%) male, 4 (40%) HIV positive, average CD4 count 520 (range 219-1049 cells/mm3), 7 (70%) HPV+ status, 2 (20%) patients with VIN or CIN, and 1 (10%) patient was a prior transplant recipient. Final pathology per patient was normal = 2 (20%), LSIL/HSIL = 5 (50%), ASCC = 3 (30%).

Normal tissue was characterized by infiltration of few lymphocytes, which were not activated, in either CD4 or CD8 T cell subset populations. In cancer lesions, the CD4 T cell subset contained significantly more regulatory T cells (Tregs) as assessed by the FoxP3 stain. While LSIL/HSIL contained a much higher number of immune cells, Tregs were virtually absent (Figure 1). Conventional CD4+ T cells (Tconv) in the malignant lesions expressed high levels of ICOS (inducible-costimulator) and PD-1, reflective of tumor-recognition and activation. CD4 Tconv in the premalignant infiltrates also expressed these markers but lacked co-expression of chronic activation markers such as CD39 and markers of proliferation such as Ki67 and CD38 (Figure 2).

CD8 TIL with the phenotype, CD103+ CD39+ (DP) T cells, indicative of chronic stimulation and tissue residency, were highly expressed in ASCC. These cells expressed markers of activation such as PD-1, TIGIT and Ki67. In premalignant lesions, DP CD8 T cells were not as prevalent.

Conclusion/Discussion: The phenotyping data demonstrates that T cell tumor infiltrates differ considerably between normal, low-and high-grade and malignant lesions – with the TIL from the cancer patients containing the highest number of activated, chronically stimulated CD4 and CD8 T cells. Development of a serum T cell biomarker test would be a game changer in anal cancer screening.

FoxP3 Regulatory T cell subsets within different tissue specimens
(A) Flow cytometric analysis for CD4 and CD8 T cells including the gating strategy for the normal tissue specimen. The normal specimen has overall lower lymphocyte infiltration. There are very few regulatory T cells (Tregs) outlined in green. (B) Flow cytometric analysis and gating strategy for the tissue with dysplasia. There are more lymphocytes present within the tissue sample but very few Tregs are noted within the CD4 T cell populations. (C) Flow cytometric analysis and gating strategy for anal squamous cell carcinoma. There are significantly more Treg cells present within the CD4 subset. The Treg cells are identified by the FOXP3 stain outlined above in red.

Flow cytometric comparison of activated memory CD8 and memory CD4 T conventional cells within normal, dysplastic and malignant specimens
(A) Flow cytometric analysis of the normal specimen contains very few CD103+ CD39+ [double positive (DP)] memory CD8+ T cells and very few DP ICOS+ PD-1+ CD4 T cells. Given the lack of double positive cells further analysis was not performed for activation markers. (B) Flow cytometric analysis of the tissue with dysplasia contains very few DP CD103+ CD39+ memory CD8+ T cells. There are a large proportion of DP ICOS+ PD-1+ CD4 T cells within this population, outlined in green, but once the DP cells are gated they do not appear to express the activation marker Ki-67. (C) Flow cytometric analysis of the cancer specimen demonstrates a higher presence of both DP CD103+ CD39+ CD8 and DP ICOS+ PD-1+ CD4 T cells. Gating the DP ICOS+ PD-1+ CD4 T cells demonstrates higher expression of activation markers, Ki-67 and CD39, outlined in red.