Modulation of the immune microenvironment of high-risk ductal carcinoma in situ by intralesional pembrolizumab injection

Abstract

Ductal carcinoma in situ (DCIS) is a risk factor for the subsequent development of invasive breast cancer. High-risk features include age <45 years, size >5 cm, high-grade, palpable mass, hormone receptor negativity, and HER2 positivity. We have previously shown that immune infiltrates are positively associated with these high-risk features, suggesting that manipulating the immune microenvironment in high-risk DCIS could potentially alter disease progression. Patients with high-risk DCIS were enrolled in this 3 × 3 phase 1 dose-escalation pilot study of 2, 4, and 8 mg intralesional injections of the PD-1 immune checkpoint inhibitor, pembrolizumab. Study participants received two intralesional injections, three weeks apart, prior to surgery. Tissue from pre-treatment biopsies and post-treatment surgical resections was analyzed using multiplex immunofluorescence (mIF) staining for various immune cell populations. The intralesional injections were easily administered and well-tolerated. mIF analyses demonstrated significant increases in total T cell and CD8⁺ T cell percentages in most patients after receiving pembrolizumab, even at the 2 mg dose. T cell expansion was confined primarily to the stroma rather than within DCIS-containing ducts. Neither cleaved caspase 3 (CC3) staining, a marker for apoptosis, nor DCIS volume (as measured by MRI) changed significantly following treatment. Intralesional injection of pembrolizumab is safe and feasible in patients with DCIS. Nearly all patients experienced robust total and CD8+ T cell responses. However, we did not observe evidence of cell death or tumor volume decrease by MRI, suggesting that additional strategies may be needed to elicit stronger anti-tumor immunity.

Fig. 1. T cell infiltrates in DCIS pre-treatment and post-treatment with intralesional anti-PD-1.

Representative hematoxylin and eosin (H&E) and multiplex immunofluorescence (mIF) images from pre-therapy biopsies and post-therapy surgical specimens are shown for each patient. For the mIF staining, CD3-positive cells are pseudocolored green, CD8-positive cells are magenta, and cytokeratin-positive cells are yellow. Original magnification: ×20.

Fig. 2. T cell infiltrates increased in 7 of 9 cases across all three dose cohorts.

Total T cell and CD8⁺ T cell density in response to 2 mg pembrolizumab (a, d), 4 mg pembrolizumab (b, e), and 8 mg pembrolizumab (c, f). Timepoint 0 (pre-therapy) and timepoint 1 (post-therapy) are shown. The bounds of each box represent the interquartile range (IQR). The horizontal line within a box is the median. Whiskers extend to the largest and smallest values no further than 1.5*IQR. Outlying points beyond the ends of the whiskers are plotted individually. Each pair of connected points represents one patient’s data; p-values are derived from paired t-tests.

Fig. 3. Changes in immune cell populations in response to intralesional anti-PD-1 therapy.

Cell densities pre-therapy (timepoint 0) and post-therapy (timepoint 1) are shown for (a) T cells, (b) cytotoxic T cells, (c) PD-1⁺ T cells, (d) PD-1⁺ cytotoxic T cells, (e) macrophages, (f) B cells, (g) regulatory T cells, and (h) PD-L1⁺ tumor cells. The bounds of each box represent the interquartile range (IQR). The horizontal line within a box is the median. Whiskers extend to the largest and smallest values no further than 1.5*IQR. Outlying points beyond the ends of the whiskers are plotted individually. Each pair of connected points represents one patient’s data; p-values derived from paired t-tests.

Fig. 4. Stromal T cell densities increased in response to intralesional anti-PD-1 therapy.

Total T cell (a, c) and CD8⁺ T cell (b, d) densities in the stromal regions (a, b) and intraductal regions (c, d) pre-therapy (timepoint 0) and post-therapy (timepoint 1). The bounds of each box represent the interquartile range (IQR). The horizontal line within a box is the median. Whiskers extend to the largest and smallest values no further than 1.5*IQR. Outlying points beyond the ends of the whiskers are plotted individually. Each pair of connected points represents one patient’s data; p-values are derived from paired t-tests.

Fig. 5. Changes in macrophage:T cell and Treg:T cell ratios in response to intralesional anti-PD-1 therapy.

a, b Comparison of pre-therapy (timepoint 0) and post-therapy (timepoint 1) macrophage:T cell ratio (a) and Treg:T cell ratio (b). The bounds of each box represent the interquartile range (IQR). The horizontal line within a box is the median. Whiskers extend to the largest and smallest values no further than 1.5*IQR. Outlying points beyond the ends of the whiskers are plotted individually. Each pair of connected points represents one patient’s data; p-values are derived from paired t-tests. c–f correlations between macrophage:T cell and Treg:T cell ratios and the change, pre-therapy versus post-therapy, in T cells (c, d) and in CD8+ T cells (e, f); R and p from Pearson’s correlation. Gray shaded area represents 95% confidence region.

Fig. 6. Densities of pre-treatment T cell populations correlated with pre-treatment PD-L1⁺ cell populations.

Correlation plots of pre-treatment densities of total T cells (a, b), CD8⁺ T cells (c, d), PD1⁺ T cells (e, f), and PD1⁺CD8⁺ T cells (g, h) with pre-treatment densities of PD-L1⁺ macrophages (b, d, f, h) and/or PD-L1⁺ tumor cells (a, c, e, g) are shown. R and p from Pearson’s correlation. Gray shaded area represents 95% confidence region.

Fig. 7. Intralesional anti-PD-1 did not decrease tumor cell proliferation or increase tumor cell apoptosis.

Proliferating Ki67⁺ tumor cell densities (a) and cleaved caspase 3 (CC3) positive tumor cell densities (b) at timepoint 0 (pre-therapy) and timepoint 1 (post-therapy) are shown. The bounds of each box represent the interquartile range (IQR). The horizontal line within a box is the median. Whiskers extend to the largest and smallest values no further than 1.5*IQR. Outlying points beyond the ends of the whiskers are plotted individually. Each pair of connected points represents one patient’s data; p-values are derived from paired t-tests.