Targeting myeloid-derived suppressor cells with gemcitabine to enhance efficacy of adoptive cell therapy in bladder cancer

Abstract

Background: New therapeutics in development for bladder cancer need to address the recalcitrant nature of the disease. Intravesical adoptive cell therapy (ACT) with tumor infiltrating lymphocytes (TIL) can potentially induce durable responses in bladder cancer while maximizing T cells at the tumor site. T cells infused into the bladder directly encounter immunosuppressive populations, such as myeloid derived suppressor cells (MDSCs), that can attenuate T cell responses. Intravesical instillation of gemcitabine can be used as a lymphodepleting agent to precondition the bladder microenvironment for infused T cell products.

Methods: Urine samples from bladder cancer patients and healthy donors were analyzed by flow cytometry and cytometric bead array for immune profiling and cytokine quantification. MDSCs were isolated from the urine and cocultured with stimulated T cells to assess effects on proliferation. An orthotopic murine model of bladder cancer was established using the MB49-OVA cell line and immune profiling was performed. MDSCs from tumor-bearing mice were cocultured with OT-I splenocytes to assess T cell proliferation. Mice received intravesical instillation of gemcitabine and depletion of immune cells was measured via flow cytometry. Bladder tumor growth of mice treated with intravesical gemcitabine, OT-I transgenic T cells, or combination was monitored via ultrasound measurement.

Results: In comparison to healthy donors, urine specimen from bladder cancer patients show high levels of MDSCs and cytokines associated with myeloid chemotaxis, T cell chemotaxis, and inflammation. T cells isolated from healthy donors were less proliferative when cocultured with MDSCs from the urine. Orthotopic murine bladder tumors also presented with high levels of MDSCs along with enrichment of cytokines found in the patient urine samples. MDSCs isolated from spleens of tumor-bearing mice exerted suppressive effects on the proliferation of OT-I T cells. Intravesical instillation of gemcitabine reduced overall immune cells, MDSCs, and T cells in orthotopic bladder tumors. Combination treatment with gemcitabine and OT-I T cells resulted in sustained anti-tumor responses in comparison to monotherapy treatments.

Conclusion: MDSCs are enriched within the microenvironment of bladder tumors and are suppressive to T cells. Gemcitabine can be used to lymphodeplete bladder tumors and precondition the microenvironment for intravesical ACT.

Keywords: adoptive cell therapy (ACT); gemcitabine; lymphodepletion; neoadjuvant; non-muscle invasive bladder cancer (NMIBC); tumor infiltrating lymphocytes (TIL).

Figure 1

Urine of bladder cancer patients is enriched for MDSCs. (A) Gating strategy for MDSCs using urine cells from bladder cancer patient (B) Graph showing the percentage of live MDSCs in the urine of bladder cancer patients and healthy donors. n=9-11 per group. (C) Comparison of percentages of live PMN-MDSC and M-MDSC in the urine of bladder cancer patients, n=9. Composition of (D) immune cells in bladder tumors (n=15) and (E) MDSCs in urine and blood (n=4-9). (F) CTV flow cytometry analysis of 1:1, 1:2, 1:4, 1:8 coculture ratios of urinary MDSCs to T cells from healthy donor PBMCs. (G) Measurement of incorporation of tritiated thymidine in 1:1, 1:2, and 1:4 coculture ratios of MDSCs to T cells. All statistical analysis performed using two-tailed Student’s t tests. *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001. With P values < 0.05.

Figure 2

Urine supernatants from bladder cancer patients are highly enriched for cytokines. Cytometric bead array analysis of urine supernatants from bladder cancer patients and healthy donors. n=11-12 per group. Graphs marked with † are outside the limits of detection of the assay. All statistical analysis performed using two-tailed Student’s t tests. ns, non-significant; p > 0.05; *, p ≤ 0.05; **, p ≤ 0.01. With P values < 0.05.

Figure 3

Murine bladder tumors are enriched for MDSCs that are suppressive to T cells. (A) Profile of immune populations in orthotopic MB49-OVA tumors normalized to number of live cells per milligram of tumor, n=7-8. (B) CTV flow cytometry analysis of 1:1, 1:2, 1:4, 1:8 coculture ratios of splenic MDSCs to OT-I T cells. (C) Graph showing expansion of CD8⁺ T cells in same coculture conditions. (D) Scanned tumor-bearing bladder tissue slides from mouse treated with intravesical instillation of OT-I T cells. Tumors are outlined in black, MDSCs are represented as green dots (identified via Ly6G staining), and blood vessels are represented as red circles (identified via CD31 staining). (E) Cytometric bead array analysis of urine supernatants from mice bearing orthotopic MB49-OVA tumors, n=3. All statistical analysis performed using two-tailed Student’s t tests. *, p ≤ 0.05. With P values < 0.05.

Figure 4

Intravesical instillation of gemcitabine reduces total number of immune cells. Graphs comparing total number of (A) CD45⁺ (B) MDSCs (C) CD8⁺ (D) CD4⁺ cells normalized to tumor mass in untreated mice and mice treated with intravesical instillation of 500 µg gemcitabine. n=12-13 per group. All statistical analysis performed using one-tailed Student’s t tests. *, p ≤ 0.05; ***, p ≤ 0.001. With P values < 0.05.

Figure 5

Gemcitabine treatment induces reduction of VEGF in tumor supernatants. Cytokine bead array analysis of supernatants collected from overnight cultures of tumor-bearing bladder fragments. n=8 per group. Graphs marked with † are outside the limits of detection of the assay. All statistical analysis performed using two-tailed Student’s t tests. ns, non-significant; p > 0.05; **, p ≤ 0.01. With P values < 0.05.

Figure 6

Pretreatment of gemcitabine combined with ACT improves anti-tumor responses. (A) Schematic showing timeline of treatments. (B) Individual growth curves of mice in the UNTX, GEM, OT-I, and GEM+OT-I groups. n=12-14 per group. (C) Average tumor growth curves for each treatment group. Statistical analysis performed via Compare Groups of Growth Curves (CGGC) test using 10,000 permutations. (D) Representative ultrasound images of each treatment group at day 20. *, p ≤ 0.05.