Impact of vitamin d on the colon tumor immune microenvironment: Results of a randomized clinical trial of preoperative vitamin D supplementation in patients with stage I-III colon cancer.

3607 Background: Although the anti-tumor properties of vitamin D in colorectal cancer (CRC) have been broadly reported, its potential role as a tumor microenvironment (TME) immunomodulator has not been well established. Vitamin D receptor (VDR) and CYP27B1 (1-α-hydroxylase; required to convert vitamin D to its active form) are critical factors in the vitamin D signaling pathway and harbor prognostic significance for CRC. Pre-clinical studies suggest that higher stromal VDR expression enhances chemotherapy efficacy. To characterize the effects of vitamin D on the TME in humans, we conducted a randomized placebo-controlled trial of preoperative high-dose vitamin D supplementation in stage I-III colon cancer patients. Methods: Forty-two patients were randomized to receive vitamin D3 50,000 IU/day x 7 days followed by 10,000 IU/day versus placebo prior to resection. Paired pre-treatment biopsy specimens and post-treatment surgical resection specimens from 19 patients who completed the assigned treatment and had sufficient tissue for analysis were collected for spatially-resolved immune cell profiling and evaluation of VDR and CYP27B1 expression using custom multiplex immunofluorescence. Blood samples were collected pre- and post-treatment to assess plasma 25-hydroxyvitamin D [25(OH)D] levels. Linear mixed effects models were used to compare tissue parameters; plasma 25(OH)D level changes were assessed using the Wilcoxon rank-sum test. Results: Patients randomized to high-dose vitamin D had increased post-treatment plasma 25(OH)D (median 18.0 to 63.6 (ng/ml)) whereas placebo-randomized patients did not (median 19.1 to 18.6 (ng/ml)) ( p <0.001). Analysis of both biopsy and resection tissue revealed highly heterogeneous VDR and CYP27B1 expression. VDR expression was highest in tumor cells, followed by immune and then stromal cells. Although resection tissue from vitamin D-treated patients showed lower VDR expression than specimens from placebo-treated patients in all cells ( p= 0.03), VDR was preferentially expressed by tumor cells at the tumor invasive front compared to tumor center in the vitamin D arm, while the opposite was observed in the placebo arm. Vitamin D supplementation did not alter CYP27B1 expression. While treatment did not change total immune cell density, resection tissue from vitamin D-treated patients had higher overall CD3 + CD8 + T cell density ( p= 0.04), particularly CD3 + CD8 + memory T cell density ( p= 0.02). Additionally, vitamin D supplementation led to greater co-localization between CD3 + CD8 + T cells and tumor cells ( p< 0.001) compared to placebo. Conclusions: In patients with stage I-III CRC, vitamin D supplementation led to changes in vitamin D pathway signaling and evidence of an anti-tumor effect in the TME. Further investigation into the underlying mechanism of action is warranted. Clinical trial information: NCT02172651 .