Pre-existing infiltration with T cells and CD14+ myeloid cells is associated with treatment response to imiquimod in primary and recurrent vulvar high-grade squamous intraepithelial lesions.

Imiquimod is a standard therapy option for vulvar high-grade squamous intraepithelial lesions (vHSIL). In a retrospective study, the pre-existing composition of the immune cell infiltrate was associated with clinical outcome after imiquimod treatment. To validate these findings, an in-depth analysis of the tumor microenvironment was performed on a vHSIL cohort treated with imiquimod in the prospective PITVIN randomized controlled trial (ClinicalTrials.gov identifier: NCT01861535). Pretreatment biopsies of per-protocol patients participating in the PITVIN trial allocated to the imiquimod arm were included (n=38). These were analyzed by multispectral immunofluorescence using two seven-color staining panels for T cell and myeloid cell composition. Samples were scanned with the Vectra imaging microscope. Cell phenotyping was conducted using semi-supervised machine learning. Quick complete response (qCR) (n=27) was defined as absence of clinical lesions at 16 weeks; slow complete response (n=4) as delayed clearance requiring treatment up to 6 months; and non-response (n=7) as persistent histological vHSIL after 6 months, necessitating extended treatment or surgery. Analysis of the tumor immune microenvironment revealed interpatient variability in immune cell infiltration. Immune infiltrate composition in primary and recurrent vHSIL was comparable and this was also reflected by their clinical responsiveness to imiquimod. Higher numbers of intraepithelial CD3+CD8-(CD4+) T helper cells CD4+FoxP3+ regulatory T cells and CD14+ inflammatory myeloid cells and lower numbers of intraepithelial CD33+ immature cells were detected in qCR when compared with other response groups. Importantly, the lesions of complete responders displayed a positive correlation between the numbers of CD4+, CD8+ T cells and CD14+ inflammatory myeloid cells infiltrating the stroma and epithelium, indicative of a coordinated immune response. Slow complete responders displayed increased intraepithelial infiltration of recently activated CD4+PD1+ T cells (p<0.05), but displayed a lower CD14+ and CD68+ myeloid cell infiltration when compared with qCR. The presence of a pre-existing coordinated infiltration of vHSIL lesions with CD4+ T cells, CD8+ T cells and CD14+ inflammatory myeloid cells in patients displaying a CR after imiquimod treatment confirms our previous findings and suggests their use as biomarkers to predict responsiveness, but may also function as biomarkers to reduce or extend treatment duration of imiquimod.